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 This is a security release.
10 JH/01 Bug 2999: Fix a possible OOB write in the external authenticator, which
11 could be triggered by externally-supplied input. Found by Trend Micro.
14 JH/02 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
15 be triggered by externally-controlled input. Found by Trend Micro.
18 JH/03 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
19 be triggered by externally-controlled input. Found by Trend Micro.
23 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
24 Make the rewrite never match and keep the logging. Trust the
25 admin to be using verify=header-syntax (to actually reject the message).
31 HS/01 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
37 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
38 SMTP connection" log lines.
40 JH/02 Option default value updates:
41 - queue_fast_ramp (main) true (was false)
42 - remote_max_parallel (main) 4 (was 2)
44 JH/03 Cache static regex pattern compilations, for use by ACLs.
46 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
47 Make the rewrite never match and keep the logging. Trust the
48 admin to be using verify=header-syntax (to actually reject the message).
50 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
51 (under Linux) we watch the dir containing the final file; previously
52 it would be the dir with the first symlink. We still do not monitor
55 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
56 dn_expand() is happy to pass them through.
58 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
59 the file with the proof had an unchanged name, the new proof(s) were
60 loaded on top of the old ones (and nover used; the old ones were stapled).
62 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
63 more than one message arrived in a single connection a reference from
64 the earlier message could be re-used. Often a sigsegv resulted.
65 These variables were introduced in Exim 4.87.
66 Debug help from Graeme Fowler.
68 JH/09 Fix ${filter } for conditions that modify $value. Previously the
69 modified version would be used in construction the result, and a memory
72 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
73 Find and fix by Jasen Betts.
75 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
76 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
77 the systemwide configuration to override the Exim config.
79 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
80 API changes in libopendmarc.
82 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
83 pid 1, in the normal "background daemon" mode, having to drop process-
84 group leadership also lost track of needing to create listener sockets.
86 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
87 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
89 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
90 a capture group which obtained no text (eg. "(abc)*" matching zero
91 occurrences) could cause a segfault if the corresponding $<n> was
94 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
95 included a close-brace character (eg. it itself used an expansion) an
98 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
99 starting TLS. Previously it was after, meaning that attackers on such
100 ports had to be screened using the host_reject_connection main config
101 option. The new sequence aligns better with the STARTTLS behaviour, and
102 permits defences against crypto-processing load attacks, even though it
103 is strictly an incompatible change.
104 Also, avoid sending any SMTP fail response for either the connect ACL
105 or host_reject_connection, for TLS-on-connect ports.
107 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
108 Previously this was not permitted, but it makes reasonable sense.
109 While there, restore a restriction on using it from a connect ACL; given
110 the change JH/16 it could only return false (and before 4.91 was not
113 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
114 was exactly sized compared to the log buffer, a crash occurred with the
115 misleading message "bad memory reference; pool not found".
116 Found and traced by Jasen Betts.
118 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
119 dns_again_means_nonexist included an element causing a DNS lookup which
120 iteslf returned DNS_AGAIN, unbounded recursion occurred. Possible results
121 included (though probably not limited to) a process crash from stack
122 memory limit, or from excessive open files. Replace this with a paniclog
123 whine (as this is likely a configuration error), and returning
126 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
127 this always failed, probably leading to the usual downgrade to in-clear
130 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
131 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
132 to not checking that list for these lookups.
134 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
137 JH/23 Fix crash in string expansions. Previously, if an empty variable was
138 immediately followed by an expansion operator, a null-indirection read
139 was done, killing the process.
141 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
142 include an SMTP response string which is longer than that supported
143 by the delivering transport. Alleviate by wrapping such lines before
146 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
147 chars (RFC limit). Previously a limit of 12 items was made, which with
148 a not-impossible References: in the message being bounced could still
149 be over-large and get stopped in the transport.
151 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
152 close. Previously a bare socket close was done.
154 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
157 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
158 message being considered for delivery was already being handled by
159 another process, and having an SMTP connection already open, the function
160 to close it tried to use an uninitialized variable. This would afftect
161 high-volume sites more, especially when running mailing-list-style loads.
162 Pollution of logs was the major effect, as the other process delivered
163 the message. Found and partly investigated by Graeme Fowler.
165 JH/29 Change format of the internal ID used for message identification. The old
166 version only supported 31 bits for a PID element; the new 64 (on systems
167 which can use Base-62 encoding, which is all currently supported ones
168 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
169 and must use Base-36). The new ID is 23 characters rather than 16, and is
170 visible in various places - notably logs, message headers, and spool file
171 names. Various of the ancillary utilities also have to know the format.
172 As well as the expanded PID portion, the sub-second part of the time
173 recorded in the ID is expanded to support finer precision. Theoretically
174 this permits a receive rate from a single comms channel of better than the
176 The major timestamp part of the ID is not changed; at 6 characters it is
177 usable until about year 3700.
178 Updating from previously releases is fully supported: old-format spool
179 files are still usable, and the utilities support both formats. New
180 message will use the new format. The one hints-DB file type which uses
181 message-IDs (the transport wait- DB) will be discarded if an old-format ID
182 is seen; new ones will be built with only new-format IDs.
183 Optionally, a utility can be used to convert spool files from old to new,
184 but this is only an efficiency measure not a requirement for operation
185 Downgrading from new to old requires running a provided utility, having
186 first stopped all operations. This will convert any spool files from new
187 back to old (losing time-precision and PID information) and remove any
188 wait- hints databases.
190 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
191 we treated them as item separators when parsing for a list item, but they
192 need to be protected by the doublequotes. While there, add handling for
195 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
196 Found and fixed by Jasen Betts. No testcase for this as my usual text
197 editor insists on emitting only valid UTF-8.
199 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
200 a null-indirection SIGSEGV for the receive process.
202 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
203 -bh use. Internal checking would spot this and take a panic.
205 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
206 In 4.96 this would expand to empty.
208 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
209 certificate. Find and fix by Andreas Metzler.
211 JH/36 Add ARC info to DMARC hostory records.
213 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
214 or fakedefer. Previously the sender could discover that the message
215 had in fact been accepted.
217 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
218 sequences. Previously the input was not noted as being tainted; notably
219 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
220 bad coding of authenticators.
222 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
223 and ${tr...}. Found and diagnosed by Heiko Schlichting.
225 JH/40 Support list of dkim results in the dkim_status ACL condition, making
226 it more usable in the data ACL.
228 JH/43 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
235 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
236 after reception to before a subsequent reception. This should
237 mean slightly faster delivery, and also confirmation of reception
240 JH/02 Move from using the pcre library to pcre2. The former is no longer
241 being developed or supported (by the original developer).
243 JH/03 Constification work in the filters module required a major version
244 bump for the local-scan API. Specifically, the "headers_charset"
245 global which is visible via the API is now const and may therefore
246 not be modified by local-scan code.
248 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
249 sendfile() didi not account for the way the ClamAV driver code called it.
251 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
252 time check was being done for every character; replace that with one
255 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
256 sent was prefixed with a length byte.
258 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
259 RFC 5321. Previously Dovecot (at least) would log errors during
262 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
263 in 4.61 without a reason logged. Binaries will be bigger, which might
264 matter on diskspace-constrained systems, but debug is easier.
266 JH/09 Fix macro-definition during "-be" expansion testing. The move to
267 write-protected store for macros had not accounted for these runtime
268 additions; fix by removing this protection for "-be" mode.
270 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
271 handing out large-numbered file descriptors, violating the usual Unix
272 assumption (and required by Posix) that the lowest possible number will be
273 allocated by the kernel when a new one is needed. In the daemon, and any
274 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
275 are not useable for FD_SET() [and hence select()] and overwrite the stack.
276 Assorted crashes happen.
278 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
279 main-section options or in a connect ACL, the value from the first ever
280 connection was never replaced for subsequent connections. Found by
283 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
284 though only once PCRE2 was introduced: the memory accounting used under
285 debug offset allocations by an int, giving a hard trap in early startup.
286 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
288 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
289 with underbars is given. The write-protection of configuration introduced
290 in 4.95 trapped when normalisation was applied to an option not needing
293 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
295 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
296 failing to close the certificates directory, every hour or any time it
299 JH/16 Debugging initiated by an ACL control now continues through into routing
300 and transport processes. Previously debugging stopped any time Exim
301 re-execs, or for processing a queued message.
303 JH/17 The "expand" debug selector now gives more detail, specifically on the
304 result of expansion operators and items.
306 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
307 bad comparison between the option value and the name of the file to
308 be included was done, and a mismatch was wrongly identified.
309 4.88 to 4.95 are affected.
311 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
313 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
314 supplied as the db file-pair basename. Previously, if a directory
315 path was given, for example via the autoreply "once" option, the DB
316 file.pag and file.dir files would be created in that directory's
319 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
320 "taint" log_selector. These were previously deprecated.
322 JH/22 Fix static address-list lookups to properly return the matched item.
323 Previously only the domain part was returned.
325 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
326 the call into OpenSSL to send a TLS Close was being repeated; this
327 resulted in the library waiting for the peer's Close. If that was never
328 sent we waited forever. Fix by tracking send calls.
330 JH/24 The ${run} expansion item now expands its command string elements after
331 splitting. Previously it was before; the new ordering makes handling
332 zero-length arguments simpler. The old ordering can be obtained by
333 appending a new option "preexpand", after a comma, to the "run".
335 JH/25 Taint-check exec arguments for transport-initiated external processes.
336 Previously, tainted values could be used. This affects "pipe", "lmtp" and
337 "queryprogram" transport, transport-filter, and ETRN commands.
338 The ${run} expansion is also affected: in "preexpand" mode no part of
339 the command line may be tainted, in default mode the executable name
342 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
343 the the facility was not passed across execs, and only the first message
344 passed over a connection could use BDAT; any further ones using DATA.
346 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
347 uses $sending_ip_address and an interface is specified.
348 Previously any use of the local address in the EHLO name disabled
349 PIPECONNECT, the common case being to use the rDNS of it.
351 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
352 resumption. Previously verify failed because no certificate status is
353 passed on the wire for the restarted session. Fix by using the recorded
354 ocsp status of the stored session for the new connection.
356 JH/29 TLS resumption: the key for session lookup in the client now includes
357 more info that a server could potentially use in configuring a TLS
358 session, avoiding oferring mismatching sessions to such a server.
359 Previously only the server IP was used.
361 JH/30 Fix string_copyn() for limit greater than actual string length.
362 Previously the copied amount was the limit, which could result in a
363 overlapping memcpy for newly allocated destination soon after a
364 source string shorter than the limit. Found/investigated by KM.
366 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
367 close; it may be needed for a subsequent connection. This caused a
368 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
370 JH/32 Fix CHUNKING for a second message on a connection when the first was
371 rejected. Previously we did not reset the chunking-offered state, and
372 erroneously rejected the BDAT command. Investigation help from
375 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
376 an empty address. Previously the expansion returned an error.
378 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
379 proxy. Previously these were misparsed, leading to paniclog entries.
385 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
386 related applications. Previously an "H" was used where available info
387 says that "M" should be, so change to match.
389 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
390 as arguments, so an implementation trying to copy these into a local
391 buffer was taking a taint-enforcement trap. Fix by using dynamically
392 created buffers. Similar fix for radius expansion condition.
394 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
395 reasonable, eg. to count headers. Fix by using dynamically created
396 buffers rather than a local. Do similar fixes for ACL actions "dcc",
397 "log_reject_target", "malware" and "spam"; the arguments are expanded
398 so could be handling tainted values.
400 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
401 broken the (no-op) support for this sendmail command. Restore it
402 to doing nothing, silently, and returning good status.
404 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
405 record path was given (or the default used) without a leading directory
406 path, an error occurred on trying to open it. Use the transport's working
409 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
410 transport. Previously it was the name on the DNS A-record; use instead
411 the head of the CNAME chain leading there (if there is one). This seems
412 to align better with RFC 6125.
414 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
415 smtp_accept_max_per_host allocated resources which were not released
416 when the limit was exceeded. This eventually crashed the daemon. Fix
417 by adding a release action in that path.
419 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
420 expanded; previously using tainted values was rejected. Fix by using
421 dynamically-created buffers.
423 JH/09 Relax restrictions on ACL verify condition needing access to message
424 headers. Previously they were only permitted in data and non-smtp ACLs;
425 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
426 not_blind, header_sender and header_names_ascii verification.
428 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
429 Previously a macro used one argument twice; when called with the
430 argument as an expression having side-effects, incorrect operation
431 resulted. Use an inlineable function.
433 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
434 held open for a verify callout. Previously this wan not accounted for
435 and a corrupt onward SMTP conversation resulted.
437 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
438 Previously they were embedded naively in the constructed address; when
439 needed, strip the quoting and quote the entire local_part.
440 Also make the inbound_srs expansion condition handle quoting.
442 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
443 excluded, not matching the documentation.
445 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
446 was given for the sqlite_dbfile a trap resulted.
448 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
449 "name" argument resulted in a trap. There is no reason to disallow such;
450 this was a coding error.
452 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
453 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
454 spent suspended, ignoring the POSIX definition. Previously we assumed
455 it did and a constant offset from real time could be used as a correction.
456 Change to using the same clock source for the start-of-message and the
457 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
458 exists, just to get a clock slightly more aligned to reality.
460 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
461 RFC says it is optional some validators care. The missing char was not
462 intended but triggered by a line-wrap alignment. Discovery and fix by
463 Guillaume Outters, hacked on by JH.
465 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
466 name being quoted was tainted a trap would be taken. Fix by using
467 dynamically created buffers. The routine could have been called by a
468 rewrite with the "h" flag, by using the "-F" command-line option, or
469 by using a "name=" option on a control=submission ACL modifier.
471 JH/19 SPF: change the Authentication-Results expansion component to give
472 smtp.helo when the sender domain is empty. Previously it gave
475 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
476 not in 127.0.0.0/8 to help in spotting list domains taken over by a
477 domain-parking registrar.
479 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
480 Previously when a whitespace character was specified it was not inserted
481 after removing the newline.
483 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
484 the domain part of the recipient address. This overrides any tls_sni
485 option set, which was previously used.
487 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
490 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
491 is_tainted() had an off-by-one error in the overenthusiastic direction.
492 Find and fix by Gavan. Although NetBSD is not a supported platform for
493 4.94 this bug could affect other platforms.
495 PP/01 Fix default prime selection to be consistent.
496 One path used ike23 still, instead of exim.dev.20160529.3; now both
497 execution flows will use the same DH primes (currently
498 exim.dev.20160529.3).
500 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
501 option in smtp transport, to match the documentation. Previously
502 verification was not being done in some cases where it should have been.
504 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
505 than one server was defined and depending on the platform memory layout
506 details, an internal consistency trap could be hit while walking the list
509 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
510 files. The value is used by the authresults expansion item. Previously
511 if this was used in a router or transport, a crash could result.
513 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
514 have a sufficient incidence of EINTR returns from select that an
515 interaction with other operations done by the main daemon loop exposed
516 a bug in the error-handling. This was benign apart from the log
519 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
520 for consistency with delivered "=>" and failed "**" lines. While we're
521 there, handle PRX and TFO.
523 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
524 applied. This resulted, if any header-line rewrite rules were configured,
525 in a panic-log triggerable by sending a message with a long address in
526 a header. Fix by increasing the arbitrary limit to larger than a single
527 (dewrapped) 5322 header line maximum size.
529 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
530 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
531 RFC 6648 which deprecates X- options in protocols as a general practice.
532 Changeover between the implementations is handled by the mechanisms
535 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
536 a remote which uses callout/hold. Previously the local was queued.
538 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
541 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
542 log line. The asterisk is now consistently appended to the remote IP
543 (and port, if given), and will also be provided on defer and fail log
544 lines. Previously it could be placed on the local IP if that was being
545 logged, and was only provided on delivery lines.
547 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
549 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
550 authenticator client_send option. Previously the next char, after a pair
551 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
552 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
553 documentation. There is still no way to get a leading ^ immediately
554 after a NUL (ie. for the password of a PLAIN method authenticator.
556 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
557 files. For bad sizes read, delete the record and whine to paniclog.
559 JH/38 When logging an AUTH failure, as server, do not include sensitive
560 information. Previously, the credentials would be included if given
561 as part of the AUTH command line and an ACL denied authentication.
563 JH/39 Bug 2691: fix $local_part_data. When the matching list element
564 referred to a file, bad data was returned. This likely also affected
567 JH/40 The gsasl authenticator now supports caching of the salted password
568 generated by the client-side implementation. This required the addition
569 of a new variable: $auth4.
571 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
572 left undeleted; the attempt to re-create it then failed - resulting in
573 the usual "SIGHUP tp have daemon reload configuration" to not work.
574 This affected any platform not supporting "abstract" Unix-domain
575 sockets (i.e. not Linux).
577 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
578 recipients" response to RCPT in a later response, with a 250. The
579 previous coding assumed this would not happen, and under PIPELINING
580 would result in both lost and duplicate recipients for a message.
582 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
583 Previously the weighting was incorrectly applied. Similar fix for socks
584 proxies. Found and fixed by Heiko Schlichting.
586 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
587 not handle sub-lists included using the +namedlist syntax. While
588 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
589 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
592 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
593 Previously the permanent pool was used, so the sore could not be freed.
594 This meant a connection with many messages would use continually-growing
597 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
598 per-pool so as not to waste too much space. Previously a constant size
599 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
600 DOS attacks harder. The cost is wasted memory use in the larger blocks.
602 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
603 same space repeatedly, and a smaller process footprint.
605 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
606 for GnuTLS. Previously the name started "%s" which, while not a bug,
607 looked as if if might be one.
609 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
610 "continued connection") make the $proxy_* variables available. Previously
611 the information was not passed across the exec() call for subsequent
612 transport executions. This also mean that the log lines for the
613 messages can show the proxy information.
615 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
616 receive time. With modern systems the difference is significant.
617 The historical behaviour can be restored by disabling (a new) log_selector
618 "queue_time_exclusive".
620 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
621 content) were specifically tested for. Now, also cover expansions
622 resulting in ACL names and inline ACL content.
624 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
625 making it unusable in complex expressions.
627 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
628 fail with a taint-check on the spoolfile name, and leave the message
631 HS/01 Enforce absolute PID file path name.
633 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
635 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
637 PP/02 Bug 2643: Correct TLS DH constants.
638 A missing NUL termination in our code-generation tool had led to some
639 incorrect Diffie-Hellman constants in the Exim source.
640 Reported by kylon94, code-gen tool fix by Simon Arlott.
642 PP/03 Impose security length checks on various command-line options.
643 Fixes CVE-2020-SPRSS reported by Qualys.
645 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
646 better. Reported by Qualys.
648 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
649 providing a particularly obnoxious sender full name.
652 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
654 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
657 PP/08 Change default for recipients_max from unlimited to 50,000.
659 PP/09 Fix security issue with too many recipients on a message (to remove a
660 known security problem if someone does set recipients_max to unlimited,
661 or if local additions add to the recipient list).
662 Fixes CVE-2020-RCPTL reported by Qualys.
664 PP/10 Fix security issue in SMTP verb option parsing
665 Fixes CVE-2020-EXOPT reported by Qualys.
667 PP/11 Fix security issue in BDAT state confusion.
668 Ensure we reset known-good where we know we need to not be reading BDAT
669 data, as a general case fix, and move the places where we switch to BDAT
670 mode until after various protocol state checks.
671 Fixes CVE-2020-BDATA reported by Qualys.
673 HS/03 Die on "/../" in msglog file names
675 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
676 the Exim runtime user are allowed to create files.
678 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
681 QS/03 When reading the output from interpreted forward files we do not
682 pass the pipe between the parent and the interpreting process to
683 executed child processes (if any).
685 QS/04 Always die if requested from internal logging, even is logging is
688 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
689 the API; compilation noo longer completes with DMARC support included.
690 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
691 been reported on other platforms.
693 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
695 JH/56 Make the majority of info read from config files readonly, for defence-in-
696 depth against exploits. Suggestion by Qualys.
697 Not supported on Solaris 10.
699 JH/57 Fix control=fakreject for a custom message containing tainted data.
700 Previously this resulted in a log complaint, due to a re-expansion present
701 since fakereject was originally introduced.
703 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
704 modifier was given, a loop resulted.
706 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
707 pipelined SMTP command followed the BDAT LAST then it would be
708 incorrectly treated as part of the message body, causing a verification
711 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
712 require looser validation rules than those for 821-level addresses,
713 which only permit IP addresses.
719 JH/01 Avoid costly startup code when not strictly needed. This reduces time
720 for some exim process initialisations. It does mean that the logging
721 of TLS configuration problems is only done for the daemon startup.
723 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
725 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
726 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
729 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
730 DKIM signing is being done. Previously a transport_filter would always
731 disable CHUNKING, falling back to traditional DATA.
733 JH/05 Regard command-line recipients as tainted.
735 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
737 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
738 PAM library frees one of the arguments given to it, despite the
739 documentation. Therefore a plain malloc must be used.
741 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
742 on-stack buffers were used, resulting in a taint trap when DSN information
743 copied from a received message was written into the buffer.
745 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
746 the ordering of its ARC headers. This caused a crash.
748 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
749 a new record was being constructed with information from the peer, a trap
752 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
753 installation would get error messages from DMARC verify, when it hit the
754 nonexistent file indicated by the default. Distros wanting DMARC enabled
755 should both provide the file and set the option.
756 Also enforce no DMARC verification for command-line sourced messages.
758 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
759 could, depending on the platform, hang at the STARTTLS response.
761 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
762 message on a connection. Previously if one message had ARC headers and
763 the following one did not, a crash could result when adding an
764 Authentication-Results: header.
766 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
767 Exim main code and Exim-related utities. The introduction of taint
768 tracking also did many adjustments to string handling. Since then, eximon
769 frequently terminated with an assert failure.
771 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
772 check for 452 responses. This slightly helps the inefficieny of doing
773 a large alias-expansion into a recipient-limited target. The max_rcpt
774 transport option still applies (and at the current default, will override
775 the new feature). The check is done for either cause of synch, and forces
776 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
777 connection. The new facility is not tunable at this time.
779 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
780 library live data was being used, so the results became garbage. Make
781 copies while it is still usable.
783 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
784 on delivery deferred (==) and failed (**) lines (if a delivery was
785 attemtped). Previously it was only on completion (=>) lines.
787 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
788 for the expansion of the server_scram_iter and server_scram_salt options.
790 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
791 are now specifically given a NO_DATA response without hitting the system
792 resolver. The library goes on to do the now-standard TXT lookup.
793 Use of dnsdb lookups is not affected.
795 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
796 only retrieve the errormessage once. Previously two calls to dlerror()
797 were used, and the second one (for mainlog/paniclog) retrieved null
800 JH/20 Taint checking: disallow use of tainted data for
801 - the appendfile transport file and directory options
802 - the pipe transport command
803 - the autoreply transport file, log and once options
804 - file names used by the redirect router (including filter files)
806 - paths used by single-key lookups
807 Previously this was permitted.
809 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
810 adjusted the size of a major service buffer; this failed because the
811 buffer was in use at the time. Change to a compile-time increase in the
812 buffer size, when this authenticator is compiled into exim.
814 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
815 previous fast-mode was untenable in the face of glibs using mmap to
816 support larger malloc requests.
818 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
819 New values supported, if defined on system where compiled:
820 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
821 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
823 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
824 running a limited number of proceses in parallel, a benefit is gained. The
825 amount varies with the platform hardware and load. The use of the option
826 queue_run_in_order means we cannot do this, as ordering becomes
829 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
830 had introduced a string-copy (for ensuring NUL-termination) which was not
831 appropriate for that case, which can include embedded NUL bytes in the
832 block of data. Investigation showed the copy to actually be needless, the
833 data being length-specified.
835 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
836 done during a receiving connection, and both used TLS, global info was
837 used rather than per-connection info for tracking the state of data
838 queued for transmission. This could result in a connection hang.
840 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
841 Previously, when delivering serveral messages down a single connection
842 only the first would provide a SIZE. This was due to the size information
843 not being properly tracked.
845 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
846 TAI (at 37 seconds currently), pretend to be in UTC for time-related
847 expansion and logging. Previously, spurious values such as a future
848 minute could be seen.
850 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
851 it could crash from a null-deref. This could also affect the
852 ${addresses: } operator and ${readsock } item.
854 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
855 message following a mime one, the variable was not reset.
857 JH/30 When an pipelined-connect fails at the first response, assume incorrect
858 cached capability (perhaps the peer reneged?) and immediately retry in
861 JH/31 Fix spurious detection of timeout while writing to transport filter.
863 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
864 an attempt to copy the string was made before checking it.
866 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
867 taint of the lookup key was maintained; we now regard the presence in the
868 filesystem as sufficient validation.
870 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
871 argument is supplied.
873 JH/35 The dsearch lookup now requires that the directory is an absolute path.
874 Previously this was not checked, and nonempty relative paths made an
875 access under Exim's current working directory.
877 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
878 Previously no event was raised.
880 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
881 parameter supplied by the sender MAIL FROM command. Previously it was
882 ignored, and only the check_spool_space option value for the required
885 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
886 the size of the signing public-key. Previously it was instead giving
887 the size of the signature hash.
889 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
890 the default. See the (new) dkim_verify_min_keysizes option.
892 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
893 and an ACL use a lookup for checking either the local_part or domain,
894 stale data could be accessed. Ensure that variable references are
895 dropped between messages.
897 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
898 by the client was not checked as pointing within response data before
899 being used. A malicious client could thus cause an out-of-bounds read and
900 possibly gain authentication. Fix by adding the check.
902 JH/42 Internationalisation: change the default for downconversion in the smtp
903 transport to be "if needed". Previously it was "as previously set" for
904 the message, which usually meant "if needed" for message-submission but
905 "no" for everything else. However, MTAs have been seen using SMTPUTF8
906 even when the envelope addresses did not need it, resulting in forwarding
907 failures to non-supporting MTAs. A downconvert in such cases will be
908 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
909 transport. The change does mean that addresses needing conversion will
910 be converted when previously a delivery failure would occur.
912 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
913 response was received it would be used unchecked in a fail-DSN, violating
914 standards on line-length limits. Truncate if needed.
916 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
917 doesn't work. (Additionally add a new main config option to configure the
924 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
925 side, to decode a TLS 1.3 packet capture.
927 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
928 Previously the default library behaviour applied, sending two, each in
931 JH/03 Debug output for ACL now gives the config file name and line number for
934 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
936 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
938 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
939 buffer overrun for (non-chunking) other transports.
941 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
942 TLS1.3, means that a server rejecting a client certificate is not visible
943 to the client until the first read of encrypted data (typically the
944 response to EHLO). Add detection for that case and treat it as a failed
945 TLS connection attempt, so that the normal retry-in-clear can work (if
946 suitably configured).
948 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
949 and/or domain. Found and fixed by Jason Betts.
951 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
952 configuration). If a CNAME target was not a wellformed name pattern, a
955 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
956 the OS reports them interleaved with other addresses.
958 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
959 used both for input and for a verify callout, both encrypted, SMTP
960 responses being sent by the server could be lost. This resulted in
961 dropped connections and sometimes bounces generated by a peer sending
964 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
965 string. Previously it was possible to cause undefined behaviour in a
966 library routine (usually a crash). Found by "zerons".
968 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
971 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
972 API was removed, so update to use the newer ones.
974 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
975 any timeout set, is taking a long time. Previously we would hang on to a
976 rotated logfile "forever" if the input was arriving with long gaps
977 (a previous attempt to fix addressed lack, for a long time, of initial
980 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
981 shared (NFS) environment. The length of the tempfile name is now
982 4 + 16 ("hdr.$message_exim_id") which might break on file
983 systems which restrict the file name length to lower values.
986 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
987 shared (NFS) environment.
989 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
990 did for all versions <4.90). Notably -M, -m, --invert, -I may be
993 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
994 on some platforms for bit 31.
996 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
997 to changes apparently associated with TLS1.3 handling some of the APIs
998 previously used were either nonfunctional or inappropriate. Strings
999 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1000 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1001 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1002 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1003 and the use of specific cipher names in the encrypted= ACL condition.
1005 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1007 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1008 verification result was not updated unless hosts_require_ocsp applied.
1010 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1011 queue_list_requires_admin set to false, non-admin users were denied the
1014 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1015 directory-of-certs mode. Previously they were advertised despite the
1018 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1019 A single TCP connection by a client will now hold a TLS connection open
1020 for multiple message deliveries, by default. Previously the default was to
1023 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1024 default. If built with the facility, DANE will be used. The facility
1025 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1027 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1028 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1029 must be defined and you must still, unless you define DISABLE_TLS, manage
1030 the the include-dir and library-file requirements that go with that
1031 choice. Non-TLS builds are still supported.
1033 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1036 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1037 default. If the platform supports and has the facility enabled, it will
1038 be requested on all coneections.
1040 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1041 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1043 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1045 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1046 success-DSN messages. Previously the From: header was always the default
1047 one for these; the option was ignored.
1049 JH/28 Fix the timeout on smtp response to apply to the whole response.
1050 Previously it was reset for every read, so a teergrubing peer sending
1051 single bytes within the time limit could extend the connection for a
1052 long time. Credit to Qualsys Security Advisory Team for the discovery.
1054 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1055 delivery address, which leaked information of the results of local
1056 forwarding. Change to the original envelope recipient address, per
1059 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1060 requested. Previously not bounce was generated and a log entry of
1061 error ignored was made.
1063 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1065 JH/32 Introduce a general tainting mechanism for values read from the input
1066 channel, and values derived from them. Refuse to expand any tainted
1067 values, to catch one form of exploit.
1069 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1070 was unused and the unexpanded text used for the test. Found and
1071 fixed by Ruben Jenster.
1073 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1074 an attempt to use a TLS library read routine dereffed a nul pointer,
1077 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1078 them in our smtp response.
1080 JH/36 Have the general router option retry_use_local_part default to true when
1081 any of the restrictive preconditions are set (to anything). Previously it
1082 was only for check_local user. The change removes one item of manual
1083 configuration which is required for proper retries when a remote router
1084 handles a subset of addresses for a domain.
1086 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1087 link count into consideration.
1089 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1090 caused the extension of big_buffer, the following lines were ignored.
1092 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1093 accordance with RFC 2308. Previously there was no expiry, so a longlived
1094 receive process (eg. due to ACL delays) versus a short SOA value could
1097 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1099 JH/39 Promote DMARC support to mainline.
1101 JH/40 Bug 2452: Add a References: header to DSNs.
1103 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1104 parameters. The relevant library call is documented as "Deprecated: This
1105 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1106 3.6.0, DH parameters are negotiated following RFC7919."
1108 HS/06 Change the default of dnssec_request_domains to "*"
1110 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1111 carried on and emitted a BDAT command, even when PIPELINING was not
1114 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1115 buffer was used for the filename, resulting in a trap when tainted
1116 arguments (eg. $domain) were used.
1118 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1119 recommended to avoid a possible server-load attack. The feature can be
1120 re-enabled via the openssl_options main cofiguration option.
1122 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1123 for version 4.90 - adding a "more data" boolean to the arguments.
1124 Bumped the ABI version number also, this having been missed previously;
1125 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1126 respect of usage of smtp_printf() by either local_scan code or libraries
1127 accessed via the ${dlfunc } expansion item. Both will need coding
1128 adjustment for any calls to smtp_printf() to match the new function
1129 signature; a FALSE value for the new argument is always safe.
1131 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1132 the file-offset (which the Linux syscall does, and exim expects); this
1133 resulted in an indefinite loop.
1135 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1136 to do ARC verification. The Authentication-Results: header line added
1137 by the configuration then had no ARC item.
1139 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1140 an incorrect mode was used for reading the result, resulting in it being
1147 JH/01 Remove code calling the customisable local_scan function, unless a new
1148 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1150 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1151 non-signal-safe functions being used.
1153 JH/03 Bug 2269: When presented with a received message having a stupidly large
1154 number of DKIM-Signature headers, disable DKIM verification to avoid
1155 a resource-consumption attack. The limit is set at twenty.
1157 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1158 report of oldest_pass in ${authres } in consequence, and separate out
1159 some descriptions of reasons for verification fail.
1161 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1162 files in the spool were present and unlocked. A queue-runner could spot
1163 them, resulting in a duplicate delivery. Fix that by doing the unlock
1164 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1165 add more error-checking on spoolfile handling while that code is being
1168 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1169 No known attacks, no CVE, this is defensive hardening.
1171 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1172 a queue-runner could start a delivery while other operations were ongoing.
1173 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1174 Found and investigated by Tim Stewart. Fix by using the open message data
1175 file handle rather than opening another, and not locally closing it (which
1176 releases a lock) for that case, while creating the temporary .eml format
1177 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1179 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1180 $sender_verify_failure/$recipient_verify_failure to "random".
1182 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1185 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1186 Previously this would segfault.
1188 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1191 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1192 like zero, since the resolver should be doing this for us, But we need one
1193 as a CNAME but no MX presence gets the CNAME returned; we need to check
1194 that doesn't point to an MX to declare it "no MX returned" rather than
1195 "error, loop". A new main option is added so the older capability of
1196 following some limited number of chain links is maintained.
1198 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1200 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1201 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1202 now-deprecated earlier definitions used only specified the range up to TLS
1203 1.2 (in the older-version library docs).
1205 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1207 JH/15 Rework TLS client-side context management. Stop using a global, and
1208 explicitly pass a context around. This enables future use of TLS for
1209 connections to service-daemons (eg. malware scanning) while a client smtp
1210 connection is using TLS; with cutthrough connections this is quite likely.
1212 JH/16 Fix ARC verification to do AS checks in reverse order.
1214 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1216 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1217 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1218 Previously the "utf8" would be re-prepended for every additional message.
1220 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1221 Previously thery were accepted, resulting in issues when attempting to
1222 forward messages to a non-supporting MTA.
1224 PP/02 Let -n work with printing macros too, not just options.
1226 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1227 one parent address was copied, and bogus data was used at delivery-logging
1228 time. Either a crash (after delivery) or bogus log data could result.
1229 Discovery and analysis by Tim Stewart.
1231 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1232 Previously if the string ended mid-character, we did not insert the
1233 promised '?' replacement.
1235 PP/04 Documentation: current string operators work on bytes, not codepoints.
1237 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1238 should pack well giving a smaller memory footprint so better caching and
1239 therefore performance. Group the declarations where this can't be done so
1240 that the byte-sized flag variables are not interspersed among pointer
1241 variables, giving a better chance of good packing by the compiler.
1243 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1244 non-null, to avoid issues with sites running BATV. Previously reports were
1245 sent with an empty envelope sender so looked like bounces.
1247 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1248 The ignore_error flag wasn't being returned from the filter subprocess so
1249 was not set for later routers. Investigation and fix by Matthias Kurz.
1251 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1252 and a msg:complete for the whole, when a message is manually removed using
1253 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1255 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1256 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1257 hope that is portable enough.
1259 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1260 requiring a known-CA anchor certificate; make it now rely entirely on the
1261 TLSA as an anchor. Checking the name on the leaf cert against the name
1262 on the A-record for the host is still done for TA (but not for EE mode).
1264 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1265 included in delivery lines for non-proxied connections, when compiled with
1266 SUPPORT_SOCKS and running with proxy logging enabled.
1268 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1269 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1270 move the existing event to fire before the normal logging of message
1271 failure so that custom logging is bracketed by normal logging.
1273 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1274 msg:fail:internal event. Developement by Matthias Kurz.
1276 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1277 far too small for todays use of crypto signatures stored there. Go all
1278 the way to the max DNS message size of 64kB, even though this might be
1279 overmuch for IOT constrained device use.
1281 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1282 copy a string over itself. The library routine is documented as not
1283 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1285 JH/32 For main options check_spool_space and check_inode_space, where the
1286 platform supports 64b integers, support more than the previous 2^31 kB
1287 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1288 the previous G, M, k.
1290 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1291 $authenticated_fail_id variable on authentication failure. Previously
1294 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1295 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1296 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1297 GNUTLS_SEC_PARAM_MEDIUM.
1299 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1300 side. Previously we would continue as if no SNI had been received.
1302 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1303 "<" character, which should be a list-separator specification, we walked
1304 off past the nul-terimation.
1306 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1307 causes) even when the retry time is not yet met. Previously they were
1308 not, meaning that when (say) an account was over-quota and temp-rejecting,
1309 and multiple senders' messages were queued, only one sender would get
1310 notified on each configured delay_warning cycle.
1312 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1314 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1315 cipher-suites, an error can be left on the stack even for a succeeding
1316 accept; this results in impossible error messages when a later operation
1319 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1320 return error codes indicating retry. Under TLS1.3 this becomes required.
1322 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1323 it only wrote the new authenticators, resulting in a lack of tracking of
1324 peer changes of ESMTP extensions until the next cache flush.
1326 JH/41 Fix the loop reading a message header line to check for integer overflow,
1327 and more-often against header_maxsize. Previously a crafted message could
1328 induce a crash of the recive process; now the message is cleanly rejected.
1330 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1331 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1337 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1338 When redis_servers is set to a list of > 1 element, and the Redis servers
1339 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1340 case of MOVED into a DEFER case instead, thus moving the query onto the
1341 next server in the list. For a cluster of N elements, all N servers must
1342 be defined in redis_servers.
1344 GF/02 Catch and remove uninitialized value warning in exiqsumm
1345 Check for existence of @ARGV before looking at $ARGV[0]
1347 JH/01 Replace the store_release() internal interface with store_newblock(),
1348 which internalises the check required to safely use the old one, plus
1349 the allocate and data copy operations duplicated in both (!) of the
1350 extant use locations.
1352 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1353 modifier. This matches the restriction on the commandline.
1355 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1356 Previously only the last row was returned.
1358 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1359 we assumed that tags in the header were well-formed, and parsed the
1360 element content after inspecting only the first char of the tag.
1361 Assumptions at that stage could crash the receive process on malformed
1364 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1365 While running the DKIM ACL we operate on the Permanent memory pool so that
1366 variables created with "set" persist to the DATA ACL. Also (at any time)
1367 DNS lookups that fail create cache records using the Permanent pool. But
1368 expansions release any allocations made on the current pool - so a dnsdb
1369 lookup expansion done in the DKIM ACL releases the memory used for the
1370 DNS negative-cache, and bad things result. Solution is to switch to the
1371 Main pool for expansions.
1372 While we're in that code, add checks on the DNS cache during store_reset,
1373 active in the testsuite.
1374 Problem spotted, and debugging aided, by Wolfgang Breyha.
1376 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1377 When none of the hosts presented to a transport match an already-open
1378 connection, close it and proceed with the list. Previously we would
1379 queue the message. Spotted by Lena with Yahoo, probably involving
1382 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1383 Previously a spurious "250 OK id=" response was appended to the proper
1386 JH/08 The "support for" informational output now, which built with Content
1387 Scanning support, has a line for the malware scanner interfaces compiled
1388 in. Interface can be individually included or not at build time.
1390 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1391 by the template makefile "src/EDITME". The "STREAM" support for an older
1392 ClamAV interface method is removed.
1394 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1395 rows affected is given instead).
1397 JH/11 The runtime Berkeley DB library version is now additionally output by
1398 "exim -d -bV". Previously only the compile-time version was shown.
1400 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1401 SMTP connection. Previously, when one had more recipients than the
1402 first, an abortive onward connection was made. Move to full support for
1403 multiple onward connections in sequence, handling cutthrough connection
1404 for all multi-message initiating connections.
1406 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1407 routers. Previously, a multi-recipient message would fail to match the
1408 onward-connection opened for the first recipient, and cause its closure.
1410 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1411 a timeout on read on a GnuTLS initiating connection, resulting in the
1412 initiating connection being dropped. This mattered most when the callout
1413 was marked defer_ok. Fix to keep the two timeout-detection methods
1416 JH/15 Relax results from ACL control request to enable cutthrough, in
1417 unsupported situations, from error to silently (except under debug)
1418 ignoring. This covers use with PRDR, frozen messages, queue-only and
1421 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1423 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1424 metadata, resulting in a crash in free().
1426 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1427 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1428 Broken also in d185889f4, with init system revamp.
1430 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1431 Previously we abruptly closed the connection after reading a malware-
1432 found indication; now we go on to read the "scan ok" response line,
1435 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1436 ACL. Previously, a crash would result.
1438 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1439 macro text after a replacement (previously it was only once per line) and
1440 by skipping builtin macros when searching for an uppercase lead character.
1442 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1443 for the build is renamed.
1445 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1446 was allocated for every new TLS startup, meaning one per message. Fix
1447 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1449 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1450 reported the original. Fix to report (as far as possible) the ACL
1451 result replacing the original.
1453 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1454 OpenSSL. Certificate information is loaded for every new TLS startup,
1455 and the resources needed to be freed.
1457 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1459 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1462 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1463 DATA response info to the (existing) per-recipient response info for
1464 the "C=" log element. It can have useful tracking info from the
1465 destination system. Patch from Simon Arlott.
1467 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1468 length value. Previously this would segfault.
1470 HS/02 Support Avast multiline protoocol, this allows passing flags to
1471 newer versions of the scanner.
1473 JH/28 Ensure that variables possibly set during message acceptance are marked
1474 dead before release of memory in the daemon loop. This stops complaints
1475 about them when the debug_store option is enabled. Discovered specifically
1476 for sender_rate_period, but applies to a whole set of variables.
1477 Do the same for the queue-runner and queue-list loops, for variables set
1478 from spool message files. Do the same for the SMTP per-message loop, for
1479 certain variables indirectly set in ACL operations.
1481 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1482 as a multi-recipient message from a mailinglist manager). The coding had
1483 an arbitrary cutoff number of characters while checking for more input;
1484 enforced by writing a NUL into the buffer. This corrupted long / fast
1485 input. The problem was exposed more widely when more pipelineing of SMTP
1486 responses was introduced, and one Exim system was feeding another.
1487 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1488 receiving system, and refused recipients seen by the sending system
1489 (propating to people being dropped from mailing lists).
1490 Discovered and pinpointed by David Carter.
1492 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1493 replaced by the ${authresults } expansion.
1495 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1497 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1498 allows proper process termination in container environments.
1500 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1501 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1503 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1504 and "err_perm", deprecated since 4.83 when the RFC-defined words
1505 "temperror" and "permerror" were introduced.
1507 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1508 transport-filters or DKIM-signing. The restriction was lost in the
1509 consolidation of verify-callout and delivery SMTP handling.
1510 Extend the restriction to also cover ARC-signing.
1512 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1513 in defer=pass mode supply a 450 to the initiator. Previously the message
1516 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1517 tls_require_ciphers is used as before.
1519 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1520 "pass_unscanned". Only tmpfails from the scanner are written to
1521 the paniclog, as they may require admin intervention (permission
1522 denied, license issues). Other scanner errors (like decompression
1523 bombs) do not cause a paniclog entry.
1525 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1526 Previously it was possible to log spurious information in receive log
1529 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1530 triggered odd behaviour from Outlook Express clients.
1532 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1535 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1536 since the IETF WG has not yet settled on that versus the original
1537 "bare" representation.
1539 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1540 Previously the millisecond value corrupted the output.
1541 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1542 corrupted the output.
1548 JH/01 Rework error string handling in TLS interface so that the caller in
1549 more cases is responsible for logging. This permits library-sourced
1550 string to be attached to addresses during delivery, and collapses
1551 pairs of long lines into single ones.
1553 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1554 during configuration. Wildcards are allowed and expanded.
1556 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1557 This permits better logging.
1559 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1560 do not close it down and have the child transport start it up again on
1561 the passed-on TCP connection. Instead, proxy the child (and any
1562 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1563 affected: the continued delivery log lines do not have any DNSSEC, TLS
1564 Certificate or OCSP information. TLS cipher information is still logged.
1566 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1567 identical IP addresses on different listening ports. Will also affect
1570 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1571 add noisy ifdef guards to special-case this sillyness.
1572 Patch from Bernd Kuhls.
1574 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1575 than 255 are no longer allowed.
1577 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1578 Disable the session-cache too, which might reduce our load. Since we
1579 currrectly use a new context for every connection, both as server and
1580 client, there is no benefit for these.
1581 GnuTLS appears to not support tickets server-side by default (we don't
1582 call gnutls_session_ticket_enable_server()) but client side is enabled
1583 by default on recent versions (3.1.3 +) unless the PFS priority string
1586 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1587 <https://reproducible-builds.org/specs/source-date-epoch/>.
1589 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1590 the check for any unsuccessful recipients did not notice the limit, and
1591 erroneously found still-pending ones.
1593 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1594 MSG_MORE. Only in-clear (not on TLS connections).
1596 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1597 transport-filter is involved we can buffer the headers in memory for
1598 creating the signature, and read the spool data file once for the
1599 signature and again for transmission.
1601 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1602 4.77 as the kernel support then wasn't solid, having issues in 64bit
1603 mode. Now, it's been long enough. Add support for FreeBSD also.
1605 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1606 case where the routing stage had gathered several addresses to send to
1607 a host before calling the transport for the first, we previously failed
1608 to close down TLS in the old transport process before passing the TCP
1609 connection to the new process. The new one sent a STARTTLS command
1610 which naturally failed, giving a failed delivery and bloating the retry
1611 database. Investigation and fix prototype from Wolfgang Breyha.
1613 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1614 false-negatives in the check that the sender had not preempted a response
1615 or prompt from Exim (running as a server), due to that code's lack of
1616 awareness of the SMTP input buffering.
1618 PP/04 Add commandline_checks_require_admin option.
1619 Exim drops privileges sanely, various checks such as -be aren't a
1620 security problem, as long as you trust local users with access to their
1621 own account. When invoked by services which pass untrusted data to
1622 Exim, this might be an issue. Set this option in main configuration
1623 AND make fixes to the calling application, such as using `--` to stop
1626 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1627 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1628 responses to those, into a single TLS record each way (this usually means
1629 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1632 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1633 affects you only if you're dancing at the edge of the param size limits.
1634 If you are, and this message makes sense to you, then: raise the
1635 configured limit or use OpenSSL 1.1. Nothing we can do for older
1638 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1639 cmdline element to get the documented default one. Previously it was
1642 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1643 are unsuable for non-message-specific reasons - eg. connection timeout,
1644 banner-time rejection.
1646 JH/16 Fix logging of delivery remote port, when specified by router, under
1649 PP/06 Repair manualroute's ability to take options in any order, even if one
1650 is the name of a transport.
1653 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1655 JH/17 Change the list-building routines interface to use the expanding-string
1656 triplet model, for better allocation and copying behaviour.
1658 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1659 Previously it was constructed the first time a possibly-matching string
1660 was met in the configuration file input during startup; now it is done
1663 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1664 compatible one, to avoid the (poorly documented) possibility of a config
1665 file in the working directory redirecting the DB files, possibly correpting
1666 some existing file. CVE-2017-10140 assigned for BDB.
1668 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1669 cache-hot. Previously, although the result was properly cached, the
1670 initial verify call returned a defer.
1672 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1673 the main verify for receipient in uncached-mode.
1675 JH/22 Retire historical build files to an "unsupported" subdir. These are
1676 defined as "ones for which we have no current evidence of testing".
1678 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1679 if present. Previously it was ignored.
1681 JH/24 Start using specified-initialisers in C structure init coding. This is
1682 a C99 feature (it's 2017, so now considered safe).
1684 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1685 if was a fixed-sized field and bitmask ops via macros; it is now more
1688 PP/07 GitHub PR 56: Apply MariaDB build fix.
1689 Patch provided by Jaroslav Škarvada.
1691 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1692 during Coverity cleanups [4.87 JH/47]
1693 Diagnosis and fix provided by Michael Fischer v. Mollard.
1695 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1696 the right size to place the terminating semicolon on its own folded
1697 line, the header hash was calculated to an incorrect value thanks to
1698 the (relaxed) space the fold became.
1700 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1701 and confused the parent.
1703 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1704 which could crash as a result. This could lead to undeliverable messages.
1706 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1709 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1710 matters on fast-turnover and PID-randomising systems, which were getting
1711 out-of-order delivery.
1713 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1714 a possibly-overlapping copy. The symptom was that "Remote host closed
1715 connection in response to HELO" was logged instead of the actual 4xx
1718 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1719 Previously only that bufferd was discarded, resulting in SYMTP command
1722 JH/32 DKIM: when a message has multiple signatures matching an identity given
1723 in dkim_verify_signers, run the dkim acl once for each. Previously only
1724 one run was done. Bug 2189.
1726 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1727 "panic the current process" to "deliberately defer". The panic log is
1728 still written with the problem list name; the mail and reject logs now
1729 get a temp-reject line for the message that was being handled, saying
1730 something like "domains check lookup or other defer". The SMTP 451
1731 message is still "Temporary local problem".
1733 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1734 A crafted sequence of BDAT commands could result in in-use memory beeing
1735 freed. CVE-2017-16943.
1737 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1738 from SMTP input. Previously it was always done; now only done for DATA
1739 and not BDAT commands. CVE-2017-16944.
1741 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1742 to the message (such as an overlong header line). Previously this was
1743 not done and we did not exit BDAT mode. Followon from the previous item
1744 though a different problem.
1750 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1751 than -2003 did; needs libidn2 in addition to libidn.
1753 JH/02 The path option on a pipe transport is now expanded before use.
1755 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1756 Patch provided by "Björn", documentation fix added too.
1758 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1759 missing a wire-to-host endian conversion.
1761 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1762 close after a BDAT command line could be taken as a following command,
1763 giving a synch failure. Fix by only checking for synch immediately
1764 before acknowledging the chunk.
1766 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1767 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1768 macro. Patches provided by Josh Soref.
1770 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1771 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1772 by IANA as a service extension. However, John Klensin suggests that we
1775 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1776 the dkim code may be unix-mode line endings rather than smtp wire-format
1777 CRLF, so prepend a CR to any bare LF.
1779 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1780 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1782 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1783 the first were themselves being wrongly included in the feed into dkim
1784 processing; with most chunk sizes in use this resulted in an incorrect
1785 body hash calculated value.
1787 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1788 DKIM signature block, for verification. Although advised against by
1789 standards it is specifically not ruled illegal.
1791 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1793 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1794 missing a body hash (the bh= tag).
1796 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1797 It seems that HAProxy sends the Proxy Protocol information in clear and
1798 only then does a TLS startup, so do the same.
1800 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1801 TCP connections (such as for Spamd) unless the daemon successfully set
1802 Fast Open mode on its listening sockets. This fixes breakage seen on
1803 too-old kernels or those not configured for Fast Open, at the cost of
1804 requiring both directions being enabled for TFO, and TFO never being used
1805 by non-daemon-related Exim processes.
1807 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1808 endings, at least on the first header line. Try to canonify any that get
1809 past that check, despite the cost.
1811 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1812 now limited to an arbitrary five deep, while parsing addresses with the
1813 strip_excess_angle_brackets option enabled.
1815 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1816 instead leave the unprompted TLS handshake in socket buffer for the
1817 TLS library to consume.
1819 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1821 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1823 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1824 region in one operation, for speed, and this leaves assigned pointers
1825 dangling. Add checks run only under the testsuite which checks all
1826 variables at a store-reset and panics on a dangling pointer; add code
1827 explicitly nulling out all the variables discovered. Fixes one known
1828 bug: a transport crash, where a dangling pointer for $sending_ip_address
1829 originally assigned in a verify callout, is re-used.
1831 PP/06 Drop '.' from @INC in various Perl scripts.
1833 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1835 PP/08 Reduce a number of compilation warnings under clang; building with
1836 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1837 should be warning-free.
1839 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1841 HS/01 Fix portability problems introduced by PP/08 for platforms where
1842 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1844 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1845 chunk. This allows us to accept broken chunked messages. We need a more
1846 general solution here.
1848 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1849 already-broken messages in the queue.
1851 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1853 JH/19 Fix reference counting bug in routing-generated-address tracking.
1859 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1860 supports it and a size is available (ie. the sending peer gave us one).
1862 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1863 years of being deprecated). The replacements are the ACLs
1864 acl_smtp_mime and acl_not_smtp_mime.
1866 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1867 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1868 or even an in-clear connection were permitted. Now, if the host lookup
1869 was dnssec and dane was requested then the host is only used if the
1870 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1871 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1872 if one fails this test.
1873 This means that a poorly-configured remote DNS will make it incommunicado;
1874 but it protects against a DNS-interception attack on it.
1876 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1877 non-noisy when a race steals the message being considered.
1879 JH/05 If main configuration option tls_certificate is unset, generate a
1880 self-signed certificate for inbound TLS connections.
1882 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1883 in rewrites and routers.
1885 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1886 and logged a warning sing 4.83; now they are a configuration file error.
1888 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1889 (lacking @domain). Apply the same qualification processing as RCPT.
1891 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1893 JH/10 Support ${sha256:} applied to a string (as well as the previous
1896 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1897 a cutthrough deliver is pending, as we always want to make a connection.
1898 This also avoids re-routing the message when later placing the cutthrough
1899 connection after a verify cache hit.
1900 Do not update it with the verify result either.
1902 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1903 when routing results in more than one destination address.
1905 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1906 signing (which inhibits the cutthrough capability). Previously only
1907 the presence of an option was tested; now an expansion evaluating as
1908 empty is permissible (obviously it should depend only on data available
1909 when the cutthrough connection is made).
1911 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1912 the relevant preceding SMTP command did not note the pipelining mode.
1914 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1915 Previously they were not counted.
1917 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1918 as one having no matching records. Previously we deferred the message
1919 that needed the lookup.
1921 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1922 distinguished as "(=".
1924 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1925 for missing MX records. Previously it only worked for missing A records.
1927 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1929 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1930 after the data-go-ahead and data-ack. Patch from Jason Betts.
1932 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1933 even for a "none" policy. Patch from Tony Meyer.
1935 JH/22 Fix continued use of a connection for further deliveries. If a port was
1936 specified by a router, it must also match for the delivery to be
1939 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1940 When one of the recipients of a message was unsuitable for the connection
1941 (has no matching addresses), we lost track of needing to mark it
1942 deferred. As a result mail would be lost.
1944 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1946 JH/25 Decoding ACL controls is now done using a binary search; the source code
1947 takes up less space and should be simpler to maintain. Merge the ACL
1948 condition decode tables also, with similar effect.
1950 JH/26 Fix problem with one_time used on a redirect router which returned the
1951 parent address unchanged. A retry would see the parent address marked as
1952 delivered, so not attempt the (identical) child. As a result mail would
1955 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1956 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1957 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1960 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1961 Main config options check_{log,spool}_{inodes,space} are now
1962 100 inodes, 10MB unless set otherwise in the configuration.
1964 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1965 Previously it only applied to the main-section connection policy
1968 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1970 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1971 by me. Added RFC7919 DH primes as an alternative.
1973 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1974 are not in the system include path.
1976 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1977 GnuTLS, when a session startup failed (eg because the client disconnected)
1978 Exim did stdio operations after fclose. This was exposed by a recent
1979 change which nulled out the file handle after the fclose.
1981 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1982 signed directly by the cert-signing cert, rather than an intermediate
1983 OCSP-signing cert. This is the model used by LetsEncrypt.
1985 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1987 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1988 an incoming connection.
1990 HS/02 Bug 1802: Do not half-close the connection after sending a request
1993 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1994 fallback to "prime256v1".
1996 JH/34 SECURITY: Use proper copy of DATA command in error message.
1997 Could leak key material. Remotely exploitable. CVE-2016-9963.
2003 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2004 and 3.4.4 - once the server is enabled to respond to an OCSP request
2005 it does even when not requested, resulting in a stapling non-aware
2006 client dropping the TLS connection.
2008 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2009 support variable-length bit vectors. No functional change.
2011 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2012 The I= interface field on outgoing lines is now after the H= remote
2013 host field, same as incoming lines. There is a separate
2014 outgoing_interface log selector which allows you to disable the
2017 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2018 If not running log_selector +smtp_connection the mainlog would be held
2019 open indefinitely after a "too many connections" event, including to a
2020 deleted file after a log rotate. Leave the per net connection logging
2021 leaving it open for efficiency as that will be quickly detected by the
2022 check on the next write.
2024 HS/01 Bug 1671: Fix post transport crash.
2025 Processing the wait-<transport> messages could crash the delivery
2026 process if the message IDs didn't exist for some reason. When
2027 using 'split_spool_directory=yes' the construction of the spool
2028 file name failed already, exposing the same netto behaviour.
2030 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2031 mime_regex ACL conditions.
2033 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2034 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2035 to HELO, local diagnostic string.
2037 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2038 log line to debug. Even when configured with a tls authenticator many
2039 client connections are expected to not authenticate in this way, so
2040 an authenticate fail is not an error.
2042 HS/02 Add the Exim version string to the process info. This way exiwhat
2043 gives some more detail about the running daemon.
2045 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2046 matter for fast-change records such as DNSBLs.
2048 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2049 retry record, even if constant. There may be multiple transports with
2050 different interface settings and the retry behaviour needs to be kept
2053 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2055 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2057 JH/10 Bug 840: fix log_defer_output option of pipe transport
2059 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2060 hosts_randomize. This matters a lot when combined with mua_wrapper.
2062 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2063 ${quote_pgsql:<string>} operator.
2065 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2067 JH/14 Change header file name for freeradius-client. Relevant if compiling
2068 with Radius support; from the Gentoo tree and checked under Fedora.
2070 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2072 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2073 option transport_filter, meaning no filtering.
2075 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2077 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2078 defaults to "*" (all hosts). The variable is now available when not built
2079 with TLS, default unset, mainly to enable keeping the testsuite sane.
2080 If a server certificate is not supplied (via tls_certificate) an error is
2081 logged, and clients will find TLS connections fail on startup. Presumably
2082 they will retry in-clear.
2083 Packagers of Exim are strongly encouraged to create a server certificate
2084 at installation time.
2086 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2087 with the $config_file variable.
2089 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2090 in transport context, after the attempt, and per-recipient. The latter type
2091 is per host attempted. The event data is the error message, and the errno
2092 information encodes the lookup type (A vs. MX) used for the (first) host,
2093 and the trailing two digits of the smtp 4xx response.
2095 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2096 to write to mainlog (or rejectlog, paniclog) in the window between file
2097 creation and permissions/ownership being changed. Particularly affects
2098 installations where exicyclog is run as root, rather than exim user;
2099 result is that the running daemon panics and dies.
2101 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2103 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2104 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2105 "pri" and "weight". Note that the previous implicit priority given by the
2106 list order is no longer honoured.
2108 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2109 for DKIM processing.
2111 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2112 by defining SUPPORT_SOCKS.
2114 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2115 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2116 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2117 variables are renamed to proxy_{local,external}_{address,port}.
2119 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2120 for a build by defining SUPPORT_I18N
2122 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2123 of the query string, and make ${quote_redis:} do that quoting.
2125 JH/29 Move Events support from Experimental to mainline, enabled by default
2126 and removable for a build by defining DISABLE_EVENT.
2128 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2130 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2131 cached by the daemon.
2133 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2134 by defining LOOKUP_REDIS. The libhiredis library is required.
2136 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2137 keys are given for lookup.
2139 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2140 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2141 only supported when built with TLS support. The PolarSSL SHA routines
2142 are still used when the TLS library is too old for convenient support.
2144 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2145 openssl_options), for security. OpenSSL forces this from version 1.1.0
2146 server-side so match that on older versions.
2148 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2149 allocation for $value could be released as the expansion processing
2150 concluded, but leaving the global pointer active for it.
2152 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2153 and to use the domains and local_parts ACL conditions.
2155 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2156 incorrectly not doubled on cutthrough transmission, hence seen as a
2157 body-termination at the receiving system - resulting in truncated mails.
2158 Commonly the sender saw a TCP-level error, and retransmitted the message
2159 via the normal store-and-forward channel. This could result in duplicates
2160 received - but deduplicating mailstores were liable to retain only the
2161 initial truncated version.
2163 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2165 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2167 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2168 we're in there, support oversigning also; bug 1309.
2170 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2172 HS/04 Add support for keep_environment and add_environment options.
2174 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2175 either intentional arithmetic overflow during PRNG, or testing config-
2178 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2179 delivery resulted in actual delivery. Cancel cutthrough before DATA
2182 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2183 rejects a recipient: pass the reject to the originator.
2185 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2186 Many were false-positives and ignorable, but it's worth fixing the
2189 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2190 for the new environment-manipulation done at startup. Move the routines
2191 from being local to tls.c to being global via the os.c file.
2193 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2194 an extract embedded as result-arg for a map, the first arg for extract
2195 is unavailable so we cannot tell if this is a numbered or keyed
2196 extraction. Accept either.
2202 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2205 JH/02 The smtp transport option "multi_domain" is now expanded.
2207 JH/03 The smtp transport now requests PRDR by default, if the server offers
2210 JH/04 Certificate name checking on server certificates, when exim is a client,
2211 is now done by default. The transport option tls_verify_cert_hostnames
2212 can be used to disable this per-host. The build option
2213 EXPERIMENTAL_CERTNAMES is withdrawn.
2215 JH/05 The value of the tls_verify_certificates smtp transport and main options
2216 default to the word "system" to access the system default CA bundle.
2217 For GnuTLS, only version 3.0.20 or later.
2219 JH/06 Verification of the server certificate for a TLS connection is now tried
2220 (but not required) by default. The verification status is now logged by
2221 default, for both outbound TLS and client-certificate supplying inbound
2224 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2227 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2228 Status Notification (bounce) messages are now MIME format per RFC 3464.
2229 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2230 under the control of the dsn_advertise_hosts option, and routers may
2231 have a dsn_lasthop option.
2233 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2234 default, modifiable by a malware= option. The list separator for
2235 the options can now be changed in the usual way. Bug 68.
2237 JH/10 The smtp_receive_timeout main option is now expanded before use.
2239 JH/11 The incoming_interface log option now also enables logging of the
2240 local interface on delivery outgoing connections.
2242 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2243 if the interface and destination host and port all match.
2245 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2248 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2249 Patch from Andrew Lewis.
2251 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2252 now supports optional time-restrictions, weighting, and priority
2253 modifiers per server. Patch originally by <rommer@active.by>.
2255 JH/16 The spamd_address main option now supports a mixed list of local
2256 and remote servers. Remote servers can be IPv6 addresses, and
2257 specify a port-range.
2259 JH/17 Bug 68: The spamd_address main option now supports an optional
2260 timeout value per server.
2262 JH/18 Bug 1581: Router and transport options headers_add/remove can
2263 now have the list separator specified.
2265 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2268 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2271 JH/21 Support for the A6 type of dns record is withdrawn.
2273 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2274 rather than the verbs used.
2276 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2277 from 255 to 1024 chars.
2279 JH/24 Verification callouts now attempt to use TLS by default.
2281 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2282 are generic router options now. The defaults didn't change.
2284 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2285 Original patch from Alexander Shikoff, worked over by JH.
2287 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2288 files if ZCAT_COMMAND is not executable.
2290 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2292 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2294 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2295 Normally benign, it bites when the pair was led to by a CNAME;
2296 modern usage is to not canonicalize the domain to a CNAME target
2297 (and we were inconsistent anyway for A-only vs AAAA+A).
2299 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2301 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2302 when evaluating $sender_host_dnssec.
2304 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2305 $sender_helo_dnssec variable.
2307 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2309 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2311 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2313 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2314 documented as working, but never had. Support all but $spam_report.
2316 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2317 added for tls authenticator.
2319 HS/03 Add perl_taintmode main config option
2325 TL/01 When running the test suite, the README says that variables such as
2326 no_msglog_check are global and can be placed anywhere in a specific
2327 test's script, however it was observed that placement needed to be near
2328 the beginning for it to behave that way. Changed the runtest perl
2329 script to read through the entire script once to detect and set these
2330 variables, reset to the beginning of the script, and then run through
2331 the script parsing/test process like normal.
2333 TL/02 The BSD's have an arc4random API. One of the functions to induce
2334 adding randomness was arc4random_stir(), but it has been removed in
2335 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2336 function when detected.
2338 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2339 cause callback expansion.
2341 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2342 syntax errors in an expansion can be treated as a string instead of
2343 logging or causing an error, due to the internal use of bool_lax
2344 instead of bool when processing it.
2346 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2347 server certificates when making smtp deliveries.
2349 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2351 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2353 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2355 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2356 Merged patch from Sebastian Wiedenroth.
2358 JH/05 Fix results-pipe from transport process. Several recipients, combined
2359 with certificate use, exposed issues where response data items split
2360 over buffer boundaries were not parsed properly. This eventually
2361 resulted in duplicates being sent. This issue only became common enough
2362 to notice due to the introduction of connection certificate information,
2363 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2365 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2366 size buffer was used, resulting in syntax errors when an expansion
2369 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2370 version 3.3.6 or later.
2372 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2373 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2374 both become "event_action", the variables become $event_name, $event_data
2375 and $event_defer_errno. There is a new variable $verify_mode, usable in
2376 routers, transports and related events. The tls:cert event is now also
2377 raised for inbound connections, if the main configuration event_action
2380 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2381 early OCSP support, but no stapling (appears to be less than 1.0.0).
2383 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2384 server certificate names available under the smtp transport option
2385 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2388 JH/10 Time-related extraction expansions from certificates now use the main
2389 option "timezone" setting for output formatting, and are consistent
2390 between OpenSSL and GnuTLS compilations. Bug 1541.
2392 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2393 encoded parameter in the incoming message. Bug 1558.
2395 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2396 include certificate info, eximon was claiming there were spoolfile
2399 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2401 JH/14 Log delivery-related information more consistently, using the sequence
2402 "H=<name> [<ip>]" wherever possible.
2404 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2405 are problematic for Debian distribution, omit them from the release
2408 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2410 JH/16 Fix string representation of time values on 64bit time_t architectures.
2413 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2414 output list separator was used.
2419 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2420 checkers that were complaining about end of non-void function with no
2423 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2424 This was a regression introduced in 4.83 by another bugfix.
2426 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2428 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2429 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2435 TF/01 Correctly close the server side of TLS when forking for delivery.
2437 When a message was received over SMTP with TLS, Exim failed to clear up
2438 the incoming connection properly after forking off the child process to
2439 deliver the message. In some situations the subsequent outgoing
2440 delivery connection happened to have the same fd number as the incoming
2441 connection previously had. Exim would try to use TLS and fail, logging
2442 a "Bad file descriptor" error.
2444 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2445 utilities have not been installed.
2447 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2448 temporary space as the ACL may create new global variables.
2450 TL/01 LDAP support uses per connection or global context settings, depending
2451 upon the detected version of the libraries at build time.
2453 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2454 to extract and use the src ip:port in logging and expansions as if it
2455 were a direct connection from the outside internet. PPv2 support was
2456 updated based on HAProxy spec change in May 2014.
2458 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2460 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2461 Properly escape header and check for NULL return.
2463 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2466 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2468 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2469 characters in header names, implemented as a verify condition.
2470 Contributed by Michael Fischer v. Mollard.
2472 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2473 results permerror and temperror. Previous values are deprecated but
2474 still accepted. In a future release, err_perm and err_temp will be
2475 completely removed, which will be a backward incompatibility if the
2476 ACL tests for either of these two old results. Patch contributed by
2477 user bes-internal on the mailing list.
2479 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2481 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2482 selectors, in both main and reject logs.
2484 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2487 JH/07 Add malware type "sock" for talking to simple daemon.
2489 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2491 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2492 routers/transports under cutthrough routing.
2494 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2495 numbers. Touch up "bool" conditional to keep the same definition.
2497 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2499 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2501 JH/12 Expand items in router/transport headers_add or headers_remove lists
2502 individually rather than the list as a whole. Bug 1452.
2504 Required for reasonable handling of multiple headers_ options when
2505 they may be empty; requires that headers_remove items with embedded
2506 colons must have them doubled (or the list-separator changed).
2508 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2509 view the policy declared in the DMARC record. Currently, $dmarc_status
2510 is a combined value of both the record presence and the result of the
2513 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2515 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2516 dnslookup router and the smtp transport (applying to the forward
2519 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2520 of ldap servers used for a specific lookup. Patch provided by Heiko
2523 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2524 New variable $lookup_dnssec_authenticated for observability.
2526 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2527 Patch submitted by Lars Timman.
2529 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2531 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2532 Requires trusted mode and valid format message id, aborts otherwise.
2533 Patch contributed by Heiko Schlichting.
2535 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2536 certextract with support for various fields. Bug 1358.
2538 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2539 is requested by default, modifiable by smtp transport option
2542 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2543 operate on certificate variables to give certificate fingerprints
2544 Also new ${sha256:cert_variable}.
2546 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2548 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2549 Christian Aistleitner.
2551 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2553 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2554 file. Patch from Wolfgang Breyha.
2556 JH/25 Expand the coverage of the delivery $host and $host_address to
2557 client authenticators run in verify callout. Bug 1476.
2559 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2560 align with daemon_smtp_ports. Bug 72.
2562 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2563 support and error reporting did not work properly.
2565 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2566 and is readable. Patch from Andrew Colin Kissa.
2568 TL/14 Enhance documentation of ${run expansion and how it parses the
2569 commandline after expansion, particularly in the case when an
2570 unquoted variable expansion results in an empty value.
2572 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2574 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2575 by renaming away from it. Addresses GH issue 15, reported by
2578 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2580 TL/15 SECURITY: prevent double expansion in math comparison functions
2581 (can expand unsanitized data). Not remotely exploitable.
2588 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2590 PP/02 Make -n do something, by making it not do something.
2591 When combined with -bP, the name of an option is not output.
2593 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2596 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2597 $sender_host_name and config options to manage this, and basic check
2600 PP/05 DSCP support for outbound connections and control modifier for inbound.
2602 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2603 (Only plugin which currently uses this is kerberos4, which nobody should
2604 be using, but we should make it available and other future plugins might
2605 conceivably use it, even though it would break NAT; stuff *should* be
2606 using channel bindings instead).
2608 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2609 name; added for Sendmail compatibility; requires admin caller.
2610 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2611 just ignore it); requires trusted caller.
2612 Also parse but ignore: -Ac -Am -X<logfile>
2615 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2617 TL/02 Add +smtp_confirmation as a default logging option.
2619 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2620 Patch by Magnus Holmgren from 2007-02-20.
2622 TL/04 Bugzilla 1281 - Spec typo.
2623 Bugzilla 1283 - Spec typo.
2624 Bugzilla 1290 - Spec grammar fixes.
2626 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2628 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2630 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2631 mailing list by by Dmitry Isaikin.
2633 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2635 JH/02 Support "G" suffix to numbers in ${if comparisons.
2637 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2639 NM/01 Bugzilla 1197 - Spec typo
2640 Bugzilla 1196 - Spec examples corrections
2642 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2644 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2645 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2648 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2649 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2651 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2652 "acl {{name}{arg}...}", and optional args on acl condition
2655 JH/05 Permit multiple router/transport headers_add/remove lines.
2657 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2659 JH/07 Avoid using a waiting database for a single-message-only transport.
2660 Performance patch from Paul Fisher. Bugzilla 1262.
2662 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2665 JH/09 Add $headers_added variable, with content from use of ACL modifier
2666 add_header (but not yet added to the message). Bugzilla 199.
2668 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2669 Pulled from Bugzilla 817 by Wolfgang Breyha.
2671 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2673 (nb: this is the same fix as in Exim 4.80.1)
2675 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2678 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2681 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2683 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2684 advertise SMTP AUTH mechanism to us, instead of a generic
2685 protocol violation error. Also, make Exim more robust to bad
2686 data from the Dovecot auth socket.
2688 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2690 When a queue runner is handling a message, Exim first routes the
2691 recipient addresses, during which it prunes them based on the retry
2692 hints database. After that it attempts to deliver the message to
2693 any remaining recipients. It then updates the hints database using
2696 So if a recipient address works intermittently, it can get repeatedly
2697 deferred at routing time. The retry hints record remains fresh so the
2698 address never reaches the final cutoff time.
2700 This is a fairly common occurrence when a user is bumping up against
2701 their storage quota. Exim had some logic in its local delivery code
2702 to deal with this. However it did not apply to per-recipient defers
2703 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2705 This change adds a proper retry rule check during routing so that the
2706 final cutoff time is checked against the message's age. We only do
2707 this check if there is an address retry record and there is not a
2708 domain retry record; this implies that previous attempts to handle
2709 the address had the retry_use_local_parts option turned on. We use
2710 this as an approximation for the destination being like a local
2711 delivery, as in LMTP.
2713 I suspect this new check makes the old local delivery cutoff check
2714 redundant, but I have not verified this so I left the code in place.
2716 TF/02 Correct gecos expansion when From: is a prefix of the username.
2718 Test 0254 submits a message to Exim with the header
2722 When I ran the test suite under the user fanf2, Exim expanded
2723 the header to contain my full name, whereas it should have added
2724 a Resent-Sender: header. It erroneously treats any prefix of the
2725 username as equal to the username.
2727 This change corrects that bug.
2729 GF/01 DCC debug and logging tidyup
2730 Error conditions log to paniclog rather than rejectlog.
2731 Debug lines prefixed by "DCC: " to remove any ambiguity.
2733 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2735 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2736 Bug spotted by Jeremy Harris; was flawed since initial commit.
2737 Would have resulted in OCSP responses post-SNI triggering an Exim
2738 NULL dereference and crash.
2740 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2742 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2743 Bug detection, analysis and fix by Samuel Thibault.
2744 Bugzilla 1331, Debian bug #698092.
2746 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2748 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2749 Server implementation by Todd Lyons, client by JH.
2750 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2751 config variable "prdr_enable" controls whether the server
2752 advertises the facility. If the client requests PRDR a new
2753 acl_data_smtp_prdr ACL is called once for each recipient, after
2754 the body content is received and before the acl_smtp_data ACL.
2755 The client is controlled by both of: a hosts_try_prdr option
2756 on the smtp transport, and the server advertisement.
2757 Default client logging of deliveries and rejections involving
2758 PRDR are flagged with the string "PRDR".
2760 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2761 fclose(). Diagnosis by Todd Lyons.
2763 PP/17 Update configure.default to handle IPv6 localhost better.
2764 Patch by Alain Williams (plus minor tweaks).
2767 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2768 This is now consistent with GnuTLS, and is now documented: the
2769 previous undocumented portable approach to treating the option as
2770 unset was to force an expansion failure. That still works, and
2771 an empty string is now equivalent.
2773 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2774 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2775 not performing validation itself.
2777 PP/20 Added force_command boolean option to pipe transport.
2778 Patch from Nick Koston, of cPanel Inc.
2780 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2783 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2785 PP/21 Fix eximon continuous updating with timestamped log-files.
2786 Broken in a format-string cleanup in 4.80, missed when I repaired the
2787 other false fix of the same issue.
2788 Report and fix from Heiko Schlichting.
2791 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2792 Report from Prashanth Katuri.
2794 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2795 It's SecureTransport, so affects any MacOS clients which use the
2796 system-integrated TLS libraries, including email clients.
2798 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2799 using a MIME ACL for non-SMTP local injection.
2800 Report and assistance in diagnosis by Warren Baker.
2802 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2804 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2806 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2807 last id that failed so it may be referenced in subsequent ACL's.
2809 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2812 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2813 ldap library initialization, allowing self-signed CA's to be
2814 used. Also properly sets require_cert option later in code by
2815 using NULL (global ldap config) instead of ldap handle (per
2816 session). Bug diagnosis and testing by alxgomz.
2818 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2819 the src/util/ subdirectory.
2821 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2822 renamed to Transport Post Delivery Action by Jeremy Harris, as
2825 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2826 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2827 redis_servers = needs to be configured which will be used by the redis
2828 lookup. Patch from Warren Baker, of The Packet Hub.
2830 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2832 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2833 hostname or reverse DNS when processing a host list. Used suggestions
2834 from multiple comments on this bug.
2836 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2838 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2839 Missed a few lines, added it to make the runtest require no keyboard
2842 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2843 contains upper case chars. Make router use caseful_local_part.
2845 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2846 support when GnuTLS has been built with p11-kit.
2852 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2854 This, or similar/improved, will also be change PP/11 of 4.82.
2860 PP/01 Handle short writes when writing local log-files.
2861 In practice, only affects FreeBSD (8 onwards).
2862 Bugzilla 1053, with thanks to Dmitry Isaikin.
2864 NM/01 Bugzilla 949 - Documentation tweak
2866 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2869 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2871 PP/02 Implemented gsasl authenticator.
2873 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2875 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2876 `pkg-config foo` for cflags/libs.
2878 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2879 with rest of GSASL and with heimdal_gssapi.
2881 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2882 `pkg-config foo` for cflags/libs for the TLS implementation.
2884 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2885 properties get this fed in as external SSF. A number of robustness
2886 and debugging improvements to the cyrus_sasl authenticator.
2888 PP/08 cyrus_sasl server now expands the server_realm option.
2890 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2891 Patch by Jeremy Harris.
2893 PP/10 Added dbmjz lookup type.
2895 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2897 PP/12 MAIL args handles TAB as well as SP, for better interop with
2898 non-compliant senders.
2899 Analysis and variant patch by Todd Lyons.
2901 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2902 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2903 Patch from Dirk Mueller <dmueller@suse.com>
2905 PP/13 tls_peerdn now print-escaped for spool files.
2906 Observed some $tls_peerdn in wild which contained \n, which resulted
2907 in spool file corruption.
2909 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2910 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2911 or write after TLS renegotiation, which otherwise led to messages
2914 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2915 as a tracking header (ie: a signed header comes before the signature).
2916 Patch from Wolfgang Breyha.
2918 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2919 comma-sep list; embedded commas doubled.
2921 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2923 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2925 Report and patch from Dmitry Banschikov.
2927 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2928 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2929 failures. We appear to now support TLS1.1+ with Exim.
2931 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2932 lets Exim select keys and certificates based upon TLS SNI from client.
2933 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2934 before an outbound SMTP session. New log_selector, +tls_sni.
2936 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2937 NULL dereference. Report and patch from Alun Jones.
2939 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2940 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2941 Not seeing resolver debug output on NetBSD, but suspect this is a
2942 resolver implementation change.
2944 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2945 Left warnings. Added "eximon gdb" invocation mode.
2947 PP/21 Defaulting "accept_8bitmime" to true, not false.
2949 PP/22 Added -bw for inetd wait mode support.
2951 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2952 locate the relevant includes and libraries. Made this the default.
2954 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2955 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2957 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2958 This may cause build issues on older platforms.
2960 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2961 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2962 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2963 Added SNI support via GnuTLS too.
2964 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2966 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2968 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2969 Applied second patch from Janne, implementing suggestion to default
2970 multiple-strings-in-record handling to match SPF spec.
2972 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2974 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2975 read-only, out of scope).
2976 Patch from Wolfgang Breyha, report from Stuart Northfield.
2978 PP/29 Fix three issues highlighted by clang analyser static analysis.
2979 Only crash-plausible issue would require the Cambridge-specific
2980 iplookup router and a misconfiguration.
2981 Report from Marcin Mirosław.
2983 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2985 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2986 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2987 As part of this, removing so much warning spew let me fix some minor
2988 real issues in debug logging.
2990 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2991 assignment on my part. Fixed.
2993 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2994 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2995 Janne Snabb (who went above and beyond: thank you).
2997 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2998 string otherwise requires a connection and a bunch more work and it's
2999 relatively easy to get wrong. Should also expose TLS library linkage
3002 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3003 64-bit ${eval} (JH/03).
3005 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3006 GNU libc to support some of the 64-bit stuff, should not lead to
3007 conflicts. Defined before os.h is pulled in, so if a given platform
3008 needs to override this, it can.
3010 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3011 protection layer was required, which is not implemented.
3012 Bugzilla 1254, patch from Wolfgang Breyha.
3014 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3015 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3016 tls_dhparam take prime identifiers. Also unbreak combination of
3017 OpenSSL+DH_params+TLSSNI.
3019 PP/39 Disable SSLv2 by default in OpenSSL support.
3025 PP/01 Solaris build fix for Oracle's LDAP libraries.
3026 Bugzilla 1109, patch from Stephen Usher.
3028 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3030 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3033 TF/02 Fix a couple more cases where we did not log the error message
3034 when unlink() failed. See also change 4.74-TF/03.
3036 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3037 lock up or crash if it happened to be inside a call to libc when it
3038 got a SIGUSR1 from exiwhat.
3040 The SIGUSR1 handler appends the current process status to the process
3041 log which is later printed by exiwhat. It used to use the general
3042 purpose logging code to do this, but several functions it calls are
3043 not safe for signals.
3045 The new output code in the SIGUSR1 handler is specific to the process
3046 log, and simple enough that it's easy to inspect for signal safety.
3047 Removing some special cases also simplifies the general logging code.
3048 Removing the spurious timestamps from the process log simplifies
3051 TF/04 Improved ratelimit ACL condition.
3053 The /noupdate option has been deprecated in favour of /readonly which
3054 has clearer semantics. The /leaky, /strict, and /readonly update modes
3055 are mutually exclusive. The update mode is no longer included in the
3056 database key; it just determines when the database is updated. (This
3057 means that when you upgrade Exim will forget old rate measurements.)
3059 Exim now checks that the per_* options are used with an update mode that
3060 makes sense for the current ACL. For example, when Exim is processing a
3061 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3062 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3063 must specify per_mail/readonly. If you omit the update mode it defaults to
3064 /leaky where that makes sense (as before) or /readonly where required.
3066 The /noupdate option is now undocumented but still supported for
3067 backwards compatibility. It is equivalent to /readonly except that in
3068 ACLs where /readonly is required you may specify /leaky/noupdate or
3069 /strict/noupdate which are treated the same as /readonly.
3071 A useful new feature is the /count= option. This is a generalization
3072 of the per_byte option, so that you can measure the throughput of other
3073 aggregate values. For example, the per_byte option is now equivalent
3074 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3076 The per_rcpt option has been generalized using the /count= mechanism
3077 (though it's more complicated than the per_byte equivalence). When it is
3078 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3079 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3080 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3081 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3082 Note that using per_rcpt with a non-readonly update mode in more than
3083 one ACL will cause the recipients to be double-counted. (The per_mail
3084 and per_byte options don't have this problem.)
3086 The handling of very low rates has changed slightly. If the computed rate
3087 is less than the event's count (usually one) then this event is the first
3088 after a long gap. In this case the rate is set to the same as this event's
3089 count, so that the first message of a spam run is counted properly.
3091 The major new feature is a mechanism for counting the rate of unique
3092 events. The new per_addr option counts the number of different
3093 recipients that someone has sent messages to in the last time period. It
3094 behaves like per_rcpt if all the recipient addresses are different, but
3095 duplicate recipient addresses do not increase the measured rate. Like
3096 the /count= option this is a general mechanism, so the per_addr option
3097 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3098 example, measure the rate that a client uses different sender addresses
3099 with the options per_mail/unique=$sender_address. There are further
3100 details in the main documentation.
3102 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3104 TF/06 Removed a few PCRE remnants.
3106 TF/07 Automatically extract Exim's version number from tags in the git
3107 repository when doing development or release builds.
3109 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3110 Bugzilla 879. Patch from Paul Fisher.
3112 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3113 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3116 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3118 PP/05 Variable $av_failed, true if the AV scanner deferred.
3119 Bugzilla 1078. Patch from John Horne.
3121 PP/06 Stop make process more reliably on build failure.
3122 Bugzilla 1087. Patch from Heiko Schlittermann.
3124 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3125 Bugzilla 1089. Patch from Heiko Schlittermann.
3127 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3128 Bugzilla 1131. Patch from Holger Weiß.
3130 PP/09 Handle IPv6 addresses with SPF.
3131 Bugzilla 860. Patch from Wolfgang Breyha.
3133 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3135 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3138 PP/11 match_* no longer expand right-hand-side by default.
3139 New compile-time build option, EXPAND_LISTMATCH_RHS.
3140 New expansion conditions, "inlist", "inlisti".
3142 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3144 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3146 PP/14 fix log_write() format string regression from TF/03.
3147 Bugzilla 1152. Patch from Dmitry Isaikin.
3153 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3155 PP/02 Harmonised TLS library version reporting; only show if debugging.
3156 Layout now matches that introduced for other libraries in 4.74 PP/03.
3158 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3160 PP/04 New "dns_use_edns0" global option.
3162 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3165 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3166 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3168 TK/01 Updated PolarSSL code to 0.14.2.
3169 Bugzilla 1097. Patch from Andreas Metzler.
3171 PP/07 Catch divide-by-zero in ${eval:...}.
3172 Fixes bugzilla 1102.
3174 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3177 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3178 format-string attack -- SECURITY: remote arbitrary code execution.
3180 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3181 time unintentionally subject to list matching rules, letting the header
3182 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3183 arbitrary string expansion). This allowed for information disclosure.
3185 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3186 INT_MIN/-1 -- value coerced to INT_MAX.
3192 NM/01 Workaround for PCRE version dependency in version reporting
3195 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3196 This fixes portability to compilers other than gcc, notably
3197 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3199 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3200 makefiles for portability to HP-UX and POSIX correctness.
3202 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3203 Also via indented variable definition in the Makefile.
3204 (Debugging by Oliver Heesakkers).
3206 PP/02 Restore caching of spamd results with expanded spamd_address.
3207 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3209 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3210 Improves build reliability. Fix from: Frank Elsner
3212 NM/02 Fix wide character breakage in the rfc2047 coding
3213 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3215 NM/03 Allow underscore in dnslist lookups
3216 Fixes bug 1026. Patch from Graeme Fowler
3218 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3219 Code patches from Adam Ciarcinski of NetBSD.
3221 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3224 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3225 is logged, to avoid truncation. Patch from John Horne.
3227 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3228 Patch from Jakob Hirsch.
3230 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3231 SQL string expansion failure details.
3232 Patch from Andrey Oktyabrski.
3234 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3235 Patch from Simon Arlott.
3237 PP/09 New lookups functionality failed to compile on old gcc which rejects
3238 extern declarations in function scope.
3239 Patch from Oliver Fleischmann
3241 PP/10 Use sig_atomic_t for flags set from signal handlers.
3242 Check getgroups() return and improve debugging.
3243 Fixed developed for diagnosis in bug 927 (which turned out to be
3246 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3247 Patch from Mark Zealey.
3249 PP/12 Bugzilla 1056: Improved spamd server selection.
3250 Patch from Mark Zealey.
3252 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3253 Based on patch from Heiko Schlittermann.
3255 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3256 Patch from Uwe Doering, sign-off by Michael Haardt.
3258 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3259 variable declaration deep within a block. Bug and patch from
3262 PP/15 lookups-Makefile IRIX compatibility coercion.
3264 PP/16 Make DISABLE_DKIM build knob functional.
3266 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3267 Patch by Simon Arlott
3269 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3270 variable argument macros. Our copy now differs from upstream.
3276 TF/01 Failure to get a lock on a hints database can have serious
3277 consequences so log it to the panic log.
3279 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3280 controlled using the smtp_confirmation log selector.
3282 TF/03 Include the error message when we fail to unlink a spool file.
3284 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3285 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3286 for maintaining out-of-tree patches for some time.
3288 PP/01 Bugzilla 139: Documentation and portability issues.
3289 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3290 Handle per-OS dynamic-module compilation flags.
3292 PP/02 Let /dev/null have normal permissions.
3293 The 4.73 fixes were a little too stringent and complained about the
3294 permissions on /dev/null. Exempt it from some checks.
3295 Reported by Andreas M. Kirchwitz.
3297 PP/03 Report version information for many libraries, including
3298 Exim version information for dynamically loaded libraries. Created
3299 version.h, now support a version extension string for distributors
3300 who patch heavily. Dynamic module ABI change.
3302 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3303 privilege escalation vulnerability whereby the Exim run-time user
3304 can cause root to append content of the attacker's choosing to
3307 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3310 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3311 If dropping privileges for untrusted macros, we disabled normal logging
3312 on the basis that it would fail; for the Exim run-time user, this is not
3313 the case, and it resulted in successful deliveries going unlogged.
3314 Fixed. Reported by Andreas Metzler.
3320 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3321 only prepend for the Resent-* case. Fixes regression introduced in
3322 Exim 4.70 by NM/22 for Bugzilla 607.
3324 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3325 increasing numbers of administrators be bitten by this.
3327 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3329 PP/03 Bugzilla 994: added openssl_options main configuration option.
3331 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3333 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3335 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3337 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3338 without a peer certificate, leading to a segfault because of an
3339 assumption that peers always have certificates. Be a little more
3340 paranoid. Problem reported by Martin Tscholak.
3342 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3343 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3344 NB: ClamAV planning to remove STREAM in "middle of 2010".
3345 CL also introduces -bmalware, various -d+acl logging additions and
3346 more caution in buffer sizes.
3348 PP/09 Implemented reverse_ip expansion operator.
3350 PP/10 Bugzilla 937: provide a "debug" ACL control.
3352 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3354 PP/12 Bugzilla 973: Implement --version.
3356 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3358 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3360 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3362 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3363 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3364 ignore trailing whitespace.
3366 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3368 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3371 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3372 Notification from Dr Andrew Aitchison.
3374 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3375 ExtendedDetectionInfo response format.
3376 Notification from John Horne.
3378 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3381 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3382 XSL and documented dependency on system catalogs, with examples of how
3385 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3388 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3389 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3390 configuration file which is writeable by the Exim user or group.
3392 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3393 of configuration files to cover files specified with the -C option if
3394 they are going to be used with root privileges, not just the default
3397 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3398 option (effectively making it always true).
3400 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3401 files to be used while preserving root privileges.
3403 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3404 that rogue child processes cannot use them.
3406 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3407 run-time user, instead of root.
3409 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3410 Exim run-time user without dropping privileges.
3412 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3413 result string, instead of calling string_vformat() twice with the same
3416 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3417 for other users. Others should always drop root privileges if they use
3418 -C on the command line, even for a whitelisted configure file.
3420 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3422 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3428 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3429 $data_path, and $header_path variables; fixed documentation bugs and
3432 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3433 exipick to access non-standard spools, including the "frozen" queue
3436 NM/01 Bugzilla 965: Support mysql stored procedures.
3437 Patch from Alain Williams
3439 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3441 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3442 Patch from Andreas Metzler
3444 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3445 Patch from Kirill Miazine
3447 NM/05 Bugzilla 671: Added umask to procmail example.
3449 JJ/03 installed exipick 20100323.0, fixing doc bug
3451 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3452 directory. Notification and patch from Dan Rosenberg.
3454 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3456 TK/02 Improve log output when DKIM signing operation fails.
3458 MH/01 Treat the transport option dkim_domain as a colon separated
3459 list, not as a single string, and sign the message with each element,
3460 omitting multiple occurences of the same signer.
3462 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3463 Bugzilla 985, 986. Patch by Simon Arlott
3465 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3466 Patch by Simon Arlott
3468 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3469 MBX locking. Notification from Dan Rosenberg.
3475 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3477 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3479 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3481 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3483 NM/04 Bugzilla 915: Flag broken perl installation during build.
3489 TK/01 Added patch by Johannes Berg that expands the main option
3490 "spamd_address" if it starts with a dollar sign.
3492 TK/02 Write list of recipients to X-Envelope-Sender header when building
3493 the mbox-format spool file for content scanning (suggested by Jakob
3496 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3497 (http://www.dcc-servers.net/) support via dccifd. Activated by
3498 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3500 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3501 by Mark Daniel Reidel <mr@df.eu>.
3503 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3504 When building exim an external PCRE library is now needed -
3505 PCRE is a system library on the majority of modern systems.
3506 See entry on PCRE_LIBS in EDITME file.
3508 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3509 conversation. Added nologin parameter to request.
3510 Patch contributed by Kirill Miazine.
3512 TF/01 Do not log submission mode rewrites if they do not change the address.
3514 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3516 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3517 log files in place. Contributed by Roberto Lima.
3519 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3521 TF/03 Bugzilla 615: When checking the local_parts router precondition
3522 after a local_part_suffix or local_part_prefix option, Exim now
3523 does not use the address's named list lookup cache, since this
3524 contains cached lookups for the whole local part.
3526 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3527 Robert Millan. Documentation is in experimental-spec.txt.
3529 TF/04 Bugzilla 668: Fix parallel build (make -j).
3531 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3533 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3534 Patch provided by Jan Srzednicki.
3536 TF/05 Leading white space used to be stripped from $spam_report which
3537 wrecked the formatting. Now it is preserved.
3539 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3540 that they are available at delivery time.
3542 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3544 TF/08 TLS error reporting now respects the incoming_interface and
3545 incoming_port log selectors.
3547 TF/09 Produce a more useful error message if an SMTP transport's hosts
3548 setting expands to an empty string.
3550 NM/06 Bugzilla 744: EXPN did not work under TLS.
3551 Patch provided by Phil Pennock.
3553 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3554 Patch provided by Richard Godbee.
3556 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3557 acl_smtp_notquit, added index entry.
3559 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3560 Patch provided by Eugene Bujak.
3562 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3563 accept(). Patch provided by Maxim Dounin.
3565 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3566 Patch provided by Phil Pennock.
3568 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3570 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3571 Patch provided by Brad "anomie" Jorsch.
3573 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3574 Patch provided by Dean Brooks.
3576 TK/05 Add native DKIM support (does not depend on external libraries).
3578 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3579 Patch provided by Graeme Fowler.
3581 NM/16 Bugzilla 851: Documentation example syntax fix.
3583 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3585 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3588 NM/19 Bugzilla 745: TLS version reporting.
3589 Patch provided by Phil Pennock.
3591 NM/20 Bugzilla 167: bool: condition support.
3592 Patch provided by Phil Pennock.
3594 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3595 clients. Patch provided by Phil Pennock.
3597 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3598 Patch provided by Brad "anomie" Jorsch.
3600 NM/23 Bugzilla 687: Fix misparses in eximstats.
3601 Patch provided by Heiko Schlittermann.
3603 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3604 Patch provided by Heiko Schlittermann.
3606 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3607 plus update to original patch.
3609 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3611 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3612 Patch provided by David Brownlee.
3614 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3616 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3618 NM/30 Bugzilla 888: TLS documentation bugfixes.
3620 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3622 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3623 Unlike the original bugzilla I have changed all shell scripts in src tree.
3625 NM/33 Bugzilla 898: Transport filter timeout fix.
3626 Patch by Todd Rinaldo.
3628 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3629 Patch by Serge Demonchaux.
3631 NM/35 Bugzilla 39: Base64 decode bug fixes.
3632 Patch by Jakob Hirsch.
3634 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3636 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3638 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3640 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3646 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3647 ALT-N's libdkim that I have put here:
3648 http://duncanthrax.net/exim-experimental/
3650 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3651 were called 'true' and it seems that C99 defines that as a reserved
3652 keyword to be used with 'bool' variable types. That means you could
3653 not include C99-style headers which use bools without triggering
3654 build errors in sieve.c.
3656 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3657 as mailq or other aliases. Changed the --help handling significantly
3658 to do whats expected. exim_usage() emits usage/help information.
3660 SC/01 Added the -bylocaldomain option to eximstats.
3662 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3664 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3666 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3672 PH/01 Another patch from the Sieve maintainer.
3674 PH/02 When an IPv6 address is converted to a string for single-key lookup
3675 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3676 dots are used instead of colons so that keys in lsearch files need not
3677 contain colons. This was done some time before quoting was made available
3678 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3679 (notated using the quote facility) so as to distinguish them from IPv4
3680 keys. This meant that lookups for IP addresses in host lists did not work
3681 for iplsearch lookups.
3683 This has been fixed by arranging for IPv6 addresses to be expressed with
3684 colons if the lookup type is iplsearch. This is not incompatible, because
3685 previously such lookups could never work.
3687 The situation is now rather anomalous, since one *can* have colons in
3688 ordinary lsearch keys. However, making the change in all cases is
3689 incompatible and would probably break a number of configurations.
3691 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3694 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3695 conversion specification without a maximum field width, thereby enabling
3696 a rogue spamd server to cause a buffer overflow. While nobody in their
3697 right mind would setup Exim to query an untrusted spamd server, an
3698 attacker that gains access to a server running spamd could potentially
3699 exploit this vulnerability to run arbitrary code as the Exim user.
3701 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3702 $primary_hostname instead of what libspf2 thinks the hosts name is.
3704 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3705 a directory entry by the name of the lookup key. Previously, if a
3706 symlink pointed to a non-existing file or a file in a directory that
3707 Exim lacked permissions to read, a lookup for a key matching that
3708 symlink would fail. Now it is enough that a matching directory entry
3709 exists, symlink or not. (Bugzilla 503.)
3711 PH/03 The body_linecount and body_zerocount variables are now exported in the
3714 PH/04 Added the $dnslist_matched variable.
3716 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3717 This means they are set thereafter only if the connection becomes
3720 PH/06 Added the client_condition to authenticators so that some can be skipped
3721 by clients under certain conditions.
3723 PH/07 The error message for a badly-placed control=no_multiline_responses left
3724 "_responses" off the end of the name.
3726 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3728 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3729 (without spaces) instead of just copying the configuration text.
3731 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3733 PH/11 Added $max_received_linelength.
3735 PH/12 Added +ignore_defer and +include_defer to host lists.
3737 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3738 way in which PCRE > 7.0 is built.
3740 PH/14 Implemented queue_only_load_latch.
3742 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3743 MAIL command. The effect was to mangle the value on 64-bit systems.
3745 PH/16 Another patch from the Sieve maintainer.
3747 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3749 PH/18 If a system quota error occurred while trying to create the file for
3750 a maildir delivery, the message "Mailbox is full" was not appended to the
3751 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3752 only to a quota excession during the actual writing of the file.
3754 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3755 characters?) which causes problems in log lines. The DN values are now
3756 passed through string_printing() before being added to log lines.
3758 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3759 and InterBase are left for another time.)
3761 PH/21 Added message_body_newlines option.
3763 PH/22 Guard against possible overflow in moan_check_errorcopy().
3765 PH/23 POSIX allows open() to be a macro; guard against that.
3767 PH/24 If the recipient of an error message contained an @ in the local part
3768 (suitably quoted, of course), incorrect values were put in $domain and
3769 $local_part during the evaluation of errors_copy.
3775 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3776 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3779 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3780 issue a MAIL command.
3782 PH/02 In an ACL statement such as
3784 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3786 if a client was not listed at all, or was listed with a value other than
3787 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3788 the condition was not true (as it should be), so access was not denied.
3789 The bug was that the ! inversion was incorrectly passed on to the second
3790 item. This has been fixed.
3792 PH/03 Added additional dnslists conditions == and =& which are different from
3793 = and & when the dns lookup returns more than one IP address.
3795 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3796 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3798 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3799 FSYNC, which compiles an option called disable_fsync that allows for
3800 bypassing fsync(). The documentation is heavily laced with warnings.
3802 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3804 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3805 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3806 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3807 including adding "make clean"; (3) Added -fPIC when compiling the test
3808 dynamically loaded module, to get rid of a warning.
3810 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3811 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3812 = 0s. The bug is otherwise harmless.
3814 PH/07 There was a bug in the dovecot authenticator such that the value of
3815 $auth1 could be overwritten, and so not correctly preserved, after a
3816 successful authentication. This usually meant that the value preserved by
3817 the server_setid option was incorrect.
3819 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3821 PH/09 Installed PCRE release 7.0.
3823 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3824 run for batched SMTP input. It is now run at the start of every message
3825 in the batch. While fixing this I discovered that the process information
3826 (output by running exiwhat) was not always getting set for -bs and -bS
3827 input. This is fixed, and it now also says "batched" for BSMTP.
3829 PH/11 Added control=no_pipelining.
3831 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3832 patch, slightly modified), and move the expansion of helo_data till after
3833 the connection is made in the smtp transport (so it can use these
3836 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3838 PH/14 Added log_selector = +pid.
3840 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3842 PH/16 Add ${if forany and ${if forall.
3844 PH/17 Added dsn_from option to vary the From: line in DSNs.
3846 PH/18 Flush SMTP output before performing a callout, unless control =
3847 no_callout_flush is set.
3849 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3850 was true (the default) a successful delivery failed to delete the retry
3851 item, thus causing premature timeout of the address. The bug is now
3854 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3856 PH/21 Long custom messages for fakedefer and fakereject are now split up
3857 into multiline responses in the same way that messages for "deny" and
3858 other ACL rejections are.
3860 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3861 with slight modification.
3863 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3864 draft, changing the syntax and factoring some duplicate code".
3866 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3867 for deliveries of the second and subsequent messages over the same SMTP
3870 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3871 ${reduce, with only minor "tidies".
3873 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3875 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3876 expansion side effects.
3878 PH/27 When a message times out after an over-quota error from an Exim-imposed
3879 quota, the bounce message says "mailbox is full". This message was not
3880 being given when it was a system quota that was exceeded. It now should
3883 MH/03 Made $recipients available in local_scan(). local_scan() already has
3884 better access to the recipient list through recipients_list[], but
3885 $recipients can be useful in postmaster-provided expansion strings.
3887 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3888 in the case of a MAIL command with additional options following the
3889 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3890 were accidentally chopped off.
3892 PH/29 SMTP synchronization checks are implemented when a command is read -
3893 there is a check that no more input is waiting when there shouldn't be
3894 any. However, for some commands, a delay in an ACL can mean that it is
3895 some time before the response is written. In this time, more input might
3896 arrive, invalidly. So now there are extra checks after an ACL has run for
3897 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3898 pipelining has not been advertised.
3900 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3902 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3903 returned in the SMTP response when smtp_return_error_details was false.
3904 This has been fixed.
3906 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3907 instead of the C I/O that was originally supplied, because problems were
3908 reported on Solaris.
3910 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3911 Exim which did not show up earlier: it was assuming that a call to
3912 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3913 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3914 was a macro that became an assignment, so it seemed to work. This has
3915 changed to a proper function call with a void return, hence the compile
3916 error. Exim's code has been fixed.
3918 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3921 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3923 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3925 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3926 because it turns out to be too big or there is a timeout) there is no
3927 "Completed" line in the log. When some messages of this type were
3928 selected by exigrep, they were listed as "not completed". Others were
3929 picked up by some special patterns. I have improved the selection
3930 criteria to be more general.
3932 PH/38 The host_find_failed option in the manualroute router can now be set
3933 to "ignore", to completely ignore a host whose IP address cannot be
3934 found. If all hosts are ignored, the behaviour is controlled by the new
3935 host_all_ignored option.
3937 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3938 homing or because of multiple MX records with /mx) generated more than
3939 one IP address, and the following item turned out to be the local host,
3940 all the secondary addresses of the first item were incorrectly removed
3941 from the list, along with the local host and any following hosts (which
3942 is what is supposed to happen).
3944 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3945 whoever called Exim into the -H file. In the case of the daemon it was
3946 behaving confusingly. When first started, it used values for whoever
3947 started the daemon, but after a SIGHUP it used the Exim user (because it
3948 calls itself on a restart). I have changed the code so that it now always
3951 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3952 message are rejected with the same error (e.g. no authentication or bad
3953 sender address), and a DATA command is nevertheless sent (as can happen
3954 with PIPELINING or a stupid MUA), the error message that was given to the
3955 RCPT commands is included in the rejection of the DATA command. This is
3956 intended to be helpful for MUAs that show only the final error to their
3959 PH/42 Another patch from the Sieve maintainer.
3961 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3962 Eximstats - Fixed some broken HTML links and added missing column headers
3964 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3965 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3967 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3969 PH/43 Yet another patch from the Sieve maintainer.
3971 PH/44 I found a way to check for a TCP/IP connection going away before sending
3972 the response to the final '.' that terminates a message, but only in the
3973 case where the client has not sent further data following the '.'
3974 (unfortunately, this is allowed). However, in many cases there won't be
3975 any further data because there won't be any more messages to send. A call
3976 to select() can be used: if it shows that the input is "ready", there is
3977 either input waiting, or the socket has been closed. An attempt to read
3978 the next input character can distinguish the two cases. Previously, Exim
3979 would have sent an OK response which the client would never have see.
3980 This could lead to message repetition. This fix should cure that, at
3981 least in a lot of common cases.
3983 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3984 advertised in response to EHLO.
3990 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3991 fixed by 4.65/MH/01 (is this a record?) are fixed:
3993 (i) An empty string was always treated as zero by the numeric comparison
3994 operators. This behaviour has been restored.
3996 (ii) It is documented that the numeric comparison operators always treat
3997 their arguments as decimal numbers. This was broken in that numbers
3998 starting with 0 were being interpreted as octal.
4000 While fixing these problems I realized that there was another issue that
4001 hadn't been noticed. Values of message_size_limit (both the global option
4002 and the transport option) were treated as octal if they started with 0.
4003 The documentation was vague. These values are now always treated as
4004 decimal, and I will make that clear in the documentation.
4010 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4011 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4014 MH/01 Don't check that the operands of numeric comparison operators are
4015 integers when their expansion is in "skipping" mode (fixes bug
4016 introduced by 4.64-PH/07).
4018 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4019 child addresses, Exim now panics and dies. Previously, because the count
4020 is held in a short int, deliveries were likely to be lost. As such a
4021 large number of recipients for a single message is ridiculous
4022 (performance will be very, very poor), I have chosen to impose a limit
4023 rather than extend the field.
4029 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4030 leftover -K file (the existence of which was triggered by #402).
4031 While we were at it, introduced process PID as part of the -K
4032 filename. This should rule out race conditions when creating
4035 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4036 processing considerably. Previous code took too long for large mails,
4037 triggering a timeout which in turn triggers #401.
4039 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4040 in the DK code in transports.c. sendfile() is not really portable,
4041 hence the _LINUX specificness.
4043 TF/01 In the add_headers option to the mail command in an Exim filter,
4044 there was a bug that Exim would claim a syntax error in any
4045 header after the first one which had an odd number of characters
4048 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4049 callout verification, Exim cached a "reject" for the entire domain. This
4050 is correct for most verifications, but it is not correct for a recipient
4051 verification with use_sender or use_postmaster set, because in that case
4052 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4053 case of MAIL FROM:<> rejection from other early rejections (e.g.
4054 rejection of HELO). When verifying a recipient using a non-null MAIL
4055 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4056 Whatever the result of the callout, the value of the domain cache is
4057 left unchanged (for any other kind of callout, getting as far as trying
4058 RCPT means that the domain itself is ok).
4060 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4063 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4064 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4065 session, the daemon ignores ECONNECT errors and logs others; it now
4066 ignores EPIPE as well.
4068 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4069 (quoted-printable decoding).
4071 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4072 later the small subsequent patch to fix an introduced bug.
4074 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4076 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4078 PH/08 An error is now given if message_size_limit is specified negative.
4080 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4081 to be given (somewhat) arbitrary names.
4083 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4086 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4087 miscellaneous code fixes
4089 PH/10 Added the log_reject_target ACL modifier to specify where to log
4092 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4093 hostname. This is wrong, because it relates to the incoming message (and
4094 probably the interface on which it is arriving) and not to the outgoing
4095 callout (which could be using a different interface). This has been
4096 changed to use the value of the helo_data option from the smtp transport
4097 instead - this is what is used when a message is actually being sent. If
4098 there is no remote transport (possible with a router that sets up host
4099 addresses), $smtp_active_hostname is used.
4101 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4102 tweaks were necessary in order to get it to work (see also 21 below):
4103 (a) The code assumed that strncpy() returns a negative number on buffer
4104 overflow, which isn't the case. Replaced with Exim's string_format()
4106 (b) There were several signed/unsigned issues. I just did the minimum
4107 hacking in of casts. There is scope for a larger refactoring.
4108 (c) The code used strcasecmp() which is not a standard C function.
4109 Replaced with Exim's strcmpic() function.
4110 (d) The code set only $1; it now sets $auth1 as well.
4111 (e) A simple test gave the error "authentication client didn't specify
4112 service in request". It would seem that Dovecot has changed its
4113 interface. Fortunately there's a specification; I followed it and
4114 changed what the client sends and it appears to be working now.
4116 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4119 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4120 address A is aliased to B and C, where B exists and C does not. Without
4121 -v the output is "A verified" because verification stops after a
4122 successful redirection if more than one address is generated. However,
4123 with -v the child addresses are also verified. Exim was outputting "A
4124 failed to verify" and then showing the successful verification for C,
4125 with its parentage. It now outputs "B failed to verify", showing B's
4126 parentage before showing the successful verification of C.
4128 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4129 look up a TXT record in a specific list after matching in a combined
4132 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4133 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4134 they consult the DNS. I had assumed they would set it the way they
4135 wanted; and indeed my experiments on Linux seem to show that in some
4136 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4137 To be on the safe side, however, I have now made the interface to
4138 host_find_byname() similar to host_find_bydns(), with an argument
4139 containing the DNS resolver options. The host_find_byname() function now
4140 sets these options at its start, just as host_find_bydns() does. The smtp
4141 transport options dns_qualify_single and dns_search_parents are passed to
4142 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4143 of host_find_byname() use the default settings of RES_DEFNAMES
4144 (qualify_single) but not RES_DNSRCH (search_parents).
4146 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4147 spool_read_header() do less string testing, by means of a preliminary
4148 switch on the second character of optional "-foo" lines. (This is
4149 overdue, caused by the large number of possibilities that now exist.
4150 Originally there were few.) While I was there, I also converted the
4151 str(n)cmp tests so they don't re-test the leading "-" and the first
4152 character, in the hope this might squeeze out yet more improvement.
4154 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4155 flag allowing group syntax was set by the header_syntax check but not
4156 turned off, possible causing trouble later; (2) The flag was not being
4157 set at all for the header_verify test, causing "group"-style headers to
4158 be rejected. I have now set it in this case, and also caused header_
4159 verify to ignore an empty address taken from a group. While doing this, I
4160 came across some other cases where the code for allowing group syntax
4161 while scanning a header line wasn't quite right (mostly, not resetting
4162 the flag correctly in the right place). These bugs could have caused
4163 trouble for malformed header lines. I hope it is now all correct.
4165 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4166 with the "reply" argument non-NULL. The code, however (which originally
4167 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4168 but it didn't always do it. This confused somebody who was copying the
4169 code for some other use. I have removed all the tests.
4171 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4172 feature that was used to support insecure browsers during the U.S. crypto
4173 embargo. It requires special client support, and Exim is probably the
4174 only MTA that supported it -- and would never use it because real RSA is
4175 always available. This code has been removed, because it had the bad
4176 effect of slowing Exim down by computing (never used) parameters for the
4177 RSA_EXPORT functionality.
4179 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4180 authenticator to fail if there's a tab character in the incoming data
4181 (there should never be unless someone is messing about, as it's supposed
4182 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4183 if the connection is using TLS or if the remote IP is the same as the
4184 local IP, and the "valid-client-cert option" if a client certificate has
4187 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4188 authenticators. This can be used for authorization after authentication
4189 succeeds. (In the case of plaintext, it servers for both authentication
4192 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4193 if any retry times were supplied.
4195 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4196 connection, where there is no client IP address to check. In this
4197 situation, the verify now always succeeds.
4199 PH/25 Applied John Jetmore's -Mset patch.
4201 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4203 PH/27 In a string expansion for a processed (not raw) header when multiple
4204 headers of the same name were present, leading whitespace was being
4205 removed from all of them, but trailing whitespace was being removed only
4206 from the last one. Now trailing whitespace is removed from each header
4207 before concatenation. Completely empty headers in a concatenation (as
4208 before) are ignored.
4210 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4211 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4213 PH/29 [Removed. This was a change that I later backed out, and forgot to
4214 correct the ChangeLog entry (that I had efficiently created) before
4215 committing the later change.]
4217 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4218 address errors (4xx response to RCPT) over the same connection as other
4219 messages routed to the same hosts. Such deliveries are always "forced",
4220 so retry times are not inspected. This resulted in far too many retries
4221 for the affected addresses. The effect occurred only when there were more
4222 hosts than the hosts_max_try setting in the smtp transport when it had
4223 the 4xx errors. Those hosts that it had tried were not added to the list
4224 of hosts for which the message was waiting, so if all were tried, there
4225 was no problem. Two fixes have been applied:
4227 (i) If there are any address or message errors in an SMTP delivery, none
4228 of the hosts (tried or untried) are now added to the list of hosts
4229 for which the message is waiting, so the message should not be a
4230 candidate for sending over the same connection that was used for a
4231 successful delivery of some other message. This seems entirely
4232 reasonable: after all the message is NOT "waiting for some host".
4233 This is so "obvious" that I'm not sure why it wasn't done
4234 previously. Hope I haven't missed anything, but it can't do any
4235 harm, as the worst effect is to miss an optimization.
4237 (ii) If, despite (i), such a delivery is accidentally attempted, the
4238 routing retry time is respected, so at least it doesn't keep
4239 hammering the server.
4241 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4242 in ${readsocket because some servers need this prod.
4244 PH/32 Added some extra debug output when updating a wait-xxx database.
4246 PH/33 The hint "could be header name not terminated by colon", which has been
4247 given for certain expansion errors for a long time, was not being given
4248 for the ${if def:h_colon_omitted{... case.
4250 PH/34 The spec says: "With one important exception, whenever a domain list is
4251 being scanned, $domain contains the subject domain." There was at least
4252 one case where this was not true.
4254 PH/35 The error "getsockname() failed: connection reset by peer" was being
4255 written to the panic log as well as the main log, but it isn't really
4256 panic-worthy as it just means the connection died rather early on. I have
4257 removed the panic log writing for the ECONNRESET error when getsockname()
4260 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4261 runs only) independently of the message's sender address. This meant
4262 that, if the 4xx error was in fact related to the sender, a different
4263 message to the same recipient with a different sender could confuse
4264 things. In particular, this can happen when sending to a greylisting
4265 server, but other circumstances could also provoke similar problems.
4266 I have changed the default so that the retry time for these errors is now
4267 based a combination of the sender and recipient addresses. This change
4268 can be overridden by setting address_retry_include_sender=false in the
4271 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4272 remote server are returned as part of bounce messages. This was not
4273 happening for LMTP over a pipe (the lmtp transport), but now it is the
4274 same for both kinds of LMTP.
4276 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4277 in header lines that were in fact CNAMEs. This is no longer the case.
4279 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4280 and queue runs started by the daemon processed all messages. This has
4281 been fixed so that -R and -S can now usefully be given with -q<time>.
4283 PH/40 Import PCRE release 6.7 (fixes some bugs).
4285 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4287 PH/42 Give an error if -q is specified more than once.
4289 PH/43 Renamed the variables $interface_address and $interface_port as
4290 $received_ip_address and $received_port, to make it clear that these
4291 values apply to message reception, and not to the outgoing interface when
4292 a message is delivered. (The old names remain recognized, of course.)
4294 PH/44 There was no timeout on the connect() call when using a Unix domain
4295 socket in the ${readsocket expansion. There now is.
4297 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4298 be meaningful with "accept".
4300 SC/01 Eximstats V1.43
4301 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4303 SC/02 Eximstats V1.44
4304 Use a glob alias rather than an array ref in the generated
4305 parser. This improves both readability and performance.
4307 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4308 Collect SpamAssassin and rejection statistics.
4309 Don't display local sender or destination tables unless
4310 there is data to show.
4311 Added average volumes into the top table text output.
4313 SC/04 Eximstats V1.46
4314 Collect data on the number of addresses (recipients)
4315 as well as the number of messages.
4317 SC/05 Eximstats V1.47
4318 Added 'Message too big' to the list of mail rejection
4319 reasons (thanks to Marco Gaiarin).
4321 SC/06 Eximstats V1.48
4322 Mainlog lines which have GMT offsets and are too short to
4323 have a flag are now skipped.
4325 SC/07 Eximstats V1.49 (Alain Williams)
4326 Added the -emptyok flag.
4328 SC/08 Eximstats V1.50
4329 Fixes for obtaining the IP address from reject messages.
4331 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4332 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4333 whitespace changes from 4.64-PH/27
4335 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4338 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4339 are found, allow negative numbers in numeric criteria)
4341 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4343 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4344 to match changes made in 4.64-PH/43
4346 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4348 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4349 "domain" is actually an IP address, and force a failure. This locks out
4350 those revolvers/nameservers that support "A-for-A" lookups, in
4351 contravention of the specifications.
4353 PH/48 When a host name was looked up from an IP address, and the subsequent
4354 forward lookup of the name timed out, the host name was left in
4355 $sender_host_name, contrary to the specification.
4357 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4358 restricted to single-key lookups, Exim was not diagnosing an error if
4359 * or *@ was used with a query-style lookup.
4361 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4363 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4364 long ago, but noone interested enough thought of it. Let's just say that
4365 the "1.1" means that there are some new functions that weren't there at
4366 some point in the past.
4368 PH/51 Error processing for expansion failure of helo_data from an smtp
4369 transport during callout processing was broken.
4371 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4372 tested/used via the -bh/-bhc/-bs options.
4374 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4375 bug, fixed in subsequent PCRE releases).
4377 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4378 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4380 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4386 SC/01 Use a glob alias rather than an array ref in eximstats generated
4387 parser. This improves both readability and performance.
4389 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4390 Don't display local sender or destination tables in eximstats unless
4391 there is data to show.
4392 Added average volumes into the eximstats top table text output.
4394 SC/03 Collect data on the number of addresses (recipients) as well
4395 as the number of messages in eximstats.
4397 TF/01 Correct an error in the documentation for the redirect router. Exim
4398 does (usually) call initgroups() when daemonizing.
4400 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4401 with consistent privilege compared to when running as a daemon.
4403 TF/03 Note in the spec that $authenticated_id is not set for local
4404 submissions from trusted users.
4406 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4407 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4409 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4410 by adding some example configuration directives to the default
4411 configuration file. A little bit of work is required to uncomment the
4412 directives and define how usernames and passwords are checked, but
4413 there is now a framework to start from.
4415 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4416 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4417 without this. I don't know how relevant this is to other LDAP libraries.
4419 PH/02 Add the verb name to the "unknown ACL verb" error.
4421 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4423 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4425 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4426 directory not expanded when it should be if an expanded home directory
4427 was set for the address (which is overridden by the transport).
4429 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4432 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4433 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4434 because it is too late at that time, and has no effect.
4436 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4437 security issue with \' (bugzilla #107). I could not use the
4438 PQescapeStringConn() function, because it needs a PGconn value as one of
4441 PH/08 When testing addresses using -bt, indicate those final addresses that
4442 are duplicates that would not cause an additional delivery. At least one
4443 person was confused, thinking that -bt output corresponded to deliveries.
4444 (Suppressing duplicates isn't a good idea as you lose the information
4445 about possibly different redirections that led to the duplicates.)
4447 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4448 systems where poll() doesn't work, in particular OS X.
4450 PH/10 Added more information to debugging output for retry time not reached.
4452 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4453 operations in malware.c.
4455 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4458 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4459 syslog_duplication set false, log lines that would normally be written
4460 both the the main log and to the reject log were not written to syslog at
4463 PH/14 In the default configuration, change the use of "message" in ACL warn
4464 statements to "add_header".
4466 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4467 not followed by a command (e.g. "seen endif").
4469 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4470 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4473 PH/17 Added extra conditions to the default value of delay_warning_condition
4477 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4478 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4479 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4482 The Auto-Submitted: and various List- headers are standardised, whereas I
4483 don't think Precedence: ever was.
4485 PH/18 Refactored debugging code in route_finduser() to show more information,
4486 in particular, the error code if getpwnam() issues one.
4488 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4489 This is apparently needed in addition to the PH/07 change above to avoid
4490 any possible encoding problems.
4492 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4493 but not after initializing Perl.
4495 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4496 output them only if debugging. By default they are written stderr,
4497 apparently, which is not desirable.
4499 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4502 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4505 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4507 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4508 authenticated or an ident call has been made. Suppress the default
4509 values for $authenticated_id and $authenticated_sender (but permit -oMai
4510 and -oMas) when testing with -bh.
4512 PH/24 Re-jigged the order of the tests in the default configuration so that the
4513 tests for valid domains and recipients precede the DNS black list and CSA
4514 tests, on the grounds that those ones are more expensive.
4516 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4517 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4518 This bug exists in every version of Exim that I still have, right back to
4521 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4522 However, an attempt to turn on case-sensitivity in a regex key by
4523 including (?-i) didn't work because the subject string was already
4524 lowercased, and the effects were non-intuitive. It turns out that a
4525 one-line patch can be used to allow (?-i) to work as expected.
4531 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4532 other effects) broke the use of negated acl sub-conditions.
4534 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4537 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4538 "Deny" causes Exim to reject the incoming connection with a 554 error.
4539 Unfortunately, if there is a major crisis, such as a disk failure,
4540 tcp-wrappers gives "deny", whereas what one would like would be some
4541 kind of temporary error. A kludge has been added to help with this.
4542 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4543 554 error is used if errno is still zero or contains ENOENT (which occurs
4544 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4547 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4549 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4550 errors. Otherwise a message that provokes a temporary error (when other
4551 messages do not) can cause a whole host to time out.
4553 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4554 addresses were routed directly to files or pipes from a redirect router.
4555 File deliveries just didn't batch; pipe deliveries might have suffered
4558 PH/06 A failure to get a lock for a hints database would erroneously always say
4559 "Failed to get write lock", even when it was really a read lock.
4561 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4562 of 0600. This has been changed to use the value of the lockfile_mode
4563 option (which defaults to 0600).
4565 PH/08 Applied small patch from the Sieve maintainer.
4567 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4568 folder from quota calculations, a direct delivery into this folder messed
4569 up the contents of the maildirsize file. This was because the regex was
4570 used only to exclude .Trash (or whatever) when the size of the mailbox
4571 was calculated. There was no check that a delivery was happening into an
4572 excluded directory. This bug has been fixed by ignoring all quota
4573 processing for deliveries into excluded directories.
4575 PH/10 Added the maildirfolder_create_regex option to appendfile.
4581 PH/01 The code for finding all the local interface addresses on a FreeBSD
4582 system running IPv6 was broken. This may well have applied to all BSD
4583 systems, as well as to others that have similar system calls. The broken
4584 code found IPv4 interfaces correctly, but gave incorrect values for the
4585 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4586 that it would not match correctly against @[] and not recognize the IPv6
4589 PH/02 The ipliteral router was not recognizing addresses of the form user@
4590 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4592 PH/03 Added disable_ipv6.
4594 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4595 decoded form, because it is most often used to construct To: headers
4596 lines in autoreplies, and the decoded form may well be syntactically
4597 invalid. However, $reply_address has leading white space removed, and all
4598 newlines turned into spaces so that the autoreply transport does not
4601 PH/05 If group was specified without a user on a router, and no group or user
4602 was specified on a transport, the group from the router was ignored.
4604 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4605 for visible compile-time settings that can be used to change these
4606 numbers, for those that want even more. Backwards compatibility with old
4607 spool files has been maintained. However, going back to a previous Exim
4608 release will lost any variables that are in spool files.
4610 PH/07 Two small changes when running in the test harness: increase delay when
4611 passing a TCP/IP connection to a new process, in case the original
4612 process has to generate a bounce, and remove special handling of
4613 127.0.0.2 (sic), which is no longer necessary.
4615 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4616 be the same on different OS.
4618 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4621 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4622 whether --show-vars was specified or not
4624 JJ/02 exipick: Added support for new ACL variable spool format introduced
4627 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4628 syntactically invalid From: or Reply-to: line, and a filter used this to
4629 generate an autoreply, and therefore failed to obtain an address for the
4630 autoreply, Exim could try to deliver to a non-existent relative file
4631 name, causing unrelated and misleading errors. What now happens is that
4632 it logs this as a hard delivery error, but does not attempt to create a
4635 PH/11 The exinext utility has a -C option for testing purposes, but although
4636 the given file was scanned by exinext itself; it wasn't being passed on
4637 when Exim was called.
4639 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4640 an end-of-file indication when reading a command response.
4642 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4643 compiled. In many other places in Exim, IPv6 addresses are always
4644 recognized, so I have changed this. It also means that IPv4 domain
4645 literals of the form [IPV4:n.n.n.n] are now always recognized.
4647 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4648 used if the router is not running as root, for example, when verifying at
4649 ACL time, or when using -bh. The debugging output from this situation was
4650 non-existent - all you got was a failure to exec. I have made two
4653 (a) Failures to set uid/gid, the current directory, or a process leader
4654 in a subprocess such as that created by queryprogram now generate
4655 suitable debugging output when -d is set.
4657 (b) The queryprogram router detects when it is not running as root,
4658 outputs suitable debugging information if -d is set, and then runs
4659 the subprocess without attempting to change uid/gid.
4661 PH/15 Minor change to Makefile for building test_host (undocumented testing
4664 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4665 additional section of a DNS packet that returns MX or SRV records.
4666 Instead, it always explicitly searches for A/AAAA records. This avoids
4667 major problems that occur when a DNS server includes only records of one
4668 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4669 fixed another bug: if SRV records were looked up and the corresponding
4670 address records were *not* found in the additional section, the port
4671 values from the SRV records were lost.
4673 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4674 using the correct key (the original address) when searching the retry
4675 rules in order to find which one to use for generating the retry hint.
4677 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4678 adding the default one. Similarly, if it contains a Reply-To: header, the
4679 errors_reply_to option, if set, is not used.
4681 PH/19 When calculating a retry time, Exim used to measure the "time since
4682 failure" by looking at the "first failed" field in the retry record. Now
4683 it does not use this if it is later than than the arrival time of the
4684 message. Instead it uses the arrival time. This makes for better
4685 behaviour in cases where some deliveries succeed, thus re-setting the
4686 "first failed" field. An example is a quota failure for a huge message
4687 when small messages continue to be delivered. Without this change, the
4688 "time since failure" will always be short, possible causing more frequent
4689 delivery attempts for the huge message than are intended.
4690 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4692 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4693 $1, $2, $3) because the numerical variables can be reset during some
4694 expansion items (e.g. "match"), thereby losing the authentication data.
4696 PH/21 Make -bV show the size of off_t variables so that the test suite can
4697 decide whether to run tests for quotas > 2G.
4699 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4700 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4701 filecount value is greater than 2G or if a quota value is greater than 2G
4702 on a system where the size of off_t is not greater than 4, a panic error
4705 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4706 never match. The debug and -bh output now contains an explicit error
4707 message indicating a malformed IPv4 address or mask.
4709 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4710 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4711 PH/23 above applies.
4713 PH/25 Do not write to syslog when running in the test harness. The only
4714 occasion when this arises is a failure to open the main or panic logs
4715 (for which there is an explicit test).
4717 PH/26 Added the /no_tell option to "control=freeze".
4719 PH/27 If a host name lookup failed very early in a connection, for example, if
4720 the IP address matched host_lookup and the reverse lookup yielded a name
4721 that did not have a forward lookup, an error message of the form "no IP
4722 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4723 could be logged. Now it outputs the IP address instead of "NULL".
4725 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4726 allows the sender and the authenticated sender to be set when
4727 submitting a message from within Exim. Since child_open_exim() is
4728 documented for local_scan(), the new function should be too.
4730 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4731 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4732 results in an empty string is now treated as unset.
4734 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4736 PH/31 Added sender_verify_fail logging option.
4738 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4739 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4740 tidied the source and removed it altogether.
4742 PH/33 When a queue run was abandoned because the load average was too high, a
4743 log line was always written; now it is written only if the queue_run log
4744 selector is set. In addition, the log line for abandonment now contains
4745 information about the queue run such as the pid. This is always present
4746 in "start" and "stop" lines but was omitted from the "abandon" line.
4748 PH/34 Omit spaces between a header name and the colon in the error message that
4749 is given when verify = headers_syntax fails (if there are lots of them,
4750 the message gets confusing).
4752 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4753 names, as there are now some PTR records that contain slashes. This check
4754 is only to protect against broken name servers that fall over on strange
4755 characters, so the fact that it applies to all lookups doesn't matter.
4757 PH/36 Now that the new test suite is complete, we can remove some of the
4758 special code in Exim that was needed for the old test suite. For example,
4759 sorting DNS records because real resolvers return them in an arbitrary
4760 order. The new test suite's fake resolver always returns records in the
4763 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4764 bounces) except when queue_only is set, to avoid logging races between
4765 the different processes.
4767 PH/38 Panic-die if .include specifies a non-absolute path.
4769 PH/39 A tweak to the "H" retry rule from its user.
4771 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4772 a label. They prevented compilation on older perls.
4774 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4775 a warning to be raised on newish perls.
4777 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4778 on queue. Changes to match documented behaviour of showing count of
4779 messages matching specified criteria.
4781 PH/40 Changed the default ident timeout from 30s to 5s.
4783 PH/41 Added support for the use of login_cap features, on those BSD systems
4784 that have them, for controlling the resources used by pipe deliveries.
4786 PH/42 The content-scanning code uses fopen() to create files in which to put
4787 message data. Previously it was not paying any attention to the mode of
4788 the files. Exim runs with umask(0) because the rest of the code creates
4789 files with open(), and sets the required mode explicitly. Thus, these
4790 files were ending up world-writeable. This was not a big issue, because,
4791 being within the spool directory, they were not world-accessible. I have
4792 created a function called modefopen, which takes an additional mode
4793 argument. It sets umask(777), creates the file, chmods it to the required
4794 mode, then resets the umask. All the relevant calls to fopen() in the
4795 content scanning code have been changed to use this function.
4797 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4798 to 24 hours. This avoids potential overflow problems when processing G
4799 and H retry rules. I suspect nobody ever tinkers with this value.
4801 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4803 PH/45 When the plaintext authenticator is running as a client, the server's
4804 challenges are checked to ensure they are valid base64 strings. By
4805 default, the authentication attempt is cancelled if an invalid string is
4806 received. Setting client_ignore_invalid_base64 true ignores these errors.
4807 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4808 they are received. Thus, the responses can be made to depend on the
4809 challenges. If an invalid string is ignored, an empty string is placed in
4812 PH/46 Messages that are created by the autoreply transport now contains a
4813 References: header, in accordance with RFCs 2822 and 3834.
4815 PH/47 Added authenticated_sender_force to the smtp transport.
4817 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4819 PH/49 Installed latest patch from the Sieve maintainer.
4821 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4822 was also set, the appendfile transport was unnecessarily scanning a
4823 directory of message files (e.g. for maildir delivery) to find the count
4824 of files (along with the size), even though it did not need this
4825 information. It now does the scan only if it needs to find either the
4826 size of the count of files.
4828 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4830 PH/52 Two bugs concerned with error handling when the smtp transport is
4833 (i) Exim was not creating retry information for temporary errors given
4834 for individual recipients after the DATA command when the smtp transport
4835 was used in LMTP mode. This meant that they could be retried too
4836 frequently, and not timed out correctly.
4838 (ii) Exim was setting the flag that allows error details to be returned
4839 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4840 recipients that were returned after the DATA command.
4842 PH/53 This is related to PH/52, but is more general: for any failing address,
4843 when detailed error information was permitted to be returned to the
4844 sender, but the error was temporary, then after the final timeout, only
4845 "retry timeout exceeded" was returned. Now it returns the full error as
4846 well as "retry timeout exceeded".
4848 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4849 do this, and (what is worse) MTAs that accept it.
4851 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4852 will now be deprecated.
4854 PH/56 New os.c-cygwin from the Cygwin maintainer.
4856 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4857 formats (previously only available in exim formats via -bpr, -bpru,
4858 and -bpra. Now also available in native and exiqgrep formats)
4860 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4861 with very large, slow to parse queues
4863 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4865 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4867 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4868 responses to authentication challenges, though it was showing the
4869 challenges; (ii) I've removed the CR characters from the debug output for
4872 PH/58 Allow for the insertion of a newline as well as a space when a string
4873 is turned into more than one encoded-word during RFC 2047 encoding. The
4874 Sieve code now uses this.
4876 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4877 data_4xx, lost_connection, tls_required.
4879 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4880 message was being sent as an SMTP response.
4882 PH/61 Add -l and -k options to exicyclog.
4884 PH/62 When verifying, if an address was redirected to one new address, so that
4885 verification continued, and the new address failed or deferred after
4886 having set something in $address_data, the value of $address_data was not
4887 passed back to the ACL. This was different to the case when no
4888 redirection occurred. The value is now passed back in both cases.
4890 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4891 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4892 use login_cap.h, so on its own it isn't the distinguishing feature. The
4893 new name refers directly to the setclassresources() function.
4895 PH/65 Added configuration files for NetBSD3.
4897 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4899 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4900 is preferred over IPv4.
4902 PH/68 The bounce_return_message and bounce_return_body options were not being
4903 honoured for bounces generated during the reception of non-SMTP messages.
4904 In particular, this applied to messages rejected by the ACL. This bug has
4905 been fixed. However, if bounce_return_message is true and bounce_return_
4906 body is false, the headers that are returned for a non-SMTP message
4907 include only those that have been read before the error was detected.
4908 (In the case of an ACL rejection, they have all been read.)
4910 PH/69 The HTML version of the specification is now built in a directory called
4911 spec_html instead of spec.html, because the latter looks like a path with
4912 a MIME-type, and this confuses some software.
4914 PH/70 Catch two compiler warnings in sieve.c.
4916 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4917 function verify_get_ident() calls ip_connect() to connect a socket, but
4918 if the "connect()" function timed out, ip_connect() used to close the
4919 socket. However, verify_get_ident() also closes the socket later, and in
4920 between Exim writes to the log, which may get opened at this point. When
4921 the socket was closed in ip_connect(), the log could get the same file
4922 descriptor number as the socket. This naturally causes chaos. The fix is
4923 not to close the socket in ip_connect(); the socket should be closed by
4924 the function that creates it. There was only one place in the code where
4925 this was missing, in the iplookup router, which I don't think anybody now
4926 uses, but I've fixed it anyway.
4928 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4929 well as to direct DNS lookups. Otherwise the handling of names in host
4930 lists is inconsistent and therefore confusing.
4936 PH/01 Two changes to the default runtime configuration:
4938 (1) Move the checks for relay_from_hosts and authenticated clients from
4939 after to before the (commented out) DNS black list checks.
4941 (2) Add control=submission to the relay_from_hosts and authenticated
4942 clients checks, on the grounds that messages accepted by these
4943 statements are most likely to be submissions.
4945 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4947 (1) Generate an error if the third argument for the ${prvs expansion is
4950 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4953 (3) Reset the variables that are obtained from the first argument of
4954 ${prvscheck and used in the second argument before leaving the code,
4955 because their memory is reclaimed, so using them afterwards may do
4958 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4959 one (it's much easier than Tom thought :-).
4961 (5) Because of (4), we can now allow for the use of $prvscheck_result
4962 inside the third argument.
4964 PH/03 For some reason, the default setting of PATH when running a command from
4965 a pipe transport was just "/usr/bin". I have changed it to
4968 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4969 anything to be listed in the output from -bV.
4971 PH/05 When a filter generated an autoreply, the entire To: header line was
4972 quoted in the delivery log line, like this:
4974 => >A.N.Other <ano@some.domain> <original@ddress> ...
4976 This has been changed so that it extracts the operative address. There
4977 may be more than one such address. If so, they are comma-separated, like
4980 => >ano@some.domain,ona@other.domain <original@ddress> ...
4982 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4983 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4984 not being looked up in the rDNS to get a host name, Exim was showing the
4985 IP address twice in Received: lines, even though the IP addresses were
4986 identical. For example:
4988 Received: from [1.2.3.4] (helo=[1.2.3.4])
4990 However, if the real host name was known, it was omitting the HELO data
4991 if it matched the actual IP address. This has been tidied up so that it
4992 doesn't show the same IP address twice.
4994 PH/07 When both +timestamp and +memory debugging was on, the value given by
4995 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4996 called by the debug printing, thereby overwriting the timestamp buffer.
4997 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4999 PH/08 When the original message was included in an autoreply transport, it
5000 always said "this is a copy of the message, including all the headers",
5001 even if body_only or headers_only was set. It now gives an appropriate
5004 PH/09 Applied a patch from the Sieve maintainer which:
5006 o fixes some comments
5007 o adds the (disabled) notify extension core
5008 o adds some debug output for the result of if/elsif tests
5009 o points to the current vacation draft in the documentation
5010 and documents the missing references header update
5014 o fixes a bug in processing the envelope test (when testing
5015 multiple envelope elements, the last element determined the
5018 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5019 Electronic Mail") by including:
5021 Auto-submitted: auto-generated
5023 in the messages that it generates (bounce messages and others, such as
5024 warnings). In the case of bounce messages for non-SMTP messages, there was
5025 also a typo: it was using "Auto_submitted" (underscore instead of
5026 hyphen). Since every message generated by Exim is necessarily in response
5027 to another message, thes have all been changed to:
5029 Auto-Submitted: auto-replied
5031 in accordance with these statements in the RFC:
5033 The auto-replied keyword:
5035 - SHOULD be used on messages sent in direct response to another
5036 message by an automatic process,
5038 - MUST NOT be used on manually-generated messages,
5040 - MAY be used on Delivery Status Notifications (DSNs) and Message
5041 Disposition Notifications (MDNs),
5043 - MUST NOT be used on messages generated by automatic or periodic
5044 processes, except for messages which are automatic responses to
5047 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5048 to the default Received: header definition.
5050 PH/12 Added log selector acl_warn_skipped (default on).
5052 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5053 variables because (a) they are in the wrong storage pool and (b) even if
5054 they were copied, it wouldn't work properly because of the caching.
5056 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5057 checking when decoding. Apparently there are clients that generate
5058 overlong encoded strings. Why am I not surprised?
5060 PH/15 If the first argument of "${if match_address" was not empty, but did not
5061 contain an "@" character, Exim crashed. Now it writes a panic log message
5062 and treats the condition as false.
5064 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5066 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5067 extension "envelope-auth". The code is finished and in agreement with
5068 other implementations, but there is no documentation so far and in fact,
5069 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5070 not changing the active code.
5072 Print executed "if" and "elsif" statements when debugging is used. This
5073 helps a great deal to understand what a filter does.
5075 Document more things not specified clearly in RFC3028. I had all this
5076 sorted out, when out of a sudden new issues came to my mind. Oops."
5078 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5081 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5082 canonical form (as documented). However, after a host name lookup from
5083 the IP address, check_host() was doing a simple string comparison with
5084 addresses acquired from the DNS when checking that the found name did
5085 have the original IP as one of its addresses. Since any found IPv6
5086 addresses are likely to be in abbreviated form, the comparison could
5087 fail. Luckily, there already exists a function for doing the comparison
5088 by converting both addresses to binary, so now that is used instead of
5089 the text comparison.
5091 PH/20 There was another similar case to PH/19, when a complete host name was
5092 given in a host list; looking up its IP address could give an abbreviated
5093 form, whereas the current host's name might or might not be abbreviated.
5094 The same fix has been applied.
5100 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5101 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5104 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5105 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5107 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5109 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5110 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5111 Also, for "Auto-Submitted", treat anything other than "no" as
5112 non-personal, in accordance with RFC 3834. (Previously it treated
5113 anything starting "auto-" as non-personal.)
5115 TF/01 The control=submission/name=... option had a problem with syntax
5116 errors if the name included a slash character. The /name= option
5117 now slurps the rest of the string, so it can include any characters
5118 but it must come last in the list of options (after /sender_retain
5121 PH/05 Some modifications to the interface to the fake nameserver for the new
5129 TK/01 Added the "success_on_redirect" address verification option. See
5130 NewStuff for rationale and an example.
5132 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5134 PH/02 Patch to exigrep to allow it to work on syslog lines.
5136 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5137 fread() to skip over the body file's header line, because in Cygwin the
5138 header line is locked and is inaccessible.
5140 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5141 co-exist for some time) to make it clear that it is the Exim ID that is
5142 referenced, not the Message-ID: header line.
5144 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5145 string_format() function, because snprintf() does not exist on all
5148 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5149 ${stat: expansion item.
5151 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5152 protocol synchronization error", to keep the pedants happy.
5154 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5155 well as for IRIX systems, when gcc is being used. See the host.c source
5158 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5160 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5163 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5164 transport and to the smtp transport in LMTP mode.
5166 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5168 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5169 run a filter in a subprocess. This could lead to confusion in subsequent
5170 lookups in the parent process. There should also be a search_tidyup() at
5171 the end of the subprocess.
5173 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5174 only if the host matched helo_try_verify_hosts, which caused the
5175 verification to occur when the EHLO/HELO command was issued. The ACL just
5176 tested the remembered result. Now, if a previous verification attempt has
5177 not happened, "verify = helo" does it there and then.
5179 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5181 TK/03 Fix log output including CR from clamd.
5183 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5184 exist provoked a memory error which could cause a segfault.
5186 PH/15 Installed PCRE 6.2
5188 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5190 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5191 of the problem. Specifically, suggested +O2 rather than +O1 for the
5194 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5196 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5197 record, the port was not correctly logged when the outgoing_port log
5198 selector was set (it logged the transort's default port).
5200 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5201 fallback_hosts, and "hosts" in the smtp transport.
5203 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5204 host errors such as "Connection refused".
5206 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5207 authentication with radiusclient 0.4.9:
5209 - Error returned from rc_read_config was caught wrongly
5210 - Username/password not passed on to radius server due to wrong length.
5212 The presumption is that some radiusclient API changes for 4.51/PH/17
5213 were not taken care of correctly. The code is still untested by me (my
5214 Linux distribution still has 0.3.2 of radiusclient), but it was
5215 contributed by a Radius user.
5217 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5218 expanding the "port" option of the smtp transport.
5220 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5221 while reading a MIME header. Thanks to Tom Hughes for a patch.
5223 PH/24 Include config.h inside local_scan.h so that configuration settings are
5226 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5227 that in an ACL for RCPT (for example), you can examine exactly what was
5230 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5231 commands, but it was not correctly comparing the address with the actual
5232 client host address. Thus, it would show the EHLO address in Received:
5233 header lines when this was not necessary.
5235 PH/27 Added the % operator to ${eval:}.
5237 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5238 it should be ignoring failures (because with -C, for example, it has lost
5239 privilege). It wasn't ignoring creation failures other than "already
5242 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5245 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5246 because an input line was too long, either on its own, or by virtue of
5247 too many continuations, the temporary file was not being removed, and the
5248 return code was incorrect.
5250 PH/31 Missing "BOOL" in function definition in filtertest.c.
5252 PH/32 Applied Sieve patches from the maintainer.
5254 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5256 PH/33 Added "verify = not_blind".
5258 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5259 Local/Makefile (with some defaults set). These are used in built scripts
5260 such as exicyclog, but they have never been used in the exim_install
5261 script (though there are many overriding facilities there). I have
5262 arranged that the exim_install script now takes note of these two
5265 PH/35 Installed configuration files for Dragonfly.
5267 PH/36 When a locally submitted message by a trusted user did not contain a
5268 From: header, and the sender address was obtained from -f or from an SMTP
5269 MAIL command, and the trusted user did not use -F to supply a sender
5270 name, $originator_name was incorrectly used when constructing a From:
5271 header. Furthermore, $originator_name was used for submission mode
5272 messages from external hosts without From: headers in a similar way,
5273 which is clearly wrong.
5275 PH/37 Added control=suppress_local_fixups.
5277 PH/38 When log_selector = +received_sender was set, and the addition of the
5278 sender made the log line's construction buffer exactly full, or one byte
5279 less than full, an overflow happened when the terminating "\n" was
5282 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5283 when the result of a list match is failure because a DNS lookup failed.
5285 PH/40 RM_COMMAND is now used in the building process.
5287 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5288 the "build-* directories that it finds.
5290 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5291 address in a domain literal was a prefix of an interface address.
5293 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5294 when verifying a sender address, unless rewrite_headers is false.
5296 PH/44 Wrote a long comment about why errors_to addresses are verified as
5297 recipients, not senders.
5299 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5300 the ratelimit ACL was added.
5302 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5304 PH/46 Added extra information about PostgreSQL errors to the error string.
5306 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5307 avoiding the need to install special zones in a real server. This is
5308 backwards compatible; if it can't find the fake resolver, it drops back.
5309 Thus, both old and new test suites can be run.
5311 TF/02 Added util/ratelimit.pl
5313 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5314 clock is set back in time.
5316 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5317 Candler <B.Candler@pobox.com>.
5319 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5320 OK for addresses that are the result of redirections.
5322 PH/48 A number of further additions for the benefit of the new test suite,
5323 including a fake gethostbyname() that interfaces to the fake DNS resolver
5326 TF/06 The fix for widen_domains has also been applied to qualify_single and
5327 search_parents which are the other dnslookup options that can cause
5330 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5333 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5335 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5336 service" if it encountered an MX record with an empty target hostname.
5337 The message is now "an MX or SRV record indicated no SMTP service".
5339 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5340 if the DNS of the sending site is misconfigured. This is quite a
5341 common situation. This change restores the behaviour of treating a
5342 helo verification defer as a failure.
5344 PH/51 If self=fail was set on a router, the bounce message did not include the
5345 actual error message.
5351 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5353 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5354 command itself ended in error, the underlying message about the transport
5355 filter timeout was being overwritten with the pipe command error. Now the
5356 underlying error message should be appended to the second error message.
5358 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5360 PH/02 Reduce the amount of output that "make" produces by default. Full output
5361 can still be requested.
5363 PH/03 The warning log line about a condition test deferring for a "warn" verb
5364 was being output only once per connection, rather than after each
5365 occurrence (because it was using the same function as for successful
5366 "warn" verbs). This seems wrong, so I have changed it.
5368 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5369 it should not have, which might have caused a crash in the right
5370 circumstances, but probably never did.
5372 PH/04 Installed a modified version of Tony Finch's patch to make submission
5373 mode fix the return path as well as the Sender: header line, and to
5374 add a /name= option so that you can make the user's friendly name appear
5377 TF/03 Added the control = fakedefer ACL modifier.
5379 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5380 Mark Lowes for thorough testing.
5382 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5384 TK/03 Merged latest SRS patch from Miles Wilton.
5386 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5387 with the definition in sysexits.h (which is #included earlier).
5388 Fortunately, Exim does not actually use EX_OK. The code used to try to
5389 preserve the sysexits.h value, by assuming that macro definitions were
5390 scanned for macro replacements. I have been disabused of this notion,
5391 so now the code just undefines EX_OK before #including unistd.h.
5393 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5394 in the smtp transport. When a block could not be written in a single
5395 write() function, the timeout was being re-applied to each part-write.
5396 This seems wrong - if the receiver was accepting one byte at a time it
5397 would take for ever. The timeout is now adjusted when this happens. It
5398 doesn't have to be particularly precise.
5400 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5401 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5403 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5404 without a domain if the check to <postmaster@domain> fails.
5406 SC/01 Eximstats: added -xls and the ability to specify output files
5407 (patch written by Frank Heydlauf).
5409 SC/02 Eximstats: use FileHandles for outputting results.
5411 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5413 SC/04 Eximstats: fixed display of large numbers with -nvr option
5415 SC/05 Eximstats: fixed merging of reports with empty tables.
5417 SC/06 Eximstats: added the -include_original_destination flag
5419 SC/07 Eximstats: removed tabs and trailing whitespace.
5421 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5423 TK/06 MBOX spool code: Add real "From " MBOX separator line
5424 so the .eml file is really in mbox format (even though
5425 most programs do not really care). Patch from Alex Miller.
5427 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5428 The latter is generated from $received_to and is only set if the
5429 message has one envelope recipient. SA can use these headers,
5430 obviously out-of-the-box. Patch from Alex Miller.
5432 PH/08 The ${def test on a variable was returning false if the variable's
5433 value was "0", contrary to what the specification has always said!
5434 The result should be true unless the variable is empty.
5436 PH/09 The syntax error of a character other than { following "${if
5437 def:variable_name" (after optional whitespace) was not being diagnosed.
5438 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5439 accidental colon was present, for example, could give incorrect results.
5441 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5442 result is used (not including appendfile, where other changes are about
5445 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5446 This involved changing a lot of size variables from int to off_t. It
5447 should work with maildirs and everything.
5449 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5450 spamd dying while we are connected to it.
5452 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5455 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5456 function for BDB 4.3.
5458 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5460 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5461 the sources for PCRE have been re-arranged and more files are now
5464 PH/15 The code I had for printing potentially long long variables in PH/11
5465 above was not the best (it lost precision). The length of off_t variables
5466 is now inspected at build time, and an appropriate printing format (%ld
5467 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5468 to be "long long int" or "long int". This is needed for the internal
5469 formatting function string_vformat().
5471 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5472 the configuration file to be ":syslog", then the script "guesses" where
5473 the logs files are, rather than using the compiled in default. In our
5474 case the guess is not the same as the compiled default, so the script
5475 suddenly stopped working when I started to use syslog. The patch checks
5476 to see if log_file_path is "". If so, it attempts to read it from exim
5477 with no configuration file to get the compiled in version, before it
5478 falls back to the previous guessing code."
5480 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5481 implementing BATV in an Exim configuration. See NewStuff for the gory
5484 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5485 Makefile that are specific to HP-UX.
5487 PH/18 If the "use_postmaster" option was set for a recipient callout together
5488 with the "random" option, the postmaster address was used as the MAIL
5489 FROM address for the random test, but not for the subsequent recipient
5490 test. It is now used for both.
5492 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5493 patch removes a few documentation additions to RFC 3028, because the
5494 latest draft now contains them. It adds the new en;ascii-case comparator
5495 and a new error check for 8bit text in MIME parts. Comparator and
5496 require names are now matched exactly. I enabled the subaddress
5497 extension, but it is not well tested yet (read: it works for me)."
5499 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5500 rework some of the code of TK/09 above to avoid the hardwired use of
5501 "%lld" and "long long". Replaced the call to snprintf() with a call to
5504 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5505 records point to non-existent hosts", "retry timeout exceeded", and
5506 "retry time not reached for any host after a long failure period".
5508 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5509 experimental DomainKeys support:
5511 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5512 (2) On an error such as an illegally used "control", the wrong name for
5513 the control was given.
5515 These problems did NOT occur unless DomainKeys support was compiled.
5517 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5519 PH/24 Added ${if match_ip condition.
5521 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5522 databases so that it will be absolutely obvious if a crash occurs in the
5523 DB library. This is a regular occurrence (often caused by mis-matched
5526 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5527 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5528 picked up on a user's system that detects such things. There doesn't seem
5529 to be a gcc warning option for this - only an attribute that has to be
5530 put on the function's prototype. It seems that in Fedora Core 4 they have
5531 set this on a number of new functions. No doubt there will be more in due
5534 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5535 specify a transport. However, if an address that was verified by such a
5536 router was the subject of a callout, Exim crashed because it tried to
5537 read the rcpt_include_affixes from the non-existent transport. Now it
5538 just assumes that the setting of that option is false. This bug was
5539 introduced by 4.51/PH/31.
5541 PH/28 Changed -d+all to exclude +memory, because that information is very
5542 rarely of interest, but it makes the output a lot bigger. People tend to
5543 do -d+all out of habit.
5545 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5546 code in os-type was giving problems when libc.so lives in lib64, like on
5549 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5550 aren't the modern standard, and it seems that some systems' include files
5551 don't always have them. Exim was already checking for some of the newer
5552 ones like T_AAAA, and defining it itself. I've added checks for all the
5553 record types that Exim uses.
5555 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5556 not automatically generating a new one, as it is supposed to. This
5557 prevented TLS from working. If the file did exist, but contained invalid
5558 data, a new version was generated, as expected. It was only the case of a
5559 non-existent file that was broken.
5561 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5562 with a change in libdomainkeys > 0.64.
5564 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5565 from DNS. If the selector record carries the flag, it now has
5566 precedence over the domain-wide flag.
5568 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5570 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5571 the use of a port name that isn't defined in /etc/services) occurred, the
5572 message was deferred as in a normal delivery, and thus remained on the
5573 spool, instead of being failed because of the mua_wrapper setting. This
5574 is now fixed, and I tidied up some of the mua_wrapper messages at the
5577 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5578 the messages in a hash of arrays rather than using individual hashes.
5579 This is a bit cleaner and results in dramatic memory savings, albeit
5580 at a slight CPU cost.
5582 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5583 as requested by Marc Sherman.
5585 SC/10 Eximstats: added histograms for user specified patterns as requested
5588 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5590 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5591 fopen() in the content-scanning modules that did not already have it.
5597 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5598 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5600 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5602 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5604 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5605 to test Sieve filters that use "vacation".
5607 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5608 that changes the way the GnuTLS parameters are stored in the cache file.
5609 The new format can be generated externally. For backward compatibility,
5610 if the data in the cache doesn't make sense, Exim assumes it has read an
5611 old-format file, and it generates new data and writes a new file. This
5612 means that you can't go back to an older release without removing the
5615 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5616 work if there are any delivery delays because "one_time" forces the
5617 parent to be marked "delivered", so its unseen clone is never tried
5618 again. For this reason, Exim now forbids the simultaneous setting of
5621 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5622 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5623 entry if you want to know the details. The fix, however, neglected to
5624 consider the case when local delivery batching is involved. The test for
5625 "previously delivered" was not happening when checking to see if an
5626 address could be batched with a previous (undelivered) one; under
5627 certain circumstances this could lead to multiple deliveries to the same
5630 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5631 in its include files, and this causes problems building Exim.
5633 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5634 header_syntax) but Exim was just ignoring anything given after a slash.
5635 In particular, this caused confusion with an attempt to use "verify =
5636 reverse_host_lookup/defer_ok". An error is now given when options are
5637 supplied for verify items that do not have them. (Maybe reverse_host_
5638 lookup should have a defer_ok option, but that's a different point.)
5640 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5641 defined by RFC 821) to 2048, because there were problems with some AUTH
5642 commands, and RFC 1869 says the size should be increased for extended
5643 SMTP commands that take arguments.
5645 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5648 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5649 "unknown" error; now it says that the functionality isn't in the binary.
5651 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5652 an address' error message when a string expansion fails (syntax or
5653 whatever). Otherwise the password may appear in the log. Following change
5654 PH/42 below, there is no longer a chance of it appearing in a bounce
5657 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5659 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5660 of its addresses was ever tried. (Bugzilla bug #2.)
5662 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5663 the result incorrectly in the debug output. (It correctly added a newline
5664 to what was transported.)
5666 TF/01 Added $received_time.
5668 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5669 commented out examples of how to interface to a virus scanner and to
5670 SpamAssassin. Also added commented examples of av_scanner and
5671 spamd_address settings.
5673 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5674 and controls are allowed in which ACLs. There were a couple of minor
5675 errors. Some of the entries in the conditions table (which is a table of
5676 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5677 negation of where the condition IS allowed.
5679 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5681 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5682 header file does not have a version number, so I've had to invent a new
5683 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5684 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5685 radiusclient), but it was contributed by a Radius user.
5687 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5688 files or directories, for OpenSSL.
5690 PH/19 When an Exim process that is running as root has to create an Exim log
5691 file, it does so in a subprocess that runs as exim:exim so as to get the
5692 ownership right at creation (otherwise, other Exim processes might see
5693 the file with the wrong ownership). There was no test for failure of this
5694 fork() call, which would lead to the process getting stuck as it waited
5695 for a non-existent subprocess. Forks do occasionally fail when resources
5696 run out. I reviewed all the other calls to fork(); they all seem to check
5699 PH/20 When checking for unexpected SMTP input at connect time (before writing
5700 the banner), Exim was not dealing correctly with a non-positive return
5701 from the read() function. If the client had disconnected by this time,
5702 the result was a log entry for a synchronization error with an empty
5703 string after "input=" when read() returned zero. If read() returned -1
5704 (an event I could not check), uninitialized data bytes were printed.
5705 There were reports of junk text (parts of files, etc) appearing after
5708 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5710 PH/22 Added support for macro redefinition, and (re)definition in between
5711 driver and ACL definitions.
5713 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5714 forgetting to use the resulting value; it was using the unexpanded value.
5716 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5717 hadn't been configured. The fix is from Juergen Kreileder, who
5718 understands it better than I do:
5720 "Here's what I see happening with three configured cyrus_sasl
5721 authenticators configured (plain, login, cram-md5):
5723 On startup auth_cyrus_sasl_init() gets called for each of these.
5724 This means three calls to sasl_listmech() without a specified mech_list.
5725 => SASL tests which mechs of all available mechs actually work
5726 => three warnings about OTP not working
5727 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5729 With the patch, sasl_listmech() also gets called three times. But now
5730 SASL's mech_list option is set to the server_mech specified in the the
5731 authenticator. Or in other words, the answer from sasl_listmech()
5732 gets limited to just the mech you're testing for (which is different
5734 => the return list contains just 'plain' or 'login', 'cram-md5' or
5735 nothing depending on the value of ob->server_mech.
5737 I've just tested the patch: Authentication still works fine,
5738 unavailable mechs specified in the exim configuration are still
5739 caught, and the auth.log warnings about OTP are gone."
5741 PH/25 When debugging is enabled, the contents of the command line are added
5742 to the debugging output, even when log_selector=+arguments is not
5745 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5746 answer is "GNU", and only if the return is "GNU/something" is the answer
5749 PH/27 $acl_verify_message is now set immediately after the failure of a
5750 verification in an ACL, and so is available in subsequent modifiers. In
5751 particular, the message can be preserved by coding like this:
5753 warn !verify = sender
5754 set acl_m0 = $acl_verify_message
5756 Previously, $acl_verify_message was set only while expanding "message"
5757 and "log_message" when a very denied access.
5759 PH/28 Modified OS/os.c-Linux with
5761 -#ifndef OS_LOAD_AVERAGE
5762 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5764 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5765 nomenclature these days.)
5767 PH/29 Installed patch from the Sieve maintainer that adds the options
5768 sieve_useraddress and sieve_subaddress to the redirect router.
5770 PH/30 In these circumstances:
5771 . Two addresses routed to the same list of hosts;
5772 . First host does not offer TLS;
5773 . First host accepts first address;
5774 . First host gives temporary error to second address;
5775 . Second host offers TLS and a TLS session is established;
5776 . Second host accepts second address.
5777 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5778 and peerdn, if requested) that were in fact used only for the second
5781 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5782 attention to any local part prefix or suffix that was matched by the
5783 router that accepted the address. It now behaves in the same way as it
5784 does for delivery: the affixes are removed from the local part unless
5785 rcpt_include_affixes is set on the transport.
5787 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5788 timeout during the DATA phase of an incoming message.
5790 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5791 applied a patch sanctioned by the Sieve maintainer.
5793 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5794 the uid or gid is negative. A case of a negative gid caused this to be
5795 noticed. The fix allows for either to be negative.
5797 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5798 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5801 PH/36 The change PH/12 above was broken. Fixed it.
5803 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5804 the grounds that routing the same address twice would always produce the
5805 same answer. This might have been true once, but it is certainly no
5806 longer true now. Routing a child address may depend on the previous
5807 routing that produced that child. Some complicated redirection strategies
5808 went wrong when messages had multiple recipients, and made Exim's
5809 behaviour dependent on the order in which the addresses were given.
5811 I have moved the duplicate checking until after the routing is complete.
5812 Exim scans the addresses that are assigned to local and remote
5813 transports, and removes any duplicates. This means that more work will be
5814 done, as duplicates will always all be routed, but duplicates are
5815 presumably rare, so I don't expect this is of any significance.
5817 For deliveries to pipes, files, and autoreplies, the duplicate checking
5818 still happens during the routing process, since they are not going to be
5821 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5822 It corrects a timeout issue with spamd. This is Ian's comment: "The
5823 background is that sometimes spamd either never reads data from a
5824 connection it has accepted, or it never writes response data. The exiscan
5825 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5826 blindly assumes that writes won't block so it may never time out."
5828 PH/39 Allow G after quota size as well as K and M.
5830 PH/40 The value set for $authenticated_id in an authenticator may not contain
5831 binary zeroes or newlines because the value is written to log lines and
5832 to spool files. There was no check on this. Now the value is run through
5833 the string_printing() function so that such characters are converted to
5834 printable escape sequences.
5836 PH/41 $message_linecount is a new variable that contains the total number of
5837 lines in the message. Compare $body_linecount, which is the count for the
5840 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5841 bounce and delay warning messages, except in certain special cases, which
5844 (a) An SMTP error message from a remote host;
5845 (b) A message specified in a :fail: redirection;
5846 (c) A message specified in a "fail" command in a system filter;
5847 (d) A message specified in a FAIL return from the queryprogram router;
5848 (e) A message specified by the cannot_route_message router option.
5850 In these cases only, Exim does include the error details in bounce and
5851 warning messages. There are also a few cases where bland messages such
5852 as "unrouteable address" or "local delivery error" are given.
5854 PH/43 $value is now also set for the "else" part of a ${run expansion.
5856 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5857 being worked on, but at least Exim now implements the latest version to
5860 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5861 process to complete was treated as a delivery failure, a timeout while
5862 writing the message to the pipe was logged, but erroneously treated as a
5863 successful delivery. Such timeouts include transport filter timeouts. For
5864 consistency with the overall process timeout, these timeouts are now
5865 treated as errors, giving rise to delivery failures by default. However,
5866 there is now a new Boolean option for the pipe transport called
5867 timeout_defer, which, if set TRUE, converts the failures into defers for
5868 both kinds of timeout. A transport filter timeout is now identified in
5871 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5872 systems where "make" and "gmake" are different, calling "gmake" at top
5873 level broke things. I've arranged for the value of $(MAKE) to be passed
5874 from the Makefile to this script so that it can call the same version of
5878 A note about Exim versions 4.44 and 4.50
5879 ----------------------------------------
5881 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5882 changes of various kinds. As a consequence, a big documentation update was
5883 needed. This delayed the release for rather longer than seemed good, especially
5884 in the light of a couple of (minor) security issues. Therefore, the changes
5885 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5886 release. So 4.44 and 4.50 are in effect two different branches that both start
5889 I have left the 4.50 change log unchanged; it contains all the changes since
5890 4.43. The change log for 4.44 is below; many of its items are identical to
5891 those for 4.50. This seems to be the most sensible way to preserve the
5892 historical information.
5898 1. Minor wording change to the doc/README.SIEVE file.
5900 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5901 computation of the current number of files was incorrect.
5903 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5904 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5907 4. Give more explanation in the error message when the command for a transport
5908 filter fails to execute.
5910 5. There are several places where Exim runs a non-Exim command in a
5911 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5912 was being done only for the command run by the queryprogram router. It is
5913 now done for all such subprocesses. The other cases are: ${run, transport
5914 filters, and the commands run by the lmtp and pipe transports.
5916 6. Added CONFIGURE_GROUP build-time option.
5918 7. Some older OS have a limit of 256 on the maximum number of file
5919 descriptors. Exim was using setrlimit() to set 1000 as a large value
5920 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5921 systems. I've change it so that if it can't get 1000, it tries for 256.
5923 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5924 was an oversight, and furthermore, ever since the addition of extra
5925 controls (e.g. 4.43/32), the checks on when to allow different forms of
5926 "control" were broken. There should now be diagnostics for all cases when a
5927 control that does not make sense is encountered.
5929 9. Added the /retain_sender option to "control=submission".
5931 10. $recipients is now available in the predata ACL (oversight).
5933 11. Tidy the search cache before the fork to do a delivery from a message
5934 received from the command line. Otherwise the child will trigger a lookup
5935 failure and thereby defer the delivery if it tries to use (for example) a
5936 cached ldap connection that the parent has called unbind on.
5938 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5939 of $address_data from the recipient verification was clobbered by the
5940 sender verification.
5942 13. The value of address_data from a sender verification is now available in
5943 $sender_address_data in subsequent conditions in the ACL statement.
5945 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5947 15. Added a new option "connect=<time>" to callout options, to set a different
5950 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5951 was its contents. (It was OK if the option was not defined at all.)
5953 17. A "Completed" log line is now written for messages that are removed from
5954 the spool by the -Mrm option.
5956 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5957 information about exactly what failed.
5959 19. Added -dd to debug only the daemon process.
5961 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5962 handles timeouts, both on the server side and network timeouts. Renamed the
5963 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5965 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5966 It is now set to "smtps".
5968 22. $host_address is now set to the target address during the checking of
5969 ignore_target_hosts.
5971 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5972 being passed; this would have caused $sender_host_name to have been used if
5973 matching the list had actually called for a host name (not very likely,
5974 since this list is usually IP addresses). A host name is now passed as
5977 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5978 code that specifies a non-restarting handler (typically sigaction() in
5979 modern systems) in an attempt to fix a rare and obscure crash bug.
5981 25. Narrowed the window for a race in the daemon that could cause it to ignore
5982 SIGCHLD signals. This is not a major problem, because they are used only to
5983 wake it up if nothing else does.
5985 26. A malformed maildirsize file could cause Exim to calculate negative values
5986 for the mailbox size or file count. Odd effects could occur as a result.
5987 The maildirsize information is now recalculated if the size or filecount
5990 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5991 support for a long time. Removed HAVE_SYS_VFS_H.
5993 28. Installed the latest version of exipick from John Jetmore.
5995 29. In an address list, if the pattern was not a regular expression, an empty
5996 subject address (from a bounce message) matched only if the pattern was an
5997 empty string. Non-empty patterns were not even tested. This was the wrong
5998 because it is perfectly reasonable to use an empty address as part of a
5999 database query. An empty address is now tested by patterns that are
6000 lookups. However, all the other forms of pattern expect the subject to
6001 contain a local part and a domain, and therefore, for them, an empty
6002 address still always fails if the pattern is not itself empty.
6004 30. Exim went into a mad DNS loop when attempting to do a callout where the
6005 host was specified on an smtp transport, and looking it up yielded more
6006 than one IP address.
6008 31. Re-factored the code for checking spool and log partition space into a
6009 function that finds that data and another that does the check. The former
6010 is then used to implement four new variables: $spool_space, $log_space,
6011 $spool_inodes, and $log_inodes.
6013 32. The RFC2047 encoding function was originally intended for short strings
6014 such as real names; it was not keeping to the 75-character limit for
6015 encoded words that the RFC imposes. It now respects the limit, and
6016 generates multiple encoded words if necessary. To be on the safe side, I
6017 have increased the buffer size for the ${rfc2047: expansion operator from
6020 33. It is now permitted to omit both strings after an "if" condition; if the
6021 condition is true, the result is "true". As before, when the second string
6022 is omitted, a false condition yields an empty string. This makes it less
6023 cumbersome to write custom ACL and router conditions.
6025 34. Failure to deliver a bounce message always caused it to be frozen, even if
6026 there was an errors_to setting on the router. The errors_to setting is now
6029 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6030 canonical form (fully expanded) before being placed in
6031 $sender_host_address.
6033 36. The table in the code that translates DNS record types into text (T_A to
6034 "A" for instance) was missing entries for NS and CNAME. It is just possible
6035 that this could have caused confusion if both these types were looked up
6036 for the same domain, because the text type is used as part of Exim's
6037 per-process caching. But the chance of anyone hitting this buglet seems
6040 37. The dnsdb lookup has been extended in a number of ways.
6042 (1) There is a new type, "zns", which walks up the domain tree until it
6043 finds some nameserver records. It should be used with care.
6045 (2) There is a new type, "mxh", which is like "mx" except that it returns
6046 just the host names, not the priorities.
6048 (3) It is now possible to give a list of domains (or IP addresses) to be
6049 looked up. The behaviour when one of the lookups defers can be
6050 controlled by a keyword.
6052 (4) It is now possible to specify the separator character for use when
6053 multiple records are returned.
6055 38. The dnslists ACL condition has been extended: it is now possible to supply
6056 a list of IP addresses and/or domains to be looked up in a particular DNS
6059 39. Added log_selector=+queue_time_overall.
6061 40. When running the queue in the test harness, wait just a tad after forking a
6062 delivery process, to get repeatability of debugging output.
6064 41. Include certificate and key file names in error message when GnuTLS fails
6065 to set them up, because the GnuTLS error message doesn't include the name
6066 of the failing file when there is a problem reading it.
6068 42. Allow both -bf and -bF in the same test run.
6070 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6072 44. The "Exiscan patch" is now merged into the mainline Exim source.
6074 45. Sometimes the final signoff response after QUIT could fail to get
6075 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6076 < 0 before doing a fflush(). This bug looks as though it goes back to the
6077 introduction of TLS in release 3.20, but "sometimes" must have been rare
6078 because the tests only now provoked it.
6080 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6081 (this can affect the format of dates).
6083 47. exim_tidydb, when checking for the continued existence of a message for
6084 which it has found a message-specific retry record, was not finding
6085 messages that were in split spool directories. Consequently, it was
6086 deleting retry records that should have stayed in existence.
6088 48. Steve fixed some bugs in eximstats.
6090 49. The SPA authentication driver was not abandoning authentication and moving
6091 on to the next authenticator when an expansion was forced to fail,
6092 contradicting the general specification for all authenticators. Instead it
6093 was generating a temporary error. It now behaves as specified.
6095 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6096 (the order specifies the preference for clients). The order is now AES256,
6097 AES128, 3DES, ARCFOUR128.
6099 51. Small patch to Sieve code - explicitly set From: when generating an
6102 52. Exim crashed if a remote delivery caused a very long error message to be
6103 recorded - for instance if somebody sent an entire SpamAssassin report back
6104 as a large number of 550 error lines. This bug was coincidentally fixed by
6105 increasing the size of one of Exim's internal buffers (big_buffer) that
6106 happened as part of the Exiscan merge. However, to be on the safe side, I
6107 have made the code more robust (and fixed the comments that describe what
6110 53. Now that there can be additional text after "Completed" in log lines (if
6111 the queue_time_overall log selector is set), a one-byte patch to exigrep
6112 was needed to allow it to recognize "Completed" as not the last thing in
6115 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6116 patch that reportedly fixes this has been added. I am not expert enough to
6117 create a test for it. This is what the patch creator wrote:
6119 "I found a little strange behaviour of ldap code when working with
6120 Windows 2003 AD Domain, where users was placed in more than one
6121 Organization Units. When I tried to give exim partial DN, the exit code
6122 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6123 But simultaneously result of request was absolutely normal ldap result,
6124 so I produce this patch..."
6126 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6127 REFERENCE, so I have modified the code to exclude the patch when that macro
6130 55. Some experimental protocols are using DNS PTR records for new purposes. The
6131 keys for these records are domain names, not reversed IP addresses. The
6132 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6133 leaves it alone. Component reversal etc. now happens only for IP addresses.
6136 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6138 57. Double the size of the debug message buffer (to 2048) so that more of very
6139 long debug lines gets shown.
6141 58. The exicyclog utility now does better if the number of log files to keep
6142 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6144 59. Two changes related to the smtp_active_hostname option:
6146 (1) $smtp_active_hostname is now available as a variable.
6147 (2) The default for smtp_banner uses $smtp_active_hostname instead
6148 of $primary_hostname.
6150 60. The host_aton() function is supposed to be passed a string that is known
6151 to be a valid IP address. However, in the case of IPv6 addresses, it was
6152 not checking this. This is a hostage to fortune. Exim now panics and dies
6153 if the condition is not met. A case was found where this could be provoked
6154 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6155 components; fortuitously, this particular loophole had already been fixed
6156 by change 4.50/55 above.
6158 If there are any other similar loopholes, the new check in host_aton()
6159 itself should stop them being exploited. The report I received stated that
6160 data on the command line could provoke the exploit when Exim was running as
6161 exim, but did not say which command line option was involved. All I could
6162 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6163 running as the user.
6166 61. There was a buffer overflow vulnerability in the SPA authentication code
6167 (which came originally from the Samba project). I have added a test to the
6168 spa_base64_to_bits() function which I hope fixes it.
6171 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6172 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6174 63. The daemon start-up calls getloadavg() while still root for those OS that
6175 need the first call to be done as root, but it missed one case: when
6176 deliver_queue_load_max is set with deliver_drop_privilege. This is
6177 necessary for the benefit of the queue runner, because there is no re-exec
6178 when deliver_drop_privilege is set.
6180 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6181 This has been fixed.
6183 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6184 was in use, was not putting the data itself into the right store pool;
6185 consequently, it could be overwritten for a subsequent message in the same
6186 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6189 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6191 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6192 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6193 the function were treating the return as a boolean value, which happened to
6194 work because 0=false and not-0=true, but is not correct code.
6196 68. The host_aton() function was not handling scoped IPv6 addresses (those
6197 with, for example, "%eth0" on the end) correctly.
6199 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6200 negated items (that is, ~something) in unsigned ints. Some compilers
6201 apparently mutter when there is no cast.
6203 70. If an address verification called from an ACL failed, and did not produce a
6204 user-specific message (i.e. there was only a "system" message), nothing was
6205 put in $acl_verify_message. In this situation, it now puts the system
6208 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6209 session; change 4.31/43 added the unwanted input to the log line - except
6210 that it did not do this in the start of session case. It now does.
6212 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6213 This is wrong and can cause the other end to generate a synchronization
6214 error if it is another Exim or anything else that does the synchronization
6215 check. A QUIT command is no longer sent after a timeout.
6217 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6218 during host lookups.
6220 74. The defer_ok option of callout verification was not working if it was used
6221 when verifying addresses in header lines, that is, for this case:
6223 verify = header_sender/callout=defer_ok
6225 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6226 those file descriptors could be used for SMTP connections. If anything
6227 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6228 could be sent to the SMTP client, causing chaos. The daemon now opens
6229 stdin, stdout, and stderr to /dev/null when it puts itself into the
6232 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6233 log by default. The user can override this with suitable Perl magic.
6235 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6236 the log message when discard triggers, was not working for the DATA ACL or
6237 for the non-SMTP ACL.
6239 78. Error message wording change in sieve.c.
6241 79. If smtp_accept_max_per_host was set, the number of connections could be
6242 restricted to fewer than expected, because the daemon was trying to set up
6243 a new connection before checking whether the processes handling previous
6244 connections had finished. The check for completed processes is now done
6245 earlier. On busy systems, this bug wouldn't be noticed because something
6246 else would have woken the daemon, and it would have reaped the completed
6249 80. If a message was submitted locally by a user whose login name contained one
6250 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6251 It caused a spool format error. I have fixed the spool reading code. A
6252 related problem was that the "from" clause in the Received: line became
6253 illegal because of the space(s). It is now covered by ${quote_local_part.
6255 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6258 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6260 83. Patch from Sieve maintainer to fix unterminated string problem in
6261 "vacation" handling.
6263 84. Some minor changes to the Linux configuration files to help with other
6264 OS variants using glibc.
6266 85. One more patch for Sieve to update vacation handling to latest spec.
6269 ----------------------------------------------------
6270 See the note above about the 4.44 and 4.50 releases.
6271 ----------------------------------------------------
6277 1. Change 4.43/35 introduced a bug that caused file counts to be
6278 incorrectly computed when quota_filecount was set in an appendfile
6281 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6282 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6285 3. Give more explanation in the error message when the command for a transport
6286 filter fails to execute.
6288 4. There are several places where Exim runs a non-Exim command in a
6289 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6290 was being done only for the command run by the queryprogram router. It is
6291 now done for all such subprocesses. The other cases are: ${run, transport
6292 filters, and the commands run by the lmtp and pipe transports.
6294 5. Some older OS have a limit of 256 on the maximum number of file
6295 descriptors. Exim was using setrlimit() to set 1000 as a large value
6296 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6297 systems. I've change it so that if it can't get 1000, it tries for 256.
6299 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6300 was an oversight, and furthermore, ever since the addition of extra
6301 controls (e.g. 4.43/32), the checks on when to allow different forms of
6302 "control" were broken. There should now be diagnostics for all cases when a
6303 control that does not make sense is encountered.
6305 7. $recipients is now available in the predata ACL (oversight).
6307 8. Tidy the search cache before the fork to do a delivery from a message
6308 received from the command line. Otherwise the child will trigger a lookup
6309 failure and thereby defer the delivery if it tries to use (for example) a
6310 cached ldap connection that the parent has called unbind on.
6312 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6313 of $address_data from the recipient verification was clobbered by the
6314 sender verification.
6316 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6317 was its contents. (It was OK if the option was not defined at all.)
6319 11. A "Completed" log line is now written for messages that are removed from
6320 the spool by the -Mrm option.
6322 12. $host_address is now set to the target address during the checking of
6323 ignore_target_hosts.
6325 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6326 being passed; this would have caused $sender_host_name to have been used if
6327 matching the list had actually called for a host name (not very likely,
6328 since this list is usually IP addresses). A host name is now passed as
6331 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6332 code that specifies a non-restarting handler (typically sigaction() in
6333 modern systems) in an attempt to fix a rare and obscure crash bug.
6335 15. Narrowed the window for a race in the daemon that could cause it to ignore
6336 SIGCHLD signals. This is not a major problem, because they are used only to
6337 wake it up if nothing else does.
6339 16. A malformed maildirsize file could cause Exim to calculate negative values
6340 for the mailbox size or file count. Odd effects could occur as a result.
6341 The maildirsize information is now recalculated if the size or filecount
6344 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6345 support for a long time. Removed HAVE_SYS_VFS_H.
6347 18. Updated exipick to current release from John Jetmore.
6349 19. Allow an empty sender to be matched against a lookup in an address list.
6350 Previously the only cases considered were a regular expression, or an
6353 20. Exim went into a mad DNS lookup loop when doing a callout where the
6354 host was specified on the transport, if the DNS lookup yielded more than
6357 21. The RFC2047 encoding function was originally intended for short strings
6358 such as real names; it was not keeping to the 75-character limit for
6359 encoded words that the RFC imposes. It now respects the limit, and
6360 generates multiple encoded words if necessary. To be on the safe side, I
6361 have increased the buffer size for the ${rfc2047: expansion operator from
6364 22. Failure to deliver a bounce message always caused it to be frozen, even if
6365 there was an errors_to setting on the router. The errors_to setting is now
6368 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6369 canonical form (fully expanded) before being placed in
6370 $sender_host_address.
6372 24. Updated eximstats to version 1.33
6374 25. Include certificate and key file names in error message when GnuTLS fails
6375 to set them up, because the GnuTLS error message doesn't include the name
6376 of the failing file when there is a problem reading it.
6378 26. Expand error message when OpenSSL has problems setting up cert/key files.
6381 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6382 (this can affect the format of dates).
6384 28. exim_tidydb, when checking for the continued existence of a message for
6385 which it has found a message-specific retry record, was not finding
6386 messages that were in split spool directories. Consequently, it was
6387 deleting retry records that should have stayed in existence.
6389 29. eximstats updated to version 1.35
6390 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6391 1.35 - bugfix such that pie charts by volume are generated correctly
6393 30. The SPA authentication driver was not abandoning authentication and moving
6394 on to the next authenticator when an expansion was forced to fail,
6395 contradicting the general specification for all authenticators. Instead it
6396 was generating a temporary error. It now behaves as specified.
6398 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6399 (the order specifies the preference for clients). The order is now AES256,
6400 AES128, 3DES, ARCFOUR128.
6402 31. Small patch to Sieve code - explicitly set From: when generating an
6405 32. Exim crashed if a remote delivery caused a very long error message to be
6406 recorded - for instance if somebody sent an entire SpamAssassin report back
6407 as a large number of 550 error lines. This bug was coincidentally fixed by
6408 increasing the size of one of Exim's internal buffers (big_buffer) that
6409 happened as part of the Exiscan merge. However, to be on the safe side, I
6410 have made the code more robust (and fixed the comments that describe what
6413 33. Some experimental protocols are using DNS PTR records for new purposes. The
6414 keys for these records are domain names, not reversed IP addresses. The
6415 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6416 leaves it alone. Component reversal etc. now happens only for IP addresses.
6419 34. The host_aton() function is supposed to be passed a string that is known
6420 to be a valid IP address. However, in the case of IPv6 addresses, it was
6421 not checking this. This is a hostage to fortune. Exim now panics and dies
6422 if the condition is not met. A case was found where this could be provoked
6423 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6424 components; fortuitously, this particular loophole had already been fixed
6425 by change 4.50/55 or 4.44/33 above.
6427 If there are any other similar loopholes, the new check in host_aton()
6428 itself should stop them being exploited. The report I received stated that
6429 data on the command line could provoke the exploit when Exim was running as
6430 exim, but did not say which command line option was involved. All I could
6431 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6432 running as the user.
6435 35. There was a buffer overflow vulnerability in the SPA authentication code
6436 (which came originally from the Samba project). I have added a test to the
6437 spa_base64_to_bits() function which I hope fixes it.
6440 36. The daemon start-up calls getloadavg() while still root for those OS that
6441 need the first call to be done as root, but it missed one case: when
6442 deliver_queue_load_max is set with deliver_drop_privilege. This is
6443 necessary for the benefit of the queue runner, because there is no re-exec
6444 when deliver_drop_privilege is set.
6446 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6447 was in use, was not putting the data itself into the right store pool;
6448 consequently, it could be overwritten for a subsequent message in the same
6449 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6452 38. Sometimes the final signoff response after QUIT could fail to get
6453 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6454 < 0 before doing a fflush(). This bug looks as though it goes back to the
6455 introduction of TLS in release 3.20, but "sometimes" must have been rare
6456 because the tests only now provoked it.
6462 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6463 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6464 Now it returns an integer. A number of calls were still expecting a T/F
6465 return. Fortuitously, in all cases, the tests worked in OK situations,
6466 which is the norm. However, things would have gone wrong on any write
6467 failures on the smtp file descriptor. This function is used when sending
6468 messages over SMTP and also when doing verify callouts.
6470 2. When Exim is called to do synchronous delivery of a locally submitted
6471 message (the -odf or -odi options), it no longer closes stderr before doing
6474 3. Implemented the mua_wrapper option.
6476 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6478 5. Implemented the functions header_remove(), header_testname(),
6479 header_add_at_position(), and receive_remove_recipient(), and exported them
6482 6. If an ACL "warn" statement specified the addition of headers, Exim already
6483 inserted X-ACL-Warn: at the start if there was no header name. However, it
6484 was not making this test for the second and subsequent header lines if
6485 there were newlines in the string. This meant that an invalid header could
6486 be inserted if Exim was badly configured.
6488 7. Allow an ACL "warn" statement to add header lines at the start or after all
6489 the Received: headers, as well as at the end.
6491 8. Added the rcpt_4xx retry error code.
6493 9. Added postmaster_mailfrom=xxx to callout verification option.
6495 10. Added mailfrom=xxxx to the callout verification option, for verify=
6498 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6499 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6501 12. Inserted some casts to stop certain compilers warning when using pointer
6502 differences as field lengths or precisions in printf-type calls (mostly
6503 affecting debugging statements).
6505 13. Added optional readline() support for -be (dynamically loaded).
6507 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6508 same clock tick as a message's arrival, so that its received time was the
6509 same as the "first fail" time on the retry record, and that message
6510 remained on the queue past the ultimate address timeout, every queue runner
6511 would try a delivery (because it was past the ultimate address timeout) but
6512 after another failure, the ultimate address timeout, which should have then
6513 bounced the address, did not kick in. This was a "< instead of <=" error;
6514 in most cases the first failure would have been in the next clock tick
6515 after the received time, and all would be well.
6517 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6518 being recognized when the domain list was tested by the match_domain
6519 condition in an expansion string.
6521 16. Added the ${str2b64: operator.
6523 17. Exim was always calling setrlimit() to set a large limit for the number of
6524 processes, without checking whether the existing limit was already
6525 adequate. (It did check for the limit on file descriptors.) Furthermore,
6526 errors from getrlimit() and setrlimit() were being ignored. Now they are
6527 logged to the main and panic logs, but Exim does carry on, to try to do its
6528 job under whatever limits there are.
6530 18. Imported PCRE 5.0.
6532 19. Trivial typo in log message " temporarily refused connection" (the leading
6535 20. If the log selector return_path_on_delivery was set and an address was
6536 redirected to /dev/null, the delivery process crashed because it assumed
6537 that a return path would always be set for a "successful" delivery. In this
6538 case, the whole delivery is bypassed as an optimization, and therefore no
6541 21. Internal re-arrangement: the function for sending a challenge and reading
6542 a response while authentication was assuming a zero-terminated challenge
6543 string. It's now changed to take a pointer and a length, to allow for
6544 binary data in such strings.
6546 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6548 23. Exim was not respecting finduser_retries when seeking the login of the
6549 uid under which it was called; it was always trying 10 times. (The default
6550 setting of finduser_retries is zero.) Also, it was sleeping after the final
6551 failure, which is pointless.
6553 24. Implemented tls_on_connect_ports.
6555 25. Implemented acl_smtp_predata.
6557 26. If the domain in control=submission is set empty, Exim assumes that the
6558 authenticated id is a complete email address when it generates From: or
6559 Sender: header lines.
6561 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6562 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6563 chown and chgrp in /bin and hostname in /usr/bin.
6565 28. Exim was keeping the "process log" file open after each use, just as it
6566 does for the main log. This opens the possibility of it remaining open for
6567 long periods when the USR1 signal hits a daemon. Occasional processlog
6568 errors were reported, that could have been caused by this. Anyway, it seems
6569 much more sensible not to leave this file open at all, so that is what now
6572 29. The long-running daemon process does not normally write to the log once it
6573 has entered its main loop, and it closes the log before doing so. This is
6574 so that log files can straightforwardly be renamed and moved. However,
6575 there are a couple of unusual error situations where the daemon does write
6576 log entries, and I had neglected to close the log afterwards.
6578 30. The text of an SMTP error response that was received during a remote
6579 delivery was being truncated at 512 bytes. This is too short for some of
6580 the long messages that one sometimes sees. I've increased the limit to
6583 31. It is now possible to make retry rules that apply only when a message has a
6584 specific sender, in particular, an empty sender.
6586 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6587 it possible to be selective about when SMTP synchronization is enforced.
6589 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6591 32. Implemented hosts_connection_nolog.
6593 33. Added an ACL for QUIT.
6595 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6598 35. Added mailbox_size and mailbox_filecount to appendfile.
6600 36. Added control = no_multiline_responses to ACLs.
6602 37. There was a bug in the logic of the code that waits for the clock to tick
6603 in the case where the clock went backwards by a substantial amount such
6604 that the microsecond fraction of "now" was more than the microsecond
6605 fraction of "then" (but the whole seconds number was less).
6607 38. Added support for the libradius Radius client library this is found on
6608 FreeBSD (previously only the radiusclient library was supported).
6614 1. When certain lookups returned multiple values in the form name=value, the
6615 quoting of the values was not always being done properly. Specifically:
6616 (a) If the value started with a double quote, but contained no whitespace,
6618 (b) If the value contained whitespace other than a space character (i.e.
6619 tabs or newlines or carriage returns) it was not quoted.
6620 This fix has been applied to the mysql and pgsql lookups by writing a
6621 separate quoting function and calling it from the lookup code. The fix
6622 should probably also be applied to nisplus, ibase and oracle lookups, but
6623 since I cannot test any of those, I have not disturbed their existing code.
6625 2. A hit in the callout cache for a specific address caused a log line with no
6626 reason for rejecting RCPT. Now it says "Previous (cached) callout
6627 verification failure".
6629 3. There was an off-by-one bug in the queryprogram router. An over-long
6630 return line was truncated at 256 instead of 255 characters, thereby
6631 overflowing its buffer with the terminating zero. As well as fixing this, I
6632 have increased the buffer size to 1024 (and made a note to document this).
6634 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6635 when Exim is waiting for an SMTP response from a remote server, Exim
6636 restarts its select() call on the socket, thereby resetting its timeout.
6637 This is not a problem when such interrupts are rare. Somebody set up a cron
6638 job to run exiwhat every 2 minutes, which is less than the normal select()
6639 timeout (5 or 10 minutes). This meant that the select() timeout never
6640 kicked in because it was always reset. I have fixed this by comparing the
6641 time when an interrupt arrives with the time at the start of the first call
6642 to select(). If more time than the timeout has elapsed, the interrupt is
6643 treated as a timeout.
6645 5. Some internal re-factoring in preparation for the addition of Sieve
6646 extensions (by MH). In particular, the "personal" test is moved to a
6647 separate function, and given an option for scanning Cc: and Bcc: (which is
6648 not set for Exim filters).
6650 6. When Exim created an email address using the login of the caller as the
6651 local part (e.g. when creating a From: or Sender: header line), it was not
6652 quoting the local part when it contained special characters such as @.
6654 7. Installed new OpenBSD configuration files.
6656 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6657 try to make them clearer.
6659 9. Callout options, other than the timeout value, were being ignored when
6660 verifying sender addresses in header lines. For example, when using
6662 verify = header_sender/callout=no_cache
6664 the cache was (incorrectly) being used.
6666 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6667 only the Cygwin environment.
6669 11. When return_path_on_delivery was set as a log selector, if different remote
6670 addresses in the same message used different return paths and parallel
6671 remote delivery occurred, the wrong values would sometimes be logged.
6672 (Whenever a remote delivery process finished, the return path value from
6673 the most recently started remote delivery process was logged.)
6675 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6676 lines when AUTH and/or TLS are in use. This is the "received protocol"
6677 field. Exim used to use "asmtp" for authenticated SMTP, without any
6678 indication (in the protocol name) for TLS use. Now it follows the RFC and
6679 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6680 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6681 names appear in log lines as well as in Received: header lines.
6683 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6684 extensions, and comparison tests, and to fix some bugs.
6686 14. Changes to the "personal" filter test:
6688 (1) The test was buggy in that it was just doing the equivalent of
6689 "contains" tests on header lines. For example, if a user's address was
6690 anne@some.where, the "personal" test would incorrectly be true for
6692 To: susanne@some.where
6694 This test is now done by extracting each address from the header in turn,
6695 and checking the entire address. Other tests that are part of "personal"
6696 are now done using regular expressions (for example, to check local parts
6697 of addresses in From: header lines).
6699 (2) The list of non-personal local parts in From: addresses has been
6700 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6701 taken from the Sieve specification recommendations.
6703 (3) If the message contains any header line starting with "List-" it is
6704 treated as non-personal.
6706 (4) The test for "circular" in the Subject: header line has been removed
6707 because it now seems ill-conceived.
6709 15. Minor typos in src/EDITME comments corrected.
6711 16. Installed latest exipick from John Jetmore.
6713 17. If headers_add on a router specified a text string that was too long for
6714 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6715 of string_sprintf() is now avoided.
6717 18. $message_body_size was not set (it was always zero) when running the DATA
6718 ACL and the local_scan() function.
6720 19. For the "mail" command in an Exim filter, no default was being set for
6721 the once_repeat time, causing a random time value to be used if "once" was
6722 specified. (If the value happened to be <= 0, no repeat happened.) The
6723 default is now 0s, meaning "never repeat". The "vacation" command was OK
6724 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6725 (I found it when inspecting the code).
6727 20. There is now an overall timeout for performing a callout verification. It
6728 defaults to 4 times the callout timeout, which applies to individual SMTP
6729 commands during the callout. The overall timeout applies when there is more
6730 than one host that can be tried. The timeout is checked before trying the
6731 next host. This prevents very long delays if there are a large number of
6732 hosts and all are timing out (e.g. when the network connections are timing
6733 out). The value of the overall timeout can be changed by specifying an
6734 additional sub-option for "callout", called "maxwait". For example:
6736 verify = sender/callout=5s,maxwait=20s
6738 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6739 to the end before writing, but this should make it even safer).
6741 22. Exim was forgetting that it had advertised PIPELINING for the second and
6742 subsequent messages on an SMTP connection. It was also not resetting its
6743 memory on STARTTLS and an internal HELO.
6745 23. When Exim logs an SMTP synchronization error within a session, it now
6746 records whether PIPELINING has been advertised or not.
6748 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6749 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6750 rather than long int.
6752 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6754 26. Added the never_mail option to autoreply.
6760 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6761 crash if the getsockname() call failed; this can happen if a connection is
6762 closed very soon after it is established. The problem was simply in the
6763 order in which certain operations were done, causing Exim to try to write
6764 to the SMTP stream before it had set up the file descriptor. The bug has
6765 been fixed by making things happen in the correct order.
6771 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6772 before the connection was closed, thus losing the rejection response.
6774 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6775 some early Solaris releases, but causes trouble in current releases where
6776 socklen_t is defined.
6778 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6781 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6784 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6785 that releasing the top of it at the end releases what was used for sub-
6786 expansions (unless the block got too big). However, discard this block if
6787 the first thing is a variable or header, so that we can use its block when
6788 it is dynamic (useful for very large $message_headers, for example).
6790 6. Lookups now cache *every* query, not just the most recent. A new, separate
6791 store pool is used for this. It can be recovered when all lookup caches are
6792 flushed. Lookups now release memory at the end of their result strings.
6793 This has involved some general refactoring of the lookup sources.
6795 7. Some code has been added to the store_xxx() functions to reduce the amount
6796 of flapping under certain conditions.
6798 8. log_incoming_interface used to affect only the <= reception log lines. Now
6799 it causes the local interface and port to be added to several more SMTP log
6800 lines, for example "SMTP connection from", and rejection lines.
6802 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6804 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6806 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6807 could be overwritten at the end of the current message (or the start of a
6808 new message if it was set in a HELO ACL). The value is now preserved for
6809 the duration of the SMTP connection.
6811 12. If a transport had a headers_rewrite setting, and a matching header line
6812 contained an unqualified address, that address was qualified, even if it
6813 did not match any rewriting rules. The underlying bug was that the values
6814 of the flags that permit the existence of unqualified sender and recipient
6815 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6816 non-local messages, and by -bnq for local messages) were not being
6817 preserved with the message after it was received.
6819 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6820 "next input=" as part of the text comprising the host identity instead of
6821 the correct text. The code was using the same buffer for two different
6822 strings. However, depending on which order the printing function evaluated
6823 its arguments, the bug did not always show up. Under Linux, for example, my
6824 test suite worked just fine.
6826 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6827 doesn't work with some older versions of Perl. It has been changed to "my",
6828 which in any case is probably the better facility to use.
6830 15. A really picky compiler found some instances of statements for creating
6831 error messages that either had too many or two few arguments for the format
6834 16. The size of the buffer for calls to the DNS resolver has been increased
6835 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6836 for addresses that have a lot of PTR records. This alleviates a problem; it
6837 does not fully solve it.
6839 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6840 buffer now truncates the list and logs the incident, which is the same
6841 action as happens when Exim is looking up a host name and its aliases.
6842 Previously in this situation something unpredictable would happen;
6843 sometimes it was "internal error: store_reset failed".
6845 18. If a server dropped the connection unexpectedly when an Exim client was
6846 using GnuTLS and trying to read a response, the client delivery process
6847 crashed while trying to generate an error log message.
6849 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6850 string, for example:
6852 warn message = H1: something\nH2: something
6854 the text was added as a single header line from Exim's point of view
6855 though it ended up OK in the delivered message. However, searching for the
6856 second and subsequent header lines using $h_h2: did not work. This has been
6857 fixed. Similarly, if a system filter added multiple headers in this way,
6858 the routers could not see them.
6860 20. Expanded the error message when iplsearch is called with an invalid key to
6861 suggest using net-iplsearch in a host list.
6863 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6864 ACLs are no longer actually imposed (and a message to that effect is
6867 22. If a "gecos" field in a passwd entry contained escaped characters, in
6868 particular, if it contained a \" sequence, Exim got it wrong when building
6869 a From: or a Sender: header from that name. A second bug also caused
6870 incorrect handling when an unquoted " was present following a character
6871 that needed quoting.
6873 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6874 was not being matched caselessly.
6876 24. Arranged for all hyphens in the exim.8 source to be escaped with
6879 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6880 properly. Recipient callout cache records were still being keyed to include
6881 the sender, even when use_sender was set false. This led to far more
6882 callouts that were necessary. The sender is no longer included in the key
6883 when use_sender is false.
6885 26. Added "control = submission" modifier to ACLs.
6887 27. Added the ${base62d: operator to decode base 62 numbers.
6889 28. dnsdb lookups can now access SRV records.
6891 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6892 the configuration file.
6894 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6895 (-v) mode. This makes the output for a verbose queue run more intelligible.
6897 31. Added a use_postmaster feature to recipient callouts.
6899 32. Added the $body_zerocount variable, containing the number of binary zero
6900 bytes in the message body.
6902 33. The time of last modification of the "new" subdirectory is now used as the
6903 "mailbox time last read" when there is a quota error for a maildir
6906 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6908 35. Added +ignore_unknown as a special item in host lists.
6910 36. Code for decoding IPv6 addresses in host lists is now included, even if
6911 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6912 address was recognized as an IP address, but was then not correctly decoded
6913 into binary, causing unexpected and incorrect effects when compared with
6920 1. Very minor rewording of debugging text in manualroute to say "list of
6921 hosts" instead of "hostlist".
6923 2. If verify=header_syntax was set, and a header line with an unqualified
6924 address (no domain) and a large number of spaces between the end of the
6925 name and the colon was received, the reception process suffered a buffer
6926 overflow, and (when I tested it) crashed. This was caused by some obsolete
6927 code that should have been removed. The fix is to remove it!
6929 3. When running in the test harness, delay a bit after writing a bounce
6930 message to get a bit more predictability in the log output.
6932 4. Added a call to search_tidyup() just before forking a reception process. In
6933 theory, someone could use a lookup in the expansion of smtp_accept_max_
6934 per_host which, without the tidyup, could leave open a database connection.
6936 5. Added the variables $recipient_data and $sender_data which get set from a
6937 lookup success in an ACL "recipients" or "senders" condition, or a router
6938 "senders" option, similar to $domain_data and $local_part_data.
6940 6. Moved the writing of debug_print from before to after the "senders" test
6943 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6944 problems for message scanning, either using a data ACL, or using
6945 local_scan() because the Received: header was not generated till after they
6946 were called (in order to set the time as the time of reception completion).
6947 I have revised the way this works. The header is now generated after the
6948 body is received, but before the ACL or local_scan() are called. After they
6949 are run, the timestamp in the header is updated.
6955 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6956 before starting a queue runner without re-exec. This happened only when
6957 deliver_drop_privilege was set or when the Exim user was set to root. The
6958 effect of the bug was that timeouts during subsequent deliveries caused
6959 crashes instead of being properly handled. The handler is now left at its
6960 default (and expected) setting.
6962 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6963 message, again when deliver_drop_privilege is set or Exim is run as root.
6964 The bug described in (1) was not present in this case, but the tidying up
6965 of the other signals was missing. I have made the two cases consistent.
6967 3. The ignore_target_hosts setting on a manualroute router was being ignored
6968 for hosts that were looked up using the /MX notation.
6970 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6973 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6974 operated on the sender address. After changing the $sender_address to <>
6975 for the sender address verify, Exim was re-instated it as the original
6976 (before rewriting) address, but remembering that it had rewritten it, so it
6977 wasn't rewriting it again. This bug also had the effect of breaking the
6978 sender address verification caching when the sender address was rewritten.
6980 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6981 This has been changed so that if the ip literal address matches
6982 ignore_target_hosts, the router declines.
6984 7. Added expansion conditions match_domain, match_address, and match_local_
6985 part (NOT match_host).
6987 8. The placeholder for the Received: header didn't have a length field set.
6989 9. Added code to Exim itself and to exim_lock to test for a specific race
6990 condition that could lead to file corruption when using MBX delivery. The
6991 issue is with the lockfile that is created in /tmp. If this file is removed
6992 after a process has opened it but before that process has acquired a lock,
6993 there is the potential for a second process to recreate the file and also
6994 acquire a lock. This could lead to two Exim processes writing to the file
6995 at the same time. The added code performs the same test as UW imapd; it
6996 checks after acquiring the lock that its file descriptor still refers to
6997 the same named file.
6999 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7000 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7003 11. Added the smtp_active_hostname option. If used, this will typically be made
7004 to depend on the incoming interface address. Because $interface_address is
7005 not set up until the daemon has forked a reception process, error responses
7006 that can happen earlier (such as "too many connections") no longer contain
7009 12. If an expansion in a condition on a "warn" statement fails because a lookup
7010 defers, the "warn" statement is abandoned, and the next ACL statement is
7011 processed. Previously this caused the whole ACL to be aborted.
7013 13. Added the iplsearch lookup type.
7015 14. Added ident_timeout as a log selector.
7017 15. Added tls_certificate_verified as a log selector.
7019 16. Added a global option tls_require_ciphers (compare the smtp transport
7020 option of the same name). This controls incoming TLS connections.
7022 17. I finally figured out how to make tls_require_ciphers do a similar thing
7023 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7024 before starting the TLS session.
7026 18. Tabs are now shown as \t in -bP output.
7028 19. If the log selector return_path_on_delivery was set, Exim crashed when
7029 bouncing a message because it had too many Received: header lines.
7031 20. If two routers both had headers_remove settings, and the first one included
7032 a superfluous trailing colon, the final name in the first list and the
7033 first name in the second list were incorrectly joined into one item (with a
7034 colon in the middle).
7040 1. Added -C and -D options to the exinext utility, mainly to make it easier
7041 to include in the automated testing, but these could be helpful when
7042 multiple configurations are in use.
7044 2. The exinext utility was not formatting the output nicely when there was
7045 an alternate port involved in the retry record key, nor when there was a
7046 message id as well (for retries that were specific to a specific message
7047 and a specific host). It was also confused by IPv6 addresses, because of
7048 the additional colons they contain. I have fixed the IPv4 problem, and
7049 patched it up to do a reasonable job for IPv6.
7051 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7052 delivery, the log line now contains "pipelined" if PIPELINING was used.
7054 4. An SMTP transport process used to panic and die if the bind() call to set
7055 an explicit outgoing interface failed. This has been changed; it is now
7056 treated in the same way as a connect() failure.
7058 5. A reference to $sender_host_name in the part of a conditional expansion
7059 that was being skipped was still causing a DNS lookup. This no longer
7062 6. The def: expansion condition was not recognizing references to header lines
7063 that used bh_ and bheader_.
7065 7. Added the _cache feature to named lists.
7067 8. The code for checking quota_filecount in the appendfile transport was
7068 allowing one more file than it should have been.
7070 9. For compatibility with Sendmail, the command line option
7078 and sets the incoming protocol and host name (for trusted callers). The
7079 host name and its colon can be omitted when only the protocol is to be set.
7080 Note the Exim already has two private options, -pd and -ps, that refer to
7081 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7082 "s", but I don't think that's a major issue.
7084 10. A number of refactoring changes to the code, none of which should affect
7087 (a) The number of logging options was getting close to filling up the
7088 32-bit word that was used as a bit map. I have split them into two classes:
7089 those that are passed in the argument to log_write(), and those that are
7090 only ever tested independently outside of that function. These are now in
7091 separate 32-bit words, so there is plenty of room for expansion again.
7092 There is no change in the user interface or the logging behaviour.
7094 (b) When building, for example, log lines, the code previously used a
7095 macro that called string_cat() twice, in order to add two strings. This is
7096 not really sufficiently general. Furthermore, there was one instance where
7097 it was actually wrong because one of the argument was used twice, and in
7098 one call a function was used. (As it happened, calling the function twice
7099 did not affect the overall behaviour.) The macro has been replaced by a
7100 function that can join an arbitrary number of extra strings onto a growing
7103 (c) The code for expansion conditions now uses a table and a binary chop
7104 instead of a serial search (which was left over from when there were very
7105 few conditions). Also, it now recognizes conditions like "pam" even when
7106 the relevant support is not compiled in: a suitably worded error message is
7107 given if an attempt is made to use such a condition.
7109 11. Added ${time_interval:xxxxx}.
7111 12. A bug was causing one of the ddress fields not to be passed back correctly
7112 from remote delivery subprocesses. The field in question was not being
7113 subsequently used, so this caused to problems in practice.
7115 13. Added new log selectors queue_time and deliver_time.
7117 14. Might have fixed a bug in maildirsizefile handling that threw up
7118 "unexpected character" debug warnings, and recalculated the data
7119 unnecessarily. In any case, I expanded the warning message to give more
7122 15. Added the message "Restricted characters in address" to the statements in
7123 the default ACL that block characters like @ and % in local parts.
7125 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7126 Three changes have been made:
7128 (a) There was a serious bug; a negative response to MAIL caused the whole
7129 recipient domain to be cached as invalid, thereby blocking all messages
7130 to all local parts at the same domain, from all senders. This bug has
7131 been fixed. The domain is no longer cached after a negative response to
7132 MAIL if the sender used is not empty.
7134 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7137 (c) A new callout option, "use_sender" has been added for people who want
7138 the modified behaviour.
7144 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7147 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7148 indeed breaks things for older releases.
7150 3. Added additional logging to the case where there is a problem reading data
7151 from a filter that is running in a subprocess using a pipe, in order to
7152 try to track down a specific problem.
7154 4. Testing facility fudge: when running in the test harness and attempting
7155 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7156 getting "No route to host". Convert this to a timeout.
7158 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7161 6. Some OS don't have socklen_t but use size_t instead. This affects the
7162 fifth argument of getsockopt() amongst other things. This is now
7163 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7164 can be set for individual OS. I have set it for SunOS5, OSF1, and
7165 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7166 some earlier ones do not.
7168 7. Change 4.30/15 was not doing the test caselessly.
7170 8. The standard form for an IPv6 address literal was being rejected by address
7171 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7172 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7173 this, as well as the form without the "IPv6" on the front (but only when
7174 address literals are enabled, of course).
7176 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7178 10. Exim crashed if a message with an empty sender address specified by -f
7179 encountered a router with an errors_to setting. This could be provoked only
7180 by a command such as
7184 where an empty string was supplied; "<>" did not hit this bug.
7186 11. Installed PCRE release 4.5.
7188 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7189 remained set. It is now erased.
7191 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7192 times from message ids (which are base 36 rather than the normal 62).
7194 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7195 were being counted as actual protocol errors, and logged if the log
7196 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7197 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7198 and DATA following a set of rejected RCPTs do not count as protocol errors.
7199 In other words, Exim assumes they were pipelined, though this may not
7200 actually be the case. Of course, in all cases the client gets an
7201 appropriate error code.
7203 15. If a lookup fails in an ACL condition, a message about the failure may
7204 be available; it is used if testing the ACL cannot continue, because most
7205 such messages specify what the cause of the deferral is. However, some
7206 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7207 that caused an old message to be retained and used if a later statement
7208 caused a defer, replacing the real cause of the deferral.
7210 16. If an IP address had so many PTR records that the DNS lookup buffer
7211 was not large enough to hold them, Exim could crash while trying to process
7212 the truncated data. It now detects and logs this case.
7214 17. Further to 4.21/58, another change has been made: if (and only if) the
7215 first line of a message (the first header line) ends with CRLF, a bare LF
7216 in a subsequent header line has a space inserted after it, so as not to
7217 terminate the header.
7219 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7220 unnecessarily, used atoi() instead of strtol(), and didn't check the
7221 termination when getting file sizes from file names by regex.
7223 19. Completely re-implemented the support for maildirsize files, in the light
7224 of a number of problems with the previous contributed implementation
7225 (4.30/29). In particular:
7227 . If the quota is zero, the maildirsize file is maintained, but no quota is
7230 . If the maildir directory does not exist, it is created before any attempt
7231 to write a maildirsize file.
7233 . The quota value in the file is just a cache; if the quota is changed in
7234 the transport, the new value overrides.
7236 . A regular expression is available for excluding directories from the
7239 20. The autoreply transport checks the characters in options that define the
7240 message's headers; it allows continued headers, but it was checking with
7241 isspace() after an embedded newline instead of explicitly looking for a
7244 21. If all the "regular" hosts to which an address was routed had passed their
7245 expiry times, and had not reached their retry times, the address was
7246 bounced, even if fallback hosts were defined. Now Exim should go on to try
7249 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7250 equivalent code in the SMTP transport. Some hosts send humungous responses
7251 to HELO/EHLO, more than 1024 it seems.
7253 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7254 gives compiler warnings in some environments. I've now done it "properly",
7257 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7258 (because of problems with the built-in one) was declared to return uschar *
7259 instead of char *, causing compiler failure.
7261 25. Fixed a file descriptor leak when processing alias/forward files.
7263 26. Fixed a minor format string issue in dbfn.c.
7265 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7267 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7268 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7271 29. When a sender address is verified, it is cached, to save repeating the test
7272 when there is more than one recipient in a message. However, when the
7273 verification involves a callout, it is possible for different callout
7274 options to be set for different recipients. It is too complicated to keep
7275 track of this in the cache, so now Exim always runs a verification when a
7276 callout is required, relying on the callout cache for the optimization.
7277 The overhead is duplication of the address routing, but this should not be
7280 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7281 to be verified with callout=postmaster, and the main callout worked but the
7282 postmaster check failed, the verification correctly failed. However, if a
7283 subsequent RCPT command asked for sender verification *without* the
7284 postmaster check, incorrect caching caused this verification also to fail,
7287 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7288 it was not caching the DNS options (qualify_single, search_parents) that
7289 were used when the lookup failed. A subsequent lookup with different
7290 options therefore always gave the same answer, though there were cases
7291 where it should not have. (Example: a "domains = !$mx_any" option on a
7292 dnslookup router: the "domains" option is always processed without any
7293 widening, but the router might have qualify_single set.) Now Exim uses the
7294 cached value only when the same options are set.
7296 32. Added John Jetmore's "exipick" utility to the distribution.
7298 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7299 than a timeout (e.g. a certificate is required, and is not provided), an
7300 Exim server now closes the connection immediately. Previously it waited for
7301 the client to close - but if the client is SSL, it seems that they each
7302 wait for each other, leading to a delay before one of them times out.
7304 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7305 maintained 0.8.x compatibility because I don't think many are using it, and
7306 it is clearly obsolete.
7308 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7311 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7312 client certificate was expired. A simple patch fixes this, though I don't
7313 understand the full logic of why the verify callback is called multiple
7316 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7317 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7318 which causes problems with some clients (such as the Certicom SSL Plus
7319 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7320 disables the coutermeasure allowing Eudora to connect."
7322 38. Exim was not checking that a write() to a log file succeeded. This could
7323 lead to Bad Things if a log got too big, in particular if it hit a file
7324 size limit. Exim now panics and dies if it cannot write to a log file, just
7325 as it does if it cannot open a log file.
7327 39. Modified OS/Makefile-Linux so that it now contains
7329 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7331 The two -D definitions ensure that Exim is compiled with large file
7332 support, which makes it possible to handle log files that are bigger than
7335 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7336 instance) a domain was checked against a named list that involved a lookup,
7337 causing $domain_data to be set, then another domain was checked against the
7338 same list, then the first domain was re-checked, the value of $domain_data
7339 after the final check could be wrong. In particular, if the second check
7340 failed, it could be set empty. This bug probably also applied to
7343 41. The strip_trailing_dot option was not being applied to the address given
7344 with the -f command-line option.
7346 42. The code for reading a message's header from the spool was incrementing
7347 $received_count, but never initializing it. This meant that the value was
7348 incorrect (doubled) while delivering a message in the same process in which
7349 it was received. In the most common configuration of Exim, this never
7350 happens - a fresh exec is done - but it can happen when
7351 deliver_drop_privilege is set.
7353 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7354 it now includes up to 150 characters of the unexpected data in the log
7357 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7358 and building data strings. The size of both of these buffers was 10 000
7359 bytes - far larger than anybody would *ever* want, thought I. Needless to
7360 say, somebody hit the limit. I have increased the maximum line length to
7361 20 000 and the maximum data length of concatenated lines to 100 000. I have
7362 also fixed two bugs, because there was no checking on these buffers. Tsk,
7363 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7364 buffer is too small.
7366 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7367 lsearch lookups. Now it does.
7369 46. When parsing a route_list item in a manualroute router, a fixed-length
7370 buffer was used for the list of hosts. I made this 1024 bytes long,
7371 thinking that nobody would ever have a list of hosts that long. Wrong.
7372 Somebody had a whole pile of complicated expansion conditions, and the
7373 string was silently truncated, leading to an expansion error. It turns out
7374 that it is easier to change to an unlimited length (owing to other changes
7375 that have happened since this code was originally written) than to build
7376 structure for giving a limitation error. The length of the item that
7377 expands into the list of hosts is now unlimited.
7379 47. The lsearch lookup could not handle data where the length of text line was
7380 more than 4095 characters. Such lines were truncated, leading to shortened
7381 data being returned. It should now handle lines of any length.
7383 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7384 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7387 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7388 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7389 no longer happens in comment lines. A list of replacements is now placed
7390 at the head of all of the source files, except those whose only change is
7391 to replace PERL_COMMAND in the very first #! line.
7393 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7394 messages on the queue, with a bottom-up merge sort, using code contributed
7395 by Michael Haardt. This should make operations like -bp somewhat faster on
7396 large queues. It won't affect queue runners, except when queue_run_in_order
7399 51. Installed eximstats 1.31 in the distribution.
7401 52. Added support for SRV lookups to the dnslookup router.
7403 53. If an ACL referred to $message_body or $message_body_end, the value was not
7404 reset for any messages that followed in the same SMTP session.
7406 54. The store-handling optimization for building very long strings was not
7407 differentiating between the different store pools. I don't think this
7408 actually made any difference in practice, but I've tidied it.
7410 55. While running the routers to verify a sender address, $sender_address
7411 was still set to the sender address. This is wrong, because when routing to
7412 send a bounce to the sender, it would be empty. Therefore, I have changed
7413 it so that, while verifying a sender address, $sender_address is set to <>.
7414 (There is no change to what happens when verifying a recipient address.)
7416 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7417 target A or AAAA records (if not already returned) without resetting the
7418 qualify_single or search_parents options of the DNS resolver. These are
7419 inappropriate in this case because the targets of MX and SRV records must
7420 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7421 target that, when qualified, matched something in the local domain. These
7422 two options are now turned off when doing these lookups.
7424 57. It seems that at least some releases of Reiserfs (which does not have the
7425 concept of a fixed number of inodes) returns zero and not -1 for the
7426 number of available inodes. This interacted badly with check_spool_inodes,
7427 which assumed that -1 was the "no such thing" setting. What I have done is
7428 to check that the total number of inodes is greater than zero before doing
7429 the test of how many are available.
7431 58. When a "warn" ACL statement has a log_message modifier, the message is
7432 remembered, and not repeated. This is to avoid a lot of repetition when a
7433 message has many recipients that cause the same warning to be written.
7434 However, Exim was preserving the list of already written lines for an
7435 entire SMTP session, which doesn't seem right. The memory is now reset if a
7436 new message is started.
7438 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7439 debugging output unless log_rewrite was also set.
7441 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7442 of (int)(handle) when we know that handle contains (void *)(-1).
7444 61. The Exim daemon panic-logs an error return when it closes the incoming
7445 connection. However "connection reset by peer" seems to be common, and
7446 isn't really an error worthy of noting specially, so that particular error
7449 62. When Exim is trying to find all the local interfaces, it used to panic and
7450 die if the ioctl to get the interface flags failed. However, it seems that
7451 on at least one OS (Solaris 9) it is possible to have an interface that is
7452 included in the list of interfaces, but for which you get a failure error
7453 for this call. This happens when the interface is not "plumbed" into a
7454 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7455 failure of the "get flags" call assumes that the interface is down.
7457 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7458 makes life easier for people who are doing arithmetic on fields extracted
7459 from dates, where you often get leading zeros that should not be
7460 interpreted as octal.
7462 64. Added qualify_domain to the redirect router, to override the global
7465 65. If a pathologically long header line contained very many addresses (the
7466 report of this problem mentioned 10 000) and each of them was rewritten,
7467 Exim could use up a very large amount of memory. (It kept on making new
7468 copies of the header line as it rewrote, and never released the old ones.)
7469 At the expense of a bit more processing, the header rewriting function has
7470 been changed so that it no longer eats memory in this way.
7472 66. The generation of the Received: header has been moved from the time that a
7473 message starts to be received, to the time that it finishes. The timestamp
7474 in the Received: header should now be very close to that of the <= log
7475 line. There are two side-effects of this change:
7477 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7478 logged header lines no longer include the local Received: line, because
7479 it has not yet been created. The same applies to a copy of the message
7480 that is returned to a non-SMTP sender when a message is rejected.
7482 (b) When a filter file is tested using -bf, no additional Received: header
7483 is added to the test message. After some thought, I decided that this
7486 This change does not affect the value of $received_for. It is still set
7487 after address rewriting, but before local_scan() is called.
7489 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7491 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7492 gave an unhelpful panic error message, and a defer error. I have managed to
7493 change this behaviour so that it now rejects any supplied certificate,
7494 which seems right, as the list of acceptable certificates is empty.
7496 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7497 gave an unhelpful defer error. I have not managed to make this reject any
7498 supplied certificates, but the error message it gives is "no certificate
7499 supplied", which is not helpful.
7501 70. exigrep's output now also includes lines that are not associated with any
7502 message, but which match the given pattern. Implemented by a patch from
7503 Martin Sluka, which also tidied up the Perl a bit.
7505 71. Recipient callout verification, like sender verification, was using <> in
7506 the MAIL FROM command. This isn't really the right thing, since the actual
7507 sender may affect whether the remote host accepts the recipient or not. I
7508 have changed it to use the actual sender in the callout; this means that
7509 the cache record is now keyed on a recipient/sender pair, not just the
7510 recipient address. There doesn't seem to be a real danger of callout loops,
7511 since a callout by the remote host to check the sender would use <>.
7512 [SEE ABOVE: changed after hitting problems.]
7514 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7515 temporary errors. However, in the case of such a code being given after
7516 the end of a data transmission (i.e. after ".") Exim was failing to write
7517 a retry record for the message. (Yes, there was some broken host that was
7518 actually sending 8xx at this point.)
7520 73. An unknown lookup type in a host list could cause Exim to panic-die when
7521 the list was checked. (An example that provoked this was putting <; in the
7522 middle of a list instead of at the start.) If this happened during a DATA
7523 ACL check, a -D file could be left lying around. This kind of configuration
7524 error no longer causes Exim to die; instead it causes a defer error. The
7525 incident is still logged to the main and panic logs.
7527 74. Buglet left over from Exim 3 conversion. The message "too many messages
7528 in one connection" was written to the rejectlog but not the mainlog, except
7529 when address rewriting (yes!) was being logged.
7531 75. Added write_rejectlog option.
7533 76. When a system filter was run not as root (that is, when system_filter_user
7534 was set), the values of the $n variables were not being returned to the
7535 main process; thus, they were not subsequently available in the $sn
7538 77. Added +return_path_on_delivery log selector.
7540 78. A connection timeout was being treated differently from recipients deferred
7541 when testing hosts_max_try with a message that was older than the host's
7542 retry timeout. (The host should not be counted, thus allowing all hosts to
7543 be tried at least once before bouncing.) This may have been the cause of an
7544 occasionally reported bug whereby a message would remain on the queue
7545 longer than the retry timeout, but would be bounced if a delivery was
7546 forced. I say "may" because I never totally pinned down the problem;
7547 setting up timeout/retry tests is difficult. See also the next item.
7549 79. The ultimate address timeout was not being applied to errors that involved
7550 a combination of host plus message (for example, a timeout on a MAIL
7551 command). When an address resolved to a number of possible hosts, and they
7552 were not all tried for each delivery (e.g. because of hosts_max_try), a
7553 message could remain on the queue longer than the retry timeout.
7555 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7558 81. Fixed an obscure SMTP outgoing bug which required at least the following
7559 conditions: (a) there was another message waiting for the same server;
7560 (b) the server returned 5xx to all RCPT commands in the first message so
7561 that the message was not completed; (c) the server dropped the connection
7562 or gave a negative response to the RSET that Exim sends to abort the
7563 transaction. The observed case was a dropped connection after DATA that had
7564 been sent in pipelining mode. That is, the server had advertised PIPELINING
7565 but was not implementing it correctly. The effect of the bug was incorrect
7566 behaviour, such as trying another host, and this could lead to a crash.
7572 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7573 and daemon.c were passed as pointers to ints; they should have been
7574 pointers to socklen_t variables (which are typically unsigned ints).
7576 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7579 3. Fixed a really odd bug that affected only the testing scheme; patching a
7580 certain fixed string in the binary changed the value of another string that
7581 happened to be identical to the end of the original first string.
7583 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7584 name", it returns that address as the IP address. On some operating
7585 systems (e.g. Solaris), it also passes back the IP address string as the
7586 "host name". However, on others (e.g. Linux), it passes back an empty
7587 string. Exim wasn't checking for this, and was changing the host name to an
7588 empty string, assuming it had been canonicalized.
7590 5. Although rare, it is permitted to have more than one PTR record for a given
7591 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7592 all the names associated with an address, because they do in Solaris.
7593 However, it seems that they do not in Linux for data that comes from the
7594 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7595 I found this out when I moved to a new Linux workstation and tried to run
7596 the Exim test suite.
7598 To get round this problem I have changed the code so that it now does its
7599 own call to the DNS to look up PTR records when searching for a host name.
7600 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7601 addresses that are only in /etc/hosts are still found.
7603 This behaviour is, however, controlled by an option called host_lookup_
7604 order, which defaults to "bydns:byaddr". If people want to use the other
7605 order, or indeed, just use one or the other means of lookup, they can
7606 specify it in this variable.
7608 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7609 some operating systems, this comes back from gethostbyaddr() as an empty
7610 string, and this is what Exim used to test for. However, it seems that in
7611 other systems, "." is yielded. Exim now tests for this case too.
7613 7. The values of check_spool_space and check_log_space are now held internally
7614 as a number of kilobytes instead of an absolute number of bytes. If a
7615 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7616 kilobyte. This means that much larger values can be stored.
7618 8. Exim monitor: an attempt to get the action menu when not actually pointing
7619 at a message produces an empty menu entitled "No message selected". This
7620 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7621 no entries in it ("Shell widget menu has zero width and/or height"). So I
7622 have added a single, blank menu entry in this case.
7624 9. Added ${quote_local_part.
7626 10. MIME decoding is now applied to the contents of Subject: header lines when
7629 11. Now that a reference to $sender_host_address automatically causes a reverse
7630 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7631 host lookup before query-style lookups in lists that might use this
7632 variable. This has therefore been abolished, and the "net-" prefix is no
7633 longer necessary for query-style lookups.
7635 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7636 have been a typo for LFLAGS, so it has been changed.
7638 13. The install script calls Exim with "-C /dev/null" in order to find the
7639 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7640 to be output. However, since Exim outputs its version number before the
7641 error, it didn't break the script. It just looked ugly. I fixed this by
7642 always allowing "-C /dev/null" if the caller is root.
7644 14. Ignore overlarge ACL variable number when reading spool file - insurance
7645 against a later release with more variables having written the file.
7647 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7648 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7649 this, as well as the form without the "IPv6" on the front.
7651 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7652 OS/Makefile-Darwin file.
7654 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7655 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7657 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7658 variable. However, some OS use ints for inodes. I've added cast to long int
7659 to get rid of the compiler warning.
7661 19. I had forgotten to lock out "/../" in configuration file names when
7662 ALT_CONFIG_PREFIX was set.
7664 20. Routers used for verification do not need to specify transports. However,
7665 if such a router generated a host list, and callout was configured, Exim
7666 crashed, because it could not find a port number from the (non-existent)
7667 transport. It now assumes port 25 in this circumstance.
7669 21. Added the -t option to exigrep.
7671 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7672 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7673 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7674 them is set, LOOKUP_LSEARCH is forced.
7676 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7678 24. Added sender and host information to the "rejected by local_scan()" log
7679 line; previously there was no indication of these.
7681 25. Added .include_if_exists.
7683 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7684 Linux. It turns out that not all file systems support this. Apparently some
7685 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7686 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7687 not supported on the file descriptor, is now ignored when Exim is trying to
7688 sync a directory. This applies only to Linux.
7690 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7692 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7693 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7696 29. Added support for maildirsize files from supplied patch (modified a bit).
7698 30. The use of :fail: followed by an empty string could lead Exim to respond to
7699 sender verification failures with (e.g.):
7701 550 Verification failed for <xxx>
7702 550 Sender verify failed
7704 where the first response line was missing the '-' that indicates it is not
7705 the final line of the response.
7707 31. The loop for finding the name of the user that called Exim had a hardwired
7708 limit of 10; it now uses the value of finduser_retries, which is used for
7709 all other user lookups.
7711 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7714 33. Exim was neglecting to zero errno before one call of strtol() when
7715 expanding a string and expecting an integer value. On some systems this
7716 resulted in spurious "integer overflow" errors. Also, it was casting the
7717 result into an int without checking.
7719 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7720 did not work. The code looks as if it has *never* worked, though it appears
7721 to have been documented since at least release 1.62. I have made it work.
7723 35. The "timeout_DNS" error in retry rules, also documented since at least
7724 1.62, also never worked. As it isn't clear exactly what this means, and
7725 clearly it isn't a major issue, I have abolished the feature by treating it
7726 as "timeout", and writing a warning to the main and panic logs.
7728 36. The display of retry rules for -brt wasn't always showing the error code
7731 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7732 timeout_connect_A, timeout_connect_MX.
7734 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7735 to the empty sender.
7737 39. The daemon was not analysing the content of -oX till after it had closed
7738 stderr and disconnected from the controlling terminal. This meant that any
7739 syntax errors were only noted on the panic log, and the return code from
7740 the command was 0. By re-arranging the code a little, I've made the
7741 decoding happen first, so such errors now appear on stderr, and the return
7742 code is 1. However, the actual setting up of the sockets still happens in
7743 the disconnected process, so errors there are still only recorded on the
7746 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7747 connections (as happens on some IP stacks) was logged at start up time as
7748 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7749 it from "IPv6 and IPv4", which means that two separate sockets are being
7752 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7753 says whether AF_INET or AF_INET6 was passed as an argument.
7755 42. Exiwhat output was messed up when time zones were included in log
7758 43. Exiwhat now gives more information about the daemon's listening ports,
7759 and whether -tls-on-connect was used.
7761 44. The "port" option of the smtp transport is now expanded.
7763 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7764 silently ignored. Now an error message is written to the main and panic
7767 46. There's a new ACL modifier called "logwrite" which writes to a log file
7768 as soon as it is encountered.
7770 47. Added $local_user_uid and $local_user_gid at routing time.
7772 48. Exim crashed when trying to verify a sender address that was being
7775 49. Exim was recognizing only a space character after ".include". It now also
7776 recognizes a tab character.
7778 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7779 extracting the relevant information from the specification. The man page no
7780 longer contains scrambled data for the -d option, and I've added a section
7781 at the front about calling Exim under different names.
7783 51. Added "extra_headers" argument to the "mail" command in filter files.
7785 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7788 53. Installed eximstats 1.29.
7790 54. Added transport_filter_timeout as a generic transport option.
7792 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7793 Cc: header lines. This was required by RFC 822, but it not required by RFC
7796 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7797 incoming messages that did not have them. Now it does so only if the
7798 message originates locally, that is, if there is no associated remote host
7799 address. When Resent- header lines are present, this applies to the Resent-
7800 lines rather than the non-Resent- lines.
7802 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7803 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7805 58. Messages for configuration errors now include the name of the main
7806 configuration file - useful now that there may be more than one file in a
7807 list (.included file names were always shown).
7809 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7810 for those rare installations that do not start the daemon as root or run it
7811 setuid root. I've cut out the call to initgroups() if the daemon is not
7814 60. The Exim user and group can now be bound into the binary as text strings
7815 that are looked up at the start of Exim's processing.
7817 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7819 62. Added $mailstore_basename variable.
7821 63. Installed patch to sieve.c from Michael Haardt.
7823 64. When Exim failed to open the panic log after failing to open the main log,
7824 the original message it was trying to log was written to stderr and debug
7825 output, but if they were not available (the usual case in production), it
7826 was lost. Now it is written to syslog before the two lines that record the
7827 failures to open the logs.
7829 65. Users' Exim filters run in subprocesses under the user's uid. It is
7830 possible for a "deliver" command or an alias in a "personal" command to
7831 provoke an address rewrite. If logging of address rewriting is configured,
7832 this fails because the process is not running as root or exim. There may be
7833 a better way of dealing with this, but for the moment (because 4.30 needs
7834 to be released), I have disabled address rewrite logging when running a
7835 filter in a non-root, non-exim process.
7841 1. The buildconfig auxiliary program wasn't quoting the value set for
7842 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7843 not defined. This bug was masked in 4.22 by the effect that was fixed in
7846 2. Some messages that were rejected after a message id was allocated were
7847 shown as "incomplete" by exigrep. It no longer does this for messages that
7848 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7850 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7851 have allow_domain_literals set, the ID did not get logged in the <= line.
7852 Domain literals are now always recognized in Message-ID: header lines.
7854 4. The first argument for a ${extract expansion item is the key name or field
7855 number. Leading and trailing spaces in this item were not being ignored,
7856 causing some misleading effects.
7858 5. When deliver_drop_privilege was set, single queue runner processes started
7859 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7860 same command in the process it spins off) were not dropping privilege.
7862 6. When the daemon running as "exim" started a queue runner, it always
7863 re-executed Exim in the spun-off process. This is a waste of effort when
7864 deliver_drop_privilege is set. The new process now just calls the
7865 queue-runner function directly.
7871 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7874 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7875 ignored. Though the use of -oP was forcing the writing of a pid file, it
7876 was always written to the default place.
7878 3. If the message "no IP address found for host xxxx" is generated during
7879 incoming verification, it is now followed by identification of the incoming
7880 connection (so you can more easily find what provoked it).
7882 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7884 5. Added some features to "harden" Exim a bit more against certain attacks:
7886 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7887 be put in Local/Makefile. This is like the never_users runtime option,
7888 but it cannot be overridden. The default setting is "root".
7890 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7891 prefix string with which any file named in a -C command line option
7894 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7895 is retained for -C and -D only if the caller of Exim is root. Without
7896 it, the exim user may also use -C and -D and retain privilege.
7898 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7899 command line option is disabled.
7901 6. Macro names set by the -D option must start with an upper case letter, just
7902 like macro names defined in the configuration file.
7904 7. Added "dereference=" facility to LDAP.
7906 8. Two instances of the typo "uknown" in the source files are fixed.
7908 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7909 the Configure-Makefile script screwed up while processing it.
7911 10. Incorporated PCRE 4.4.
7913 11. The SMTP synchronization check was not operating right at the start of an
7914 SMTP session. For example, it could not catch a HELO sent before the client
7915 waited for the greeting. There is now a check for outstanding input at the
7916 point when the greeting is written. Because of the duplex, asynchronous
7917 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7918 way, but not yet received, when the check is performed.
7920 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7921 on TCP/IP sockets, because this apparently causes some broken clients to
7924 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7925 unchanged) from the Cygwin maintainer.
7927 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7928 was supported instead of testing for mbx. Effectively a typo.
7930 15. The spa authenticator server code was not checking that the input it
7931 received was valid base64.
7933 16. The debug output line for the "set" modifier in ACLs was not showing the
7934 name of the variable that was being set.
7936 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7938 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7939 lookup on its own. Something else was needed to trigger the lookup. For
7940 example, a match in host_lookup or the need for a host name in a host list.
7941 Now, if $sender_host_name is referenced and the host name has not yet been
7942 looked up, a lookup is performed. If the lookup fails, the variable remains
7943 empty, and $host_lookup_failed is set to "1".
7945 19. Added "eqi" as a case-independent comparison operator.
7947 20. The saslauthd authentication condition could segfault if neither service
7948 nor realm was specified.
7950 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7951 error message that was logged was misleading, and incoming SMTP
7952 connections were dropped. The message is now more accurate, and temporary
7953 errors are given to SMTP connections.
7955 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7956 (except RSET) with a 421 error, until QUIT is received. However, it was
7957 failing to send a response to QUIT.
7959 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7960 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7961 ACL is now run after the helo_try_verify_hosts code.
7963 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7964 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7965 case-independent, so other case variants are also recognized). Apparently
7966 some systems use these upper case variants.
7968 25. If more than two messages were waiting for the same host, and a transport
7969 filter was specified for the transport, Exim sent two messages over the
7970 same TCP/IP connection, and then failed with "socket operation on non-
7971 socket" when it tried to send the third.
7973 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7975 27. The extern definition of crypt16() in expand.c was not being excluded when
7976 the OS had its own crypt16() function.
7978 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7979 as a preferred synonym for return_size_limit, both as an option and as an
7982 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7984 30. Changed the default configuration ACL to relax the local part checking rule
7985 for addresses that are not in any local domains. For these addresses,
7986 slashes and pipe symbols are allowed within local parts, but the sequence
7987 /../ is explicitly forbidden.
7989 31. SPA server authentication was not clearing the challenge buffer before
7992 32. log_message in a "warn" ACL statement was writing to the reject log as
7993 well as to the main log, which contradicts the documentation and doesn't
7994 seem right (because no rejection is happening). So I have stopped it.
7996 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7997 However, I am unable to do any testing of this.
7999 34. Fixed an infelicity in the appendfile transport. When checking directories
8000 for a mailbox, to see if any needed to be created, it was accidentally
8001 using path names with one or more superfluous leading slashes; tracing
8002 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8004 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8005 contents are added to the log line that is written for every discarded
8006 recipient. (Previously a log_message setting was ignored.)
8008 36. The ${quote: operator now quotes the string if it is empty.
8010 37. The install script runs exim in order to find its version number. If for
8011 some reason other than non-existence or emptiness, which it checks, it
8012 could not run './exim', it was installing it with an empty version number,
8013 i.e. as "exim-". This error state is now caught, and the installation is
8016 38. An argument was missing from the function that creates an error message
8017 when Exim fails to connect to the socket for saslauthd authentication.
8018 This could cause Exim to crash, or give a corrupted message.
8020 39. Added isip, isip4, and isip6 to ${if conditions.
8022 40. The ACL variables $acl_xx are now saved with the message, and can be
8023 accessed later in routers, transports, and filters.
8025 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8026 strings in the file are not string-expanded.
8028 42. If a MAIL command specified a SIZE value that was too large to fit into an
8029 int variable, the check against message_size_limit failed. Such values are
8030 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8031 day this will have to be increased, but I don't think I want to be around
8032 when emails are that large.
8039 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8040 iconv() is not standard in FreeBSD.
8042 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8043 IPv6 enabled. The observed symptom was a segmentation fault on return from
8044 the function os_common_find_running_interfaces() in src/os.c.
8046 3. In the check_special_case() function in daemon.c I had used "errno" as an
8047 argument name, which causes warnings on some systems. This was basically a
8048 typo, since it was named "eno" in the comments!
8050 4. The code that waits for the clock to tick (at a resolution of some fraction
8051 of a second) so as to ensure message-id uniqueness was always waiting for
8052 at least one whole tick, when it could have waited for less. [This is
8053 almost certainly not relevant at current processor speeds, where it is
8054 unlikely to ever wait at all. But we try to future-proof.]
8056 5. The function that sleeps for a time interval that includes fractions of a
8057 second contained a race. It did not block SIGALRM between setting the
8058 timer, and suspending (a couple of lines later). If the interval was short
8059 and the sigsuspend() was delayed until after it had expired, the suspension
8060 never ended. On busy systems this could lead to processes getting stuck for
8063 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8064 process, before it forks any delivery processes. The open lookup caching
8065 mechanism meant that the open file or database connection was passed into
8066 the delivery process. The problem was that delivery processes always tidy
8067 up cached lookup data. This could cause a problem for the next delivery
8068 process started by the queue runner, because the external queue runner
8069 process does not know about the closure. So the next delivery process
8070 still has data in the lookup cache. In the case of a file lookup, there was
8071 no problem because closing a file descriptor in a subprocess doesn't affect
8072 the parent. However, if the lookup was caching a connection to a database,
8073 the connection was closed, and the second delivery process was likely to
8074 see errors such as "PGSQL: query failed: server closed the connection
8075 unexpectedly". The problem has been fixed by closing all cached lookups
8076 in a queue runner before running a delivery process.
8078 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8079 seem to have the "const" qualifier which it has on other OS. I've
8082 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8083 *of the same type* (client or server) with the same public name that an
8084 error should be diagnosed.
8086 9. When Exim looked up a host name for an IP address, but failed to find the
8087 original IP address when looking up the host name (a safety check), it
8088 output the message "<ip address> does not match any IP for NULL", which was
8089 confusing, to say the least. The bug was that the host name should have
8090 appeared instead of "NULL".
8092 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8093 user that is built into the binary, and the -C or -D options is used, root
8094 privilege is dropped before the configuration file is read. In addition,
8095 logging is switched to stderr instead of the normal log files. If the
8096 configuration then re-defines the Exim user, the unprivileged environment
8097 is probably not what is expected, so Exim logs a panic warning message (but
8100 However, if deliver_drop_privilege is set, the unprivileged state may well
8101 be exactly what is intended, so the warning has been cut out in that case,
8102 and Exim is allowed to try to write to its normal log files.
8108 1. smtp_return_error_details was not giving details for temporary sender
8109 or receiver verification errors.
8111 2. Diagnose a configuration error if two authenticators have the same public
8114 3. Exim used not to create the message log file for a message until the first
8115 delivery attempt. This could be confusing when incoming messages were held
8116 for policy or load reasons. The message log file is now created at the time
8117 the message is received, and an initial "Received" line is written to it.
8119 4. The automatically generated man page for command line options had a minor
8120 bug that caused no ill effects; however, a more serious problem was that
8121 the procedure for building the man page automatically didn't always
8122 operate. Consequently, release 4.20 contains an out-of-date version. This
8123 shouldn't happen again.
8125 5. When building Exim with embedded Perl support, the script that builds the
8126 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8127 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8129 6. The freeze_tell option was not being used for messages that were frozen on
8130 arrival, either by an ACL or by local_scan().
8132 7. Added the smtp_incomplete_transaction log selector.
8134 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8135 was accepting AUTH without a new EHLO.
8137 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8138 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8141 10. Logging of TCP/IP connections (when configured) now happens in the main
8142 daemon process instead of the child process, so that the TCP/IP connection
8143 count is more accurate (but it can never be perfect).
8145 11. The use of "drop" in a nested ACL was not being handled correctly in the
8146 outer ACL. Now, if condition failure induced by the nested "drop" causes
8147 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8148 or "require"), the connection is dropped.
8150 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8151 that yield "discard" can now be used with an "accept" or a "discard" verb,
8152 but an error is generated for any others (because I can't see a useful way
8153 to define what should happen).
8155 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8156 are now processed in the same way as lines in the Exim configuration file.
8157 In particular, continuation lines are supported.
8159 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8161 15. Added -ti meaning -t -i.
8163 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8164 domains, and warn by logging if others are found.
8166 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8167 returned from GIFCONF when Exim is trying to find the list of interfaces on
8168 a host. The code in os.c has been modified to copy each ifreq to an aligned
8169 structure in all cases.
8171 Also, in some cases, the returned ifreq's were being copied to a 'struct
8172 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8173 means the last couple of bytes of an IPv6 address could be chopped if the
8174 ifreq contained only a normal sockaddr (14 bytes storage).
8176 18. Named domain lists were not supported in the hosts_treat_as_local option.
8177 An entry such as +xxxx was not recognized, and was treated as a literal
8180 19. Ensure that header lines added by a DATA ACL are included in the reject log
8181 if the ACL subsequently rejects the message.
8183 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8184 MD5 (which is deprecated).
8186 21. When testing a filter file using -bf, Exim was writing a message when it
8187 took the sender from a "From " line in the message, but it was not doing so
8188 when it took $return_path from a Return-Path: header line. It now does.
8190 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8191 with a valid header line field name (a series of printing characters
8192 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8194 23. Changed "disc" in the source to "disk" to conform to the documentation and
8195 the book and for uniformity.
8197 24. Ignore Sendmail's -Ooption=value command line item.
8199 25. When execve() failed while trying to run a command in a pipe transport,
8200 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8201 could be confused with a return value of 69 from the command itself. This
8202 has been changed to 127, the value the shell returns if it is asked to run
8203 a non-existent command. The wording for the related log line suggests a
8204 non-existent command as the problem.
8206 26. If received_header_text expands to an empty string, do not add a Received:
8207 header line to the message. (Well, it adds a token one on the spool, but
8208 marks it "old" so that it doesn't get used or transmitted.)
8210 27. Installed eximstats 1.28 (addition of -nt option).
8212 28. There was no check for failure on the call to getsockname() in the daemon
8213 code. This can fail if there is a shortage of resources on the system, with
8214 ENOMEM, for example. A temporary error is now given on failure.
8216 29. Contrary to the C standard, it seems that in some environments, the
8217 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8218 program. Exim now does this explicitly; it affects the formatting of
8219 timestamps using strftime().
8221 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8222 complaints. I've now initialized all the variables, to avoid this.
8224 32. Header lines added by a system filter were not being "seen" during
8225 transport-time rewrites.
8227 33. The info_callback() function passed to OpenSSL is set up with type void
8228 (*)(SSL *, int, int), as described somewhere. However, when calling the
8229 function (actually a macro) that sets it up, the type void(*)() is
8230 expected. I've put in a cast to prevent warnings from picky compilers.
8232 34. If a DNS black list lookup found a CNAME record, but there were no A
8233 records associated with the domain it pointed at, Exim crashed.
8235 35. If a DNS black list lookup returned more than one A record, Exim ignored
8236 all but the first. It now scans all returned addresses if a particular IP
8237 value is being sought. In this situation, the contents of the
8238 $dnslist_value variable are a list of all the addresses, separated by a
8241 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8242 to accept a host name and all its aliases if the forward lookup for any of
8243 them yielded the IP address of the incoming connection. Now it accepts only
8244 those names whose forward lookup yields the correct IP address. Any other
8245 names are discarded. This closes a loophole whereby a rogue DNS
8246 administrator could create reverse DNS records to break through a
8247 wildcarded host restriction in an ACL.
8249 37. If a user filter or a system filter that ran in a subprocess used any of
8250 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8251 the wrong values were passed to the pipe command ($thisaddress had the
8252 value of $0, $0 had the value of $1, etc). This bug was introduced by
8253 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8255 38. Improved the line breaking for long SMTP error messages from ACLs.
8256 Previously, if there was no break point between 40 and 75 characters, Exim
8257 left the rest of the message alone. Two changes have been made: (a) I've
8258 reduced the minimum length to 35 characters; (b) if it can't find a break
8259 point between 35 and 75 characters, it looks ahead and uses the first one
8260 that it finds. This may give the occasional overlong line, but at least the
8261 remaining text gets split now.
8263 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8264 file descriptors might be low, and that setting 1000 would always raise it.
8265 It turns out that in some environments, the limit is already over 1000 and
8266 that lowering it causes trouble. So now Exim takes care not to decrease it.
8268 40. When delivering a message, the value of $return_path is set to $sender_
8269 address at the start of routing (routers may change the value). By an
8270 oversight, this default was not being set up when an address was tested by
8271 -bt or -bv, which affected the outcome if any router or filter referred to
8274 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8275 log only when "message" or "log_message" are set. However, if one of the
8276 conditions was an address verification, or a call to a nested ACL, the
8277 messages generated by the underlying test were being passed through. This
8278 no longer happens. The underlying message is available in $acl_verify_
8279 message for both "message" and "log_message" expansions, so it can be
8280 passed through if needed.
8282 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8283 new expansion $bh_ to give the encoded byte string without charset
8284 translation. Translation happens only if iconv() is available; HAVE_ICONV
8285 indicates this at build time. HEADERS_CHARSET gives the charset to
8286 translate to; headers_charset can change it in the configuration, and
8287 "headers charset" can change it in an individual filter file.
8289 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8290 that creates RFC 2047 encoded "words" labels them as that charset instead
8291 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8292 expansion operator; (ii) when Exim creates a From: line for a local
8293 message; (iii) when a header line is rewritten to include a "phrase" part.
8295 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8296 buggy, causing it to skip the first lines of messages whose message ID
8297 ended in 'D'. This would not have bitten before Exim release 4.14, because
8298 message IDs were unlikely to end in 'D' before then. The effect was to have
8299 incorrect size information for certain domains.
8301 45. #include "config.h" was missing at the start of the crypt16.c module. This
8302 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8305 46. If there was a timeout during a "random" callout check, Exim treated it as
8306 a failure of the random address, and carried on sending RSET and the real
8307 address. If the delay was just some slowness somewhere, the response to the
8308 original RCPT would be taken as a response to RSET and so on, causing
8309 mayhem of various kinds.
8311 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8312 when I implemented it. It didn't allow for the fact that some option values
8313 may legitimately be negative (e.g. size_addition), and it didn't even do
8314 the right test for positive values.
8316 48. Domain names in DNS records are case-independent. Exim always looks them up
8317 in lower case. Some resolvers return domain names in exactly the case they
8318 appear in the zone file, that is, they may contain uppercase letters. Not
8319 all resolvers do this - some return always lower case. Exim was treating a
8320 change of case by a resolver as a change of domain, similar to a widening
8321 of a domain abbreviation. This triggered its re-routing code and so it was
8322 trying to route what was effectively the same domain again. This normally
8323 caused routing to fail (because the router wouldn't handle the domain
8324 twice). Now Exim checks for this case specially, and just changes the
8325 casing of the domain that it ultimately uses when it transmits the message
8328 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8331 50. If a filter generated a file delivery with a non-absolute name (possible if
8332 no home directory exists for the router), the forbid_file option was not
8335 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8336 the existing equality matching.
8338 52. Exim was using ints instead of ino_t variables in some places where it was
8339 dealing with inode numbers.
8341 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8342 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8343 called TMPDIR, and if it finds it is different, it changes its value.
8345 54. The smtp_printf() function is now made available to local_scan() so
8346 additional output lines can be written before returning. There is also an
8347 smtp_fflush() function to enable the detection of a dropped connection.
8348 The variables smtp_input and smtp_batched_input are exported to
8351 55. Changed the default runtime configuration: the message "Unknown user"
8352 has been removed from the ACL, and instead placed on the localuser router,
8353 using the cannot_route_message feature. This means that any verification
8354 failures that generate their own messages won't get overridden. Similarly,
8355 the "Unrouteable address" message that was in the ACL for unverifiable
8356 relay addresses has also been removed.
8358 56. Added hosts_avoid_esmtp to the smtp transport.
8360 57. The exicyclog script was not checking for the esoteric option
8361 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8362 will work only if exicyclog is run under the appropriate euid.
8364 58. Following a discussion on the list, the rules by which Exim recognises line
8365 endings on incoming messages have been changed. The -dropcr and drop_cr
8366 options are now no-ops, retained only for backwards compatibility. The
8367 following line terminators are recognized: LF CRLF CR. However, special
8368 processing applies to CR:
8370 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8371 nor a local message in the state where . is a terminator.
8373 (ii) If a bare CR is encountered in a header line, an extra space is added
8374 after the line terminator so as not to end the header. The reasoning
8375 behind this is that bare CRs in header lines are most likely either
8376 to be mistakes, or people trying to play silly games.
8378 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8379 was being incorrectly given as 18 bytes larger than it should have been.
8380 This is a VOB (very old bug).
8382 60. This may never have affected anything current, but just in case it has:
8383 When the local host is found other than at the start of a list of hosts,
8384 the local host, those with the same MX, and any that follow, are discarded.
8385 When the list in question was part of a longer list of hosts, the following
8386 hosts (not currently being processed) were also being discarded. This no
8387 longer happens. I'm not sure if this situation could ever has previously
8390 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8393 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8394 header. This is something that is recommended in a new Internet Draft, and
8395 is something that is documented as being done by Sendmail. There are two
8396 possible values. For messages generated by the autoreply transport, Exim
8399 Auto-Submitted: auto-replied
8401 whereas for all other generated messages (e.g. bounces) it adds
8403 Auto-Submitted: auto-generated
8405 63. The "personal" condition in filters now includes a test for the
8406 Auto-Submitted: header. If it contains the string "auto-" the message it
8407 not considered personal.
8409 64. Added rcpt_include_affixes as a generic transport option.
8411 65. Added queue_only_override (default true).
8413 66. Added the syslog_duplication option.
8415 67. If what should have been the first header line of a message consisted of
8416 a space followed by a colon, Exim was mis-interpreting it as a header line.
8417 It isn't of course - it is syntactically invalid and should therefore be
8418 treated as the start of the message body. The misbehaviour could have
8419 caused a number of strange effects, including loss of data in subsequent
8420 header lines, and spool format errors.
8422 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8423 client host had authenticated. This control can now be exercised by an ACL
8424 for more flexibility.
8426 69. By default, callouts do not happen when testing with -bh. There is now a
8427 variant, -bhc, which does actually run the callout code, including
8428 consulting and updating the callout cache.
8430 70. Added support for saslauthd authentication, courtesy of Alexander
8433 71. If statvfs() failed on the spool or log directories while checking their
8434 size for availability, Exim confusingly gave the error "space shortage".
8435 Furthermore, in debugging mode it crashed with a floating point exception.
8436 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8437 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8438 serious problem, Exim now writes to the main and panic logs when this
8439 happens, with details of the failure. It then refuses to accept the
8440 incoming message, giving the message "spool directory problem" or "log
8441 directory problem" with a 421 code for SMTP messages.
8443 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8444 0, 1, and 2 exist, because some OS complain for execs without them (see
8445 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8446 descriptors. However, the code omitted to check that the open succeeded,
8447 causing mysterious errors if for some reason the permissions on /dev/null
8448 got screwed. Now Exim writes a message to the main and panic logs, and
8449 bombs out if it can't open /dev/null.
8451 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8452 interact so that it is all more flexible. It is supposed to remain
8453 backwards compatible. Also added extra_local_interfaces.
8455 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8456 to bomb out with an assertion failure - to the client this appears as a
8457 connection drop. This problem occurs in the part of the code that was taken
8458 from the Samba project. Fortunately, the assertion is in a very simple
8459 function, so I have fixed this by reproducing the function inline in the
8460 one place where it is called, and arranging for authentication to fail
8461 instead of killing the process with assert().
8463 75. The SPA client code was not working when the server requested OEM rather
8464 than Unicode encoding.
8466 76. Added code to make require_files with a specific uid setting more usable in
8467 the case where statting the file as root fails - usually a non-root-mounted
8468 NFS file system. When this happens and the failure is EACCES, Exim now
8469 forks a subprocess and does the per-uid checking as the relevant uid.
8471 77. Added process_log_path.
8473 78. If log_file_path was not explicitly set, a setting of check_log_space or
8474 check_log_inodes was ignored.
8476 79. If a space check for the spool or log partitions fails, the incident is now
8477 logged. Of course, in the latter case the data may get lost...
8479 80. Added the %p formatting code to string_format() so that it can be used to
8480 print addresses in debug_print(). Adjusted all the address printing in the
8481 debugging in store.c to use %p rather than %d.
8483 81. There was a concern that a line of code in smtp_in.c could overflow a
8484 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8485 initially expressed, the concern was not well-founded, because trailing
8486 spaces are removed early. However, if the trailing spaces were followed by
8487 a NULL, they did not get removed, so the overflow was possible. Two fixes
8490 (a) I re-wrote the offending code in a cleaner fashion.
8491 (b) If an incoming SMTP command contains a NULL character, it is rejected
8494 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8495 runs initgroups(), so that if the Exim user is in any additional groups,
8496 they will be used during message reception.
8502 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob