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.
232 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
233 after reception to before a subsequent reception. This should
234 mean slightly faster delivery, and also confirmation of reception
237 JH/02 Move from using the pcre library to pcre2. The former is no longer
238 being developed or supported (by the original developer).
240 JH/03 Constification work in the filters module required a major version
241 bump for the local-scan API. Specifically, the "headers_charset"
242 global which is visible via the API is now const and may therefore
243 not be modified by local-scan code.
245 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
246 sendfile() didi not account for the way the ClamAV driver code called it.
248 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
249 time check was being done for every character; replace that with one
252 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
253 sent was prefixed with a length byte.
255 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
256 RFC 5321. Previously Dovecot (at least) would log errors during
259 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
260 in 4.61 without a reason logged. Binaries will be bigger, which might
261 matter on diskspace-constrained systems, but debug is easier.
263 JH/09 Fix macro-definition during "-be" expansion testing. The move to
264 write-protected store for macros had not accounted for these runtime
265 additions; fix by removing this protection for "-be" mode.
267 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
268 handing out large-numbered file descriptors, violating the usual Unix
269 assumption (and required by Posix) that the lowest possible number will be
270 allocated by the kernel when a new one is needed. In the daemon, and any
271 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
272 are not useable for FD_SET() [and hence select()] and overwrite the stack.
273 Assorted crashes happen.
275 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
276 main-section options or in a connect ACL, the value from the first ever
277 connection was never replaced for subsequent connections. Found by
280 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
281 though only once PCRE2 was introduced: the memory accounting used under
282 debug offset allocations by an int, giving a hard trap in early startup.
283 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
285 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
286 with underbars is given. The write-protection of configuration introduced
287 in 4.95 trapped when normalisation was applied to an option not needing
290 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
292 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
293 failing to close the certificates directory, every hour or any time it
296 JH/16 Debugging initiated by an ACL control now continues through into routing
297 and transport processes. Previously debugging stopped any time Exim
298 re-execs, or for processing a queued message.
300 JH/17 The "expand" debug selector now gives more detail, specifically on the
301 result of expansion operators and items.
303 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
304 bad comparison between the option value and the name of the file to
305 be included was done, and a mismatch was wrongly identified.
306 4.88 to 4.95 are affected.
308 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
310 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
311 supplied as the db file-pair basename. Previously, if a directory
312 path was given, for example via the autoreply "once" option, the DB
313 file.pag and file.dir files would be created in that directory's
316 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
317 "taint" log_selector. These were previously deprecated.
319 JH/22 Fix static address-list lookups to properly return the matched item.
320 Previously only the domain part was returned.
322 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
323 the call into OpenSSL to send a TLS Close was being repeated; this
324 resulted in the library waiting for the peer's Close. If that was never
325 sent we waited forever. Fix by tracking send calls.
327 JH/24 The ${run} expansion item now expands its command string elements after
328 splitting. Previously it was before; the new ordering makes handling
329 zero-length arguments simpler. The old ordering can be obtained by
330 appending a new option "preexpand", after a comma, to the "run".
332 JH/25 Taint-check exec arguments for transport-initiated external processes.
333 Previously, tainted values could be used. This affects "pipe", "lmtp" and
334 "queryprogram" transport, transport-filter, and ETRN commands.
335 The ${run} expansion is also affected: in "preexpand" mode no part of
336 the command line may be tainted, in default mode the executable name
339 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
340 the the facility was not passed across execs, and only the first message
341 passed over a connection could use BDAT; any further ones using DATA.
343 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
344 uses $sending_ip_address and an interface is specified.
345 Previously any use of the local address in the EHLO name disabled
346 PIPECONNECT, the common case being to use the rDNS of it.
348 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
349 resumption. Previously verify failed because no certificate status is
350 passed on the wire for the restarted session. Fix by using the recorded
351 ocsp status of the stored session for the new connection.
353 JH/29 TLS resumption: the key for session lookup in the client now includes
354 more info that a server could potentially use in configuring a TLS
355 session, avoiding oferring mismatching sessions to such a server.
356 Previously only the server IP was used.
358 JH/30 Fix string_copyn() for limit greater than actual string length.
359 Previously the copied amount was the limit, which could result in a
360 overlapping memcpy for newly allocated destination soon after a
361 source string shorter than the limit. Found/investigated by KM.
363 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
364 close; it may be needed for a subsequent connection. This caused a
365 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
367 JH/32 Fix CHUNKING for a second message on a connection when the first was
368 rejected. Previously we did not reset the chunking-offered state, and
369 erroneously rejected the BDAT command. Investigation help from
372 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
373 an empty address. Previously the expansion returned an error.
375 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
376 proxy. Previously these were misparsed, leading to paniclog entries.
382 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
383 related applications. Previously an "H" was used where available info
384 says that "M" should be, so change to match.
386 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
387 as arguments, so an implementation trying to copy these into a local
388 buffer was taking a taint-enforcement trap. Fix by using dynamically
389 created buffers. Similar fix for radius expansion condition.
391 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
392 reasonable, eg. to count headers. Fix by using dynamically created
393 buffers rather than a local. Do similar fixes for ACL actions "dcc",
394 "log_reject_target", "malware" and "spam"; the arguments are expanded
395 so could be handling tainted values.
397 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
398 broken the (no-op) support for this sendmail command. Restore it
399 to doing nothing, silently, and returning good status.
401 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
402 record path was given (or the default used) without a leading directory
403 path, an error occurred on trying to open it. Use the transport's working
406 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
407 transport. Previously it was the name on the DNS A-record; use instead
408 the head of the CNAME chain leading there (if there is one). This seems
409 to align better with RFC 6125.
411 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
412 smtp_accept_max_per_host allocated resources which were not released
413 when the limit was exceeded. This eventually crashed the daemon. Fix
414 by adding a release action in that path.
416 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
417 expanded; previously using tainted values was rejected. Fix by using
418 dynamically-created buffers.
420 JH/09 Relax restrictions on ACL verify condition needing access to message
421 headers. Previously they were only permitted in data and non-smtp ACLs;
422 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
423 not_blind, header_sender and header_names_ascii verification.
425 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
426 Previously a macro used one argument twice; when called with the
427 argument as an expression having side-effects, incorrect operation
428 resulted. Use an inlineable function.
430 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
431 held open for a verify callout. Previously this wan not accounted for
432 and a corrupt onward SMTP conversation resulted.
434 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
435 Previously they were embedded naively in the constructed address; when
436 needed, strip the quoting and quote the entire local_part.
437 Also make the inbound_srs expansion condition handle quoting.
439 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
440 excluded, not matching the documentation.
442 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
443 was given for the sqlite_dbfile a trap resulted.
445 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
446 "name" argument resulted in a trap. There is no reason to disallow such;
447 this was a coding error.
449 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
450 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
451 spent suspended, ignoring the POSIX definition. Previously we assumed
452 it did and a constant offset from real time could be used as a correction.
453 Change to using the same clock source for the start-of-message and the
454 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
455 exists, just to get a clock slightly more aligned to reality.
457 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
458 RFC says it is optional some validators care. The missing char was not
459 intended but triggered by a line-wrap alignment. Discovery and fix by
460 Guillaume Outters, hacked on by JH.
462 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
463 name being quoted was tainted a trap would be taken. Fix by using
464 dynamically created buffers. The routine could have been called by a
465 rewrite with the "h" flag, by using the "-F" command-line option, or
466 by using a "name=" option on a control=submission ACL modifier.
468 JH/19 SPF: change the Authentication-Results expansion component to give
469 smtp.helo when the sender domain is empty. Previously it gave
472 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
473 not in 127.0.0.0/8 to help in spotting list domains taken over by a
474 domain-parking registrar.
476 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
477 Previously when a whitespace character was specified it was not inserted
478 after removing the newline.
480 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
481 the domain part of the recipient address. This overrides any tls_sni
482 option set, which was previously used.
484 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
487 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
488 is_tainted() had an off-by-one error in the overenthusiastic direction.
489 Find and fix by Gavan. Although NetBSD is not a supported platform for
490 4.94 this bug could affect other platforms.
492 PP/01 Fix default prime selection to be consistent.
493 One path used ike23 still, instead of exim.dev.20160529.3; now both
494 execution flows will use the same DH primes (currently
495 exim.dev.20160529.3).
497 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
498 option in smtp transport, to match the documentation. Previously
499 verification was not being done in some cases where it should have been.
501 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
502 than one server was defined and depending on the platform memory layout
503 details, an internal consistency trap could be hit while walking the list
506 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
507 files. The value is used by the authresults expansion item. Previously
508 if this was used in a router or transport, a crash could result.
510 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
511 have a sufficient incidence of EINTR returns from select that an
512 interaction with other operations done by the main daemon loop exposed
513 a bug in the error-handling. This was benign apart from the log
516 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
517 for consistency with delivered "=>" and failed "**" lines. While we're
518 there, handle PRX and TFO.
520 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
521 applied. This resulted, if any header-line rewrite rules were configured,
522 in a panic-log triggerable by sending a message with a long address in
523 a header. Fix by increasing the arbitrary limit to larger than a single
524 (dewrapped) 5322 header line maximum size.
526 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
527 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
528 RFC 6648 which deprecates X- options in protocols as a general practice.
529 Changeover between the implementations is handled by the mechanisms
532 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
533 a remote which uses callout/hold. Previously the local was queued.
535 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
538 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
539 log line. The asterisk is now consistently appended to the remote IP
540 (and port, if given), and will also be provided on defer and fail log
541 lines. Previously it could be placed on the local IP if that was being
542 logged, and was only provided on delivery lines.
544 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
546 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
547 authenticator client_send option. Previously the next char, after a pair
548 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
549 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
550 documentation. There is still no way to get a leading ^ immediately
551 after a NUL (ie. for the password of a PLAIN method authenticator.
553 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
554 files. For bad sizes read, delete the record and whine to paniclog.
556 JH/38 When logging an AUTH failure, as server, do not include sensitive
557 information. Previously, the credentials would be included if given
558 as part of the AUTH command line and an ACL denied authentication.
560 JH/39 Bug 2691: fix $local_part_data. When the matching list element
561 referred to a file, bad data was returned. This likely also affected
564 JH/40 The gsasl authenticator now supports caching of the salted password
565 generated by the client-side implementation. This required the addition
566 of a new variable: $auth4.
568 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
569 left undeleted; the attempt to re-create it then failed - resulting in
570 the usual "SIGHUP tp have daemon reload configuration" to not work.
571 This affected any platform not supporting "abstract" Unix-domain
572 sockets (i.e. not Linux).
574 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
575 recipients" response to RCPT in a later response, with a 250. The
576 previous coding assumed this would not happen, and under PIPELINING
577 would result in both lost and duplicate recipients for a message.
579 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
580 Previously the weighting was incorrectly applied. Similar fix for socks
581 proxies. Found and fixed by Heiko Schlichting.
583 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
584 not handle sub-lists included using the +namedlist syntax. While
585 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
586 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
589 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
590 Previously the permanent pool was used, so the sore could not be freed.
591 This meant a connection with many messages would use continually-growing
594 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
595 per-pool so as not to waste too much space. Previously a constant size
596 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
597 DOS attacks harder. The cost is wasted memory use in the larger blocks.
599 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
600 same space repeatedly, and a smaller process footprint.
602 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
603 for GnuTLS. Previously the name started "%s" which, while not a bug,
604 looked as if if might be one.
606 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
607 "continued connection") make the $proxy_* variables available. Previously
608 the information was not passed across the exec() call for subsequent
609 transport executions. This also mean that the log lines for the
610 messages can show the proxy information.
612 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
613 receive time. With modern systems the difference is significant.
614 The historical behaviour can be restored by disabling (a new) log_selector
615 "queue_time_exclusive".
617 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
618 content) were specifically tested for. Now, also cover expansions
619 resulting in ACL names and inline ACL content.
621 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
622 making it unusable in complex expressions.
624 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
625 fail with a taint-check on the spoolfile name, and leave the message
628 HS/01 Enforce absolute PID file path name.
630 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
632 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
634 PP/02 Bug 2643: Correct TLS DH constants.
635 A missing NUL termination in our code-generation tool had led to some
636 incorrect Diffie-Hellman constants in the Exim source.
637 Reported by kylon94, code-gen tool fix by Simon Arlott.
639 PP/03 Impose security length checks on various command-line options.
640 Fixes CVE-2020-SPRSS reported by Qualys.
642 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
643 better. Reported by Qualys.
645 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
646 providing a particularly obnoxious sender full name.
649 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
651 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
654 PP/08 Change default for recipients_max from unlimited to 50,000.
656 PP/09 Fix security issue with too many recipients on a message (to remove a
657 known security problem if someone does set recipients_max to unlimited,
658 or if local additions add to the recipient list).
659 Fixes CVE-2020-RCPTL reported by Qualys.
661 PP/10 Fix security issue in SMTP verb option parsing
662 Fixes CVE-2020-EXOPT reported by Qualys.
664 PP/11 Fix security issue in BDAT state confusion.
665 Ensure we reset known-good where we know we need to not be reading BDAT
666 data, as a general case fix, and move the places where we switch to BDAT
667 mode until after various protocol state checks.
668 Fixes CVE-2020-BDATA reported by Qualys.
670 HS/03 Die on "/../" in msglog file names
672 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
673 the Exim runtime user are allowed to create files.
675 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
678 QS/03 When reading the output from interpreted forward files we do not
679 pass the pipe between the parent and the interpreting process to
680 executed child processes (if any).
682 QS/04 Always die if requested from internal logging, even is logging is
685 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
686 the API; compilation noo longer completes with DMARC support included.
687 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
688 been reported on other platforms.
690 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
692 JH/56 Make the majority of info read from config files readonly, for defence-in-
693 depth against exploits. Suggestion by Qualys.
694 Not supported on Solaris 10.
696 JH/57 Fix control=fakreject for a custom message containing tainted data.
697 Previously this resulted in a log complaint, due to a re-expansion present
698 since fakereject was originally introduced.
700 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
701 modifier was given, a loop resulted.
703 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
704 pipelined SMTP command followed the BDAT LAST then it would be
705 incorrectly treated as part of the message body, causing a verification
708 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
709 require looser validation rules than those for 821-level addresses,
710 which only permit IP addresses.
716 JH/01 Avoid costly startup code when not strictly needed. This reduces time
717 for some exim process initialisations. It does mean that the logging
718 of TLS configuration problems is only done for the daemon startup.
720 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
722 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
723 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
726 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
727 DKIM signing is being done. Previously a transport_filter would always
728 disable CHUNKING, falling back to traditional DATA.
730 JH/05 Regard command-line recipients as tainted.
732 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
734 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
735 PAM library frees one of the arguments given to it, despite the
736 documentation. Therefore a plain malloc must be used.
738 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
739 on-stack buffers were used, resulting in a taint trap when DSN information
740 copied from a received message was written into the buffer.
742 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
743 the ordering of its ARC headers. This caused a crash.
745 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
746 a new record was being constructed with information from the peer, a trap
749 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
750 installation would get error messages from DMARC verify, when it hit the
751 nonexistent file indicated by the default. Distros wanting DMARC enabled
752 should both provide the file and set the option.
753 Also enforce no DMARC verification for command-line sourced messages.
755 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
756 could, depending on the platform, hang at the STARTTLS response.
758 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
759 message on a connection. Previously if one message had ARC headers and
760 the following one did not, a crash could result when adding an
761 Authentication-Results: header.
763 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
764 Exim main code and Exim-related utities. The introduction of taint
765 tracking also did many adjustments to string handling. Since then, eximon
766 frequently terminated with an assert failure.
768 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
769 check for 452 responses. This slightly helps the inefficieny of doing
770 a large alias-expansion into a recipient-limited target. The max_rcpt
771 transport option still applies (and at the current default, will override
772 the new feature). The check is done for either cause of synch, and forces
773 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
774 connection. The new facility is not tunable at this time.
776 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
777 library live data was being used, so the results became garbage. Make
778 copies while it is still usable.
780 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
781 on delivery deferred (==) and failed (**) lines (if a delivery was
782 attemtped). Previously it was only on completion (=>) lines.
784 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
785 for the expansion of the server_scram_iter and server_scram_salt options.
787 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
788 are now specifically given a NO_DATA response without hitting the system
789 resolver. The library goes on to do the now-standard TXT lookup.
790 Use of dnsdb lookups is not affected.
792 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
793 only retrieve the errormessage once. Previously two calls to dlerror()
794 were used, and the second one (for mainlog/paniclog) retrieved null
797 JH/20 Taint checking: disallow use of tainted data for
798 - the appendfile transport file and directory options
799 - the pipe transport command
800 - the autoreply transport file, log and once options
801 - file names used by the redirect router (including filter files)
803 - paths used by single-key lookups
804 Previously this was permitted.
806 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
807 adjusted the size of a major service buffer; this failed because the
808 buffer was in use at the time. Change to a compile-time increase in the
809 buffer size, when this authenticator is compiled into exim.
811 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
812 previous fast-mode was untenable in the face of glibs using mmap to
813 support larger malloc requests.
815 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
816 New values supported, if defined on system where compiled:
817 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
818 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
820 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
821 running a limited number of proceses in parallel, a benefit is gained. The
822 amount varies with the platform hardware and load. The use of the option
823 queue_run_in_order means we cannot do this, as ordering becomes
826 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
827 had introduced a string-copy (for ensuring NUL-termination) which was not
828 appropriate for that case, which can include embedded NUL bytes in the
829 block of data. Investigation showed the copy to actually be needless, the
830 data being length-specified.
832 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
833 done during a receiving connection, and both used TLS, global info was
834 used rather than per-connection info for tracking the state of data
835 queued for transmission. This could result in a connection hang.
837 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
838 Previously, when delivering serveral messages down a single connection
839 only the first would provide a SIZE. This was due to the size information
840 not being properly tracked.
842 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
843 TAI (at 37 seconds currently), pretend to be in UTC for time-related
844 expansion and logging. Previously, spurious values such as a future
845 minute could be seen.
847 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
848 it could crash from a null-deref. This could also affect the
849 ${addresses: } operator and ${readsock } item.
851 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
852 message following a mime one, the variable was not reset.
854 JH/30 When an pipelined-connect fails at the first response, assume incorrect
855 cached capability (perhaps the peer reneged?) and immediately retry in
858 JH/31 Fix spurious detection of timeout while writing to transport filter.
860 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
861 an attempt to copy the string was made before checking it.
863 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
864 taint of the lookup key was maintained; we now regard the presence in the
865 filesystem as sufficient validation.
867 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
868 argument is supplied.
870 JH/35 The dsearch lookup now requires that the directory is an absolute path.
871 Previously this was not checked, and nonempty relative paths made an
872 access under Exim's current working directory.
874 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
875 Previously no event was raised.
877 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
878 parameter supplied by the sender MAIL FROM command. Previously it was
879 ignored, and only the check_spool_space option value for the required
882 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
883 the size of the signing public-key. Previously it was instead giving
884 the size of the signature hash.
886 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
887 the default. See the (new) dkim_verify_min_keysizes option.
889 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
890 and an ACL use a lookup for checking either the local_part or domain,
891 stale data could be accessed. Ensure that variable references are
892 dropped between messages.
894 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
895 by the client was not checked as pointing within response data before
896 being used. A malicious client could thus cause an out-of-bounds read and
897 possibly gain authentication. Fix by adding the check.
899 JH/42 Internationalisation: change the default for downconversion in the smtp
900 transport to be "if needed". Previously it was "as previously set" for
901 the message, which usually meant "if needed" for message-submission but
902 "no" for everything else. However, MTAs have been seen using SMTPUTF8
903 even when the envelope addresses did not need it, resulting in forwarding
904 failures to non-supporting MTAs. A downconvert in such cases will be
905 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
906 transport. The change does mean that addresses needing conversion will
907 be converted when previously a delivery failure would occur.
909 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
910 response was received it would be used unchecked in a fail-DSN, violating
911 standards on line-length limits. Truncate if needed.
913 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
914 doesn't work. (Additionally add a new main config option to configure the
921 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
922 side, to decode a TLS 1.3 packet capture.
924 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
925 Previously the default library behaviour applied, sending two, each in
928 JH/03 Debug output for ACL now gives the config file name and line number for
931 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
933 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
935 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
936 buffer overrun for (non-chunking) other transports.
938 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
939 TLS1.3, means that a server rejecting a client certificate is not visible
940 to the client until the first read of encrypted data (typically the
941 response to EHLO). Add detection for that case and treat it as a failed
942 TLS connection attempt, so that the normal retry-in-clear can work (if
943 suitably configured).
945 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
946 and/or domain. Found and fixed by Jason Betts.
948 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
949 configuration). If a CNAME target was not a wellformed name pattern, a
952 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
953 the OS reports them interleaved with other addresses.
955 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
956 used both for input and for a verify callout, both encrypted, SMTP
957 responses being sent by the server could be lost. This resulted in
958 dropped connections and sometimes bounces generated by a peer sending
961 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
962 string. Previously it was possible to cause undefined behaviour in a
963 library routine (usually a crash). Found by "zerons".
965 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
968 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
969 API was removed, so update to use the newer ones.
971 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
972 any timeout set, is taking a long time. Previously we would hang on to a
973 rotated logfile "forever" if the input was arriving with long gaps
974 (a previous attempt to fix addressed lack, for a long time, of initial
977 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
978 shared (NFS) environment. The length of the tempfile name is now
979 4 + 16 ("hdr.$message_exim_id") which might break on file
980 systems which restrict the file name length to lower values.
983 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
984 shared (NFS) environment.
986 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
987 did for all versions <4.90). Notably -M, -m, --invert, -I may be
990 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
991 on some platforms for bit 31.
993 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
994 to changes apparently associated with TLS1.3 handling some of the APIs
995 previously used were either nonfunctional or inappropriate. Strings
996 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
997 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
998 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
999 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1000 and the use of specific cipher names in the encrypted= ACL condition.
1002 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1004 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1005 verification result was not updated unless hosts_require_ocsp applied.
1007 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1008 queue_list_requires_admin set to false, non-admin users were denied the
1011 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1012 directory-of-certs mode. Previously they were advertised despite the
1015 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1016 A single TCP connection by a client will now hold a TLS connection open
1017 for multiple message deliveries, by default. Previously the default was to
1020 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1021 default. If built with the facility, DANE will be used. The facility
1022 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1024 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1025 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1026 must be defined and you must still, unless you define DISABLE_TLS, manage
1027 the the include-dir and library-file requirements that go with that
1028 choice. Non-TLS builds are still supported.
1030 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1033 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1034 default. If the platform supports and has the facility enabled, it will
1035 be requested on all coneections.
1037 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1038 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1040 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1042 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1043 success-DSN messages. Previously the From: header was always the default
1044 one for these; the option was ignored.
1046 JH/28 Fix the timeout on smtp response to apply to the whole response.
1047 Previously it was reset for every read, so a teergrubing peer sending
1048 single bytes within the time limit could extend the connection for a
1049 long time. Credit to Qualsys Security Advisory Team for the discovery.
1051 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1052 delivery address, which leaked information of the results of local
1053 forwarding. Change to the original envelope recipient address, per
1056 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1057 requested. Previously not bounce was generated and a log entry of
1058 error ignored was made.
1060 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1062 JH/32 Introduce a general tainting mechanism for values read from the input
1063 channel, and values derived from them. Refuse to expand any tainted
1064 values, to catch one form of exploit.
1066 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1067 was unused and the unexpanded text used for the test. Found and
1068 fixed by Ruben Jenster.
1070 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1071 an attempt to use a TLS library read routine dereffed a nul pointer,
1074 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1075 them in our smtp response.
1077 JH/36 Have the general router option retry_use_local_part default to true when
1078 any of the restrictive preconditions are set (to anything). Previously it
1079 was only for check_local user. The change removes one item of manual
1080 configuration which is required for proper retries when a remote router
1081 handles a subset of addresses for a domain.
1083 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1084 link count into consideration.
1086 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1087 caused the extension of big_buffer, the following lines were ignored.
1089 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1090 accordance with RFC 2308. Previously there was no expiry, so a longlived
1091 receive process (eg. due to ACL delays) versus a short SOA value could
1094 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1096 JH/39 Promote DMARC support to mainline.
1098 JH/40 Bug 2452: Add a References: header to DSNs.
1100 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1101 parameters. The relevant library call is documented as "Deprecated: This
1102 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1103 3.6.0, DH parameters are negotiated following RFC7919."
1105 HS/06 Change the default of dnssec_request_domains to "*"
1107 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1108 carried on and emitted a BDAT command, even when PIPELINING was not
1111 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1112 buffer was used for the filename, resulting in a trap when tainted
1113 arguments (eg. $domain) were used.
1115 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1116 recommended to avoid a possible server-load attack. The feature can be
1117 re-enabled via the openssl_options main cofiguration option.
1119 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1120 for version 4.90 - adding a "more data" boolean to the arguments.
1121 Bumped the ABI version number also, this having been missed previously;
1122 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1123 respect of usage of smtp_printf() by either local_scan code or libraries
1124 accessed via the ${dlfunc } expansion item. Both will need coding
1125 adjustment for any calls to smtp_printf() to match the new function
1126 signature; a FALSE value for the new argument is always safe.
1128 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1129 the file-offset (which the Linux syscall does, and exim expects); this
1130 resulted in an indefinite loop.
1132 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1133 to do ARC verification. The Authentication-Results: header line added
1134 by the configuration then had no ARC item.
1136 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1137 an incorrect mode was used for reading the result, resulting in it being
1144 JH/01 Remove code calling the customisable local_scan function, unless a new
1145 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1147 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1148 non-signal-safe functions being used.
1150 JH/03 Bug 2269: When presented with a received message having a stupidly large
1151 number of DKIM-Signature headers, disable DKIM verification to avoid
1152 a resource-consumption attack. The limit is set at twenty.
1154 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1155 report of oldest_pass in ${authres } in consequence, and separate out
1156 some descriptions of reasons for verification fail.
1158 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1159 files in the spool were present and unlocked. A queue-runner could spot
1160 them, resulting in a duplicate delivery. Fix that by doing the unlock
1161 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1162 add more error-checking on spoolfile handling while that code is being
1165 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1166 No known attacks, no CVE, this is defensive hardening.
1168 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1169 a queue-runner could start a delivery while other operations were ongoing.
1170 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1171 Found and investigated by Tim Stewart. Fix by using the open message data
1172 file handle rather than opening another, and not locally closing it (which
1173 releases a lock) for that case, while creating the temporary .eml format
1174 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1176 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1177 $sender_verify_failure/$recipient_verify_failure to "random".
1179 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1182 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1183 Previously this would segfault.
1185 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1188 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1189 like zero, since the resolver should be doing this for us, But we need one
1190 as a CNAME but no MX presence gets the CNAME returned; we need to check
1191 that doesn't point to an MX to declare it "no MX returned" rather than
1192 "error, loop". A new main option is added so the older capability of
1193 following some limited number of chain links is maintained.
1195 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1197 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1198 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1199 now-deprecated earlier definitions used only specified the range up to TLS
1200 1.2 (in the older-version library docs).
1202 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1204 JH/15 Rework TLS client-side context management. Stop using a global, and
1205 explicitly pass a context around. This enables future use of TLS for
1206 connections to service-daemons (eg. malware scanning) while a client smtp
1207 connection is using TLS; with cutthrough connections this is quite likely.
1209 JH/16 Fix ARC verification to do AS checks in reverse order.
1211 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1213 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1214 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1215 Previously the "utf8" would be re-prepended for every additional message.
1217 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1218 Previously thery were accepted, resulting in issues when attempting to
1219 forward messages to a non-supporting MTA.
1221 PP/02 Let -n work with printing macros too, not just options.
1223 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1224 one parent address was copied, and bogus data was used at delivery-logging
1225 time. Either a crash (after delivery) or bogus log data could result.
1226 Discovery and analysis by Tim Stewart.
1228 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1229 Previously if the string ended mid-character, we did not insert the
1230 promised '?' replacement.
1232 PP/04 Documentation: current string operators work on bytes, not codepoints.
1234 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1235 should pack well giving a smaller memory footprint so better caching and
1236 therefore performance. Group the declarations where this can't be done so
1237 that the byte-sized flag variables are not interspersed among pointer
1238 variables, giving a better chance of good packing by the compiler.
1240 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1241 non-null, to avoid issues with sites running BATV. Previously reports were
1242 sent with an empty envelope sender so looked like bounces.
1244 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1245 The ignore_error flag wasn't being returned from the filter subprocess so
1246 was not set for later routers. Investigation and fix by Matthias Kurz.
1248 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1249 and a msg:complete for the whole, when a message is manually removed using
1250 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1252 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1253 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1254 hope that is portable enough.
1256 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1257 requiring a known-CA anchor certificate; make it now rely entirely on the
1258 TLSA as an anchor. Checking the name on the leaf cert against the name
1259 on the A-record for the host is still done for TA (but not for EE mode).
1261 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1262 included in delivery lines for non-proxied connections, when compiled with
1263 SUPPORT_SOCKS and running with proxy logging enabled.
1265 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1266 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1267 move the existing event to fire before the normal logging of message
1268 failure so that custom logging is bracketed by normal logging.
1270 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1271 msg:fail:internal event. Developement by Matthias Kurz.
1273 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1274 far too small for todays use of crypto signatures stored there. Go all
1275 the way to the max DNS message size of 64kB, even though this might be
1276 overmuch for IOT constrained device use.
1278 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1279 copy a string over itself. The library routine is documented as not
1280 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1282 JH/32 For main options check_spool_space and check_inode_space, where the
1283 platform supports 64b integers, support more than the previous 2^31 kB
1284 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1285 the previous G, M, k.
1287 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1288 $authenticated_fail_id variable on authentication failure. Previously
1291 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1292 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1293 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1294 GNUTLS_SEC_PARAM_MEDIUM.
1296 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1297 side. Previously we would continue as if no SNI had been received.
1299 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1300 "<" character, which should be a list-separator specification, we walked
1301 off past the nul-terimation.
1303 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1304 causes) even when the retry time is not yet met. Previously they were
1305 not, meaning that when (say) an account was over-quota and temp-rejecting,
1306 and multiple senders' messages were queued, only one sender would get
1307 notified on each configured delay_warning cycle.
1309 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1311 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1312 cipher-suites, an error can be left on the stack even for a succeeding
1313 accept; this results in impossible error messages when a later operation
1316 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1317 return error codes indicating retry. Under TLS1.3 this becomes required.
1319 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1320 it only wrote the new authenticators, resulting in a lack of tracking of
1321 peer changes of ESMTP extensions until the next cache flush.
1323 JH/41 Fix the loop reading a message header line to check for integer overflow,
1324 and more-often against header_maxsize. Previously a crafted message could
1325 induce a crash of the recive process; now the message is cleanly rejected.
1327 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1328 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1334 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1335 When redis_servers is set to a list of > 1 element, and the Redis servers
1336 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1337 case of MOVED into a DEFER case instead, thus moving the query onto the
1338 next server in the list. For a cluster of N elements, all N servers must
1339 be defined in redis_servers.
1341 GF/02 Catch and remove uninitialized value warning in exiqsumm
1342 Check for existence of @ARGV before looking at $ARGV[0]
1344 JH/01 Replace the store_release() internal interface with store_newblock(),
1345 which internalises the check required to safely use the old one, plus
1346 the allocate and data copy operations duplicated in both (!) of the
1347 extant use locations.
1349 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1350 modifier. This matches the restriction on the commandline.
1352 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1353 Previously only the last row was returned.
1355 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1356 we assumed that tags in the header were well-formed, and parsed the
1357 element content after inspecting only the first char of the tag.
1358 Assumptions at that stage could crash the receive process on malformed
1361 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1362 While running the DKIM ACL we operate on the Permanent memory pool so that
1363 variables created with "set" persist to the DATA ACL. Also (at any time)
1364 DNS lookups that fail create cache records using the Permanent pool. But
1365 expansions release any allocations made on the current pool - so a dnsdb
1366 lookup expansion done in the DKIM ACL releases the memory used for the
1367 DNS negative-cache, and bad things result. Solution is to switch to the
1368 Main pool for expansions.
1369 While we're in that code, add checks on the DNS cache during store_reset,
1370 active in the testsuite.
1371 Problem spotted, and debugging aided, by Wolfgang Breyha.
1373 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1374 When none of the hosts presented to a transport match an already-open
1375 connection, close it and proceed with the list. Previously we would
1376 queue the message. Spotted by Lena with Yahoo, probably involving
1379 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1380 Previously a spurious "250 OK id=" response was appended to the proper
1383 JH/08 The "support for" informational output now, which built with Content
1384 Scanning support, has a line for the malware scanner interfaces compiled
1385 in. Interface can be individually included or not at build time.
1387 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1388 by the template makefile "src/EDITME". The "STREAM" support for an older
1389 ClamAV interface method is removed.
1391 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1392 rows affected is given instead).
1394 JH/11 The runtime Berkeley DB library version is now additionally output by
1395 "exim -d -bV". Previously only the compile-time version was shown.
1397 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1398 SMTP connection. Previously, when one had more recipients than the
1399 first, an abortive onward connection was made. Move to full support for
1400 multiple onward connections in sequence, handling cutthrough connection
1401 for all multi-message initiating connections.
1403 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1404 routers. Previously, a multi-recipient message would fail to match the
1405 onward-connection opened for the first recipient, and cause its closure.
1407 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1408 a timeout on read on a GnuTLS initiating connection, resulting in the
1409 initiating connection being dropped. This mattered most when the callout
1410 was marked defer_ok. Fix to keep the two timeout-detection methods
1413 JH/15 Relax results from ACL control request to enable cutthrough, in
1414 unsupported situations, from error to silently (except under debug)
1415 ignoring. This covers use with PRDR, frozen messages, queue-only and
1418 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1420 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1421 metadata, resulting in a crash in free().
1423 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1424 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1425 Broken also in d185889f4, with init system revamp.
1427 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1428 Previously we abruptly closed the connection after reading a malware-
1429 found indication; now we go on to read the "scan ok" response line,
1432 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1433 ACL. Previously, a crash would result.
1435 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1436 macro text after a replacement (previously it was only once per line) and
1437 by skipping builtin macros when searching for an uppercase lead character.
1439 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1440 for the build is renamed.
1442 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1443 was allocated for every new TLS startup, meaning one per message. Fix
1444 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1446 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1447 reported the original. Fix to report (as far as possible) the ACL
1448 result replacing the original.
1450 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1451 OpenSSL. Certificate information is loaded for every new TLS startup,
1452 and the resources needed to be freed.
1454 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1456 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1459 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1460 DATA response info to the (existing) per-recipient response info for
1461 the "C=" log element. It can have useful tracking info from the
1462 destination system. Patch from Simon Arlott.
1464 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1465 length value. Previously this would segfault.
1467 HS/02 Support Avast multiline protoocol, this allows passing flags to
1468 newer versions of the scanner.
1470 JH/28 Ensure that variables possibly set during message acceptance are marked
1471 dead before release of memory in the daemon loop. This stops complaints
1472 about them when the debug_store option is enabled. Discovered specifically
1473 for sender_rate_period, but applies to a whole set of variables.
1474 Do the same for the queue-runner and queue-list loops, for variables set
1475 from spool message files. Do the same for the SMTP per-message loop, for
1476 certain variables indirectly set in ACL operations.
1478 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1479 as a multi-recipient message from a mailinglist manager). The coding had
1480 an arbitrary cutoff number of characters while checking for more input;
1481 enforced by writing a NUL into the buffer. This corrupted long / fast
1482 input. The problem was exposed more widely when more pipelineing of SMTP
1483 responses was introduced, and one Exim system was feeding another.
1484 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1485 receiving system, and refused recipients seen by the sending system
1486 (propating to people being dropped from mailing lists).
1487 Discovered and pinpointed by David Carter.
1489 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1490 replaced by the ${authresults } expansion.
1492 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1494 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1495 allows proper process termination in container environments.
1497 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1498 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1500 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1501 and "err_perm", deprecated since 4.83 when the RFC-defined words
1502 "temperror" and "permerror" were introduced.
1504 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1505 transport-filters or DKIM-signing. The restriction was lost in the
1506 consolidation of verify-callout and delivery SMTP handling.
1507 Extend the restriction to also cover ARC-signing.
1509 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1510 in defer=pass mode supply a 450 to the initiator. Previously the message
1513 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1514 tls_require_ciphers is used as before.
1516 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1517 "pass_unscanned". Only tmpfails from the scanner are written to
1518 the paniclog, as they may require admin intervention (permission
1519 denied, license issues). Other scanner errors (like decompression
1520 bombs) do not cause a paniclog entry.
1522 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1523 Previously it was possible to log spurious information in receive log
1526 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1527 triggered odd behaviour from Outlook Express clients.
1529 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1532 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1533 since the IETF WG has not yet settled on that versus the original
1534 "bare" representation.
1536 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1537 Previously the millisecond value corrupted the output.
1538 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1539 corrupted the output.
1545 JH/01 Rework error string handling in TLS interface so that the caller in
1546 more cases is responsible for logging. This permits library-sourced
1547 string to be attached to addresses during delivery, and collapses
1548 pairs of long lines into single ones.
1550 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1551 during configuration. Wildcards are allowed and expanded.
1553 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1554 This permits better logging.
1556 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1557 do not close it down and have the child transport start it up again on
1558 the passed-on TCP connection. Instead, proxy the child (and any
1559 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1560 affected: the continued delivery log lines do not have any DNSSEC, TLS
1561 Certificate or OCSP information. TLS cipher information is still logged.
1563 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1564 identical IP addresses on different listening ports. Will also affect
1567 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1568 add noisy ifdef guards to special-case this sillyness.
1569 Patch from Bernd Kuhls.
1571 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1572 than 255 are no longer allowed.
1574 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1575 Disable the session-cache too, which might reduce our load. Since we
1576 currrectly use a new context for every connection, both as server and
1577 client, there is no benefit for these.
1578 GnuTLS appears to not support tickets server-side by default (we don't
1579 call gnutls_session_ticket_enable_server()) but client side is enabled
1580 by default on recent versions (3.1.3 +) unless the PFS priority string
1583 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1584 <https://reproducible-builds.org/specs/source-date-epoch/>.
1586 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1587 the check for any unsuccessful recipients did not notice the limit, and
1588 erroneously found still-pending ones.
1590 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1591 MSG_MORE. Only in-clear (not on TLS connections).
1593 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1594 transport-filter is involved we can buffer the headers in memory for
1595 creating the signature, and read the spool data file once for the
1596 signature and again for transmission.
1598 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1599 4.77 as the kernel support then wasn't solid, having issues in 64bit
1600 mode. Now, it's been long enough. Add support for FreeBSD also.
1602 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1603 case where the routing stage had gathered several addresses to send to
1604 a host before calling the transport for the first, we previously failed
1605 to close down TLS in the old transport process before passing the TCP
1606 connection to the new process. The new one sent a STARTTLS command
1607 which naturally failed, giving a failed delivery and bloating the retry
1608 database. Investigation and fix prototype from Wolfgang Breyha.
1610 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1611 false-negatives in the check that the sender had not preempted a response
1612 or prompt from Exim (running as a server), due to that code's lack of
1613 awareness of the SMTP input buffering.
1615 PP/04 Add commandline_checks_require_admin option.
1616 Exim drops privileges sanely, various checks such as -be aren't a
1617 security problem, as long as you trust local users with access to their
1618 own account. When invoked by services which pass untrusted data to
1619 Exim, this might be an issue. Set this option in main configuration
1620 AND make fixes to the calling application, such as using `--` to stop
1623 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1624 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1625 responses to those, into a single TLS record each way (this usually means
1626 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1629 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1630 affects you only if you're dancing at the edge of the param size limits.
1631 If you are, and this message makes sense to you, then: raise the
1632 configured limit or use OpenSSL 1.1. Nothing we can do for older
1635 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1636 cmdline element to get the documented default one. Previously it was
1639 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1640 are unsuable for non-message-specific reasons - eg. connection timeout,
1641 banner-time rejection.
1643 JH/16 Fix logging of delivery remote port, when specified by router, under
1646 PP/06 Repair manualroute's ability to take options in any order, even if one
1647 is the name of a transport.
1650 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1652 JH/17 Change the list-building routines interface to use the expanding-string
1653 triplet model, for better allocation and copying behaviour.
1655 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1656 Previously it was constructed the first time a possibly-matching string
1657 was met in the configuration file input during startup; now it is done
1660 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1661 compatible one, to avoid the (poorly documented) possibility of a config
1662 file in the working directory redirecting the DB files, possibly correpting
1663 some existing file. CVE-2017-10140 assigned for BDB.
1665 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1666 cache-hot. Previously, although the result was properly cached, the
1667 initial verify call returned a defer.
1669 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1670 the main verify for receipient in uncached-mode.
1672 JH/22 Retire historical build files to an "unsupported" subdir. These are
1673 defined as "ones for which we have no current evidence of testing".
1675 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1676 if present. Previously it was ignored.
1678 JH/24 Start using specified-initialisers in C structure init coding. This is
1679 a C99 feature (it's 2017, so now considered safe).
1681 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1682 if was a fixed-sized field and bitmask ops via macros; it is now more
1685 PP/07 GitHub PR 56: Apply MariaDB build fix.
1686 Patch provided by Jaroslav Škarvada.
1688 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1689 during Coverity cleanups [4.87 JH/47]
1690 Diagnosis and fix provided by Michael Fischer v. Mollard.
1692 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1693 the right size to place the terminating semicolon on its own folded
1694 line, the header hash was calculated to an incorrect value thanks to
1695 the (relaxed) space the fold became.
1697 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1698 and confused the parent.
1700 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1701 which could crash as a result. This could lead to undeliverable messages.
1703 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1706 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1707 matters on fast-turnover and PID-randomising systems, which were getting
1708 out-of-order delivery.
1710 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1711 a possibly-overlapping copy. The symptom was that "Remote host closed
1712 connection in response to HELO" was logged instead of the actual 4xx
1715 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1716 Previously only that bufferd was discarded, resulting in SYMTP command
1719 JH/32 DKIM: when a message has multiple signatures matching an identity given
1720 in dkim_verify_signers, run the dkim acl once for each. Previously only
1721 one run was done. Bug 2189.
1723 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1724 "panic the current process" to "deliberately defer". The panic log is
1725 still written with the problem list name; the mail and reject logs now
1726 get a temp-reject line for the message that was being handled, saying
1727 something like "domains check lookup or other defer". The SMTP 451
1728 message is still "Temporary local problem".
1730 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1731 A crafted sequence of BDAT commands could result in in-use memory beeing
1732 freed. CVE-2017-16943.
1734 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1735 from SMTP input. Previously it was always done; now only done for DATA
1736 and not BDAT commands. CVE-2017-16944.
1738 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1739 to the message (such as an overlong header line). Previously this was
1740 not done and we did not exit BDAT mode. Followon from the previous item
1741 though a different problem.
1747 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1748 than -2003 did; needs libidn2 in addition to libidn.
1750 JH/02 The path option on a pipe transport is now expanded before use.
1752 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1753 Patch provided by "Björn", documentation fix added too.
1755 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1756 missing a wire-to-host endian conversion.
1758 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1759 close after a BDAT command line could be taken as a following command,
1760 giving a synch failure. Fix by only checking for synch immediately
1761 before acknowledging the chunk.
1763 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1764 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1765 macro. Patches provided by Josh Soref.
1767 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1768 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1769 by IANA as a service extension. However, John Klensin suggests that we
1772 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1773 the dkim code may be unix-mode line endings rather than smtp wire-format
1774 CRLF, so prepend a CR to any bare LF.
1776 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1777 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1779 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1780 the first were themselves being wrongly included in the feed into dkim
1781 processing; with most chunk sizes in use this resulted in an incorrect
1782 body hash calculated value.
1784 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1785 DKIM signature block, for verification. Although advised against by
1786 standards it is specifically not ruled illegal.
1788 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1790 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1791 missing a body hash (the bh= tag).
1793 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1794 It seems that HAProxy sends the Proxy Protocol information in clear and
1795 only then does a TLS startup, so do the same.
1797 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1798 TCP connections (such as for Spamd) unless the daemon successfully set
1799 Fast Open mode on its listening sockets. This fixes breakage seen on
1800 too-old kernels or those not configured for Fast Open, at the cost of
1801 requiring both directions being enabled for TFO, and TFO never being used
1802 by non-daemon-related Exim processes.
1804 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1805 endings, at least on the first header line. Try to canonify any that get
1806 past that check, despite the cost.
1808 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1809 now limited to an arbitrary five deep, while parsing addresses with the
1810 strip_excess_angle_brackets option enabled.
1812 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1813 instead leave the unprompted TLS handshake in socket buffer for the
1814 TLS library to consume.
1816 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1818 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1820 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1821 region in one operation, for speed, and this leaves assigned pointers
1822 dangling. Add checks run only under the testsuite which checks all
1823 variables at a store-reset and panics on a dangling pointer; add code
1824 explicitly nulling out all the variables discovered. Fixes one known
1825 bug: a transport crash, where a dangling pointer for $sending_ip_address
1826 originally assigned in a verify callout, is re-used.
1828 PP/06 Drop '.' from @INC in various Perl scripts.
1830 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1832 PP/08 Reduce a number of compilation warnings under clang; building with
1833 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1834 should be warning-free.
1836 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1838 HS/01 Fix portability problems introduced by PP/08 for platforms where
1839 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1841 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1842 chunk. This allows us to accept broken chunked messages. We need a more
1843 general solution here.
1845 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1846 already-broken messages in the queue.
1848 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1850 JH/19 Fix reference counting bug in routing-generated-address tracking.
1856 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1857 supports it and a size is available (ie. the sending peer gave us one).
1859 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1860 years of being deprecated). The replacements are the ACLs
1861 acl_smtp_mime and acl_not_smtp_mime.
1863 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1864 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1865 or even an in-clear connection were permitted. Now, if the host lookup
1866 was dnssec and dane was requested then the host is only used if the
1867 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1868 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1869 if one fails this test.
1870 This means that a poorly-configured remote DNS will make it incommunicado;
1871 but it protects against a DNS-interception attack on it.
1873 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1874 non-noisy when a race steals the message being considered.
1876 JH/05 If main configuration option tls_certificate is unset, generate a
1877 self-signed certificate for inbound TLS connections.
1879 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1880 in rewrites and routers.
1882 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1883 and logged a warning sing 4.83; now they are a configuration file error.
1885 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1886 (lacking @domain). Apply the same qualification processing as RCPT.
1888 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1890 JH/10 Support ${sha256:} applied to a string (as well as the previous
1893 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1894 a cutthrough deliver is pending, as we always want to make a connection.
1895 This also avoids re-routing the message when later placing the cutthrough
1896 connection after a verify cache hit.
1897 Do not update it with the verify result either.
1899 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1900 when routing results in more than one destination address.
1902 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1903 signing (which inhibits the cutthrough capability). Previously only
1904 the presence of an option was tested; now an expansion evaluating as
1905 empty is permissible (obviously it should depend only on data available
1906 when the cutthrough connection is made).
1908 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1909 the relevant preceding SMTP command did not note the pipelining mode.
1911 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1912 Previously they were not counted.
1914 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1915 as one having no matching records. Previously we deferred the message
1916 that needed the lookup.
1918 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1919 distinguished as "(=".
1921 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1922 for missing MX records. Previously it only worked for missing A records.
1924 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1926 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1927 after the data-go-ahead and data-ack. Patch from Jason Betts.
1929 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1930 even for a "none" policy. Patch from Tony Meyer.
1932 JH/22 Fix continued use of a connection for further deliveries. If a port was
1933 specified by a router, it must also match for the delivery to be
1936 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1937 When one of the recipients of a message was unsuitable for the connection
1938 (has no matching addresses), we lost track of needing to mark it
1939 deferred. As a result mail would be lost.
1941 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1943 JH/25 Decoding ACL controls is now done using a binary search; the source code
1944 takes up less space and should be simpler to maintain. Merge the ACL
1945 condition decode tables also, with similar effect.
1947 JH/26 Fix problem with one_time used on a redirect router which returned the
1948 parent address unchanged. A retry would see the parent address marked as
1949 delivered, so not attempt the (identical) child. As a result mail would
1952 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1953 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1954 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1957 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1958 Main config options check_{log,spool}_{inodes,space} are now
1959 100 inodes, 10MB unless set otherwise in the configuration.
1961 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1962 Previously it only applied to the main-section connection policy
1965 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1967 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1968 by me. Added RFC7919 DH primes as an alternative.
1970 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1971 are not in the system include path.
1973 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1974 GnuTLS, when a session startup failed (eg because the client disconnected)
1975 Exim did stdio operations after fclose. This was exposed by a recent
1976 change which nulled out the file handle after the fclose.
1978 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1979 signed directly by the cert-signing cert, rather than an intermediate
1980 OCSP-signing cert. This is the model used by LetsEncrypt.
1982 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1984 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1985 an incoming connection.
1987 HS/02 Bug 1802: Do not half-close the connection after sending a request
1990 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1991 fallback to "prime256v1".
1993 JH/34 SECURITY: Use proper copy of DATA command in error message.
1994 Could leak key material. Remotely exploitable. CVE-2016-9963.
2000 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2001 and 3.4.4 - once the server is enabled to respond to an OCSP request
2002 it does even when not requested, resulting in a stapling non-aware
2003 client dropping the TLS connection.
2005 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2006 support variable-length bit vectors. No functional change.
2008 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2009 The I= interface field on outgoing lines is now after the H= remote
2010 host field, same as incoming lines. There is a separate
2011 outgoing_interface log selector which allows you to disable the
2014 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2015 If not running log_selector +smtp_connection the mainlog would be held
2016 open indefinitely after a "too many connections" event, including to a
2017 deleted file after a log rotate. Leave the per net connection logging
2018 leaving it open for efficiency as that will be quickly detected by the
2019 check on the next write.
2021 HS/01 Bug 1671: Fix post transport crash.
2022 Processing the wait-<transport> messages could crash the delivery
2023 process if the message IDs didn't exist for some reason. When
2024 using 'split_spool_directory=yes' the construction of the spool
2025 file name failed already, exposing the same netto behaviour.
2027 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2028 mime_regex ACL conditions.
2030 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2031 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2032 to HELO, local diagnostic string.
2034 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2035 log line to debug. Even when configured with a tls authenticator many
2036 client connections are expected to not authenticate in this way, so
2037 an authenticate fail is not an error.
2039 HS/02 Add the Exim version string to the process info. This way exiwhat
2040 gives some more detail about the running daemon.
2042 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2043 matter for fast-change records such as DNSBLs.
2045 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2046 retry record, even if constant. There may be multiple transports with
2047 different interface settings and the retry behaviour needs to be kept
2050 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2052 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2054 JH/10 Bug 840: fix log_defer_output option of pipe transport
2056 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2057 hosts_randomize. This matters a lot when combined with mua_wrapper.
2059 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2060 ${quote_pgsql:<string>} operator.
2062 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2064 JH/14 Change header file name for freeradius-client. Relevant if compiling
2065 with Radius support; from the Gentoo tree and checked under Fedora.
2067 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2069 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2070 option transport_filter, meaning no filtering.
2072 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2074 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2075 defaults to "*" (all hosts). The variable is now available when not built
2076 with TLS, default unset, mainly to enable keeping the testsuite sane.
2077 If a server certificate is not supplied (via tls_certificate) an error is
2078 logged, and clients will find TLS connections fail on startup. Presumably
2079 they will retry in-clear.
2080 Packagers of Exim are strongly encouraged to create a server certificate
2081 at installation time.
2083 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2084 with the $config_file variable.
2086 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2087 in transport context, after the attempt, and per-recipient. The latter type
2088 is per host attempted. The event data is the error message, and the errno
2089 information encodes the lookup type (A vs. MX) used for the (first) host,
2090 and the trailing two digits of the smtp 4xx response.
2092 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2093 to write to mainlog (or rejectlog, paniclog) in the window between file
2094 creation and permissions/ownership being changed. Particularly affects
2095 installations where exicyclog is run as root, rather than exim user;
2096 result is that the running daemon panics and dies.
2098 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2100 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2101 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2102 "pri" and "weight". Note that the previous implicit priority given by the
2103 list order is no longer honoured.
2105 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2106 for DKIM processing.
2108 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2109 by defining SUPPORT_SOCKS.
2111 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2112 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2113 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2114 variables are renamed to proxy_{local,external}_{address,port}.
2116 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2117 for a build by defining SUPPORT_I18N
2119 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2120 of the query string, and make ${quote_redis:} do that quoting.
2122 JH/29 Move Events support from Experimental to mainline, enabled by default
2123 and removable for a build by defining DISABLE_EVENT.
2125 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2127 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2128 cached by the daemon.
2130 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2131 by defining LOOKUP_REDIS. The libhiredis library is required.
2133 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2134 keys are given for lookup.
2136 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2137 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2138 only supported when built with TLS support. The PolarSSL SHA routines
2139 are still used when the TLS library is too old for convenient support.
2141 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2142 openssl_options), for security. OpenSSL forces this from version 1.1.0
2143 server-side so match that on older versions.
2145 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2146 allocation for $value could be released as the expansion processing
2147 concluded, but leaving the global pointer active for it.
2149 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2150 and to use the domains and local_parts ACL conditions.
2152 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2153 incorrectly not doubled on cutthrough transmission, hence seen as a
2154 body-termination at the receiving system - resulting in truncated mails.
2155 Commonly the sender saw a TCP-level error, and retransmitted the message
2156 via the normal store-and-forward channel. This could result in duplicates
2157 received - but deduplicating mailstores were liable to retain only the
2158 initial truncated version.
2160 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2162 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2164 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2165 we're in there, support oversigning also; bug 1309.
2167 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2169 HS/04 Add support for keep_environment and add_environment options.
2171 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2172 either intentional arithmetic overflow during PRNG, or testing config-
2175 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2176 delivery resulted in actual delivery. Cancel cutthrough before DATA
2179 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2180 rejects a recipient: pass the reject to the originator.
2182 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2183 Many were false-positives and ignorable, but it's worth fixing the
2186 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2187 for the new environment-manipulation done at startup. Move the routines
2188 from being local to tls.c to being global via the os.c file.
2190 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2191 an extract embedded as result-arg for a map, the first arg for extract
2192 is unavailable so we cannot tell if this is a numbered or keyed
2193 extraction. Accept either.
2199 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2202 JH/02 The smtp transport option "multi_domain" is now expanded.
2204 JH/03 The smtp transport now requests PRDR by default, if the server offers
2207 JH/04 Certificate name checking on server certificates, when exim is a client,
2208 is now done by default. The transport option tls_verify_cert_hostnames
2209 can be used to disable this per-host. The build option
2210 EXPERIMENTAL_CERTNAMES is withdrawn.
2212 JH/05 The value of the tls_verify_certificates smtp transport and main options
2213 default to the word "system" to access the system default CA bundle.
2214 For GnuTLS, only version 3.0.20 or later.
2216 JH/06 Verification of the server certificate for a TLS connection is now tried
2217 (but not required) by default. The verification status is now logged by
2218 default, for both outbound TLS and client-certificate supplying inbound
2221 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2224 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2225 Status Notification (bounce) messages are now MIME format per RFC 3464.
2226 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2227 under the control of the dsn_advertise_hosts option, and routers may
2228 have a dsn_lasthop option.
2230 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2231 default, modifiable by a malware= option. The list separator for
2232 the options can now be changed in the usual way. Bug 68.
2234 JH/10 The smtp_receive_timeout main option is now expanded before use.
2236 JH/11 The incoming_interface log option now also enables logging of the
2237 local interface on delivery outgoing connections.
2239 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2240 if the interface and destination host and port all match.
2242 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2245 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2246 Patch from Andrew Lewis.
2248 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2249 now supports optional time-restrictions, weighting, and priority
2250 modifiers per server. Patch originally by <rommer@active.by>.
2252 JH/16 The spamd_address main option now supports a mixed list of local
2253 and remote servers. Remote servers can be IPv6 addresses, and
2254 specify a port-range.
2256 JH/17 Bug 68: The spamd_address main option now supports an optional
2257 timeout value per server.
2259 JH/18 Bug 1581: Router and transport options headers_add/remove can
2260 now have the list separator specified.
2262 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2265 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2268 JH/21 Support for the A6 type of dns record is withdrawn.
2270 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2271 rather than the verbs used.
2273 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2274 from 255 to 1024 chars.
2276 JH/24 Verification callouts now attempt to use TLS by default.
2278 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2279 are generic router options now. The defaults didn't change.
2281 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2282 Original patch from Alexander Shikoff, worked over by JH.
2284 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2285 files if ZCAT_COMMAND is not executable.
2287 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2289 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2291 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2292 Normally benign, it bites when the pair was led to by a CNAME;
2293 modern usage is to not canonicalize the domain to a CNAME target
2294 (and we were inconsistent anyway for A-only vs AAAA+A).
2296 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2298 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2299 when evaluating $sender_host_dnssec.
2301 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2302 $sender_helo_dnssec variable.
2304 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2306 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2308 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2310 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2311 documented as working, but never had. Support all but $spam_report.
2313 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2314 added for tls authenticator.
2316 HS/03 Add perl_taintmode main config option
2322 TL/01 When running the test suite, the README says that variables such as
2323 no_msglog_check are global and can be placed anywhere in a specific
2324 test's script, however it was observed that placement needed to be near
2325 the beginning for it to behave that way. Changed the runtest perl
2326 script to read through the entire script once to detect and set these
2327 variables, reset to the beginning of the script, and then run through
2328 the script parsing/test process like normal.
2330 TL/02 The BSD's have an arc4random API. One of the functions to induce
2331 adding randomness was arc4random_stir(), but it has been removed in
2332 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2333 function when detected.
2335 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2336 cause callback expansion.
2338 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2339 syntax errors in an expansion can be treated as a string instead of
2340 logging or causing an error, due to the internal use of bool_lax
2341 instead of bool when processing it.
2343 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2344 server certificates when making smtp deliveries.
2346 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2348 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2350 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2352 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2353 Merged patch from Sebastian Wiedenroth.
2355 JH/05 Fix results-pipe from transport process. Several recipients, combined
2356 with certificate use, exposed issues where response data items split
2357 over buffer boundaries were not parsed properly. This eventually
2358 resulted in duplicates being sent. This issue only became common enough
2359 to notice due to the introduction of connection certificate information,
2360 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2362 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2363 size buffer was used, resulting in syntax errors when an expansion
2366 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2367 version 3.3.6 or later.
2369 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2370 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2371 both become "event_action", the variables become $event_name, $event_data
2372 and $event_defer_errno. There is a new variable $verify_mode, usable in
2373 routers, transports and related events. The tls:cert event is now also
2374 raised for inbound connections, if the main configuration event_action
2377 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2378 early OCSP support, but no stapling (appears to be less than 1.0.0).
2380 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2381 server certificate names available under the smtp transport option
2382 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2385 JH/10 Time-related extraction expansions from certificates now use the main
2386 option "timezone" setting for output formatting, and are consistent
2387 between OpenSSL and GnuTLS compilations. Bug 1541.
2389 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2390 encoded parameter in the incoming message. Bug 1558.
2392 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2393 include certificate info, eximon was claiming there were spoolfile
2396 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2398 JH/14 Log delivery-related information more consistently, using the sequence
2399 "H=<name> [<ip>]" wherever possible.
2401 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2402 are problematic for Debian distribution, omit them from the release
2405 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2407 JH/16 Fix string representation of time values on 64bit time_t architectures.
2410 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2411 output list separator was used.
2416 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2417 checkers that were complaining about end of non-void function with no
2420 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2421 This was a regression introduced in 4.83 by another bugfix.
2423 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2425 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2426 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2432 TF/01 Correctly close the server side of TLS when forking for delivery.
2434 When a message was received over SMTP with TLS, Exim failed to clear up
2435 the incoming connection properly after forking off the child process to
2436 deliver the message. In some situations the subsequent outgoing
2437 delivery connection happened to have the same fd number as the incoming
2438 connection previously had. Exim would try to use TLS and fail, logging
2439 a "Bad file descriptor" error.
2441 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2442 utilities have not been installed.
2444 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2445 temporary space as the ACL may create new global variables.
2447 TL/01 LDAP support uses per connection or global context settings, depending
2448 upon the detected version of the libraries at build time.
2450 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2451 to extract and use the src ip:port in logging and expansions as if it
2452 were a direct connection from the outside internet. PPv2 support was
2453 updated based on HAProxy spec change in May 2014.
2455 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2457 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2458 Properly escape header and check for NULL return.
2460 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2463 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2465 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2466 characters in header names, implemented as a verify condition.
2467 Contributed by Michael Fischer v. Mollard.
2469 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2470 results permerror and temperror. Previous values are deprecated but
2471 still accepted. In a future release, err_perm and err_temp will be
2472 completely removed, which will be a backward incompatibility if the
2473 ACL tests for either of these two old results. Patch contributed by
2474 user bes-internal on the mailing list.
2476 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2478 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2479 selectors, in both main and reject logs.
2481 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2484 JH/07 Add malware type "sock" for talking to simple daemon.
2486 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2488 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2489 routers/transports under cutthrough routing.
2491 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2492 numbers. Touch up "bool" conditional to keep the same definition.
2494 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2496 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2498 JH/12 Expand items in router/transport headers_add or headers_remove lists
2499 individually rather than the list as a whole. Bug 1452.
2501 Required for reasonable handling of multiple headers_ options when
2502 they may be empty; requires that headers_remove items with embedded
2503 colons must have them doubled (or the list-separator changed).
2505 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2506 view the policy declared in the DMARC record. Currently, $dmarc_status
2507 is a combined value of both the record presence and the result of the
2510 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2512 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2513 dnslookup router and the smtp transport (applying to the forward
2516 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2517 of ldap servers used for a specific lookup. Patch provided by Heiko
2520 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2521 New variable $lookup_dnssec_authenticated for observability.
2523 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2524 Patch submitted by Lars Timman.
2526 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2528 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2529 Requires trusted mode and valid format message id, aborts otherwise.
2530 Patch contributed by Heiko Schlichting.
2532 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2533 certextract with support for various fields. Bug 1358.
2535 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2536 is requested by default, modifiable by smtp transport option
2539 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2540 operate on certificate variables to give certificate fingerprints
2541 Also new ${sha256:cert_variable}.
2543 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2545 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2546 Christian Aistleitner.
2548 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2550 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2551 file. Patch from Wolfgang Breyha.
2553 JH/25 Expand the coverage of the delivery $host and $host_address to
2554 client authenticators run in verify callout. Bug 1476.
2556 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2557 align with daemon_smtp_ports. Bug 72.
2559 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2560 support and error reporting did not work properly.
2562 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2563 and is readable. Patch from Andrew Colin Kissa.
2565 TL/14 Enhance documentation of ${run expansion and how it parses the
2566 commandline after expansion, particularly in the case when an
2567 unquoted variable expansion results in an empty value.
2569 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2571 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2572 by renaming away from it. Addresses GH issue 15, reported by
2575 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2577 TL/15 SECURITY: prevent double expansion in math comparison functions
2578 (can expand unsanitized data). Not remotely exploitable.
2585 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2587 PP/02 Make -n do something, by making it not do something.
2588 When combined with -bP, the name of an option is not output.
2590 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2593 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2594 $sender_host_name and config options to manage this, and basic check
2597 PP/05 DSCP support for outbound connections and control modifier for inbound.
2599 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2600 (Only plugin which currently uses this is kerberos4, which nobody should
2601 be using, but we should make it available and other future plugins might
2602 conceivably use it, even though it would break NAT; stuff *should* be
2603 using channel bindings instead).
2605 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2606 name; added for Sendmail compatibility; requires admin caller.
2607 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2608 just ignore it); requires trusted caller.
2609 Also parse but ignore: -Ac -Am -X<logfile>
2612 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2614 TL/02 Add +smtp_confirmation as a default logging option.
2616 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2617 Patch by Magnus Holmgren from 2007-02-20.
2619 TL/04 Bugzilla 1281 - Spec typo.
2620 Bugzilla 1283 - Spec typo.
2621 Bugzilla 1290 - Spec grammar fixes.
2623 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2625 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2627 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2628 mailing list by by Dmitry Isaikin.
2630 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2632 JH/02 Support "G" suffix to numbers in ${if comparisons.
2634 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2636 NM/01 Bugzilla 1197 - Spec typo
2637 Bugzilla 1196 - Spec examples corrections
2639 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2641 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2642 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2645 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2646 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2648 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2649 "acl {{name}{arg}...}", and optional args on acl condition
2652 JH/05 Permit multiple router/transport headers_add/remove lines.
2654 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2656 JH/07 Avoid using a waiting database for a single-message-only transport.
2657 Performance patch from Paul Fisher. Bugzilla 1262.
2659 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2662 JH/09 Add $headers_added variable, with content from use of ACL modifier
2663 add_header (but not yet added to the message). Bugzilla 199.
2665 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2666 Pulled from Bugzilla 817 by Wolfgang Breyha.
2668 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2670 (nb: this is the same fix as in Exim 4.80.1)
2672 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2675 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2678 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2680 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2681 advertise SMTP AUTH mechanism to us, instead of a generic
2682 protocol violation error. Also, make Exim more robust to bad
2683 data from the Dovecot auth socket.
2685 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2687 When a queue runner is handling a message, Exim first routes the
2688 recipient addresses, during which it prunes them based on the retry
2689 hints database. After that it attempts to deliver the message to
2690 any remaining recipients. It then updates the hints database using
2693 So if a recipient address works intermittently, it can get repeatedly
2694 deferred at routing time. The retry hints record remains fresh so the
2695 address never reaches the final cutoff time.
2697 This is a fairly common occurrence when a user is bumping up against
2698 their storage quota. Exim had some logic in its local delivery code
2699 to deal with this. However it did not apply to per-recipient defers
2700 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2702 This change adds a proper retry rule check during routing so that the
2703 final cutoff time is checked against the message's age. We only do
2704 this check if there is an address retry record and there is not a
2705 domain retry record; this implies that previous attempts to handle
2706 the address had the retry_use_local_parts option turned on. We use
2707 this as an approximation for the destination being like a local
2708 delivery, as in LMTP.
2710 I suspect this new check makes the old local delivery cutoff check
2711 redundant, but I have not verified this so I left the code in place.
2713 TF/02 Correct gecos expansion when From: is a prefix of the username.
2715 Test 0254 submits a message to Exim with the header
2719 When I ran the test suite under the user fanf2, Exim expanded
2720 the header to contain my full name, whereas it should have added
2721 a Resent-Sender: header. It erroneously treats any prefix of the
2722 username as equal to the username.
2724 This change corrects that bug.
2726 GF/01 DCC debug and logging tidyup
2727 Error conditions log to paniclog rather than rejectlog.
2728 Debug lines prefixed by "DCC: " to remove any ambiguity.
2730 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2732 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2733 Bug spotted by Jeremy Harris; was flawed since initial commit.
2734 Would have resulted in OCSP responses post-SNI triggering an Exim
2735 NULL dereference and crash.
2737 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2739 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2740 Bug detection, analysis and fix by Samuel Thibault.
2741 Bugzilla 1331, Debian bug #698092.
2743 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2745 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2746 Server implementation by Todd Lyons, client by JH.
2747 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2748 config variable "prdr_enable" controls whether the server
2749 advertises the facility. If the client requests PRDR a new
2750 acl_data_smtp_prdr ACL is called once for each recipient, after
2751 the body content is received and before the acl_smtp_data ACL.
2752 The client is controlled by both of: a hosts_try_prdr option
2753 on the smtp transport, and the server advertisement.
2754 Default client logging of deliveries and rejections involving
2755 PRDR are flagged with the string "PRDR".
2757 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2758 fclose(). Diagnosis by Todd Lyons.
2760 PP/17 Update configure.default to handle IPv6 localhost better.
2761 Patch by Alain Williams (plus minor tweaks).
2764 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2765 This is now consistent with GnuTLS, and is now documented: the
2766 previous undocumented portable approach to treating the option as
2767 unset was to force an expansion failure. That still works, and
2768 an empty string is now equivalent.
2770 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2771 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2772 not performing validation itself.
2774 PP/20 Added force_command boolean option to pipe transport.
2775 Patch from Nick Koston, of cPanel Inc.
2777 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2780 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2782 PP/21 Fix eximon continuous updating with timestamped log-files.
2783 Broken in a format-string cleanup in 4.80, missed when I repaired the
2784 other false fix of the same issue.
2785 Report and fix from Heiko Schlichting.
2788 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2789 Report from Prashanth Katuri.
2791 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2792 It's SecureTransport, so affects any MacOS clients which use the
2793 system-integrated TLS libraries, including email clients.
2795 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2796 using a MIME ACL for non-SMTP local injection.
2797 Report and assistance in diagnosis by Warren Baker.
2799 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2801 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2803 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2804 last id that failed so it may be referenced in subsequent ACL's.
2806 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2809 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2810 ldap library initialization, allowing self-signed CA's to be
2811 used. Also properly sets require_cert option later in code by
2812 using NULL (global ldap config) instead of ldap handle (per
2813 session). Bug diagnosis and testing by alxgomz.
2815 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2816 the src/util/ subdirectory.
2818 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2819 renamed to Transport Post Delivery Action by Jeremy Harris, as
2822 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2823 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2824 redis_servers = needs to be configured which will be used by the redis
2825 lookup. Patch from Warren Baker, of The Packet Hub.
2827 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2829 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2830 hostname or reverse DNS when processing a host list. Used suggestions
2831 from multiple comments on this bug.
2833 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2835 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2836 Missed a few lines, added it to make the runtest require no keyboard
2839 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2840 contains upper case chars. Make router use caseful_local_part.
2842 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2843 support when GnuTLS has been built with p11-kit.
2849 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2851 This, or similar/improved, will also be change PP/11 of 4.82.
2857 PP/01 Handle short writes when writing local log-files.
2858 In practice, only affects FreeBSD (8 onwards).
2859 Bugzilla 1053, with thanks to Dmitry Isaikin.
2861 NM/01 Bugzilla 949 - Documentation tweak
2863 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2866 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2868 PP/02 Implemented gsasl authenticator.
2870 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2872 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2873 `pkg-config foo` for cflags/libs.
2875 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2876 with rest of GSASL and with heimdal_gssapi.
2878 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2879 `pkg-config foo` for cflags/libs for the TLS implementation.
2881 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2882 properties get this fed in as external SSF. A number of robustness
2883 and debugging improvements to the cyrus_sasl authenticator.
2885 PP/08 cyrus_sasl server now expands the server_realm option.
2887 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2888 Patch by Jeremy Harris.
2890 PP/10 Added dbmjz lookup type.
2892 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2894 PP/12 MAIL args handles TAB as well as SP, for better interop with
2895 non-compliant senders.
2896 Analysis and variant patch by Todd Lyons.
2898 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2899 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2900 Patch from Dirk Mueller <dmueller@suse.com>
2902 PP/13 tls_peerdn now print-escaped for spool files.
2903 Observed some $tls_peerdn in wild which contained \n, which resulted
2904 in spool file corruption.
2906 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2907 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2908 or write after TLS renegotiation, which otherwise led to messages
2911 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2912 as a tracking header (ie: a signed header comes before the signature).
2913 Patch from Wolfgang Breyha.
2915 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2916 comma-sep list; embedded commas doubled.
2918 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2920 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2922 Report and patch from Dmitry Banschikov.
2924 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2925 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2926 failures. We appear to now support TLS1.1+ with Exim.
2928 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2929 lets Exim select keys and certificates based upon TLS SNI from client.
2930 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2931 before an outbound SMTP session. New log_selector, +tls_sni.
2933 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2934 NULL dereference. Report and patch from Alun Jones.
2936 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2937 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2938 Not seeing resolver debug output on NetBSD, but suspect this is a
2939 resolver implementation change.
2941 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2942 Left warnings. Added "eximon gdb" invocation mode.
2944 PP/21 Defaulting "accept_8bitmime" to true, not false.
2946 PP/22 Added -bw for inetd wait mode support.
2948 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2949 locate the relevant includes and libraries. Made this the default.
2951 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2952 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2954 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2955 This may cause build issues on older platforms.
2957 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2958 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2959 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2960 Added SNI support via GnuTLS too.
2961 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2963 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2965 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2966 Applied second patch from Janne, implementing suggestion to default
2967 multiple-strings-in-record handling to match SPF spec.
2969 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2971 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2972 read-only, out of scope).
2973 Patch from Wolfgang Breyha, report from Stuart Northfield.
2975 PP/29 Fix three issues highlighted by clang analyser static analysis.
2976 Only crash-plausible issue would require the Cambridge-specific
2977 iplookup router and a misconfiguration.
2978 Report from Marcin Mirosław.
2980 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2982 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2983 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2984 As part of this, removing so much warning spew let me fix some minor
2985 real issues in debug logging.
2987 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2988 assignment on my part. Fixed.
2990 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2991 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2992 Janne Snabb (who went above and beyond: thank you).
2994 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2995 string otherwise requires a connection and a bunch more work and it's
2996 relatively easy to get wrong. Should also expose TLS library linkage
2999 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3000 64-bit ${eval} (JH/03).
3002 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3003 GNU libc to support some of the 64-bit stuff, should not lead to
3004 conflicts. Defined before os.h is pulled in, so if a given platform
3005 needs to override this, it can.
3007 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3008 protection layer was required, which is not implemented.
3009 Bugzilla 1254, patch from Wolfgang Breyha.
3011 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3012 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3013 tls_dhparam take prime identifiers. Also unbreak combination of
3014 OpenSSL+DH_params+TLSSNI.
3016 PP/39 Disable SSLv2 by default in OpenSSL support.
3022 PP/01 Solaris build fix for Oracle's LDAP libraries.
3023 Bugzilla 1109, patch from Stephen Usher.
3025 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3027 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3030 TF/02 Fix a couple more cases where we did not log the error message
3031 when unlink() failed. See also change 4.74-TF/03.
3033 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3034 lock up or crash if it happened to be inside a call to libc when it
3035 got a SIGUSR1 from exiwhat.
3037 The SIGUSR1 handler appends the current process status to the process
3038 log which is later printed by exiwhat. It used to use the general
3039 purpose logging code to do this, but several functions it calls are
3040 not safe for signals.
3042 The new output code in the SIGUSR1 handler is specific to the process
3043 log, and simple enough that it's easy to inspect for signal safety.
3044 Removing some special cases also simplifies the general logging code.
3045 Removing the spurious timestamps from the process log simplifies
3048 TF/04 Improved ratelimit ACL condition.
3050 The /noupdate option has been deprecated in favour of /readonly which
3051 has clearer semantics. The /leaky, /strict, and /readonly update modes
3052 are mutually exclusive. The update mode is no longer included in the
3053 database key; it just determines when the database is updated. (This
3054 means that when you upgrade Exim will forget old rate measurements.)
3056 Exim now checks that the per_* options are used with an update mode that
3057 makes sense for the current ACL. For example, when Exim is processing a
3058 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3059 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3060 must specify per_mail/readonly. If you omit the update mode it defaults to
3061 /leaky where that makes sense (as before) or /readonly where required.
3063 The /noupdate option is now undocumented but still supported for
3064 backwards compatibility. It is equivalent to /readonly except that in
3065 ACLs where /readonly is required you may specify /leaky/noupdate or
3066 /strict/noupdate which are treated the same as /readonly.
3068 A useful new feature is the /count= option. This is a generalization
3069 of the per_byte option, so that you can measure the throughput of other
3070 aggregate values. For example, the per_byte option is now equivalent
3071 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3073 The per_rcpt option has been generalized using the /count= mechanism
3074 (though it's more complicated than the per_byte equivalence). When it is
3075 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3076 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3077 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3078 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3079 Note that using per_rcpt with a non-readonly update mode in more than
3080 one ACL will cause the recipients to be double-counted. (The per_mail
3081 and per_byte options don't have this problem.)
3083 The handling of very low rates has changed slightly. If the computed rate
3084 is less than the event's count (usually one) then this event is the first
3085 after a long gap. In this case the rate is set to the same as this event's
3086 count, so that the first message of a spam run is counted properly.
3088 The major new feature is a mechanism for counting the rate of unique
3089 events. The new per_addr option counts the number of different
3090 recipients that someone has sent messages to in the last time period. It
3091 behaves like per_rcpt if all the recipient addresses are different, but
3092 duplicate recipient addresses do not increase the measured rate. Like
3093 the /count= option this is a general mechanism, so the per_addr option
3094 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3095 example, measure the rate that a client uses different sender addresses
3096 with the options per_mail/unique=$sender_address. There are further
3097 details in the main documentation.
3099 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3101 TF/06 Removed a few PCRE remnants.
3103 TF/07 Automatically extract Exim's version number from tags in the git
3104 repository when doing development or release builds.
3106 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3107 Bugzilla 879. Patch from Paul Fisher.
3109 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3110 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3113 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3115 PP/05 Variable $av_failed, true if the AV scanner deferred.
3116 Bugzilla 1078. Patch from John Horne.
3118 PP/06 Stop make process more reliably on build failure.
3119 Bugzilla 1087. Patch from Heiko Schlittermann.
3121 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3122 Bugzilla 1089. Patch from Heiko Schlittermann.
3124 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3125 Bugzilla 1131. Patch from Holger Weiß.
3127 PP/09 Handle IPv6 addresses with SPF.
3128 Bugzilla 860. Patch from Wolfgang Breyha.
3130 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3132 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3135 PP/11 match_* no longer expand right-hand-side by default.
3136 New compile-time build option, EXPAND_LISTMATCH_RHS.
3137 New expansion conditions, "inlist", "inlisti".
3139 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3141 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3143 PP/14 fix log_write() format string regression from TF/03.
3144 Bugzilla 1152. Patch from Dmitry Isaikin.
3150 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3152 PP/02 Harmonised TLS library version reporting; only show if debugging.
3153 Layout now matches that introduced for other libraries in 4.74 PP/03.
3155 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3157 PP/04 New "dns_use_edns0" global option.
3159 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3162 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3163 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3165 TK/01 Updated PolarSSL code to 0.14.2.
3166 Bugzilla 1097. Patch from Andreas Metzler.
3168 PP/07 Catch divide-by-zero in ${eval:...}.
3169 Fixes bugzilla 1102.
3171 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3174 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3175 format-string attack -- SECURITY: remote arbitrary code execution.
3177 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3178 time unintentionally subject to list matching rules, letting the header
3179 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3180 arbitrary string expansion). This allowed for information disclosure.
3182 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3183 INT_MIN/-1 -- value coerced to INT_MAX.
3189 NM/01 Workaround for PCRE version dependency in version reporting
3192 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3193 This fixes portability to compilers other than gcc, notably
3194 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3196 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3197 makefiles for portability to HP-UX and POSIX correctness.
3199 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3200 Also via indented variable definition in the Makefile.
3201 (Debugging by Oliver Heesakkers).
3203 PP/02 Restore caching of spamd results with expanded spamd_address.
3204 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3206 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3207 Improves build reliability. Fix from: Frank Elsner
3209 NM/02 Fix wide character breakage in the rfc2047 coding
3210 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3212 NM/03 Allow underscore in dnslist lookups
3213 Fixes bug 1026. Patch from Graeme Fowler
3215 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3216 Code patches from Adam Ciarcinski of NetBSD.
3218 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3221 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3222 is logged, to avoid truncation. Patch from John Horne.
3224 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3225 Patch from Jakob Hirsch.
3227 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3228 SQL string expansion failure details.
3229 Patch from Andrey Oktyabrski.
3231 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3232 Patch from Simon Arlott.
3234 PP/09 New lookups functionality failed to compile on old gcc which rejects
3235 extern declarations in function scope.
3236 Patch from Oliver Fleischmann
3238 PP/10 Use sig_atomic_t for flags set from signal handlers.
3239 Check getgroups() return and improve debugging.
3240 Fixed developed for diagnosis in bug 927 (which turned out to be
3243 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3244 Patch from Mark Zealey.
3246 PP/12 Bugzilla 1056: Improved spamd server selection.
3247 Patch from Mark Zealey.
3249 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3250 Based on patch from Heiko Schlittermann.
3252 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3253 Patch from Uwe Doering, sign-off by Michael Haardt.
3255 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3256 variable declaration deep within a block. Bug and patch from
3259 PP/15 lookups-Makefile IRIX compatibility coercion.
3261 PP/16 Make DISABLE_DKIM build knob functional.
3263 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3264 Patch by Simon Arlott
3266 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3267 variable argument macros. Our copy now differs from upstream.
3273 TF/01 Failure to get a lock on a hints database can have serious
3274 consequences so log it to the panic log.
3276 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3277 controlled using the smtp_confirmation log selector.
3279 TF/03 Include the error message when we fail to unlink a spool file.
3281 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3282 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3283 for maintaining out-of-tree patches for some time.
3285 PP/01 Bugzilla 139: Documentation and portability issues.
3286 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3287 Handle per-OS dynamic-module compilation flags.
3289 PP/02 Let /dev/null have normal permissions.
3290 The 4.73 fixes were a little too stringent and complained about the
3291 permissions on /dev/null. Exempt it from some checks.
3292 Reported by Andreas M. Kirchwitz.
3294 PP/03 Report version information for many libraries, including
3295 Exim version information for dynamically loaded libraries. Created
3296 version.h, now support a version extension string for distributors
3297 who patch heavily. Dynamic module ABI change.
3299 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3300 privilege escalation vulnerability whereby the Exim run-time user
3301 can cause root to append content of the attacker's choosing to
3304 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3307 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3308 If dropping privileges for untrusted macros, we disabled normal logging
3309 on the basis that it would fail; for the Exim run-time user, this is not
3310 the case, and it resulted in successful deliveries going unlogged.
3311 Fixed. Reported by Andreas Metzler.
3317 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3318 only prepend for the Resent-* case. Fixes regression introduced in
3319 Exim 4.70 by NM/22 for Bugzilla 607.
3321 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3322 increasing numbers of administrators be bitten by this.
3324 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3326 PP/03 Bugzilla 994: added openssl_options main configuration option.
3328 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3330 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3332 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3334 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3335 without a peer certificate, leading to a segfault because of an
3336 assumption that peers always have certificates. Be a little more
3337 paranoid. Problem reported by Martin Tscholak.
3339 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3340 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3341 NB: ClamAV planning to remove STREAM in "middle of 2010".
3342 CL also introduces -bmalware, various -d+acl logging additions and
3343 more caution in buffer sizes.
3345 PP/09 Implemented reverse_ip expansion operator.
3347 PP/10 Bugzilla 937: provide a "debug" ACL control.
3349 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3351 PP/12 Bugzilla 973: Implement --version.
3353 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3355 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3357 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3359 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3360 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3361 ignore trailing whitespace.
3363 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3365 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3368 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3369 Notification from Dr Andrew Aitchison.
3371 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3372 ExtendedDetectionInfo response format.
3373 Notification from John Horne.
3375 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3378 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3379 XSL and documented dependency on system catalogs, with examples of how
3382 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3385 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3386 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3387 configuration file which is writeable by the Exim user or group.
3389 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3390 of configuration files to cover files specified with the -C option if
3391 they are going to be used with root privileges, not just the default
3394 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3395 option (effectively making it always true).
3397 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3398 files to be used while preserving root privileges.
3400 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3401 that rogue child processes cannot use them.
3403 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3404 run-time user, instead of root.
3406 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3407 Exim run-time user without dropping privileges.
3409 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3410 result string, instead of calling string_vformat() twice with the same
3413 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3414 for other users. Others should always drop root privileges if they use
3415 -C on the command line, even for a whitelisted configure file.
3417 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3419 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3425 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3426 $data_path, and $header_path variables; fixed documentation bugs and
3429 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3430 exipick to access non-standard spools, including the "frozen" queue
3433 NM/01 Bugzilla 965: Support mysql stored procedures.
3434 Patch from Alain Williams
3436 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3438 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3439 Patch from Andreas Metzler
3441 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3442 Patch from Kirill Miazine
3444 NM/05 Bugzilla 671: Added umask to procmail example.
3446 JJ/03 installed exipick 20100323.0, fixing doc bug
3448 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3449 directory. Notification and patch from Dan Rosenberg.
3451 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3453 TK/02 Improve log output when DKIM signing operation fails.
3455 MH/01 Treat the transport option dkim_domain as a colon separated
3456 list, not as a single string, and sign the message with each element,
3457 omitting multiple occurences of the same signer.
3459 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3460 Bugzilla 985, 986. Patch by Simon Arlott
3462 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3463 Patch by Simon Arlott
3465 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3466 MBX locking. Notification from Dan Rosenberg.
3472 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3474 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3476 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3478 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3480 NM/04 Bugzilla 915: Flag broken perl installation during build.
3486 TK/01 Added patch by Johannes Berg that expands the main option
3487 "spamd_address" if it starts with a dollar sign.
3489 TK/02 Write list of recipients to X-Envelope-Sender header when building
3490 the mbox-format spool file for content scanning (suggested by Jakob
3493 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3494 (http://www.dcc-servers.net/) support via dccifd. Activated by
3495 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3497 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3498 by Mark Daniel Reidel <mr@df.eu>.
3500 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3501 When building exim an external PCRE library is now needed -
3502 PCRE is a system library on the majority of modern systems.
3503 See entry on PCRE_LIBS in EDITME file.
3505 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3506 conversation. Added nologin parameter to request.
3507 Patch contributed by Kirill Miazine.
3509 TF/01 Do not log submission mode rewrites if they do not change the address.
3511 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3513 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3514 log files in place. Contributed by Roberto Lima.
3516 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3518 TF/03 Bugzilla 615: When checking the local_parts router precondition
3519 after a local_part_suffix or local_part_prefix option, Exim now
3520 does not use the address's named list lookup cache, since this
3521 contains cached lookups for the whole local part.
3523 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3524 Robert Millan. Documentation is in experimental-spec.txt.
3526 TF/04 Bugzilla 668: Fix parallel build (make -j).
3528 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3530 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3531 Patch provided by Jan Srzednicki.
3533 TF/05 Leading white space used to be stripped from $spam_report which
3534 wrecked the formatting. Now it is preserved.
3536 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3537 that they are available at delivery time.
3539 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3541 TF/08 TLS error reporting now respects the incoming_interface and
3542 incoming_port log selectors.
3544 TF/09 Produce a more useful error message if an SMTP transport's hosts
3545 setting expands to an empty string.
3547 NM/06 Bugzilla 744: EXPN did not work under TLS.
3548 Patch provided by Phil Pennock.
3550 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3551 Patch provided by Richard Godbee.
3553 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3554 acl_smtp_notquit, added index entry.
3556 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3557 Patch provided by Eugene Bujak.
3559 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3560 accept(). Patch provided by Maxim Dounin.
3562 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3563 Patch provided by Phil Pennock.
3565 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3567 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3568 Patch provided by Brad "anomie" Jorsch.
3570 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3571 Patch provided by Dean Brooks.
3573 TK/05 Add native DKIM support (does not depend on external libraries).
3575 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3576 Patch provided by Graeme Fowler.
3578 NM/16 Bugzilla 851: Documentation example syntax fix.
3580 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3582 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3585 NM/19 Bugzilla 745: TLS version reporting.
3586 Patch provided by Phil Pennock.
3588 NM/20 Bugzilla 167: bool: condition support.
3589 Patch provided by Phil Pennock.
3591 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3592 clients. Patch provided by Phil Pennock.
3594 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3595 Patch provided by Brad "anomie" Jorsch.
3597 NM/23 Bugzilla 687: Fix misparses in eximstats.
3598 Patch provided by Heiko Schlittermann.
3600 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3601 Patch provided by Heiko Schlittermann.
3603 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3604 plus update to original patch.
3606 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3608 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3609 Patch provided by David Brownlee.
3611 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3613 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3615 NM/30 Bugzilla 888: TLS documentation bugfixes.
3617 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3619 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3620 Unlike the original bugzilla I have changed all shell scripts in src tree.
3622 NM/33 Bugzilla 898: Transport filter timeout fix.
3623 Patch by Todd Rinaldo.
3625 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3626 Patch by Serge Demonchaux.
3628 NM/35 Bugzilla 39: Base64 decode bug fixes.
3629 Patch by Jakob Hirsch.
3631 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3633 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3635 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3637 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3643 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3644 ALT-N's libdkim that I have put here:
3645 http://duncanthrax.net/exim-experimental/
3647 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3648 were called 'true' and it seems that C99 defines that as a reserved
3649 keyword to be used with 'bool' variable types. That means you could
3650 not include C99-style headers which use bools without triggering
3651 build errors in sieve.c.
3653 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3654 as mailq or other aliases. Changed the --help handling significantly
3655 to do whats expected. exim_usage() emits usage/help information.
3657 SC/01 Added the -bylocaldomain option to eximstats.
3659 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3661 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3663 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3669 PH/01 Another patch from the Sieve maintainer.
3671 PH/02 When an IPv6 address is converted to a string for single-key lookup
3672 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3673 dots are used instead of colons so that keys in lsearch files need not
3674 contain colons. This was done some time before quoting was made available
3675 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3676 (notated using the quote facility) so as to distinguish them from IPv4
3677 keys. This meant that lookups for IP addresses in host lists did not work
3678 for iplsearch lookups.
3680 This has been fixed by arranging for IPv6 addresses to be expressed with
3681 colons if the lookup type is iplsearch. This is not incompatible, because
3682 previously such lookups could never work.
3684 The situation is now rather anomalous, since one *can* have colons in
3685 ordinary lsearch keys. However, making the change in all cases is
3686 incompatible and would probably break a number of configurations.
3688 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3691 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3692 conversion specification without a maximum field width, thereby enabling
3693 a rogue spamd server to cause a buffer overflow. While nobody in their
3694 right mind would setup Exim to query an untrusted spamd server, an
3695 attacker that gains access to a server running spamd could potentially
3696 exploit this vulnerability to run arbitrary code as the Exim user.
3698 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3699 $primary_hostname instead of what libspf2 thinks the hosts name is.
3701 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3702 a directory entry by the name of the lookup key. Previously, if a
3703 symlink pointed to a non-existing file or a file in a directory that
3704 Exim lacked permissions to read, a lookup for a key matching that
3705 symlink would fail. Now it is enough that a matching directory entry
3706 exists, symlink or not. (Bugzilla 503.)
3708 PH/03 The body_linecount and body_zerocount variables are now exported in the
3711 PH/04 Added the $dnslist_matched variable.
3713 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3714 This means they are set thereafter only if the connection becomes
3717 PH/06 Added the client_condition to authenticators so that some can be skipped
3718 by clients under certain conditions.
3720 PH/07 The error message for a badly-placed control=no_multiline_responses left
3721 "_responses" off the end of the name.
3723 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3725 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3726 (without spaces) instead of just copying the configuration text.
3728 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3730 PH/11 Added $max_received_linelength.
3732 PH/12 Added +ignore_defer and +include_defer to host lists.
3734 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3735 way in which PCRE > 7.0 is built.
3737 PH/14 Implemented queue_only_load_latch.
3739 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3740 MAIL command. The effect was to mangle the value on 64-bit systems.
3742 PH/16 Another patch from the Sieve maintainer.
3744 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3746 PH/18 If a system quota error occurred while trying to create the file for
3747 a maildir delivery, the message "Mailbox is full" was not appended to the
3748 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3749 only to a quota excession during the actual writing of the file.
3751 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3752 characters?) which causes problems in log lines. The DN values are now
3753 passed through string_printing() before being added to log lines.
3755 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3756 and InterBase are left for another time.)
3758 PH/21 Added message_body_newlines option.
3760 PH/22 Guard against possible overflow in moan_check_errorcopy().
3762 PH/23 POSIX allows open() to be a macro; guard against that.
3764 PH/24 If the recipient of an error message contained an @ in the local part
3765 (suitably quoted, of course), incorrect values were put in $domain and
3766 $local_part during the evaluation of errors_copy.
3772 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3773 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3776 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3777 issue a MAIL command.
3779 PH/02 In an ACL statement such as
3781 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3783 if a client was not listed at all, or was listed with a value other than
3784 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3785 the condition was not true (as it should be), so access was not denied.
3786 The bug was that the ! inversion was incorrectly passed on to the second
3787 item. This has been fixed.
3789 PH/03 Added additional dnslists conditions == and =& which are different from
3790 = and & when the dns lookup returns more than one IP address.
3792 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3793 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3795 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3796 FSYNC, which compiles an option called disable_fsync that allows for
3797 bypassing fsync(). The documentation is heavily laced with warnings.
3799 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3801 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3802 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3803 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3804 including adding "make clean"; (3) Added -fPIC when compiling the test
3805 dynamically loaded module, to get rid of a warning.
3807 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3808 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3809 = 0s. The bug is otherwise harmless.
3811 PH/07 There was a bug in the dovecot authenticator such that the value of
3812 $auth1 could be overwritten, and so not correctly preserved, after a
3813 successful authentication. This usually meant that the value preserved by
3814 the server_setid option was incorrect.
3816 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3818 PH/09 Installed PCRE release 7.0.
3820 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3821 run for batched SMTP input. It is now run at the start of every message
3822 in the batch. While fixing this I discovered that the process information
3823 (output by running exiwhat) was not always getting set for -bs and -bS
3824 input. This is fixed, and it now also says "batched" for BSMTP.
3826 PH/11 Added control=no_pipelining.
3828 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3829 patch, slightly modified), and move the expansion of helo_data till after
3830 the connection is made in the smtp transport (so it can use these
3833 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3835 PH/14 Added log_selector = +pid.
3837 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3839 PH/16 Add ${if forany and ${if forall.
3841 PH/17 Added dsn_from option to vary the From: line in DSNs.
3843 PH/18 Flush SMTP output before performing a callout, unless control =
3844 no_callout_flush is set.
3846 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3847 was true (the default) a successful delivery failed to delete the retry
3848 item, thus causing premature timeout of the address. The bug is now
3851 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3853 PH/21 Long custom messages for fakedefer and fakereject are now split up
3854 into multiline responses in the same way that messages for "deny" and
3855 other ACL rejections are.
3857 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3858 with slight modification.
3860 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3861 draft, changing the syntax and factoring some duplicate code".
3863 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3864 for deliveries of the second and subsequent messages over the same SMTP
3867 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3868 ${reduce, with only minor "tidies".
3870 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3872 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3873 expansion side effects.
3875 PH/27 When a message times out after an over-quota error from an Exim-imposed
3876 quota, the bounce message says "mailbox is full". This message was not
3877 being given when it was a system quota that was exceeded. It now should
3880 MH/03 Made $recipients available in local_scan(). local_scan() already has
3881 better access to the recipient list through recipients_list[], but
3882 $recipients can be useful in postmaster-provided expansion strings.
3884 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3885 in the case of a MAIL command with additional options following the
3886 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3887 were accidentally chopped off.
3889 PH/29 SMTP synchronization checks are implemented when a command is read -
3890 there is a check that no more input is waiting when there shouldn't be
3891 any. However, for some commands, a delay in an ACL can mean that it is
3892 some time before the response is written. In this time, more input might
3893 arrive, invalidly. So now there are extra checks after an ACL has run for
3894 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3895 pipelining has not been advertised.
3897 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3899 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3900 returned in the SMTP response when smtp_return_error_details was false.
3901 This has been fixed.
3903 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3904 instead of the C I/O that was originally supplied, because problems were
3905 reported on Solaris.
3907 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3908 Exim which did not show up earlier: it was assuming that a call to
3909 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3910 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3911 was a macro that became an assignment, so it seemed to work. This has
3912 changed to a proper function call with a void return, hence the compile
3913 error. Exim's code has been fixed.
3915 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3918 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3920 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3922 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3923 because it turns out to be too big or there is a timeout) there is no
3924 "Completed" line in the log. When some messages of this type were
3925 selected by exigrep, they were listed as "not completed". Others were
3926 picked up by some special patterns. I have improved the selection
3927 criteria to be more general.
3929 PH/38 The host_find_failed option in the manualroute router can now be set
3930 to "ignore", to completely ignore a host whose IP address cannot be
3931 found. If all hosts are ignored, the behaviour is controlled by the new
3932 host_all_ignored option.
3934 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3935 homing or because of multiple MX records with /mx) generated more than
3936 one IP address, and the following item turned out to be the local host,
3937 all the secondary addresses of the first item were incorrectly removed
3938 from the list, along with the local host and any following hosts (which
3939 is what is supposed to happen).
3941 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3942 whoever called Exim into the -H file. In the case of the daemon it was
3943 behaving confusingly. When first started, it used values for whoever
3944 started the daemon, but after a SIGHUP it used the Exim user (because it
3945 calls itself on a restart). I have changed the code so that it now always
3948 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3949 message are rejected with the same error (e.g. no authentication or bad
3950 sender address), and a DATA command is nevertheless sent (as can happen
3951 with PIPELINING or a stupid MUA), the error message that was given to the
3952 RCPT commands is included in the rejection of the DATA command. This is
3953 intended to be helpful for MUAs that show only the final error to their
3956 PH/42 Another patch from the Sieve maintainer.
3958 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3959 Eximstats - Fixed some broken HTML links and added missing column headers
3961 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3962 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3964 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3966 PH/43 Yet another patch from the Sieve maintainer.
3968 PH/44 I found a way to check for a TCP/IP connection going away before sending
3969 the response to the final '.' that terminates a message, but only in the
3970 case where the client has not sent further data following the '.'
3971 (unfortunately, this is allowed). However, in many cases there won't be
3972 any further data because there won't be any more messages to send. A call
3973 to select() can be used: if it shows that the input is "ready", there is
3974 either input waiting, or the socket has been closed. An attempt to read
3975 the next input character can distinguish the two cases. Previously, Exim
3976 would have sent an OK response which the client would never have see.
3977 This could lead to message repetition. This fix should cure that, at
3978 least in a lot of common cases.
3980 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3981 advertised in response to EHLO.
3987 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3988 fixed by 4.65/MH/01 (is this a record?) are fixed:
3990 (i) An empty string was always treated as zero by the numeric comparison
3991 operators. This behaviour has been restored.
3993 (ii) It is documented that the numeric comparison operators always treat
3994 their arguments as decimal numbers. This was broken in that numbers
3995 starting with 0 were being interpreted as octal.
3997 While fixing these problems I realized that there was another issue that
3998 hadn't been noticed. Values of message_size_limit (both the global option
3999 and the transport option) were treated as octal if they started with 0.
4000 The documentation was vague. These values are now always treated as
4001 decimal, and I will make that clear in the documentation.
4007 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4008 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4011 MH/01 Don't check that the operands of numeric comparison operators are
4012 integers when their expansion is in "skipping" mode (fixes bug
4013 introduced by 4.64-PH/07).
4015 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4016 child addresses, Exim now panics and dies. Previously, because the count
4017 is held in a short int, deliveries were likely to be lost. As such a
4018 large number of recipients for a single message is ridiculous
4019 (performance will be very, very poor), I have chosen to impose a limit
4020 rather than extend the field.
4026 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4027 leftover -K file (the existence of which was triggered by #402).
4028 While we were at it, introduced process PID as part of the -K
4029 filename. This should rule out race conditions when creating
4032 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4033 processing considerably. Previous code took too long for large mails,
4034 triggering a timeout which in turn triggers #401.
4036 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4037 in the DK code in transports.c. sendfile() is not really portable,
4038 hence the _LINUX specificness.
4040 TF/01 In the add_headers option to the mail command in an Exim filter,
4041 there was a bug that Exim would claim a syntax error in any
4042 header after the first one which had an odd number of characters
4045 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4046 callout verification, Exim cached a "reject" for the entire domain. This
4047 is correct for most verifications, but it is not correct for a recipient
4048 verification with use_sender or use_postmaster set, because in that case
4049 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4050 case of MAIL FROM:<> rejection from other early rejections (e.g.
4051 rejection of HELO). When verifying a recipient using a non-null MAIL
4052 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4053 Whatever the result of the callout, the value of the domain cache is
4054 left unchanged (for any other kind of callout, getting as far as trying
4055 RCPT means that the domain itself is ok).
4057 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4060 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4061 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4062 session, the daemon ignores ECONNECT errors and logs others; it now
4063 ignores EPIPE as well.
4065 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4066 (quoted-printable decoding).
4068 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4069 later the small subsequent patch to fix an introduced bug.
4071 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4073 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4075 PH/08 An error is now given if message_size_limit is specified negative.
4077 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4078 to be given (somewhat) arbitrary names.
4080 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4083 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4084 miscellaneous code fixes
4086 PH/10 Added the log_reject_target ACL modifier to specify where to log
4089 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4090 hostname. This is wrong, because it relates to the incoming message (and
4091 probably the interface on which it is arriving) and not to the outgoing
4092 callout (which could be using a different interface). This has been
4093 changed to use the value of the helo_data option from the smtp transport
4094 instead - this is what is used when a message is actually being sent. If
4095 there is no remote transport (possible with a router that sets up host
4096 addresses), $smtp_active_hostname is used.
4098 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4099 tweaks were necessary in order to get it to work (see also 21 below):
4100 (a) The code assumed that strncpy() returns a negative number on buffer
4101 overflow, which isn't the case. Replaced with Exim's string_format()
4103 (b) There were several signed/unsigned issues. I just did the minimum
4104 hacking in of casts. There is scope for a larger refactoring.
4105 (c) The code used strcasecmp() which is not a standard C function.
4106 Replaced with Exim's strcmpic() function.
4107 (d) The code set only $1; it now sets $auth1 as well.
4108 (e) A simple test gave the error "authentication client didn't specify
4109 service in request". It would seem that Dovecot has changed its
4110 interface. Fortunately there's a specification; I followed it and
4111 changed what the client sends and it appears to be working now.
4113 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4116 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4117 address A is aliased to B and C, where B exists and C does not. Without
4118 -v the output is "A verified" because verification stops after a
4119 successful redirection if more than one address is generated. However,
4120 with -v the child addresses are also verified. Exim was outputting "A
4121 failed to verify" and then showing the successful verification for C,
4122 with its parentage. It now outputs "B failed to verify", showing B's
4123 parentage before showing the successful verification of C.
4125 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4126 look up a TXT record in a specific list after matching in a combined
4129 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4130 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4131 they consult the DNS. I had assumed they would set it the way they
4132 wanted; and indeed my experiments on Linux seem to show that in some
4133 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4134 To be on the safe side, however, I have now made the interface to
4135 host_find_byname() similar to host_find_bydns(), with an argument
4136 containing the DNS resolver options. The host_find_byname() function now
4137 sets these options at its start, just as host_find_bydns() does. The smtp
4138 transport options dns_qualify_single and dns_search_parents are passed to
4139 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4140 of host_find_byname() use the default settings of RES_DEFNAMES
4141 (qualify_single) but not RES_DNSRCH (search_parents).
4143 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4144 spool_read_header() do less string testing, by means of a preliminary
4145 switch on the second character of optional "-foo" lines. (This is
4146 overdue, caused by the large number of possibilities that now exist.
4147 Originally there were few.) While I was there, I also converted the
4148 str(n)cmp tests so they don't re-test the leading "-" and the first
4149 character, in the hope this might squeeze out yet more improvement.
4151 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4152 flag allowing group syntax was set by the header_syntax check but not
4153 turned off, possible causing trouble later; (2) The flag was not being
4154 set at all for the header_verify test, causing "group"-style headers to
4155 be rejected. I have now set it in this case, and also caused header_
4156 verify to ignore an empty address taken from a group. While doing this, I
4157 came across some other cases where the code for allowing group syntax
4158 while scanning a header line wasn't quite right (mostly, not resetting
4159 the flag correctly in the right place). These bugs could have caused
4160 trouble for malformed header lines. I hope it is now all correct.
4162 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4163 with the "reply" argument non-NULL. The code, however (which originally
4164 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4165 but it didn't always do it. This confused somebody who was copying the
4166 code for some other use. I have removed all the tests.
4168 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4169 feature that was used to support insecure browsers during the U.S. crypto
4170 embargo. It requires special client support, and Exim is probably the
4171 only MTA that supported it -- and would never use it because real RSA is
4172 always available. This code has been removed, because it had the bad
4173 effect of slowing Exim down by computing (never used) parameters for the
4174 RSA_EXPORT functionality.
4176 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4177 authenticator to fail if there's a tab character in the incoming data
4178 (there should never be unless someone is messing about, as it's supposed
4179 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4180 if the connection is using TLS or if the remote IP is the same as the
4181 local IP, and the "valid-client-cert option" if a client certificate has
4184 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4185 authenticators. This can be used for authorization after authentication
4186 succeeds. (In the case of plaintext, it servers for both authentication
4189 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4190 if any retry times were supplied.
4192 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4193 connection, where there is no client IP address to check. In this
4194 situation, the verify now always succeeds.
4196 PH/25 Applied John Jetmore's -Mset patch.
4198 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4200 PH/27 In a string expansion for a processed (not raw) header when multiple
4201 headers of the same name were present, leading whitespace was being
4202 removed from all of them, but trailing whitespace was being removed only
4203 from the last one. Now trailing whitespace is removed from each header
4204 before concatenation. Completely empty headers in a concatenation (as
4205 before) are ignored.
4207 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4208 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4210 PH/29 [Removed. This was a change that I later backed out, and forgot to
4211 correct the ChangeLog entry (that I had efficiently created) before
4212 committing the later change.]
4214 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4215 address errors (4xx response to RCPT) over the same connection as other
4216 messages routed to the same hosts. Such deliveries are always "forced",
4217 so retry times are not inspected. This resulted in far too many retries
4218 for the affected addresses. The effect occurred only when there were more
4219 hosts than the hosts_max_try setting in the smtp transport when it had
4220 the 4xx errors. Those hosts that it had tried were not added to the list
4221 of hosts for which the message was waiting, so if all were tried, there
4222 was no problem. Two fixes have been applied:
4224 (i) If there are any address or message errors in an SMTP delivery, none
4225 of the hosts (tried or untried) are now added to the list of hosts
4226 for which the message is waiting, so the message should not be a
4227 candidate for sending over the same connection that was used for a
4228 successful delivery of some other message. This seems entirely
4229 reasonable: after all the message is NOT "waiting for some host".
4230 This is so "obvious" that I'm not sure why it wasn't done
4231 previously. Hope I haven't missed anything, but it can't do any
4232 harm, as the worst effect is to miss an optimization.
4234 (ii) If, despite (i), such a delivery is accidentally attempted, the
4235 routing retry time is respected, so at least it doesn't keep
4236 hammering the server.
4238 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4239 in ${readsocket because some servers need this prod.
4241 PH/32 Added some extra debug output when updating a wait-xxx database.
4243 PH/33 The hint "could be header name not terminated by colon", which has been
4244 given for certain expansion errors for a long time, was not being given
4245 for the ${if def:h_colon_omitted{... case.
4247 PH/34 The spec says: "With one important exception, whenever a domain list is
4248 being scanned, $domain contains the subject domain." There was at least
4249 one case where this was not true.
4251 PH/35 The error "getsockname() failed: connection reset by peer" was being
4252 written to the panic log as well as the main log, but it isn't really
4253 panic-worthy as it just means the connection died rather early on. I have
4254 removed the panic log writing for the ECONNRESET error when getsockname()
4257 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4258 runs only) independently of the message's sender address. This meant
4259 that, if the 4xx error was in fact related to the sender, a different
4260 message to the same recipient with a different sender could confuse
4261 things. In particular, this can happen when sending to a greylisting
4262 server, but other circumstances could also provoke similar problems.
4263 I have changed the default so that the retry time for these errors is now
4264 based a combination of the sender and recipient addresses. This change
4265 can be overridden by setting address_retry_include_sender=false in the
4268 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4269 remote server are returned as part of bounce messages. This was not
4270 happening for LMTP over a pipe (the lmtp transport), but now it is the
4271 same for both kinds of LMTP.
4273 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4274 in header lines that were in fact CNAMEs. This is no longer the case.
4276 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4277 and queue runs started by the daemon processed all messages. This has
4278 been fixed so that -R and -S can now usefully be given with -q<time>.
4280 PH/40 Import PCRE release 6.7 (fixes some bugs).
4282 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4284 PH/42 Give an error if -q is specified more than once.
4286 PH/43 Renamed the variables $interface_address and $interface_port as
4287 $received_ip_address and $received_port, to make it clear that these
4288 values apply to message reception, and not to the outgoing interface when
4289 a message is delivered. (The old names remain recognized, of course.)
4291 PH/44 There was no timeout on the connect() call when using a Unix domain
4292 socket in the ${readsocket expansion. There now is.
4294 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4295 be meaningful with "accept".
4297 SC/01 Eximstats V1.43
4298 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4300 SC/02 Eximstats V1.44
4301 Use a glob alias rather than an array ref in the generated
4302 parser. This improves both readability and performance.
4304 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4305 Collect SpamAssassin and rejection statistics.
4306 Don't display local sender or destination tables unless
4307 there is data to show.
4308 Added average volumes into the top table text output.
4310 SC/04 Eximstats V1.46
4311 Collect data on the number of addresses (recipients)
4312 as well as the number of messages.
4314 SC/05 Eximstats V1.47
4315 Added 'Message too big' to the list of mail rejection
4316 reasons (thanks to Marco Gaiarin).
4318 SC/06 Eximstats V1.48
4319 Mainlog lines which have GMT offsets and are too short to
4320 have a flag are now skipped.
4322 SC/07 Eximstats V1.49 (Alain Williams)
4323 Added the -emptyok flag.
4325 SC/08 Eximstats V1.50
4326 Fixes for obtaining the IP address from reject messages.
4328 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4329 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4330 whitespace changes from 4.64-PH/27
4332 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4335 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4336 are found, allow negative numbers in numeric criteria)
4338 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4340 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4341 to match changes made in 4.64-PH/43
4343 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4345 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4346 "domain" is actually an IP address, and force a failure. This locks out
4347 those revolvers/nameservers that support "A-for-A" lookups, in
4348 contravention of the specifications.
4350 PH/48 When a host name was looked up from an IP address, and the subsequent
4351 forward lookup of the name timed out, the host name was left in
4352 $sender_host_name, contrary to the specification.
4354 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4355 restricted to single-key lookups, Exim was not diagnosing an error if
4356 * or *@ was used with a query-style lookup.
4358 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4360 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4361 long ago, but noone interested enough thought of it. Let's just say that
4362 the "1.1" means that there are some new functions that weren't there at
4363 some point in the past.
4365 PH/51 Error processing for expansion failure of helo_data from an smtp
4366 transport during callout processing was broken.
4368 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4369 tested/used via the -bh/-bhc/-bs options.
4371 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4372 bug, fixed in subsequent PCRE releases).
4374 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4375 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4377 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4383 SC/01 Use a glob alias rather than an array ref in eximstats generated
4384 parser. This improves both readability and performance.
4386 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4387 Don't display local sender or destination tables in eximstats unless
4388 there is data to show.
4389 Added average volumes into the eximstats top table text output.
4391 SC/03 Collect data on the number of addresses (recipients) as well
4392 as the number of messages in eximstats.
4394 TF/01 Correct an error in the documentation for the redirect router. Exim
4395 does (usually) call initgroups() when daemonizing.
4397 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4398 with consistent privilege compared to when running as a daemon.
4400 TF/03 Note in the spec that $authenticated_id is not set for local
4401 submissions from trusted users.
4403 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4404 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4406 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4407 by adding some example configuration directives to the default
4408 configuration file. A little bit of work is required to uncomment the
4409 directives and define how usernames and passwords are checked, but
4410 there is now a framework to start from.
4412 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4413 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4414 without this. I don't know how relevant this is to other LDAP libraries.
4416 PH/02 Add the verb name to the "unknown ACL verb" error.
4418 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4420 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4422 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4423 directory not expanded when it should be if an expanded home directory
4424 was set for the address (which is overridden by the transport).
4426 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4429 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4430 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4431 because it is too late at that time, and has no effect.
4433 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4434 security issue with \' (bugzilla #107). I could not use the
4435 PQescapeStringConn() function, because it needs a PGconn value as one of
4438 PH/08 When testing addresses using -bt, indicate those final addresses that
4439 are duplicates that would not cause an additional delivery. At least one
4440 person was confused, thinking that -bt output corresponded to deliveries.
4441 (Suppressing duplicates isn't a good idea as you lose the information
4442 about possibly different redirections that led to the duplicates.)
4444 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4445 systems where poll() doesn't work, in particular OS X.
4447 PH/10 Added more information to debugging output for retry time not reached.
4449 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4450 operations in malware.c.
4452 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4455 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4456 syslog_duplication set false, log lines that would normally be written
4457 both the the main log and to the reject log were not written to syslog at
4460 PH/14 In the default configuration, change the use of "message" in ACL warn
4461 statements to "add_header".
4463 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4464 not followed by a command (e.g. "seen endif").
4466 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4467 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4470 PH/17 Added extra conditions to the default value of delay_warning_condition
4474 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4475 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4476 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4479 The Auto-Submitted: and various List- headers are standardised, whereas I
4480 don't think Precedence: ever was.
4482 PH/18 Refactored debugging code in route_finduser() to show more information,
4483 in particular, the error code if getpwnam() issues one.
4485 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4486 This is apparently needed in addition to the PH/07 change above to avoid
4487 any possible encoding problems.
4489 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4490 but not after initializing Perl.
4492 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4493 output them only if debugging. By default they are written stderr,
4494 apparently, which is not desirable.
4496 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4499 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4502 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4504 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4505 authenticated or an ident call has been made. Suppress the default
4506 values for $authenticated_id and $authenticated_sender (but permit -oMai
4507 and -oMas) when testing with -bh.
4509 PH/24 Re-jigged the order of the tests in the default configuration so that the
4510 tests for valid domains and recipients precede the DNS black list and CSA
4511 tests, on the grounds that those ones are more expensive.
4513 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4514 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4515 This bug exists in every version of Exim that I still have, right back to
4518 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4519 However, an attempt to turn on case-sensitivity in a regex key by
4520 including (?-i) didn't work because the subject string was already
4521 lowercased, and the effects were non-intuitive. It turns out that a
4522 one-line patch can be used to allow (?-i) to work as expected.
4528 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4529 other effects) broke the use of negated acl sub-conditions.
4531 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4534 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4535 "Deny" causes Exim to reject the incoming connection with a 554 error.
4536 Unfortunately, if there is a major crisis, such as a disk failure,
4537 tcp-wrappers gives "deny", whereas what one would like would be some
4538 kind of temporary error. A kludge has been added to help with this.
4539 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4540 554 error is used if errno is still zero or contains ENOENT (which occurs
4541 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4544 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4546 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4547 errors. Otherwise a message that provokes a temporary error (when other
4548 messages do not) can cause a whole host to time out.
4550 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4551 addresses were routed directly to files or pipes from a redirect router.
4552 File deliveries just didn't batch; pipe deliveries might have suffered
4555 PH/06 A failure to get a lock for a hints database would erroneously always say
4556 "Failed to get write lock", even when it was really a read lock.
4558 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4559 of 0600. This has been changed to use the value of the lockfile_mode
4560 option (which defaults to 0600).
4562 PH/08 Applied small patch from the Sieve maintainer.
4564 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4565 folder from quota calculations, a direct delivery into this folder messed
4566 up the contents of the maildirsize file. This was because the regex was
4567 used only to exclude .Trash (or whatever) when the size of the mailbox
4568 was calculated. There was no check that a delivery was happening into an
4569 excluded directory. This bug has been fixed by ignoring all quota
4570 processing for deliveries into excluded directories.
4572 PH/10 Added the maildirfolder_create_regex option to appendfile.
4578 PH/01 The code for finding all the local interface addresses on a FreeBSD
4579 system running IPv6 was broken. This may well have applied to all BSD
4580 systems, as well as to others that have similar system calls. The broken
4581 code found IPv4 interfaces correctly, but gave incorrect values for the
4582 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4583 that it would not match correctly against @[] and not recognize the IPv6
4586 PH/02 The ipliteral router was not recognizing addresses of the form user@
4587 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4589 PH/03 Added disable_ipv6.
4591 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4592 decoded form, because it is most often used to construct To: headers
4593 lines in autoreplies, and the decoded form may well be syntactically
4594 invalid. However, $reply_address has leading white space removed, and all
4595 newlines turned into spaces so that the autoreply transport does not
4598 PH/05 If group was specified without a user on a router, and no group or user
4599 was specified on a transport, the group from the router was ignored.
4601 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4602 for visible compile-time settings that can be used to change these
4603 numbers, for those that want even more. Backwards compatibility with old
4604 spool files has been maintained. However, going back to a previous Exim
4605 release will lost any variables that are in spool files.
4607 PH/07 Two small changes when running in the test harness: increase delay when
4608 passing a TCP/IP connection to a new process, in case the original
4609 process has to generate a bounce, and remove special handling of
4610 127.0.0.2 (sic), which is no longer necessary.
4612 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4613 be the same on different OS.
4615 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4618 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4619 whether --show-vars was specified or not
4621 JJ/02 exipick: Added support for new ACL variable spool format introduced
4624 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4625 syntactically invalid From: or Reply-to: line, and a filter used this to
4626 generate an autoreply, and therefore failed to obtain an address for the
4627 autoreply, Exim could try to deliver to a non-existent relative file
4628 name, causing unrelated and misleading errors. What now happens is that
4629 it logs this as a hard delivery error, but does not attempt to create a
4632 PH/11 The exinext utility has a -C option for testing purposes, but although
4633 the given file was scanned by exinext itself; it wasn't being passed on
4634 when Exim was called.
4636 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4637 an end-of-file indication when reading a command response.
4639 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4640 compiled. In many other places in Exim, IPv6 addresses are always
4641 recognized, so I have changed this. It also means that IPv4 domain
4642 literals of the form [IPV4:n.n.n.n] are now always recognized.
4644 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4645 used if the router is not running as root, for example, when verifying at
4646 ACL time, or when using -bh. The debugging output from this situation was
4647 non-existent - all you got was a failure to exec. I have made two
4650 (a) Failures to set uid/gid, the current directory, or a process leader
4651 in a subprocess such as that created by queryprogram now generate
4652 suitable debugging output when -d is set.
4654 (b) The queryprogram router detects when it is not running as root,
4655 outputs suitable debugging information if -d is set, and then runs
4656 the subprocess without attempting to change uid/gid.
4658 PH/15 Minor change to Makefile for building test_host (undocumented testing
4661 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4662 additional section of a DNS packet that returns MX or SRV records.
4663 Instead, it always explicitly searches for A/AAAA records. This avoids
4664 major problems that occur when a DNS server includes only records of one
4665 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4666 fixed another bug: if SRV records were looked up and the corresponding
4667 address records were *not* found in the additional section, the port
4668 values from the SRV records were lost.
4670 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4671 using the correct key (the original address) when searching the retry
4672 rules in order to find which one to use for generating the retry hint.
4674 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4675 adding the default one. Similarly, if it contains a Reply-To: header, the
4676 errors_reply_to option, if set, is not used.
4678 PH/19 When calculating a retry time, Exim used to measure the "time since
4679 failure" by looking at the "first failed" field in the retry record. Now
4680 it does not use this if it is later than than the arrival time of the
4681 message. Instead it uses the arrival time. This makes for better
4682 behaviour in cases where some deliveries succeed, thus re-setting the
4683 "first failed" field. An example is a quota failure for a huge message
4684 when small messages continue to be delivered. Without this change, the
4685 "time since failure" will always be short, possible causing more frequent
4686 delivery attempts for the huge message than are intended.
4687 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4689 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4690 $1, $2, $3) because the numerical variables can be reset during some
4691 expansion items (e.g. "match"), thereby losing the authentication data.
4693 PH/21 Make -bV show the size of off_t variables so that the test suite can
4694 decide whether to run tests for quotas > 2G.
4696 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4697 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4698 filecount value is greater than 2G or if a quota value is greater than 2G
4699 on a system where the size of off_t is not greater than 4, a panic error
4702 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4703 never match. The debug and -bh output now contains an explicit error
4704 message indicating a malformed IPv4 address or mask.
4706 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4707 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4708 PH/23 above applies.
4710 PH/25 Do not write to syslog when running in the test harness. The only
4711 occasion when this arises is a failure to open the main or panic logs
4712 (for which there is an explicit test).
4714 PH/26 Added the /no_tell option to "control=freeze".
4716 PH/27 If a host name lookup failed very early in a connection, for example, if
4717 the IP address matched host_lookup and the reverse lookup yielded a name
4718 that did not have a forward lookup, an error message of the form "no IP
4719 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4720 could be logged. Now it outputs the IP address instead of "NULL".
4722 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4723 allows the sender and the authenticated sender to be set when
4724 submitting a message from within Exim. Since child_open_exim() is
4725 documented for local_scan(), the new function should be too.
4727 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4728 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4729 results in an empty string is now treated as unset.
4731 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4733 PH/31 Added sender_verify_fail logging option.
4735 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4736 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4737 tidied the source and removed it altogether.
4739 PH/33 When a queue run was abandoned because the load average was too high, a
4740 log line was always written; now it is written only if the queue_run log
4741 selector is set. In addition, the log line for abandonment now contains
4742 information about the queue run such as the pid. This is always present
4743 in "start" and "stop" lines but was omitted from the "abandon" line.
4745 PH/34 Omit spaces between a header name and the colon in the error message that
4746 is given when verify = headers_syntax fails (if there are lots of them,
4747 the message gets confusing).
4749 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4750 names, as there are now some PTR records that contain slashes. This check
4751 is only to protect against broken name servers that fall over on strange
4752 characters, so the fact that it applies to all lookups doesn't matter.
4754 PH/36 Now that the new test suite is complete, we can remove some of the
4755 special code in Exim that was needed for the old test suite. For example,
4756 sorting DNS records because real resolvers return them in an arbitrary
4757 order. The new test suite's fake resolver always returns records in the
4760 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4761 bounces) except when queue_only is set, to avoid logging races between
4762 the different processes.
4764 PH/38 Panic-die if .include specifies a non-absolute path.
4766 PH/39 A tweak to the "H" retry rule from its user.
4768 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4769 a label. They prevented compilation on older perls.
4771 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4772 a warning to be raised on newish perls.
4774 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4775 on queue. Changes to match documented behaviour of showing count of
4776 messages matching specified criteria.
4778 PH/40 Changed the default ident timeout from 30s to 5s.
4780 PH/41 Added support for the use of login_cap features, on those BSD systems
4781 that have them, for controlling the resources used by pipe deliveries.
4783 PH/42 The content-scanning code uses fopen() to create files in which to put
4784 message data. Previously it was not paying any attention to the mode of
4785 the files. Exim runs with umask(0) because the rest of the code creates
4786 files with open(), and sets the required mode explicitly. Thus, these
4787 files were ending up world-writeable. This was not a big issue, because,
4788 being within the spool directory, they were not world-accessible. I have
4789 created a function called modefopen, which takes an additional mode
4790 argument. It sets umask(777), creates the file, chmods it to the required
4791 mode, then resets the umask. All the relevant calls to fopen() in the
4792 content scanning code have been changed to use this function.
4794 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4795 to 24 hours. This avoids potential overflow problems when processing G
4796 and H retry rules. I suspect nobody ever tinkers with this value.
4798 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4800 PH/45 When the plaintext authenticator is running as a client, the server's
4801 challenges are checked to ensure they are valid base64 strings. By
4802 default, the authentication attempt is cancelled if an invalid string is
4803 received. Setting client_ignore_invalid_base64 true ignores these errors.
4804 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4805 they are received. Thus, the responses can be made to depend on the
4806 challenges. If an invalid string is ignored, an empty string is placed in
4809 PH/46 Messages that are created by the autoreply transport now contains a
4810 References: header, in accordance with RFCs 2822 and 3834.
4812 PH/47 Added authenticated_sender_force to the smtp transport.
4814 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4816 PH/49 Installed latest patch from the Sieve maintainer.
4818 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4819 was also set, the appendfile transport was unnecessarily scanning a
4820 directory of message files (e.g. for maildir delivery) to find the count
4821 of files (along with the size), even though it did not need this
4822 information. It now does the scan only if it needs to find either the
4823 size of the count of files.
4825 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4827 PH/52 Two bugs concerned with error handling when the smtp transport is
4830 (i) Exim was not creating retry information for temporary errors given
4831 for individual recipients after the DATA command when the smtp transport
4832 was used in LMTP mode. This meant that they could be retried too
4833 frequently, and not timed out correctly.
4835 (ii) Exim was setting the flag that allows error details to be returned
4836 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4837 recipients that were returned after the DATA command.
4839 PH/53 This is related to PH/52, but is more general: for any failing address,
4840 when detailed error information was permitted to be returned to the
4841 sender, but the error was temporary, then after the final timeout, only
4842 "retry timeout exceeded" was returned. Now it returns the full error as
4843 well as "retry timeout exceeded".
4845 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4846 do this, and (what is worse) MTAs that accept it.
4848 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4849 will now be deprecated.
4851 PH/56 New os.c-cygwin from the Cygwin maintainer.
4853 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4854 formats (previously only available in exim formats via -bpr, -bpru,
4855 and -bpra. Now also available in native and exiqgrep formats)
4857 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4858 with very large, slow to parse queues
4860 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4862 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4864 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4865 responses to authentication challenges, though it was showing the
4866 challenges; (ii) I've removed the CR characters from the debug output for
4869 PH/58 Allow for the insertion of a newline as well as a space when a string
4870 is turned into more than one encoded-word during RFC 2047 encoding. The
4871 Sieve code now uses this.
4873 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4874 data_4xx, lost_connection, tls_required.
4876 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4877 message was being sent as an SMTP response.
4879 PH/61 Add -l and -k options to exicyclog.
4881 PH/62 When verifying, if an address was redirected to one new address, so that
4882 verification continued, and the new address failed or deferred after
4883 having set something in $address_data, the value of $address_data was not
4884 passed back to the ACL. This was different to the case when no
4885 redirection occurred. The value is now passed back in both cases.
4887 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4888 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4889 use login_cap.h, so on its own it isn't the distinguishing feature. The
4890 new name refers directly to the setclassresources() function.
4892 PH/65 Added configuration files for NetBSD3.
4894 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4896 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4897 is preferred over IPv4.
4899 PH/68 The bounce_return_message and bounce_return_body options were not being
4900 honoured for bounces generated during the reception of non-SMTP messages.
4901 In particular, this applied to messages rejected by the ACL. This bug has
4902 been fixed. However, if bounce_return_message is true and bounce_return_
4903 body is false, the headers that are returned for a non-SMTP message
4904 include only those that have been read before the error was detected.
4905 (In the case of an ACL rejection, they have all been read.)
4907 PH/69 The HTML version of the specification is now built in a directory called
4908 spec_html instead of spec.html, because the latter looks like a path with
4909 a MIME-type, and this confuses some software.
4911 PH/70 Catch two compiler warnings in sieve.c.
4913 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4914 function verify_get_ident() calls ip_connect() to connect a socket, but
4915 if the "connect()" function timed out, ip_connect() used to close the
4916 socket. However, verify_get_ident() also closes the socket later, and in
4917 between Exim writes to the log, which may get opened at this point. When
4918 the socket was closed in ip_connect(), the log could get the same file
4919 descriptor number as the socket. This naturally causes chaos. The fix is
4920 not to close the socket in ip_connect(); the socket should be closed by
4921 the function that creates it. There was only one place in the code where
4922 this was missing, in the iplookup router, which I don't think anybody now
4923 uses, but I've fixed it anyway.
4925 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4926 well as to direct DNS lookups. Otherwise the handling of names in host
4927 lists is inconsistent and therefore confusing.
4933 PH/01 Two changes to the default runtime configuration:
4935 (1) Move the checks for relay_from_hosts and authenticated clients from
4936 after to before the (commented out) DNS black list checks.
4938 (2) Add control=submission to the relay_from_hosts and authenticated
4939 clients checks, on the grounds that messages accepted by these
4940 statements are most likely to be submissions.
4942 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4944 (1) Generate an error if the third argument for the ${prvs expansion is
4947 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4950 (3) Reset the variables that are obtained from the first argument of
4951 ${prvscheck and used in the second argument before leaving the code,
4952 because their memory is reclaimed, so using them afterwards may do
4955 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4956 one (it's much easier than Tom thought :-).
4958 (5) Because of (4), we can now allow for the use of $prvscheck_result
4959 inside the third argument.
4961 PH/03 For some reason, the default setting of PATH when running a command from
4962 a pipe transport was just "/usr/bin". I have changed it to
4965 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4966 anything to be listed in the output from -bV.
4968 PH/05 When a filter generated an autoreply, the entire To: header line was
4969 quoted in the delivery log line, like this:
4971 => >A.N.Other <ano@some.domain> <original@ddress> ...
4973 This has been changed so that it extracts the operative address. There
4974 may be more than one such address. If so, they are comma-separated, like
4977 => >ano@some.domain,ona@other.domain <original@ddress> ...
4979 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4980 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4981 not being looked up in the rDNS to get a host name, Exim was showing the
4982 IP address twice in Received: lines, even though the IP addresses were
4983 identical. For example:
4985 Received: from [1.2.3.4] (helo=[1.2.3.4])
4987 However, if the real host name was known, it was omitting the HELO data
4988 if it matched the actual IP address. This has been tidied up so that it
4989 doesn't show the same IP address twice.
4991 PH/07 When both +timestamp and +memory debugging was on, the value given by
4992 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4993 called by the debug printing, thereby overwriting the timestamp buffer.
4994 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4996 PH/08 When the original message was included in an autoreply transport, it
4997 always said "this is a copy of the message, including all the headers",
4998 even if body_only or headers_only was set. It now gives an appropriate
5001 PH/09 Applied a patch from the Sieve maintainer which:
5003 o fixes some comments
5004 o adds the (disabled) notify extension core
5005 o adds some debug output for the result of if/elsif tests
5006 o points to the current vacation draft in the documentation
5007 and documents the missing references header update
5011 o fixes a bug in processing the envelope test (when testing
5012 multiple envelope elements, the last element determined the
5015 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5016 Electronic Mail") by including:
5018 Auto-submitted: auto-generated
5020 in the messages that it generates (bounce messages and others, such as
5021 warnings). In the case of bounce messages for non-SMTP messages, there was
5022 also a typo: it was using "Auto_submitted" (underscore instead of
5023 hyphen). Since every message generated by Exim is necessarily in response
5024 to another message, thes have all been changed to:
5026 Auto-Submitted: auto-replied
5028 in accordance with these statements in the RFC:
5030 The auto-replied keyword:
5032 - SHOULD be used on messages sent in direct response to another
5033 message by an automatic process,
5035 - MUST NOT be used on manually-generated messages,
5037 - MAY be used on Delivery Status Notifications (DSNs) and Message
5038 Disposition Notifications (MDNs),
5040 - MUST NOT be used on messages generated by automatic or periodic
5041 processes, except for messages which are automatic responses to
5044 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5045 to the default Received: header definition.
5047 PH/12 Added log selector acl_warn_skipped (default on).
5049 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5050 variables because (a) they are in the wrong storage pool and (b) even if
5051 they were copied, it wouldn't work properly because of the caching.
5053 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5054 checking when decoding. Apparently there are clients that generate
5055 overlong encoded strings. Why am I not surprised?
5057 PH/15 If the first argument of "${if match_address" was not empty, but did not
5058 contain an "@" character, Exim crashed. Now it writes a panic log message
5059 and treats the condition as false.
5061 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5063 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5064 extension "envelope-auth". The code is finished and in agreement with
5065 other implementations, but there is no documentation so far and in fact,
5066 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5067 not changing the active code.
5069 Print executed "if" and "elsif" statements when debugging is used. This
5070 helps a great deal to understand what a filter does.
5072 Document more things not specified clearly in RFC3028. I had all this
5073 sorted out, when out of a sudden new issues came to my mind. Oops."
5075 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5078 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5079 canonical form (as documented). However, after a host name lookup from
5080 the IP address, check_host() was doing a simple string comparison with
5081 addresses acquired from the DNS when checking that the found name did
5082 have the original IP as one of its addresses. Since any found IPv6
5083 addresses are likely to be in abbreviated form, the comparison could
5084 fail. Luckily, there already exists a function for doing the comparison
5085 by converting both addresses to binary, so now that is used instead of
5086 the text comparison.
5088 PH/20 There was another similar case to PH/19, when a complete host name was
5089 given in a host list; looking up its IP address could give an abbreviated
5090 form, whereas the current host's name might or might not be abbreviated.
5091 The same fix has been applied.
5097 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5098 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5101 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5102 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5104 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5106 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5107 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5108 Also, for "Auto-Submitted", treat anything other than "no" as
5109 non-personal, in accordance with RFC 3834. (Previously it treated
5110 anything starting "auto-" as non-personal.)
5112 TF/01 The control=submission/name=... option had a problem with syntax
5113 errors if the name included a slash character. The /name= option
5114 now slurps the rest of the string, so it can include any characters
5115 but it must come last in the list of options (after /sender_retain
5118 PH/05 Some modifications to the interface to the fake nameserver for the new
5126 TK/01 Added the "success_on_redirect" address verification option. See
5127 NewStuff for rationale and an example.
5129 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5131 PH/02 Patch to exigrep to allow it to work on syslog lines.
5133 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5134 fread() to skip over the body file's header line, because in Cygwin the
5135 header line is locked and is inaccessible.
5137 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5138 co-exist for some time) to make it clear that it is the Exim ID that is
5139 referenced, not the Message-ID: header line.
5141 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5142 string_format() function, because snprintf() does not exist on all
5145 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5146 ${stat: expansion item.
5148 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5149 protocol synchronization error", to keep the pedants happy.
5151 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5152 well as for IRIX systems, when gcc is being used. See the host.c source
5155 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5157 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5160 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5161 transport and to the smtp transport in LMTP mode.
5163 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5165 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5166 run a filter in a subprocess. This could lead to confusion in subsequent
5167 lookups in the parent process. There should also be a search_tidyup() at
5168 the end of the subprocess.
5170 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5171 only if the host matched helo_try_verify_hosts, which caused the
5172 verification to occur when the EHLO/HELO command was issued. The ACL just
5173 tested the remembered result. Now, if a previous verification attempt has
5174 not happened, "verify = helo" does it there and then.
5176 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5178 TK/03 Fix log output including CR from clamd.
5180 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5181 exist provoked a memory error which could cause a segfault.
5183 PH/15 Installed PCRE 6.2
5185 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5187 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5188 of the problem. Specifically, suggested +O2 rather than +O1 for the
5191 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5193 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5194 record, the port was not correctly logged when the outgoing_port log
5195 selector was set (it logged the transort's default port).
5197 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5198 fallback_hosts, and "hosts" in the smtp transport.
5200 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5201 host errors such as "Connection refused".
5203 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5204 authentication with radiusclient 0.4.9:
5206 - Error returned from rc_read_config was caught wrongly
5207 - Username/password not passed on to radius server due to wrong length.
5209 The presumption is that some radiusclient API changes for 4.51/PH/17
5210 were not taken care of correctly. The code is still untested by me (my
5211 Linux distribution still has 0.3.2 of radiusclient), but it was
5212 contributed by a Radius user.
5214 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5215 expanding the "port" option of the smtp transport.
5217 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5218 while reading a MIME header. Thanks to Tom Hughes for a patch.
5220 PH/24 Include config.h inside local_scan.h so that configuration settings are
5223 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5224 that in an ACL for RCPT (for example), you can examine exactly what was
5227 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5228 commands, but it was not correctly comparing the address with the actual
5229 client host address. Thus, it would show the EHLO address in Received:
5230 header lines when this was not necessary.
5232 PH/27 Added the % operator to ${eval:}.
5234 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5235 it should be ignoring failures (because with -C, for example, it has lost
5236 privilege). It wasn't ignoring creation failures other than "already
5239 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5242 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5243 because an input line was too long, either on its own, or by virtue of
5244 too many continuations, the temporary file was not being removed, and the
5245 return code was incorrect.
5247 PH/31 Missing "BOOL" in function definition in filtertest.c.
5249 PH/32 Applied Sieve patches from the maintainer.
5251 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5253 PH/33 Added "verify = not_blind".
5255 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5256 Local/Makefile (with some defaults set). These are used in built scripts
5257 such as exicyclog, but they have never been used in the exim_install
5258 script (though there are many overriding facilities there). I have
5259 arranged that the exim_install script now takes note of these two
5262 PH/35 Installed configuration files for Dragonfly.
5264 PH/36 When a locally submitted message by a trusted user did not contain a
5265 From: header, and the sender address was obtained from -f or from an SMTP
5266 MAIL command, and the trusted user did not use -F to supply a sender
5267 name, $originator_name was incorrectly used when constructing a From:
5268 header. Furthermore, $originator_name was used for submission mode
5269 messages from external hosts without From: headers in a similar way,
5270 which is clearly wrong.
5272 PH/37 Added control=suppress_local_fixups.
5274 PH/38 When log_selector = +received_sender was set, and the addition of the
5275 sender made the log line's construction buffer exactly full, or one byte
5276 less than full, an overflow happened when the terminating "\n" was
5279 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5280 when the result of a list match is failure because a DNS lookup failed.
5282 PH/40 RM_COMMAND is now used in the building process.
5284 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5285 the "build-* directories that it finds.
5287 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5288 address in a domain literal was a prefix of an interface address.
5290 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5291 when verifying a sender address, unless rewrite_headers is false.
5293 PH/44 Wrote a long comment about why errors_to addresses are verified as
5294 recipients, not senders.
5296 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5297 the ratelimit ACL was added.
5299 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5301 PH/46 Added extra information about PostgreSQL errors to the error string.
5303 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5304 avoiding the need to install special zones in a real server. This is
5305 backwards compatible; if it can't find the fake resolver, it drops back.
5306 Thus, both old and new test suites can be run.
5308 TF/02 Added util/ratelimit.pl
5310 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5311 clock is set back in time.
5313 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5314 Candler <B.Candler@pobox.com>.
5316 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5317 OK for addresses that are the result of redirections.
5319 PH/48 A number of further additions for the benefit of the new test suite,
5320 including a fake gethostbyname() that interfaces to the fake DNS resolver
5323 TF/06 The fix for widen_domains has also been applied to qualify_single and
5324 search_parents which are the other dnslookup options that can cause
5327 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5330 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5332 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5333 service" if it encountered an MX record with an empty target hostname.
5334 The message is now "an MX or SRV record indicated no SMTP service".
5336 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5337 if the DNS of the sending site is misconfigured. This is quite a
5338 common situation. This change restores the behaviour of treating a
5339 helo verification defer as a failure.
5341 PH/51 If self=fail was set on a router, the bounce message did not include the
5342 actual error message.
5348 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5350 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5351 command itself ended in error, the underlying message about the transport
5352 filter timeout was being overwritten with the pipe command error. Now the
5353 underlying error message should be appended to the second error message.
5355 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5357 PH/02 Reduce the amount of output that "make" produces by default. Full output
5358 can still be requested.
5360 PH/03 The warning log line about a condition test deferring for a "warn" verb
5361 was being output only once per connection, rather than after each
5362 occurrence (because it was using the same function as for successful
5363 "warn" verbs). This seems wrong, so I have changed it.
5365 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5366 it should not have, which might have caused a crash in the right
5367 circumstances, but probably never did.
5369 PH/04 Installed a modified version of Tony Finch's patch to make submission
5370 mode fix the return path as well as the Sender: header line, and to
5371 add a /name= option so that you can make the user's friendly name appear
5374 TF/03 Added the control = fakedefer ACL modifier.
5376 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5377 Mark Lowes for thorough testing.
5379 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5381 TK/03 Merged latest SRS patch from Miles Wilton.
5383 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5384 with the definition in sysexits.h (which is #included earlier).
5385 Fortunately, Exim does not actually use EX_OK. The code used to try to
5386 preserve the sysexits.h value, by assuming that macro definitions were
5387 scanned for macro replacements. I have been disabused of this notion,
5388 so now the code just undefines EX_OK before #including unistd.h.
5390 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5391 in the smtp transport. When a block could not be written in a single
5392 write() function, the timeout was being re-applied to each part-write.
5393 This seems wrong - if the receiver was accepting one byte at a time it
5394 would take for ever. The timeout is now adjusted when this happens. It
5395 doesn't have to be particularly precise.
5397 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5398 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5400 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5401 without a domain if the check to <postmaster@domain> fails.
5403 SC/01 Eximstats: added -xls and the ability to specify output files
5404 (patch written by Frank Heydlauf).
5406 SC/02 Eximstats: use FileHandles for outputting results.
5408 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5410 SC/04 Eximstats: fixed display of large numbers with -nvr option
5412 SC/05 Eximstats: fixed merging of reports with empty tables.
5414 SC/06 Eximstats: added the -include_original_destination flag
5416 SC/07 Eximstats: removed tabs and trailing whitespace.
5418 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5420 TK/06 MBOX spool code: Add real "From " MBOX separator line
5421 so the .eml file is really in mbox format (even though
5422 most programs do not really care). Patch from Alex Miller.
5424 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5425 The latter is generated from $received_to and is only set if the
5426 message has one envelope recipient. SA can use these headers,
5427 obviously out-of-the-box. Patch from Alex Miller.
5429 PH/08 The ${def test on a variable was returning false if the variable's
5430 value was "0", contrary to what the specification has always said!
5431 The result should be true unless the variable is empty.
5433 PH/09 The syntax error of a character other than { following "${if
5434 def:variable_name" (after optional whitespace) was not being diagnosed.
5435 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5436 accidental colon was present, for example, could give incorrect results.
5438 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5439 result is used (not including appendfile, where other changes are about
5442 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5443 This involved changing a lot of size variables from int to off_t. It
5444 should work with maildirs and everything.
5446 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5447 spamd dying while we are connected to it.
5449 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5452 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5453 function for BDB 4.3.
5455 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5457 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5458 the sources for PCRE have been re-arranged and more files are now
5461 PH/15 The code I had for printing potentially long long variables in PH/11
5462 above was not the best (it lost precision). The length of off_t variables
5463 is now inspected at build time, and an appropriate printing format (%ld
5464 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5465 to be "long long int" or "long int". This is needed for the internal
5466 formatting function string_vformat().
5468 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5469 the configuration file to be ":syslog", then the script "guesses" where
5470 the logs files are, rather than using the compiled in default. In our
5471 case the guess is not the same as the compiled default, so the script
5472 suddenly stopped working when I started to use syslog. The patch checks
5473 to see if log_file_path is "". If so, it attempts to read it from exim
5474 with no configuration file to get the compiled in version, before it
5475 falls back to the previous guessing code."
5477 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5478 implementing BATV in an Exim configuration. See NewStuff for the gory
5481 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5482 Makefile that are specific to HP-UX.
5484 PH/18 If the "use_postmaster" option was set for a recipient callout together
5485 with the "random" option, the postmaster address was used as the MAIL
5486 FROM address for the random test, but not for the subsequent recipient
5487 test. It is now used for both.
5489 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5490 patch removes a few documentation additions to RFC 3028, because the
5491 latest draft now contains them. It adds the new en;ascii-case comparator
5492 and a new error check for 8bit text in MIME parts. Comparator and
5493 require names are now matched exactly. I enabled the subaddress
5494 extension, but it is not well tested yet (read: it works for me)."
5496 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5497 rework some of the code of TK/09 above to avoid the hardwired use of
5498 "%lld" and "long long". Replaced the call to snprintf() with a call to
5501 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5502 records point to non-existent hosts", "retry timeout exceeded", and
5503 "retry time not reached for any host after a long failure period".
5505 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5506 experimental DomainKeys support:
5508 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5509 (2) On an error such as an illegally used "control", the wrong name for
5510 the control was given.
5512 These problems did NOT occur unless DomainKeys support was compiled.
5514 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5516 PH/24 Added ${if match_ip condition.
5518 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5519 databases so that it will be absolutely obvious if a crash occurs in the
5520 DB library. This is a regular occurrence (often caused by mis-matched
5523 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5524 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5525 picked up on a user's system that detects such things. There doesn't seem
5526 to be a gcc warning option for this - only an attribute that has to be
5527 put on the function's prototype. It seems that in Fedora Core 4 they have
5528 set this on a number of new functions. No doubt there will be more in due
5531 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5532 specify a transport. However, if an address that was verified by such a
5533 router was the subject of a callout, Exim crashed because it tried to
5534 read the rcpt_include_affixes from the non-existent transport. Now it
5535 just assumes that the setting of that option is false. This bug was
5536 introduced by 4.51/PH/31.
5538 PH/28 Changed -d+all to exclude +memory, because that information is very
5539 rarely of interest, but it makes the output a lot bigger. People tend to
5540 do -d+all out of habit.
5542 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5543 code in os-type was giving problems when libc.so lives in lib64, like on
5546 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5547 aren't the modern standard, and it seems that some systems' include files
5548 don't always have them. Exim was already checking for some of the newer
5549 ones like T_AAAA, and defining it itself. I've added checks for all the
5550 record types that Exim uses.
5552 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5553 not automatically generating a new one, as it is supposed to. This
5554 prevented TLS from working. If the file did exist, but contained invalid
5555 data, a new version was generated, as expected. It was only the case of a
5556 non-existent file that was broken.
5558 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5559 with a change in libdomainkeys > 0.64.
5561 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5562 from DNS. If the selector record carries the flag, it now has
5563 precedence over the domain-wide flag.
5565 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5567 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5568 the use of a port name that isn't defined in /etc/services) occurred, the
5569 message was deferred as in a normal delivery, and thus remained on the
5570 spool, instead of being failed because of the mua_wrapper setting. This
5571 is now fixed, and I tidied up some of the mua_wrapper messages at the
5574 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5575 the messages in a hash of arrays rather than using individual hashes.
5576 This is a bit cleaner and results in dramatic memory savings, albeit
5577 at a slight CPU cost.
5579 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5580 as requested by Marc Sherman.
5582 SC/10 Eximstats: added histograms for user specified patterns as requested
5585 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5587 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5588 fopen() in the content-scanning modules that did not already have it.
5594 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5595 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5597 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5599 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5601 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5602 to test Sieve filters that use "vacation".
5604 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5605 that changes the way the GnuTLS parameters are stored in the cache file.
5606 The new format can be generated externally. For backward compatibility,
5607 if the data in the cache doesn't make sense, Exim assumes it has read an
5608 old-format file, and it generates new data and writes a new file. This
5609 means that you can't go back to an older release without removing the
5612 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5613 work if there are any delivery delays because "one_time" forces the
5614 parent to be marked "delivered", so its unseen clone is never tried
5615 again. For this reason, Exim now forbids the simultaneous setting of
5618 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5619 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5620 entry if you want to know the details. The fix, however, neglected to
5621 consider the case when local delivery batching is involved. The test for
5622 "previously delivered" was not happening when checking to see if an
5623 address could be batched with a previous (undelivered) one; under
5624 certain circumstances this could lead to multiple deliveries to the same
5627 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5628 in its include files, and this causes problems building Exim.
5630 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5631 header_syntax) but Exim was just ignoring anything given after a slash.
5632 In particular, this caused confusion with an attempt to use "verify =
5633 reverse_host_lookup/defer_ok". An error is now given when options are
5634 supplied for verify items that do not have them. (Maybe reverse_host_
5635 lookup should have a defer_ok option, but that's a different point.)
5637 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5638 defined by RFC 821) to 2048, because there were problems with some AUTH
5639 commands, and RFC 1869 says the size should be increased for extended
5640 SMTP commands that take arguments.
5642 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5645 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5646 "unknown" error; now it says that the functionality isn't in the binary.
5648 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5649 an address' error message when a string expansion fails (syntax or
5650 whatever). Otherwise the password may appear in the log. Following change
5651 PH/42 below, there is no longer a chance of it appearing in a bounce
5654 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5656 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5657 of its addresses was ever tried. (Bugzilla bug #2.)
5659 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5660 the result incorrectly in the debug output. (It correctly added a newline
5661 to what was transported.)
5663 TF/01 Added $received_time.
5665 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5666 commented out examples of how to interface to a virus scanner and to
5667 SpamAssassin. Also added commented examples of av_scanner and
5668 spamd_address settings.
5670 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5671 and controls are allowed in which ACLs. There were a couple of minor
5672 errors. Some of the entries in the conditions table (which is a table of
5673 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5674 negation of where the condition IS allowed.
5676 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5678 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5679 header file does not have a version number, so I've had to invent a new
5680 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5681 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5682 radiusclient), but it was contributed by a Radius user.
5684 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5685 files or directories, for OpenSSL.
5687 PH/19 When an Exim process that is running as root has to create an Exim log
5688 file, it does so in a subprocess that runs as exim:exim so as to get the
5689 ownership right at creation (otherwise, other Exim processes might see
5690 the file with the wrong ownership). There was no test for failure of this
5691 fork() call, which would lead to the process getting stuck as it waited
5692 for a non-existent subprocess. Forks do occasionally fail when resources
5693 run out. I reviewed all the other calls to fork(); they all seem to check
5696 PH/20 When checking for unexpected SMTP input at connect time (before writing
5697 the banner), Exim was not dealing correctly with a non-positive return
5698 from the read() function. If the client had disconnected by this time,
5699 the result was a log entry for a synchronization error with an empty
5700 string after "input=" when read() returned zero. If read() returned -1
5701 (an event I could not check), uninitialized data bytes were printed.
5702 There were reports of junk text (parts of files, etc) appearing after
5705 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5707 PH/22 Added support for macro redefinition, and (re)definition in between
5708 driver and ACL definitions.
5710 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5711 forgetting to use the resulting value; it was using the unexpanded value.
5713 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5714 hadn't been configured. The fix is from Juergen Kreileder, who
5715 understands it better than I do:
5717 "Here's what I see happening with three configured cyrus_sasl
5718 authenticators configured (plain, login, cram-md5):
5720 On startup auth_cyrus_sasl_init() gets called for each of these.
5721 This means three calls to sasl_listmech() without a specified mech_list.
5722 => SASL tests which mechs of all available mechs actually work
5723 => three warnings about OTP not working
5724 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5726 With the patch, sasl_listmech() also gets called three times. But now
5727 SASL's mech_list option is set to the server_mech specified in the the
5728 authenticator. Or in other words, the answer from sasl_listmech()
5729 gets limited to just the mech you're testing for (which is different
5731 => the return list contains just 'plain' or 'login', 'cram-md5' or
5732 nothing depending on the value of ob->server_mech.
5734 I've just tested the patch: Authentication still works fine,
5735 unavailable mechs specified in the exim configuration are still
5736 caught, and the auth.log warnings about OTP are gone."
5738 PH/25 When debugging is enabled, the contents of the command line are added
5739 to the debugging output, even when log_selector=+arguments is not
5742 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5743 answer is "GNU", and only if the return is "GNU/something" is the answer
5746 PH/27 $acl_verify_message is now set immediately after the failure of a
5747 verification in an ACL, and so is available in subsequent modifiers. In
5748 particular, the message can be preserved by coding like this:
5750 warn !verify = sender
5751 set acl_m0 = $acl_verify_message
5753 Previously, $acl_verify_message was set only while expanding "message"
5754 and "log_message" when a very denied access.
5756 PH/28 Modified OS/os.c-Linux with
5758 -#ifndef OS_LOAD_AVERAGE
5759 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5761 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5762 nomenclature these days.)
5764 PH/29 Installed patch from the Sieve maintainer that adds the options
5765 sieve_useraddress and sieve_subaddress to the redirect router.
5767 PH/30 In these circumstances:
5768 . Two addresses routed to the same list of hosts;
5769 . First host does not offer TLS;
5770 . First host accepts first address;
5771 . First host gives temporary error to second address;
5772 . Second host offers TLS and a TLS session is established;
5773 . Second host accepts second address.
5774 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5775 and peerdn, if requested) that were in fact used only for the second
5778 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5779 attention to any local part prefix or suffix that was matched by the
5780 router that accepted the address. It now behaves in the same way as it
5781 does for delivery: the affixes are removed from the local part unless
5782 rcpt_include_affixes is set on the transport.
5784 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5785 timeout during the DATA phase of an incoming message.
5787 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5788 applied a patch sanctioned by the Sieve maintainer.
5790 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5791 the uid or gid is negative. A case of a negative gid caused this to be
5792 noticed. The fix allows for either to be negative.
5794 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5795 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5798 PH/36 The change PH/12 above was broken. Fixed it.
5800 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5801 the grounds that routing the same address twice would always produce the
5802 same answer. This might have been true once, but it is certainly no
5803 longer true now. Routing a child address may depend on the previous
5804 routing that produced that child. Some complicated redirection strategies
5805 went wrong when messages had multiple recipients, and made Exim's
5806 behaviour dependent on the order in which the addresses were given.
5808 I have moved the duplicate checking until after the routing is complete.
5809 Exim scans the addresses that are assigned to local and remote
5810 transports, and removes any duplicates. This means that more work will be
5811 done, as duplicates will always all be routed, but duplicates are
5812 presumably rare, so I don't expect this is of any significance.
5814 For deliveries to pipes, files, and autoreplies, the duplicate checking
5815 still happens during the routing process, since they are not going to be
5818 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5819 It corrects a timeout issue with spamd. This is Ian's comment: "The
5820 background is that sometimes spamd either never reads data from a
5821 connection it has accepted, or it never writes response data. The exiscan
5822 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5823 blindly assumes that writes won't block so it may never time out."
5825 PH/39 Allow G after quota size as well as K and M.
5827 PH/40 The value set for $authenticated_id in an authenticator may not contain
5828 binary zeroes or newlines because the value is written to log lines and
5829 to spool files. There was no check on this. Now the value is run through
5830 the string_printing() function so that such characters are converted to
5831 printable escape sequences.
5833 PH/41 $message_linecount is a new variable that contains the total number of
5834 lines in the message. Compare $body_linecount, which is the count for the
5837 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5838 bounce and delay warning messages, except in certain special cases, which
5841 (a) An SMTP error message from a remote host;
5842 (b) A message specified in a :fail: redirection;
5843 (c) A message specified in a "fail" command in a system filter;
5844 (d) A message specified in a FAIL return from the queryprogram router;
5845 (e) A message specified by the cannot_route_message router option.
5847 In these cases only, Exim does include the error details in bounce and
5848 warning messages. There are also a few cases where bland messages such
5849 as "unrouteable address" or "local delivery error" are given.
5851 PH/43 $value is now also set for the "else" part of a ${run expansion.
5853 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5854 being worked on, but at least Exim now implements the latest version to
5857 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5858 process to complete was treated as a delivery failure, a timeout while
5859 writing the message to the pipe was logged, but erroneously treated as a
5860 successful delivery. Such timeouts include transport filter timeouts. For
5861 consistency with the overall process timeout, these timeouts are now
5862 treated as errors, giving rise to delivery failures by default. However,
5863 there is now a new Boolean option for the pipe transport called
5864 timeout_defer, which, if set TRUE, converts the failures into defers for
5865 both kinds of timeout. A transport filter timeout is now identified in
5868 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5869 systems where "make" and "gmake" are different, calling "gmake" at top
5870 level broke things. I've arranged for the value of $(MAKE) to be passed
5871 from the Makefile to this script so that it can call the same version of
5875 A note about Exim versions 4.44 and 4.50
5876 ----------------------------------------
5878 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5879 changes of various kinds. As a consequence, a big documentation update was
5880 needed. This delayed the release for rather longer than seemed good, especially
5881 in the light of a couple of (minor) security issues. Therefore, the changes
5882 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5883 release. So 4.44 and 4.50 are in effect two different branches that both start
5886 I have left the 4.50 change log unchanged; it contains all the changes since
5887 4.43. The change log for 4.44 is below; many of its items are identical to
5888 those for 4.50. This seems to be the most sensible way to preserve the
5889 historical information.
5895 1. Minor wording change to the doc/README.SIEVE file.
5897 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5898 computation of the current number of files was incorrect.
5900 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5901 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5904 4. Give more explanation in the error message when the command for a transport
5905 filter fails to execute.
5907 5. There are several places where Exim runs a non-Exim command in a
5908 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5909 was being done only for the command run by the queryprogram router. It is
5910 now done for all such subprocesses. The other cases are: ${run, transport
5911 filters, and the commands run by the lmtp and pipe transports.
5913 6. Added CONFIGURE_GROUP build-time option.
5915 7. Some older OS have a limit of 256 on the maximum number of file
5916 descriptors. Exim was using setrlimit() to set 1000 as a large value
5917 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5918 systems. I've change it so that if it can't get 1000, it tries for 256.
5920 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5921 was an oversight, and furthermore, ever since the addition of extra
5922 controls (e.g. 4.43/32), the checks on when to allow different forms of
5923 "control" were broken. There should now be diagnostics for all cases when a
5924 control that does not make sense is encountered.
5926 9. Added the /retain_sender option to "control=submission".
5928 10. $recipients is now available in the predata ACL (oversight).
5930 11. Tidy the search cache before the fork to do a delivery from a message
5931 received from the command line. Otherwise the child will trigger a lookup
5932 failure and thereby defer the delivery if it tries to use (for example) a
5933 cached ldap connection that the parent has called unbind on.
5935 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5936 of $address_data from the recipient verification was clobbered by the
5937 sender verification.
5939 13. The value of address_data from a sender verification is now available in
5940 $sender_address_data in subsequent conditions in the ACL statement.
5942 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5944 15. Added a new option "connect=<time>" to callout options, to set a different
5947 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5948 was its contents. (It was OK if the option was not defined at all.)
5950 17. A "Completed" log line is now written for messages that are removed from
5951 the spool by the -Mrm option.
5953 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5954 information about exactly what failed.
5956 19. Added -dd to debug only the daemon process.
5958 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5959 handles timeouts, both on the server side and network timeouts. Renamed the
5960 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5962 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5963 It is now set to "smtps".
5965 22. $host_address is now set to the target address during the checking of
5966 ignore_target_hosts.
5968 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5969 being passed; this would have caused $sender_host_name to have been used if
5970 matching the list had actually called for a host name (not very likely,
5971 since this list is usually IP addresses). A host name is now passed as
5974 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5975 code that specifies a non-restarting handler (typically sigaction() in
5976 modern systems) in an attempt to fix a rare and obscure crash bug.
5978 25. Narrowed the window for a race in the daemon that could cause it to ignore
5979 SIGCHLD signals. This is not a major problem, because they are used only to
5980 wake it up if nothing else does.
5982 26. A malformed maildirsize file could cause Exim to calculate negative values
5983 for the mailbox size or file count. Odd effects could occur as a result.
5984 The maildirsize information is now recalculated if the size or filecount
5987 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5988 support for a long time. Removed HAVE_SYS_VFS_H.
5990 28. Installed the latest version of exipick from John Jetmore.
5992 29. In an address list, if the pattern was not a regular expression, an empty
5993 subject address (from a bounce message) matched only if the pattern was an
5994 empty string. Non-empty patterns were not even tested. This was the wrong
5995 because it is perfectly reasonable to use an empty address as part of a
5996 database query. An empty address is now tested by patterns that are
5997 lookups. However, all the other forms of pattern expect the subject to
5998 contain a local part and a domain, and therefore, for them, an empty
5999 address still always fails if the pattern is not itself empty.
6001 30. Exim went into a mad DNS loop when attempting to do a callout where the
6002 host was specified on an smtp transport, and looking it up yielded more
6003 than one IP address.
6005 31. Re-factored the code for checking spool and log partition space into a
6006 function that finds that data and another that does the check. The former
6007 is then used to implement four new variables: $spool_space, $log_space,
6008 $spool_inodes, and $log_inodes.
6010 32. The RFC2047 encoding function was originally intended for short strings
6011 such as real names; it was not keeping to the 75-character limit for
6012 encoded words that the RFC imposes. It now respects the limit, and
6013 generates multiple encoded words if necessary. To be on the safe side, I
6014 have increased the buffer size for the ${rfc2047: expansion operator from
6017 33. It is now permitted to omit both strings after an "if" condition; if the
6018 condition is true, the result is "true". As before, when the second string
6019 is omitted, a false condition yields an empty string. This makes it less
6020 cumbersome to write custom ACL and router conditions.
6022 34. Failure to deliver a bounce message always caused it to be frozen, even if
6023 there was an errors_to setting on the router. The errors_to setting is now
6026 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6027 canonical form (fully expanded) before being placed in
6028 $sender_host_address.
6030 36. The table in the code that translates DNS record types into text (T_A to
6031 "A" for instance) was missing entries for NS and CNAME. It is just possible
6032 that this could have caused confusion if both these types were looked up
6033 for the same domain, because the text type is used as part of Exim's
6034 per-process caching. But the chance of anyone hitting this buglet seems
6037 37. The dnsdb lookup has been extended in a number of ways.
6039 (1) There is a new type, "zns", which walks up the domain tree until it
6040 finds some nameserver records. It should be used with care.
6042 (2) There is a new type, "mxh", which is like "mx" except that it returns
6043 just the host names, not the priorities.
6045 (3) It is now possible to give a list of domains (or IP addresses) to be
6046 looked up. The behaviour when one of the lookups defers can be
6047 controlled by a keyword.
6049 (4) It is now possible to specify the separator character for use when
6050 multiple records are returned.
6052 38. The dnslists ACL condition has been extended: it is now possible to supply
6053 a list of IP addresses and/or domains to be looked up in a particular DNS
6056 39. Added log_selector=+queue_time_overall.
6058 40. When running the queue in the test harness, wait just a tad after forking a
6059 delivery process, to get repeatability of debugging output.
6061 41. Include certificate and key file names in error message when GnuTLS fails
6062 to set them up, because the GnuTLS error message doesn't include the name
6063 of the failing file when there is a problem reading it.
6065 42. Allow both -bf and -bF in the same test run.
6067 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6069 44. The "Exiscan patch" is now merged into the mainline Exim source.
6071 45. Sometimes the final signoff response after QUIT could fail to get
6072 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6073 < 0 before doing a fflush(). This bug looks as though it goes back to the
6074 introduction of TLS in release 3.20, but "sometimes" must have been rare
6075 because the tests only now provoked it.
6077 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6078 (this can affect the format of dates).
6080 47. exim_tidydb, when checking for the continued existence of a message for
6081 which it has found a message-specific retry record, was not finding
6082 messages that were in split spool directories. Consequently, it was
6083 deleting retry records that should have stayed in existence.
6085 48. Steve fixed some bugs in eximstats.
6087 49. The SPA authentication driver was not abandoning authentication and moving
6088 on to the next authenticator when an expansion was forced to fail,
6089 contradicting the general specification for all authenticators. Instead it
6090 was generating a temporary error. It now behaves as specified.
6092 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6093 (the order specifies the preference for clients). The order is now AES256,
6094 AES128, 3DES, ARCFOUR128.
6096 51. Small patch to Sieve code - explicitly set From: when generating an
6099 52. Exim crashed if a remote delivery caused a very long error message to be
6100 recorded - for instance if somebody sent an entire SpamAssassin report back
6101 as a large number of 550 error lines. This bug was coincidentally fixed by
6102 increasing the size of one of Exim's internal buffers (big_buffer) that
6103 happened as part of the Exiscan merge. However, to be on the safe side, I
6104 have made the code more robust (and fixed the comments that describe what
6107 53. Now that there can be additional text after "Completed" in log lines (if
6108 the queue_time_overall log selector is set), a one-byte patch to exigrep
6109 was needed to allow it to recognize "Completed" as not the last thing in
6112 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6113 patch that reportedly fixes this has been added. I am not expert enough to
6114 create a test for it. This is what the patch creator wrote:
6116 "I found a little strange behaviour of ldap code when working with
6117 Windows 2003 AD Domain, where users was placed in more than one
6118 Organization Units. When I tried to give exim partial DN, the exit code
6119 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6120 But simultaneously result of request was absolutely normal ldap result,
6121 so I produce this patch..."
6123 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6124 REFERENCE, so I have modified the code to exclude the patch when that macro
6127 55. Some experimental protocols are using DNS PTR records for new purposes. The
6128 keys for these records are domain names, not reversed IP addresses. The
6129 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6130 leaves it alone. Component reversal etc. now happens only for IP addresses.
6133 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6135 57. Double the size of the debug message buffer (to 2048) so that more of very
6136 long debug lines gets shown.
6138 58. The exicyclog utility now does better if the number of log files to keep
6139 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6141 59. Two changes related to the smtp_active_hostname option:
6143 (1) $smtp_active_hostname is now available as a variable.
6144 (2) The default for smtp_banner uses $smtp_active_hostname instead
6145 of $primary_hostname.
6147 60. The host_aton() function is supposed to be passed a string that is known
6148 to be a valid IP address. However, in the case of IPv6 addresses, it was
6149 not checking this. This is a hostage to fortune. Exim now panics and dies
6150 if the condition is not met. A case was found where this could be provoked
6151 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6152 components; fortuitously, this particular loophole had already been fixed
6153 by change 4.50/55 above.
6155 If there are any other similar loopholes, the new check in host_aton()
6156 itself should stop them being exploited. The report I received stated that
6157 data on the command line could provoke the exploit when Exim was running as
6158 exim, but did not say which command line option was involved. All I could
6159 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6160 running as the user.
6163 61. There was a buffer overflow vulnerability in the SPA authentication code
6164 (which came originally from the Samba project). I have added a test to the
6165 spa_base64_to_bits() function which I hope fixes it.
6168 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6169 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6171 63. The daemon start-up calls getloadavg() while still root for those OS that
6172 need the first call to be done as root, but it missed one case: when
6173 deliver_queue_load_max is set with deliver_drop_privilege. This is
6174 necessary for the benefit of the queue runner, because there is no re-exec
6175 when deliver_drop_privilege is set.
6177 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6178 This has been fixed.
6180 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6181 was in use, was not putting the data itself into the right store pool;
6182 consequently, it could be overwritten for a subsequent message in the same
6183 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6186 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6188 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6189 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6190 the function were treating the return as a boolean value, which happened to
6191 work because 0=false and not-0=true, but is not correct code.
6193 68. The host_aton() function was not handling scoped IPv6 addresses (those
6194 with, for example, "%eth0" on the end) correctly.
6196 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6197 negated items (that is, ~something) in unsigned ints. Some compilers
6198 apparently mutter when there is no cast.
6200 70. If an address verification called from an ACL failed, and did not produce a
6201 user-specific message (i.e. there was only a "system" message), nothing was
6202 put in $acl_verify_message. In this situation, it now puts the system
6205 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6206 session; change 4.31/43 added the unwanted input to the log line - except
6207 that it did not do this in the start of session case. It now does.
6209 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6210 This is wrong and can cause the other end to generate a synchronization
6211 error if it is another Exim or anything else that does the synchronization
6212 check. A QUIT command is no longer sent after a timeout.
6214 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6215 during host lookups.
6217 74. The defer_ok option of callout verification was not working if it was used
6218 when verifying addresses in header lines, that is, for this case:
6220 verify = header_sender/callout=defer_ok
6222 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6223 those file descriptors could be used for SMTP connections. If anything
6224 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6225 could be sent to the SMTP client, causing chaos. The daemon now opens
6226 stdin, stdout, and stderr to /dev/null when it puts itself into the
6229 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6230 log by default. The user can override this with suitable Perl magic.
6232 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6233 the log message when discard triggers, was not working for the DATA ACL or
6234 for the non-SMTP ACL.
6236 78. Error message wording change in sieve.c.
6238 79. If smtp_accept_max_per_host was set, the number of connections could be
6239 restricted to fewer than expected, because the daemon was trying to set up
6240 a new connection before checking whether the processes handling previous
6241 connections had finished. The check for completed processes is now done
6242 earlier. On busy systems, this bug wouldn't be noticed because something
6243 else would have woken the daemon, and it would have reaped the completed
6246 80. If a message was submitted locally by a user whose login name contained one
6247 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6248 It caused a spool format error. I have fixed the spool reading code. A
6249 related problem was that the "from" clause in the Received: line became
6250 illegal because of the space(s). It is now covered by ${quote_local_part.
6252 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6255 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6257 83. Patch from Sieve maintainer to fix unterminated string problem in
6258 "vacation" handling.
6260 84. Some minor changes to the Linux configuration files to help with other
6261 OS variants using glibc.
6263 85. One more patch for Sieve to update vacation handling to latest spec.
6266 ----------------------------------------------------
6267 See the note above about the 4.44 and 4.50 releases.
6268 ----------------------------------------------------
6274 1. Change 4.43/35 introduced a bug that caused file counts to be
6275 incorrectly computed when quota_filecount was set in an appendfile
6278 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6279 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6282 3. Give more explanation in the error message when the command for a transport
6283 filter fails to execute.
6285 4. There are several places where Exim runs a non-Exim command in a
6286 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6287 was being done only for the command run by the queryprogram router. It is
6288 now done for all such subprocesses. The other cases are: ${run, transport
6289 filters, and the commands run by the lmtp and pipe transports.
6291 5. Some older OS have a limit of 256 on the maximum number of file
6292 descriptors. Exim was using setrlimit() to set 1000 as a large value
6293 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6294 systems. I've change it so that if it can't get 1000, it tries for 256.
6296 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6297 was an oversight, and furthermore, ever since the addition of extra
6298 controls (e.g. 4.43/32), the checks on when to allow different forms of
6299 "control" were broken. There should now be diagnostics for all cases when a
6300 control that does not make sense is encountered.
6302 7. $recipients is now available in the predata ACL (oversight).
6304 8. Tidy the search cache before the fork to do a delivery from a message
6305 received from the command line. Otherwise the child will trigger a lookup
6306 failure and thereby defer the delivery if it tries to use (for example) a
6307 cached ldap connection that the parent has called unbind on.
6309 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6310 of $address_data from the recipient verification was clobbered by the
6311 sender verification.
6313 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6314 was its contents. (It was OK if the option was not defined at all.)
6316 11. A "Completed" log line is now written for messages that are removed from
6317 the spool by the -Mrm option.
6319 12. $host_address is now set to the target address during the checking of
6320 ignore_target_hosts.
6322 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6323 being passed; this would have caused $sender_host_name to have been used if
6324 matching the list had actually called for a host name (not very likely,
6325 since this list is usually IP addresses). A host name is now passed as
6328 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6329 code that specifies a non-restarting handler (typically sigaction() in
6330 modern systems) in an attempt to fix a rare and obscure crash bug.
6332 15. Narrowed the window for a race in the daemon that could cause it to ignore
6333 SIGCHLD signals. This is not a major problem, because they are used only to
6334 wake it up if nothing else does.
6336 16. A malformed maildirsize file could cause Exim to calculate negative values
6337 for the mailbox size or file count. Odd effects could occur as a result.
6338 The maildirsize information is now recalculated if the size or filecount
6341 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6342 support for a long time. Removed HAVE_SYS_VFS_H.
6344 18. Updated exipick to current release from John Jetmore.
6346 19. Allow an empty sender to be matched against a lookup in an address list.
6347 Previously the only cases considered were a regular expression, or an
6350 20. Exim went into a mad DNS lookup loop when doing a callout where the
6351 host was specified on the transport, if the DNS lookup yielded more than
6354 21. The RFC2047 encoding function was originally intended for short strings
6355 such as real names; it was not keeping to the 75-character limit for
6356 encoded words that the RFC imposes. It now respects the limit, and
6357 generates multiple encoded words if necessary. To be on the safe side, I
6358 have increased the buffer size for the ${rfc2047: expansion operator from
6361 22. Failure to deliver a bounce message always caused it to be frozen, even if
6362 there was an errors_to setting on the router. The errors_to setting is now
6365 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6366 canonical form (fully expanded) before being placed in
6367 $sender_host_address.
6369 24. Updated eximstats to version 1.33
6371 25. Include certificate and key file names in error message when GnuTLS fails
6372 to set them up, because the GnuTLS error message doesn't include the name
6373 of the failing file when there is a problem reading it.
6375 26. Expand error message when OpenSSL has problems setting up cert/key files.
6378 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6379 (this can affect the format of dates).
6381 28. exim_tidydb, when checking for the continued existence of a message for
6382 which it has found a message-specific retry record, was not finding
6383 messages that were in split spool directories. Consequently, it was
6384 deleting retry records that should have stayed in existence.
6386 29. eximstats updated to version 1.35
6387 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6388 1.35 - bugfix such that pie charts by volume are generated correctly
6390 30. The SPA authentication driver was not abandoning authentication and moving
6391 on to the next authenticator when an expansion was forced to fail,
6392 contradicting the general specification for all authenticators. Instead it
6393 was generating a temporary error. It now behaves as specified.
6395 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6396 (the order specifies the preference for clients). The order is now AES256,
6397 AES128, 3DES, ARCFOUR128.
6399 31. Small patch to Sieve code - explicitly set From: when generating an
6402 32. Exim crashed if a remote delivery caused a very long error message to be
6403 recorded - for instance if somebody sent an entire SpamAssassin report back
6404 as a large number of 550 error lines. This bug was coincidentally fixed by
6405 increasing the size of one of Exim's internal buffers (big_buffer) that
6406 happened as part of the Exiscan merge. However, to be on the safe side, I
6407 have made the code more robust (and fixed the comments that describe what
6410 33. Some experimental protocols are using DNS PTR records for new purposes. The
6411 keys for these records are domain names, not reversed IP addresses. The
6412 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6413 leaves it alone. Component reversal etc. now happens only for IP addresses.
6416 34. The host_aton() function is supposed to be passed a string that is known
6417 to be a valid IP address. However, in the case of IPv6 addresses, it was
6418 not checking this. This is a hostage to fortune. Exim now panics and dies
6419 if the condition is not met. A case was found where this could be provoked
6420 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6421 components; fortuitously, this particular loophole had already been fixed
6422 by change 4.50/55 or 4.44/33 above.
6424 If there are any other similar loopholes, the new check in host_aton()
6425 itself should stop them being exploited. The report I received stated that
6426 data on the command line could provoke the exploit when Exim was running as
6427 exim, but did not say which command line option was involved. All I could
6428 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6429 running as the user.
6432 35. There was a buffer overflow vulnerability in the SPA authentication code
6433 (which came originally from the Samba project). I have added a test to the
6434 spa_base64_to_bits() function which I hope fixes it.
6437 36. The daemon start-up calls getloadavg() while still root for those OS that
6438 need the first call to be done as root, but it missed one case: when
6439 deliver_queue_load_max is set with deliver_drop_privilege. This is
6440 necessary for the benefit of the queue runner, because there is no re-exec
6441 when deliver_drop_privilege is set.
6443 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6444 was in use, was not putting the data itself into the right store pool;
6445 consequently, it could be overwritten for a subsequent message in the same
6446 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6449 38. Sometimes the final signoff response after QUIT could fail to get
6450 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6451 < 0 before doing a fflush(). This bug looks as though it goes back to the
6452 introduction of TLS in release 3.20, but "sometimes" must have been rare
6453 because the tests only now provoked it.
6459 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6460 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6461 Now it returns an integer. A number of calls were still expecting a T/F
6462 return. Fortuitously, in all cases, the tests worked in OK situations,
6463 which is the norm. However, things would have gone wrong on any write
6464 failures on the smtp file descriptor. This function is used when sending
6465 messages over SMTP and also when doing verify callouts.
6467 2. When Exim is called to do synchronous delivery of a locally submitted
6468 message (the -odf or -odi options), it no longer closes stderr before doing
6471 3. Implemented the mua_wrapper option.
6473 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6475 5. Implemented the functions header_remove(), header_testname(),
6476 header_add_at_position(), and receive_remove_recipient(), and exported them
6479 6. If an ACL "warn" statement specified the addition of headers, Exim already
6480 inserted X-ACL-Warn: at the start if there was no header name. However, it
6481 was not making this test for the second and subsequent header lines if
6482 there were newlines in the string. This meant that an invalid header could
6483 be inserted if Exim was badly configured.
6485 7. Allow an ACL "warn" statement to add header lines at the start or after all
6486 the Received: headers, as well as at the end.
6488 8. Added the rcpt_4xx retry error code.
6490 9. Added postmaster_mailfrom=xxx to callout verification option.
6492 10. Added mailfrom=xxxx to the callout verification option, for verify=
6495 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6496 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6498 12. Inserted some casts to stop certain compilers warning when using pointer
6499 differences as field lengths or precisions in printf-type calls (mostly
6500 affecting debugging statements).
6502 13. Added optional readline() support for -be (dynamically loaded).
6504 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6505 same clock tick as a message's arrival, so that its received time was the
6506 same as the "first fail" time on the retry record, and that message
6507 remained on the queue past the ultimate address timeout, every queue runner
6508 would try a delivery (because it was past the ultimate address timeout) but
6509 after another failure, the ultimate address timeout, which should have then
6510 bounced the address, did not kick in. This was a "< instead of <=" error;
6511 in most cases the first failure would have been in the next clock tick
6512 after the received time, and all would be well.
6514 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6515 being recognized when the domain list was tested by the match_domain
6516 condition in an expansion string.
6518 16. Added the ${str2b64: operator.
6520 17. Exim was always calling setrlimit() to set a large limit for the number of
6521 processes, without checking whether the existing limit was already
6522 adequate. (It did check for the limit on file descriptors.) Furthermore,
6523 errors from getrlimit() and setrlimit() were being ignored. Now they are
6524 logged to the main and panic logs, but Exim does carry on, to try to do its
6525 job under whatever limits there are.
6527 18. Imported PCRE 5.0.
6529 19. Trivial typo in log message " temporarily refused connection" (the leading
6532 20. If the log selector return_path_on_delivery was set and an address was
6533 redirected to /dev/null, the delivery process crashed because it assumed
6534 that a return path would always be set for a "successful" delivery. In this
6535 case, the whole delivery is bypassed as an optimization, and therefore no
6538 21. Internal re-arrangement: the function for sending a challenge and reading
6539 a response while authentication was assuming a zero-terminated challenge
6540 string. It's now changed to take a pointer and a length, to allow for
6541 binary data in such strings.
6543 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6545 23. Exim was not respecting finduser_retries when seeking the login of the
6546 uid under which it was called; it was always trying 10 times. (The default
6547 setting of finduser_retries is zero.) Also, it was sleeping after the final
6548 failure, which is pointless.
6550 24. Implemented tls_on_connect_ports.
6552 25. Implemented acl_smtp_predata.
6554 26. If the domain in control=submission is set empty, Exim assumes that the
6555 authenticated id is a complete email address when it generates From: or
6556 Sender: header lines.
6558 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6559 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6560 chown and chgrp in /bin and hostname in /usr/bin.
6562 28. Exim was keeping the "process log" file open after each use, just as it
6563 does for the main log. This opens the possibility of it remaining open for
6564 long periods when the USR1 signal hits a daemon. Occasional processlog
6565 errors were reported, that could have been caused by this. Anyway, it seems
6566 much more sensible not to leave this file open at all, so that is what now
6569 29. The long-running daemon process does not normally write to the log once it
6570 has entered its main loop, and it closes the log before doing so. This is
6571 so that log files can straightforwardly be renamed and moved. However,
6572 there are a couple of unusual error situations where the daemon does write
6573 log entries, and I had neglected to close the log afterwards.
6575 30. The text of an SMTP error response that was received during a remote
6576 delivery was being truncated at 512 bytes. This is too short for some of
6577 the long messages that one sometimes sees. I've increased the limit to
6580 31. It is now possible to make retry rules that apply only when a message has a
6581 specific sender, in particular, an empty sender.
6583 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6584 it possible to be selective about when SMTP synchronization is enforced.
6586 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6588 32. Implemented hosts_connection_nolog.
6590 33. Added an ACL for QUIT.
6592 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6595 35. Added mailbox_size and mailbox_filecount to appendfile.
6597 36. Added control = no_multiline_responses to ACLs.
6599 37. There was a bug in the logic of the code that waits for the clock to tick
6600 in the case where the clock went backwards by a substantial amount such
6601 that the microsecond fraction of "now" was more than the microsecond
6602 fraction of "then" (but the whole seconds number was less).
6604 38. Added support for the libradius Radius client library this is found on
6605 FreeBSD (previously only the radiusclient library was supported).
6611 1. When certain lookups returned multiple values in the form name=value, the
6612 quoting of the values was not always being done properly. Specifically:
6613 (a) If the value started with a double quote, but contained no whitespace,
6615 (b) If the value contained whitespace other than a space character (i.e.
6616 tabs or newlines or carriage returns) it was not quoted.
6617 This fix has been applied to the mysql and pgsql lookups by writing a
6618 separate quoting function and calling it from the lookup code. The fix
6619 should probably also be applied to nisplus, ibase and oracle lookups, but
6620 since I cannot test any of those, I have not disturbed their existing code.
6622 2. A hit in the callout cache for a specific address caused a log line with no
6623 reason for rejecting RCPT. Now it says "Previous (cached) callout
6624 verification failure".
6626 3. There was an off-by-one bug in the queryprogram router. An over-long
6627 return line was truncated at 256 instead of 255 characters, thereby
6628 overflowing its buffer with the terminating zero. As well as fixing this, I
6629 have increased the buffer size to 1024 (and made a note to document this).
6631 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6632 when Exim is waiting for an SMTP response from a remote server, Exim
6633 restarts its select() call on the socket, thereby resetting its timeout.
6634 This is not a problem when such interrupts are rare. Somebody set up a cron
6635 job to run exiwhat every 2 minutes, which is less than the normal select()
6636 timeout (5 or 10 minutes). This meant that the select() timeout never
6637 kicked in because it was always reset. I have fixed this by comparing the
6638 time when an interrupt arrives with the time at the start of the first call
6639 to select(). If more time than the timeout has elapsed, the interrupt is
6640 treated as a timeout.
6642 5. Some internal re-factoring in preparation for the addition of Sieve
6643 extensions (by MH). In particular, the "personal" test is moved to a
6644 separate function, and given an option for scanning Cc: and Bcc: (which is
6645 not set for Exim filters).
6647 6. When Exim created an email address using the login of the caller as the
6648 local part (e.g. when creating a From: or Sender: header line), it was not
6649 quoting the local part when it contained special characters such as @.
6651 7. Installed new OpenBSD configuration files.
6653 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6654 try to make them clearer.
6656 9. Callout options, other than the timeout value, were being ignored when
6657 verifying sender addresses in header lines. For example, when using
6659 verify = header_sender/callout=no_cache
6661 the cache was (incorrectly) being used.
6663 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6664 only the Cygwin environment.
6666 11. When return_path_on_delivery was set as a log selector, if different remote
6667 addresses in the same message used different return paths and parallel
6668 remote delivery occurred, the wrong values would sometimes be logged.
6669 (Whenever a remote delivery process finished, the return path value from
6670 the most recently started remote delivery process was logged.)
6672 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6673 lines when AUTH and/or TLS are in use. This is the "received protocol"
6674 field. Exim used to use "asmtp" for authenticated SMTP, without any
6675 indication (in the protocol name) for TLS use. Now it follows the RFC and
6676 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6677 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6678 names appear in log lines as well as in Received: header lines.
6680 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6681 extensions, and comparison tests, and to fix some bugs.
6683 14. Changes to the "personal" filter test:
6685 (1) The test was buggy in that it was just doing the equivalent of
6686 "contains" tests on header lines. For example, if a user's address was
6687 anne@some.where, the "personal" test would incorrectly be true for
6689 To: susanne@some.where
6691 This test is now done by extracting each address from the header in turn,
6692 and checking the entire address. Other tests that are part of "personal"
6693 are now done using regular expressions (for example, to check local parts
6694 of addresses in From: header lines).
6696 (2) The list of non-personal local parts in From: addresses has been
6697 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6698 taken from the Sieve specification recommendations.
6700 (3) If the message contains any header line starting with "List-" it is
6701 treated as non-personal.
6703 (4) The test for "circular" in the Subject: header line has been removed
6704 because it now seems ill-conceived.
6706 15. Minor typos in src/EDITME comments corrected.
6708 16. Installed latest exipick from John Jetmore.
6710 17. If headers_add on a router specified a text string that was too long for
6711 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6712 of string_sprintf() is now avoided.
6714 18. $message_body_size was not set (it was always zero) when running the DATA
6715 ACL and the local_scan() function.
6717 19. For the "mail" command in an Exim filter, no default was being set for
6718 the once_repeat time, causing a random time value to be used if "once" was
6719 specified. (If the value happened to be <= 0, no repeat happened.) The
6720 default is now 0s, meaning "never repeat". The "vacation" command was OK
6721 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6722 (I found it when inspecting the code).
6724 20. There is now an overall timeout for performing a callout verification. It
6725 defaults to 4 times the callout timeout, which applies to individual SMTP
6726 commands during the callout. The overall timeout applies when there is more
6727 than one host that can be tried. The timeout is checked before trying the
6728 next host. This prevents very long delays if there are a large number of
6729 hosts and all are timing out (e.g. when the network connections are timing
6730 out). The value of the overall timeout can be changed by specifying an
6731 additional sub-option for "callout", called "maxwait". For example:
6733 verify = sender/callout=5s,maxwait=20s
6735 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6736 to the end before writing, but this should make it even safer).
6738 22. Exim was forgetting that it had advertised PIPELINING for the second and
6739 subsequent messages on an SMTP connection. It was also not resetting its
6740 memory on STARTTLS and an internal HELO.
6742 23. When Exim logs an SMTP synchronization error within a session, it now
6743 records whether PIPELINING has been advertised or not.
6745 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6746 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6747 rather than long int.
6749 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6751 26. Added the never_mail option to autoreply.
6757 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6758 crash if the getsockname() call failed; this can happen if a connection is
6759 closed very soon after it is established. The problem was simply in the
6760 order in which certain operations were done, causing Exim to try to write
6761 to the SMTP stream before it had set up the file descriptor. The bug has
6762 been fixed by making things happen in the correct order.
6768 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6769 before the connection was closed, thus losing the rejection response.
6771 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6772 some early Solaris releases, but causes trouble in current releases where
6773 socklen_t is defined.
6775 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6778 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6781 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6782 that releasing the top of it at the end releases what was used for sub-
6783 expansions (unless the block got too big). However, discard this block if
6784 the first thing is a variable or header, so that we can use its block when
6785 it is dynamic (useful for very large $message_headers, for example).
6787 6. Lookups now cache *every* query, not just the most recent. A new, separate
6788 store pool is used for this. It can be recovered when all lookup caches are
6789 flushed. Lookups now release memory at the end of their result strings.
6790 This has involved some general refactoring of the lookup sources.
6792 7. Some code has been added to the store_xxx() functions to reduce the amount
6793 of flapping under certain conditions.
6795 8. log_incoming_interface used to affect only the <= reception log lines. Now
6796 it causes the local interface and port to be added to several more SMTP log
6797 lines, for example "SMTP connection from", and rejection lines.
6799 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6801 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6803 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6804 could be overwritten at the end of the current message (or the start of a
6805 new message if it was set in a HELO ACL). The value is now preserved for
6806 the duration of the SMTP connection.
6808 12. If a transport had a headers_rewrite setting, and a matching header line
6809 contained an unqualified address, that address was qualified, even if it
6810 did not match any rewriting rules. The underlying bug was that the values
6811 of the flags that permit the existence of unqualified sender and recipient
6812 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6813 non-local messages, and by -bnq for local messages) were not being
6814 preserved with the message after it was received.
6816 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6817 "next input=" as part of the text comprising the host identity instead of
6818 the correct text. The code was using the same buffer for two different
6819 strings. However, depending on which order the printing function evaluated
6820 its arguments, the bug did not always show up. Under Linux, for example, my
6821 test suite worked just fine.
6823 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6824 doesn't work with some older versions of Perl. It has been changed to "my",
6825 which in any case is probably the better facility to use.
6827 15. A really picky compiler found some instances of statements for creating
6828 error messages that either had too many or two few arguments for the format
6831 16. The size of the buffer for calls to the DNS resolver has been increased
6832 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6833 for addresses that have a lot of PTR records. This alleviates a problem; it
6834 does not fully solve it.
6836 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6837 buffer now truncates the list and logs the incident, which is the same
6838 action as happens when Exim is looking up a host name and its aliases.
6839 Previously in this situation something unpredictable would happen;
6840 sometimes it was "internal error: store_reset failed".
6842 18. If a server dropped the connection unexpectedly when an Exim client was
6843 using GnuTLS and trying to read a response, the client delivery process
6844 crashed while trying to generate an error log message.
6846 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6847 string, for example:
6849 warn message = H1: something\nH2: something
6851 the text was added as a single header line from Exim's point of view
6852 though it ended up OK in the delivered message. However, searching for the
6853 second and subsequent header lines using $h_h2: did not work. This has been
6854 fixed. Similarly, if a system filter added multiple headers in this way,
6855 the routers could not see them.
6857 20. Expanded the error message when iplsearch is called with an invalid key to
6858 suggest using net-iplsearch in a host list.
6860 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6861 ACLs are no longer actually imposed (and a message to that effect is
6864 22. If a "gecos" field in a passwd entry contained escaped characters, in
6865 particular, if it contained a \" sequence, Exim got it wrong when building
6866 a From: or a Sender: header from that name. A second bug also caused
6867 incorrect handling when an unquoted " was present following a character
6868 that needed quoting.
6870 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6871 was not being matched caselessly.
6873 24. Arranged for all hyphens in the exim.8 source to be escaped with
6876 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6877 properly. Recipient callout cache records were still being keyed to include
6878 the sender, even when use_sender was set false. This led to far more
6879 callouts that were necessary. The sender is no longer included in the key
6880 when use_sender is false.
6882 26. Added "control = submission" modifier to ACLs.
6884 27. Added the ${base62d: operator to decode base 62 numbers.
6886 28. dnsdb lookups can now access SRV records.
6888 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6889 the configuration file.
6891 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6892 (-v) mode. This makes the output for a verbose queue run more intelligible.
6894 31. Added a use_postmaster feature to recipient callouts.
6896 32. Added the $body_zerocount variable, containing the number of binary zero
6897 bytes in the message body.
6899 33. The time of last modification of the "new" subdirectory is now used as the
6900 "mailbox time last read" when there is a quota error for a maildir
6903 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6905 35. Added +ignore_unknown as a special item in host lists.
6907 36. Code for decoding IPv6 addresses in host lists is now included, even if
6908 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6909 address was recognized as an IP address, but was then not correctly decoded
6910 into binary, causing unexpected and incorrect effects when compared with
6917 1. Very minor rewording of debugging text in manualroute to say "list of
6918 hosts" instead of "hostlist".
6920 2. If verify=header_syntax was set, and a header line with an unqualified
6921 address (no domain) and a large number of spaces between the end of the
6922 name and the colon was received, the reception process suffered a buffer
6923 overflow, and (when I tested it) crashed. This was caused by some obsolete
6924 code that should have been removed. The fix is to remove it!
6926 3. When running in the test harness, delay a bit after writing a bounce
6927 message to get a bit more predictability in the log output.
6929 4. Added a call to search_tidyup() just before forking a reception process. In
6930 theory, someone could use a lookup in the expansion of smtp_accept_max_
6931 per_host which, without the tidyup, could leave open a database connection.
6933 5. Added the variables $recipient_data and $sender_data which get set from a
6934 lookup success in an ACL "recipients" or "senders" condition, or a router
6935 "senders" option, similar to $domain_data and $local_part_data.
6937 6. Moved the writing of debug_print from before to after the "senders" test
6940 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6941 problems for message scanning, either using a data ACL, or using
6942 local_scan() because the Received: header was not generated till after they
6943 were called (in order to set the time as the time of reception completion).
6944 I have revised the way this works. The header is now generated after the
6945 body is received, but before the ACL or local_scan() are called. After they
6946 are run, the timestamp in the header is updated.
6952 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6953 before starting a queue runner without re-exec. This happened only when
6954 deliver_drop_privilege was set or when the Exim user was set to root. The
6955 effect of the bug was that timeouts during subsequent deliveries caused
6956 crashes instead of being properly handled. The handler is now left at its
6957 default (and expected) setting.
6959 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6960 message, again when deliver_drop_privilege is set or Exim is run as root.
6961 The bug described in (1) was not present in this case, but the tidying up
6962 of the other signals was missing. I have made the two cases consistent.
6964 3. The ignore_target_hosts setting on a manualroute router was being ignored
6965 for hosts that were looked up using the /MX notation.
6967 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6970 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6971 operated on the sender address. After changing the $sender_address to <>
6972 for the sender address verify, Exim was re-instated it as the original
6973 (before rewriting) address, but remembering that it had rewritten it, so it
6974 wasn't rewriting it again. This bug also had the effect of breaking the
6975 sender address verification caching when the sender address was rewritten.
6977 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6978 This has been changed so that if the ip literal address matches
6979 ignore_target_hosts, the router declines.
6981 7. Added expansion conditions match_domain, match_address, and match_local_
6982 part (NOT match_host).
6984 8. The placeholder for the Received: header didn't have a length field set.
6986 9. Added code to Exim itself and to exim_lock to test for a specific race
6987 condition that could lead to file corruption when using MBX delivery. The
6988 issue is with the lockfile that is created in /tmp. If this file is removed
6989 after a process has opened it but before that process has acquired a lock,
6990 there is the potential for a second process to recreate the file and also
6991 acquire a lock. This could lead to two Exim processes writing to the file
6992 at the same time. The added code performs the same test as UW imapd; it
6993 checks after acquiring the lock that its file descriptor still refers to
6994 the same named file.
6996 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6997 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7000 11. Added the smtp_active_hostname option. If used, this will typically be made
7001 to depend on the incoming interface address. Because $interface_address is
7002 not set up until the daemon has forked a reception process, error responses
7003 that can happen earlier (such as "too many connections") no longer contain
7006 12. If an expansion in a condition on a "warn" statement fails because a lookup
7007 defers, the "warn" statement is abandoned, and the next ACL statement is
7008 processed. Previously this caused the whole ACL to be aborted.
7010 13. Added the iplsearch lookup type.
7012 14. Added ident_timeout as a log selector.
7014 15. Added tls_certificate_verified as a log selector.
7016 16. Added a global option tls_require_ciphers (compare the smtp transport
7017 option of the same name). This controls incoming TLS connections.
7019 17. I finally figured out how to make tls_require_ciphers do a similar thing
7020 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7021 before starting the TLS session.
7023 18. Tabs are now shown as \t in -bP output.
7025 19. If the log selector return_path_on_delivery was set, Exim crashed when
7026 bouncing a message because it had too many Received: header lines.
7028 20. If two routers both had headers_remove settings, and the first one included
7029 a superfluous trailing colon, the final name in the first list and the
7030 first name in the second list were incorrectly joined into one item (with a
7031 colon in the middle).
7037 1. Added -C and -D options to the exinext utility, mainly to make it easier
7038 to include in the automated testing, but these could be helpful when
7039 multiple configurations are in use.
7041 2. The exinext utility was not formatting the output nicely when there was
7042 an alternate port involved in the retry record key, nor when there was a
7043 message id as well (for retries that were specific to a specific message
7044 and a specific host). It was also confused by IPv6 addresses, because of
7045 the additional colons they contain. I have fixed the IPv4 problem, and
7046 patched it up to do a reasonable job for IPv6.
7048 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7049 delivery, the log line now contains "pipelined" if PIPELINING was used.
7051 4. An SMTP transport process used to panic and die if the bind() call to set
7052 an explicit outgoing interface failed. This has been changed; it is now
7053 treated in the same way as a connect() failure.
7055 5. A reference to $sender_host_name in the part of a conditional expansion
7056 that was being skipped was still causing a DNS lookup. This no longer
7059 6. The def: expansion condition was not recognizing references to header lines
7060 that used bh_ and bheader_.
7062 7. Added the _cache feature to named lists.
7064 8. The code for checking quota_filecount in the appendfile transport was
7065 allowing one more file than it should have been.
7067 9. For compatibility with Sendmail, the command line option
7075 and sets the incoming protocol and host name (for trusted callers). The
7076 host name and its colon can be omitted when only the protocol is to be set.
7077 Note the Exim already has two private options, -pd and -ps, that refer to
7078 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7079 "s", but I don't think that's a major issue.
7081 10. A number of refactoring changes to the code, none of which should affect
7084 (a) The number of logging options was getting close to filling up the
7085 32-bit word that was used as a bit map. I have split them into two classes:
7086 those that are passed in the argument to log_write(), and those that are
7087 only ever tested independently outside of that function. These are now in
7088 separate 32-bit words, so there is plenty of room for expansion again.
7089 There is no change in the user interface or the logging behaviour.
7091 (b) When building, for example, log lines, the code previously used a
7092 macro that called string_cat() twice, in order to add two strings. This is
7093 not really sufficiently general. Furthermore, there was one instance where
7094 it was actually wrong because one of the argument was used twice, and in
7095 one call a function was used. (As it happened, calling the function twice
7096 did not affect the overall behaviour.) The macro has been replaced by a
7097 function that can join an arbitrary number of extra strings onto a growing
7100 (c) The code for expansion conditions now uses a table and a binary chop
7101 instead of a serial search (which was left over from when there were very
7102 few conditions). Also, it now recognizes conditions like "pam" even when
7103 the relevant support is not compiled in: a suitably worded error message is
7104 given if an attempt is made to use such a condition.
7106 11. Added ${time_interval:xxxxx}.
7108 12. A bug was causing one of the ddress fields not to be passed back correctly
7109 from remote delivery subprocesses. The field in question was not being
7110 subsequently used, so this caused to problems in practice.
7112 13. Added new log selectors queue_time and deliver_time.
7114 14. Might have fixed a bug in maildirsizefile handling that threw up
7115 "unexpected character" debug warnings, and recalculated the data
7116 unnecessarily. In any case, I expanded the warning message to give more
7119 15. Added the message "Restricted characters in address" to the statements in
7120 the default ACL that block characters like @ and % in local parts.
7122 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7123 Three changes have been made:
7125 (a) There was a serious bug; a negative response to MAIL caused the whole
7126 recipient domain to be cached as invalid, thereby blocking all messages
7127 to all local parts at the same domain, from all senders. This bug has
7128 been fixed. The domain is no longer cached after a negative response to
7129 MAIL if the sender used is not empty.
7131 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7134 (c) A new callout option, "use_sender" has been added for people who want
7135 the modified behaviour.
7141 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7144 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7145 indeed breaks things for older releases.
7147 3. Added additional logging to the case where there is a problem reading data
7148 from a filter that is running in a subprocess using a pipe, in order to
7149 try to track down a specific problem.
7151 4. Testing facility fudge: when running in the test harness and attempting
7152 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7153 getting "No route to host". Convert this to a timeout.
7155 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7158 6. Some OS don't have socklen_t but use size_t instead. This affects the
7159 fifth argument of getsockopt() amongst other things. This is now
7160 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7161 can be set for individual OS. I have set it for SunOS5, OSF1, and
7162 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7163 some earlier ones do not.
7165 7. Change 4.30/15 was not doing the test caselessly.
7167 8. The standard form for an IPv6 address literal was being rejected by address
7168 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7169 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7170 this, as well as the form without the "IPv6" on the front (but only when
7171 address literals are enabled, of course).
7173 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7175 10. Exim crashed if a message with an empty sender address specified by -f
7176 encountered a router with an errors_to setting. This could be provoked only
7177 by a command such as
7181 where an empty string was supplied; "<>" did not hit this bug.
7183 11. Installed PCRE release 4.5.
7185 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7186 remained set. It is now erased.
7188 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7189 times from message ids (which are base 36 rather than the normal 62).
7191 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7192 were being counted as actual protocol errors, and logged if the log
7193 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7194 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7195 and DATA following a set of rejected RCPTs do not count as protocol errors.
7196 In other words, Exim assumes they were pipelined, though this may not
7197 actually be the case. Of course, in all cases the client gets an
7198 appropriate error code.
7200 15. If a lookup fails in an ACL condition, a message about the failure may
7201 be available; it is used if testing the ACL cannot continue, because most
7202 such messages specify what the cause of the deferral is. However, some
7203 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7204 that caused an old message to be retained and used if a later statement
7205 caused a defer, replacing the real cause of the deferral.
7207 16. If an IP address had so many PTR records that the DNS lookup buffer
7208 was not large enough to hold them, Exim could crash while trying to process
7209 the truncated data. It now detects and logs this case.
7211 17. Further to 4.21/58, another change has been made: if (and only if) the
7212 first line of a message (the first header line) ends with CRLF, a bare LF
7213 in a subsequent header line has a space inserted after it, so as not to
7214 terminate the header.
7216 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7217 unnecessarily, used atoi() instead of strtol(), and didn't check the
7218 termination when getting file sizes from file names by regex.
7220 19. Completely re-implemented the support for maildirsize files, in the light
7221 of a number of problems with the previous contributed implementation
7222 (4.30/29). In particular:
7224 . If the quota is zero, the maildirsize file is maintained, but no quota is
7227 . If the maildir directory does not exist, it is created before any attempt
7228 to write a maildirsize file.
7230 . The quota value in the file is just a cache; if the quota is changed in
7231 the transport, the new value overrides.
7233 . A regular expression is available for excluding directories from the
7236 20. The autoreply transport checks the characters in options that define the
7237 message's headers; it allows continued headers, but it was checking with
7238 isspace() after an embedded newline instead of explicitly looking for a
7241 21. If all the "regular" hosts to which an address was routed had passed their
7242 expiry times, and had not reached their retry times, the address was
7243 bounced, even if fallback hosts were defined. Now Exim should go on to try
7246 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7247 equivalent code in the SMTP transport. Some hosts send humungous responses
7248 to HELO/EHLO, more than 1024 it seems.
7250 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7251 gives compiler warnings in some environments. I've now done it "properly",
7254 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7255 (because of problems with the built-in one) was declared to return uschar *
7256 instead of char *, causing compiler failure.
7258 25. Fixed a file descriptor leak when processing alias/forward files.
7260 26. Fixed a minor format string issue in dbfn.c.
7262 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7264 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7265 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7268 29. When a sender address is verified, it is cached, to save repeating the test
7269 when there is more than one recipient in a message. However, when the
7270 verification involves a callout, it is possible for different callout
7271 options to be set for different recipients. It is too complicated to keep
7272 track of this in the cache, so now Exim always runs a verification when a
7273 callout is required, relying on the callout cache for the optimization.
7274 The overhead is duplication of the address routing, but this should not be
7277 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7278 to be verified with callout=postmaster, and the main callout worked but the
7279 postmaster check failed, the verification correctly failed. However, if a
7280 subsequent RCPT command asked for sender verification *without* the
7281 postmaster check, incorrect caching caused this verification also to fail,
7284 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7285 it was not caching the DNS options (qualify_single, search_parents) that
7286 were used when the lookup failed. A subsequent lookup with different
7287 options therefore always gave the same answer, though there were cases
7288 where it should not have. (Example: a "domains = !$mx_any" option on a
7289 dnslookup router: the "domains" option is always processed without any
7290 widening, but the router might have qualify_single set.) Now Exim uses the
7291 cached value only when the same options are set.
7293 32. Added John Jetmore's "exipick" utility to the distribution.
7295 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7296 than a timeout (e.g. a certificate is required, and is not provided), an
7297 Exim server now closes the connection immediately. Previously it waited for
7298 the client to close - but if the client is SSL, it seems that they each
7299 wait for each other, leading to a delay before one of them times out.
7301 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7302 maintained 0.8.x compatibility because I don't think many are using it, and
7303 it is clearly obsolete.
7305 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7308 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7309 client certificate was expired. A simple patch fixes this, though I don't
7310 understand the full logic of why the verify callback is called multiple
7313 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7314 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7315 which causes problems with some clients (such as the Certicom SSL Plus
7316 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7317 disables the coutermeasure allowing Eudora to connect."
7319 38. Exim was not checking that a write() to a log file succeeded. This could
7320 lead to Bad Things if a log got too big, in particular if it hit a file
7321 size limit. Exim now panics and dies if it cannot write to a log file, just
7322 as it does if it cannot open a log file.
7324 39. Modified OS/Makefile-Linux so that it now contains
7326 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7328 The two -D definitions ensure that Exim is compiled with large file
7329 support, which makes it possible to handle log files that are bigger than
7332 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7333 instance) a domain was checked against a named list that involved a lookup,
7334 causing $domain_data to be set, then another domain was checked against the
7335 same list, then the first domain was re-checked, the value of $domain_data
7336 after the final check could be wrong. In particular, if the second check
7337 failed, it could be set empty. This bug probably also applied to
7340 41. The strip_trailing_dot option was not being applied to the address given
7341 with the -f command-line option.
7343 42. The code for reading a message's header from the spool was incrementing
7344 $received_count, but never initializing it. This meant that the value was
7345 incorrect (doubled) while delivering a message in the same process in which
7346 it was received. In the most common configuration of Exim, this never
7347 happens - a fresh exec is done - but it can happen when
7348 deliver_drop_privilege is set.
7350 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7351 it now includes up to 150 characters of the unexpected data in the log
7354 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7355 and building data strings. The size of both of these buffers was 10 000
7356 bytes - far larger than anybody would *ever* want, thought I. Needless to
7357 say, somebody hit the limit. I have increased the maximum line length to
7358 20 000 and the maximum data length of concatenated lines to 100 000. I have
7359 also fixed two bugs, because there was no checking on these buffers. Tsk,
7360 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7361 buffer is too small.
7363 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7364 lsearch lookups. Now it does.
7366 46. When parsing a route_list item in a manualroute router, a fixed-length
7367 buffer was used for the list of hosts. I made this 1024 bytes long,
7368 thinking that nobody would ever have a list of hosts that long. Wrong.
7369 Somebody had a whole pile of complicated expansion conditions, and the
7370 string was silently truncated, leading to an expansion error. It turns out
7371 that it is easier to change to an unlimited length (owing to other changes
7372 that have happened since this code was originally written) than to build
7373 structure for giving a limitation error. The length of the item that
7374 expands into the list of hosts is now unlimited.
7376 47. The lsearch lookup could not handle data where the length of text line was
7377 more than 4095 characters. Such lines were truncated, leading to shortened
7378 data being returned. It should now handle lines of any length.
7380 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7381 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7384 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7385 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7386 no longer happens in comment lines. A list of replacements is now placed
7387 at the head of all of the source files, except those whose only change is
7388 to replace PERL_COMMAND in the very first #! line.
7390 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7391 messages on the queue, with a bottom-up merge sort, using code contributed
7392 by Michael Haardt. This should make operations like -bp somewhat faster on
7393 large queues. It won't affect queue runners, except when queue_run_in_order
7396 51. Installed eximstats 1.31 in the distribution.
7398 52. Added support for SRV lookups to the dnslookup router.
7400 53. If an ACL referred to $message_body or $message_body_end, the value was not
7401 reset for any messages that followed in the same SMTP session.
7403 54. The store-handling optimization for building very long strings was not
7404 differentiating between the different store pools. I don't think this
7405 actually made any difference in practice, but I've tidied it.
7407 55. While running the routers to verify a sender address, $sender_address
7408 was still set to the sender address. This is wrong, because when routing to
7409 send a bounce to the sender, it would be empty. Therefore, I have changed
7410 it so that, while verifying a sender address, $sender_address is set to <>.
7411 (There is no change to what happens when verifying a recipient address.)
7413 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7414 target A or AAAA records (if not already returned) without resetting the
7415 qualify_single or search_parents options of the DNS resolver. These are
7416 inappropriate in this case because the targets of MX and SRV records must
7417 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7418 target that, when qualified, matched something in the local domain. These
7419 two options are now turned off when doing these lookups.
7421 57. It seems that at least some releases of Reiserfs (which does not have the
7422 concept of a fixed number of inodes) returns zero and not -1 for the
7423 number of available inodes. This interacted badly with check_spool_inodes,
7424 which assumed that -1 was the "no such thing" setting. What I have done is
7425 to check that the total number of inodes is greater than zero before doing
7426 the test of how many are available.
7428 58. When a "warn" ACL statement has a log_message modifier, the message is
7429 remembered, and not repeated. This is to avoid a lot of repetition when a
7430 message has many recipients that cause the same warning to be written.
7431 However, Exim was preserving the list of already written lines for an
7432 entire SMTP session, which doesn't seem right. The memory is now reset if a
7433 new message is started.
7435 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7436 debugging output unless log_rewrite was also set.
7438 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7439 of (int)(handle) when we know that handle contains (void *)(-1).
7441 61. The Exim daemon panic-logs an error return when it closes the incoming
7442 connection. However "connection reset by peer" seems to be common, and
7443 isn't really an error worthy of noting specially, so that particular error
7446 62. When Exim is trying to find all the local interfaces, it used to panic and
7447 die if the ioctl to get the interface flags failed. However, it seems that
7448 on at least one OS (Solaris 9) it is possible to have an interface that is
7449 included in the list of interfaces, but for which you get a failure error
7450 for this call. This happens when the interface is not "plumbed" into a
7451 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7452 failure of the "get flags" call assumes that the interface is down.
7454 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7455 makes life easier for people who are doing arithmetic on fields extracted
7456 from dates, where you often get leading zeros that should not be
7457 interpreted as octal.
7459 64. Added qualify_domain to the redirect router, to override the global
7462 65. If a pathologically long header line contained very many addresses (the
7463 report of this problem mentioned 10 000) and each of them was rewritten,
7464 Exim could use up a very large amount of memory. (It kept on making new
7465 copies of the header line as it rewrote, and never released the old ones.)
7466 At the expense of a bit more processing, the header rewriting function has
7467 been changed so that it no longer eats memory in this way.
7469 66. The generation of the Received: header has been moved from the time that a
7470 message starts to be received, to the time that it finishes. The timestamp
7471 in the Received: header should now be very close to that of the <= log
7472 line. There are two side-effects of this change:
7474 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7475 logged header lines no longer include the local Received: line, because
7476 it has not yet been created. The same applies to a copy of the message
7477 that is returned to a non-SMTP sender when a message is rejected.
7479 (b) When a filter file is tested using -bf, no additional Received: header
7480 is added to the test message. After some thought, I decided that this
7483 This change does not affect the value of $received_for. It is still set
7484 after address rewriting, but before local_scan() is called.
7486 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7488 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7489 gave an unhelpful panic error message, and a defer error. I have managed to
7490 change this behaviour so that it now rejects any supplied certificate,
7491 which seems right, as the list of acceptable certificates is empty.
7493 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7494 gave an unhelpful defer error. I have not managed to make this reject any
7495 supplied certificates, but the error message it gives is "no certificate
7496 supplied", which is not helpful.
7498 70. exigrep's output now also includes lines that are not associated with any
7499 message, but which match the given pattern. Implemented by a patch from
7500 Martin Sluka, which also tidied up the Perl a bit.
7502 71. Recipient callout verification, like sender verification, was using <> in
7503 the MAIL FROM command. This isn't really the right thing, since the actual
7504 sender may affect whether the remote host accepts the recipient or not. I
7505 have changed it to use the actual sender in the callout; this means that
7506 the cache record is now keyed on a recipient/sender pair, not just the
7507 recipient address. There doesn't seem to be a real danger of callout loops,
7508 since a callout by the remote host to check the sender would use <>.
7509 [SEE ABOVE: changed after hitting problems.]
7511 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7512 temporary errors. However, in the case of such a code being given after
7513 the end of a data transmission (i.e. after ".") Exim was failing to write
7514 a retry record for the message. (Yes, there was some broken host that was
7515 actually sending 8xx at this point.)
7517 73. An unknown lookup type in a host list could cause Exim to panic-die when
7518 the list was checked. (An example that provoked this was putting <; in the
7519 middle of a list instead of at the start.) If this happened during a DATA
7520 ACL check, a -D file could be left lying around. This kind of configuration
7521 error no longer causes Exim to die; instead it causes a defer error. The
7522 incident is still logged to the main and panic logs.
7524 74. Buglet left over from Exim 3 conversion. The message "too many messages
7525 in one connection" was written to the rejectlog but not the mainlog, except
7526 when address rewriting (yes!) was being logged.
7528 75. Added write_rejectlog option.
7530 76. When a system filter was run not as root (that is, when system_filter_user
7531 was set), the values of the $n variables were not being returned to the
7532 main process; thus, they were not subsequently available in the $sn
7535 77. Added +return_path_on_delivery log selector.
7537 78. A connection timeout was being treated differently from recipients deferred
7538 when testing hosts_max_try with a message that was older than the host's
7539 retry timeout. (The host should not be counted, thus allowing all hosts to
7540 be tried at least once before bouncing.) This may have been the cause of an
7541 occasionally reported bug whereby a message would remain on the queue
7542 longer than the retry timeout, but would be bounced if a delivery was
7543 forced. I say "may" because I never totally pinned down the problem;
7544 setting up timeout/retry tests is difficult. See also the next item.
7546 79. The ultimate address timeout was not being applied to errors that involved
7547 a combination of host plus message (for example, a timeout on a MAIL
7548 command). When an address resolved to a number of possible hosts, and they
7549 were not all tried for each delivery (e.g. because of hosts_max_try), a
7550 message could remain on the queue longer than the retry timeout.
7552 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7555 81. Fixed an obscure SMTP outgoing bug which required at least the following
7556 conditions: (a) there was another message waiting for the same server;
7557 (b) the server returned 5xx to all RCPT commands in the first message so
7558 that the message was not completed; (c) the server dropped the connection
7559 or gave a negative response to the RSET that Exim sends to abort the
7560 transaction. The observed case was a dropped connection after DATA that had
7561 been sent in pipelining mode. That is, the server had advertised PIPELINING
7562 but was not implementing it correctly. The effect of the bug was incorrect
7563 behaviour, such as trying another host, and this could lead to a crash.
7569 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7570 and daemon.c were passed as pointers to ints; they should have been
7571 pointers to socklen_t variables (which are typically unsigned ints).
7573 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7576 3. Fixed a really odd bug that affected only the testing scheme; patching a
7577 certain fixed string in the binary changed the value of another string that
7578 happened to be identical to the end of the original first string.
7580 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7581 name", it returns that address as the IP address. On some operating
7582 systems (e.g. Solaris), it also passes back the IP address string as the
7583 "host name". However, on others (e.g. Linux), it passes back an empty
7584 string. Exim wasn't checking for this, and was changing the host name to an
7585 empty string, assuming it had been canonicalized.
7587 5. Although rare, it is permitted to have more than one PTR record for a given
7588 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7589 all the names associated with an address, because they do in Solaris.
7590 However, it seems that they do not in Linux for data that comes from the
7591 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7592 I found this out when I moved to a new Linux workstation and tried to run
7593 the Exim test suite.
7595 To get round this problem I have changed the code so that it now does its
7596 own call to the DNS to look up PTR records when searching for a host name.
7597 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7598 addresses that are only in /etc/hosts are still found.
7600 This behaviour is, however, controlled by an option called host_lookup_
7601 order, which defaults to "bydns:byaddr". If people want to use the other
7602 order, or indeed, just use one or the other means of lookup, they can
7603 specify it in this variable.
7605 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7606 some operating systems, this comes back from gethostbyaddr() as an empty
7607 string, and this is what Exim used to test for. However, it seems that in
7608 other systems, "." is yielded. Exim now tests for this case too.
7610 7. The values of check_spool_space and check_log_space are now held internally
7611 as a number of kilobytes instead of an absolute number of bytes. If a
7612 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7613 kilobyte. This means that much larger values can be stored.
7615 8. Exim monitor: an attempt to get the action menu when not actually pointing
7616 at a message produces an empty menu entitled "No message selected". This
7617 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7618 no entries in it ("Shell widget menu has zero width and/or height"). So I
7619 have added a single, blank menu entry in this case.
7621 9. Added ${quote_local_part.
7623 10. MIME decoding is now applied to the contents of Subject: header lines when
7626 11. Now that a reference to $sender_host_address automatically causes a reverse
7627 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7628 host lookup before query-style lookups in lists that might use this
7629 variable. This has therefore been abolished, and the "net-" prefix is no
7630 longer necessary for query-style lookups.
7632 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7633 have been a typo for LFLAGS, so it has been changed.
7635 13. The install script calls Exim with "-C /dev/null" in order to find the
7636 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7637 to be output. However, since Exim outputs its version number before the
7638 error, it didn't break the script. It just looked ugly. I fixed this by
7639 always allowing "-C /dev/null" if the caller is root.
7641 14. Ignore overlarge ACL variable number when reading spool file - insurance
7642 against a later release with more variables having written the file.
7644 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7645 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7646 this, as well as the form without the "IPv6" on the front.
7648 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7649 OS/Makefile-Darwin file.
7651 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7652 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7654 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7655 variable. However, some OS use ints for inodes. I've added cast to long int
7656 to get rid of the compiler warning.
7658 19. I had forgotten to lock out "/../" in configuration file names when
7659 ALT_CONFIG_PREFIX was set.
7661 20. Routers used for verification do not need to specify transports. However,
7662 if such a router generated a host list, and callout was configured, Exim
7663 crashed, because it could not find a port number from the (non-existent)
7664 transport. It now assumes port 25 in this circumstance.
7666 21. Added the -t option to exigrep.
7668 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7669 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7670 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7671 them is set, LOOKUP_LSEARCH is forced.
7673 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7675 24. Added sender and host information to the "rejected by local_scan()" log
7676 line; previously there was no indication of these.
7678 25. Added .include_if_exists.
7680 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7681 Linux. It turns out that not all file systems support this. Apparently some
7682 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7683 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7684 not supported on the file descriptor, is now ignored when Exim is trying to
7685 sync a directory. This applies only to Linux.
7687 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7689 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7690 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7693 29. Added support for maildirsize files from supplied patch (modified a bit).
7695 30. The use of :fail: followed by an empty string could lead Exim to respond to
7696 sender verification failures with (e.g.):
7698 550 Verification failed for <xxx>
7699 550 Sender verify failed
7701 where the first response line was missing the '-' that indicates it is not
7702 the final line of the response.
7704 31. The loop for finding the name of the user that called Exim had a hardwired
7705 limit of 10; it now uses the value of finduser_retries, which is used for
7706 all other user lookups.
7708 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7711 33. Exim was neglecting to zero errno before one call of strtol() when
7712 expanding a string and expecting an integer value. On some systems this
7713 resulted in spurious "integer overflow" errors. Also, it was casting the
7714 result into an int without checking.
7716 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7717 did not work. The code looks as if it has *never* worked, though it appears
7718 to have been documented since at least release 1.62. I have made it work.
7720 35. The "timeout_DNS" error in retry rules, also documented since at least
7721 1.62, also never worked. As it isn't clear exactly what this means, and
7722 clearly it isn't a major issue, I have abolished the feature by treating it
7723 as "timeout", and writing a warning to the main and panic logs.
7725 36. The display of retry rules for -brt wasn't always showing the error code
7728 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7729 timeout_connect_A, timeout_connect_MX.
7731 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7732 to the empty sender.
7734 39. The daemon was not analysing the content of -oX till after it had closed
7735 stderr and disconnected from the controlling terminal. This meant that any
7736 syntax errors were only noted on the panic log, and the return code from
7737 the command was 0. By re-arranging the code a little, I've made the
7738 decoding happen first, so such errors now appear on stderr, and the return
7739 code is 1. However, the actual setting up of the sockets still happens in
7740 the disconnected process, so errors there are still only recorded on the
7743 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7744 connections (as happens on some IP stacks) was logged at start up time as
7745 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7746 it from "IPv6 and IPv4", which means that two separate sockets are being
7749 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7750 says whether AF_INET or AF_INET6 was passed as an argument.
7752 42. Exiwhat output was messed up when time zones were included in log
7755 43. Exiwhat now gives more information about the daemon's listening ports,
7756 and whether -tls-on-connect was used.
7758 44. The "port" option of the smtp transport is now expanded.
7760 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7761 silently ignored. Now an error message is written to the main and panic
7764 46. There's a new ACL modifier called "logwrite" which writes to a log file
7765 as soon as it is encountered.
7767 47. Added $local_user_uid and $local_user_gid at routing time.
7769 48. Exim crashed when trying to verify a sender address that was being
7772 49. Exim was recognizing only a space character after ".include". It now also
7773 recognizes a tab character.
7775 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7776 extracting the relevant information from the specification. The man page no
7777 longer contains scrambled data for the -d option, and I've added a section
7778 at the front about calling Exim under different names.
7780 51. Added "extra_headers" argument to the "mail" command in filter files.
7782 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7785 53. Installed eximstats 1.29.
7787 54. Added transport_filter_timeout as a generic transport option.
7789 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7790 Cc: header lines. This was required by RFC 822, but it not required by RFC
7793 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7794 incoming messages that did not have them. Now it does so only if the
7795 message originates locally, that is, if there is no associated remote host
7796 address. When Resent- header lines are present, this applies to the Resent-
7797 lines rather than the non-Resent- lines.
7799 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7800 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7802 58. Messages for configuration errors now include the name of the main
7803 configuration file - useful now that there may be more than one file in a
7804 list (.included file names were always shown).
7806 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7807 for those rare installations that do not start the daemon as root or run it
7808 setuid root. I've cut out the call to initgroups() if the daemon is not
7811 60. The Exim user and group can now be bound into the binary as text strings
7812 that are looked up at the start of Exim's processing.
7814 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7816 62. Added $mailstore_basename variable.
7818 63. Installed patch to sieve.c from Michael Haardt.
7820 64. When Exim failed to open the panic log after failing to open the main log,
7821 the original message it was trying to log was written to stderr and debug
7822 output, but if they were not available (the usual case in production), it
7823 was lost. Now it is written to syslog before the two lines that record the
7824 failures to open the logs.
7826 65. Users' Exim filters run in subprocesses under the user's uid. It is
7827 possible for a "deliver" command or an alias in a "personal" command to
7828 provoke an address rewrite. If logging of address rewriting is configured,
7829 this fails because the process is not running as root or exim. There may be
7830 a better way of dealing with this, but for the moment (because 4.30 needs
7831 to be released), I have disabled address rewrite logging when running a
7832 filter in a non-root, non-exim process.
7838 1. The buildconfig auxiliary program wasn't quoting the value set for
7839 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7840 not defined. This bug was masked in 4.22 by the effect that was fixed in
7843 2. Some messages that were rejected after a message id was allocated were
7844 shown as "incomplete" by exigrep. It no longer does this for messages that
7845 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7847 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7848 have allow_domain_literals set, the ID did not get logged in the <= line.
7849 Domain literals are now always recognized in Message-ID: header lines.
7851 4. The first argument for a ${extract expansion item is the key name or field
7852 number. Leading and trailing spaces in this item were not being ignored,
7853 causing some misleading effects.
7855 5. When deliver_drop_privilege was set, single queue runner processes started
7856 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7857 same command in the process it spins off) were not dropping privilege.
7859 6. When the daemon running as "exim" started a queue runner, it always
7860 re-executed Exim in the spun-off process. This is a waste of effort when
7861 deliver_drop_privilege is set. The new process now just calls the
7862 queue-runner function directly.
7868 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7871 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7872 ignored. Though the use of -oP was forcing the writing of a pid file, it
7873 was always written to the default place.
7875 3. If the message "no IP address found for host xxxx" is generated during
7876 incoming verification, it is now followed by identification of the incoming
7877 connection (so you can more easily find what provoked it).
7879 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7881 5. Added some features to "harden" Exim a bit more against certain attacks:
7883 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7884 be put in Local/Makefile. This is like the never_users runtime option,
7885 but it cannot be overridden. The default setting is "root".
7887 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7888 prefix string with which any file named in a -C command line option
7891 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7892 is retained for -C and -D only if the caller of Exim is root. Without
7893 it, the exim user may also use -C and -D and retain privilege.
7895 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7896 command line option is disabled.
7898 6. Macro names set by the -D option must start with an upper case letter, just
7899 like macro names defined in the configuration file.
7901 7. Added "dereference=" facility to LDAP.
7903 8. Two instances of the typo "uknown" in the source files are fixed.
7905 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7906 the Configure-Makefile script screwed up while processing it.
7908 10. Incorporated PCRE 4.4.
7910 11. The SMTP synchronization check was not operating right at the start of an
7911 SMTP session. For example, it could not catch a HELO sent before the client
7912 waited for the greeting. There is now a check for outstanding input at the
7913 point when the greeting is written. Because of the duplex, asynchronous
7914 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7915 way, but not yet received, when the check is performed.
7917 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7918 on TCP/IP sockets, because this apparently causes some broken clients to
7921 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7922 unchanged) from the Cygwin maintainer.
7924 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7925 was supported instead of testing for mbx. Effectively a typo.
7927 15. The spa authenticator server code was not checking that the input it
7928 received was valid base64.
7930 16. The debug output line for the "set" modifier in ACLs was not showing the
7931 name of the variable that was being set.
7933 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7935 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7936 lookup on its own. Something else was needed to trigger the lookup. For
7937 example, a match in host_lookup or the need for a host name in a host list.
7938 Now, if $sender_host_name is referenced and the host name has not yet been
7939 looked up, a lookup is performed. If the lookup fails, the variable remains
7940 empty, and $host_lookup_failed is set to "1".
7942 19. Added "eqi" as a case-independent comparison operator.
7944 20. The saslauthd authentication condition could segfault if neither service
7945 nor realm was specified.
7947 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7948 error message that was logged was misleading, and incoming SMTP
7949 connections were dropped. The message is now more accurate, and temporary
7950 errors are given to SMTP connections.
7952 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7953 (except RSET) with a 421 error, until QUIT is received. However, it was
7954 failing to send a response to QUIT.
7956 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7957 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7958 ACL is now run after the helo_try_verify_hosts code.
7960 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7961 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7962 case-independent, so other case variants are also recognized). Apparently
7963 some systems use these upper case variants.
7965 25. If more than two messages were waiting for the same host, and a transport
7966 filter was specified for the transport, Exim sent two messages over the
7967 same TCP/IP connection, and then failed with "socket operation on non-
7968 socket" when it tried to send the third.
7970 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7972 27. The extern definition of crypt16() in expand.c was not being excluded when
7973 the OS had its own crypt16() function.
7975 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7976 as a preferred synonym for return_size_limit, both as an option and as an
7979 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7981 30. Changed the default configuration ACL to relax the local part checking rule
7982 for addresses that are not in any local domains. For these addresses,
7983 slashes and pipe symbols are allowed within local parts, but the sequence
7984 /../ is explicitly forbidden.
7986 31. SPA server authentication was not clearing the challenge buffer before
7989 32. log_message in a "warn" ACL statement was writing to the reject log as
7990 well as to the main log, which contradicts the documentation and doesn't
7991 seem right (because no rejection is happening). So I have stopped it.
7993 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7994 However, I am unable to do any testing of this.
7996 34. Fixed an infelicity in the appendfile transport. When checking directories
7997 for a mailbox, to see if any needed to be created, it was accidentally
7998 using path names with one or more superfluous leading slashes; tracing
7999 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8001 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8002 contents are added to the log line that is written for every discarded
8003 recipient. (Previously a log_message setting was ignored.)
8005 36. The ${quote: operator now quotes the string if it is empty.
8007 37. The install script runs exim in order to find its version number. If for
8008 some reason other than non-existence or emptiness, which it checks, it
8009 could not run './exim', it was installing it with an empty version number,
8010 i.e. as "exim-". This error state is now caught, and the installation is
8013 38. An argument was missing from the function that creates an error message
8014 when Exim fails to connect to the socket for saslauthd authentication.
8015 This could cause Exim to crash, or give a corrupted message.
8017 39. Added isip, isip4, and isip6 to ${if conditions.
8019 40. The ACL variables $acl_xx are now saved with the message, and can be
8020 accessed later in routers, transports, and filters.
8022 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8023 strings in the file are not string-expanded.
8025 42. If a MAIL command specified a SIZE value that was too large to fit into an
8026 int variable, the check against message_size_limit failed. Such values are
8027 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8028 day this will have to be increased, but I don't think I want to be around
8029 when emails are that large.
8036 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8037 iconv() is not standard in FreeBSD.
8039 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8040 IPv6 enabled. The observed symptom was a segmentation fault on return from
8041 the function os_common_find_running_interfaces() in src/os.c.
8043 3. In the check_special_case() function in daemon.c I had used "errno" as an
8044 argument name, which causes warnings on some systems. This was basically a
8045 typo, since it was named "eno" in the comments!
8047 4. The code that waits for the clock to tick (at a resolution of some fraction
8048 of a second) so as to ensure message-id uniqueness was always waiting for
8049 at least one whole tick, when it could have waited for less. [This is
8050 almost certainly not relevant at current processor speeds, where it is
8051 unlikely to ever wait at all. But we try to future-proof.]
8053 5. The function that sleeps for a time interval that includes fractions of a
8054 second contained a race. It did not block SIGALRM between setting the
8055 timer, and suspending (a couple of lines later). If the interval was short
8056 and the sigsuspend() was delayed until after it had expired, the suspension
8057 never ended. On busy systems this could lead to processes getting stuck for
8060 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8061 process, before it forks any delivery processes. The open lookup caching
8062 mechanism meant that the open file or database connection was passed into
8063 the delivery process. The problem was that delivery processes always tidy
8064 up cached lookup data. This could cause a problem for the next delivery
8065 process started by the queue runner, because the external queue runner
8066 process does not know about the closure. So the next delivery process
8067 still has data in the lookup cache. In the case of a file lookup, there was
8068 no problem because closing a file descriptor in a subprocess doesn't affect
8069 the parent. However, if the lookup was caching a connection to a database,
8070 the connection was closed, and the second delivery process was likely to
8071 see errors such as "PGSQL: query failed: server closed the connection
8072 unexpectedly". The problem has been fixed by closing all cached lookups
8073 in a queue runner before running a delivery process.
8075 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8076 seem to have the "const" qualifier which it has on other OS. I've
8079 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8080 *of the same type* (client or server) with the same public name that an
8081 error should be diagnosed.
8083 9. When Exim looked up a host name for an IP address, but failed to find the
8084 original IP address when looking up the host name (a safety check), it
8085 output the message "<ip address> does not match any IP for NULL", which was
8086 confusing, to say the least. The bug was that the host name should have
8087 appeared instead of "NULL".
8089 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8090 user that is built into the binary, and the -C or -D options is used, root
8091 privilege is dropped before the configuration file is read. In addition,
8092 logging is switched to stderr instead of the normal log files. If the
8093 configuration then re-defines the Exim user, the unprivileged environment
8094 is probably not what is expected, so Exim logs a panic warning message (but
8097 However, if deliver_drop_privilege is set, the unprivileged state may well
8098 be exactly what is intended, so the warning has been cut out in that case,
8099 and Exim is allowed to try to write to its normal log files.
8105 1. smtp_return_error_details was not giving details for temporary sender
8106 or receiver verification errors.
8108 2. Diagnose a configuration error if two authenticators have the same public
8111 3. Exim used not to create the message log file for a message until the first
8112 delivery attempt. This could be confusing when incoming messages were held
8113 for policy or load reasons. The message log file is now created at the time
8114 the message is received, and an initial "Received" line is written to it.
8116 4. The automatically generated man page for command line options had a minor
8117 bug that caused no ill effects; however, a more serious problem was that
8118 the procedure for building the man page automatically didn't always
8119 operate. Consequently, release 4.20 contains an out-of-date version. This
8120 shouldn't happen again.
8122 5. When building Exim with embedded Perl support, the script that builds the
8123 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8124 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8126 6. The freeze_tell option was not being used for messages that were frozen on
8127 arrival, either by an ACL or by local_scan().
8129 7. Added the smtp_incomplete_transaction log selector.
8131 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8132 was accepting AUTH without a new EHLO.
8134 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8135 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8138 10. Logging of TCP/IP connections (when configured) now happens in the main
8139 daemon process instead of the child process, so that the TCP/IP connection
8140 count is more accurate (but it can never be perfect).
8142 11. The use of "drop" in a nested ACL was not being handled correctly in the
8143 outer ACL. Now, if condition failure induced by the nested "drop" causes
8144 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8145 or "require"), the connection is dropped.
8147 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8148 that yield "discard" can now be used with an "accept" or a "discard" verb,
8149 but an error is generated for any others (because I can't see a useful way
8150 to define what should happen).
8152 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8153 are now processed in the same way as lines in the Exim configuration file.
8154 In particular, continuation lines are supported.
8156 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8158 15. Added -ti meaning -t -i.
8160 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8161 domains, and warn by logging if others are found.
8163 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8164 returned from GIFCONF when Exim is trying to find the list of interfaces on
8165 a host. The code in os.c has been modified to copy each ifreq to an aligned
8166 structure in all cases.
8168 Also, in some cases, the returned ifreq's were being copied to a 'struct
8169 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8170 means the last couple of bytes of an IPv6 address could be chopped if the
8171 ifreq contained only a normal sockaddr (14 bytes storage).
8173 18. Named domain lists were not supported in the hosts_treat_as_local option.
8174 An entry such as +xxxx was not recognized, and was treated as a literal
8177 19. Ensure that header lines added by a DATA ACL are included in the reject log
8178 if the ACL subsequently rejects the message.
8180 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8181 MD5 (which is deprecated).
8183 21. When testing a filter file using -bf, Exim was writing a message when it
8184 took the sender from a "From " line in the message, but it was not doing so
8185 when it took $return_path from a Return-Path: header line. It now does.
8187 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8188 with a valid header line field name (a series of printing characters
8189 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8191 23. Changed "disc" in the source to "disk" to conform to the documentation and
8192 the book and for uniformity.
8194 24. Ignore Sendmail's -Ooption=value command line item.
8196 25. When execve() failed while trying to run a command in a pipe transport,
8197 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8198 could be confused with a return value of 69 from the command itself. This
8199 has been changed to 127, the value the shell returns if it is asked to run
8200 a non-existent command. The wording for the related log line suggests a
8201 non-existent command as the problem.
8203 26. If received_header_text expands to an empty string, do not add a Received:
8204 header line to the message. (Well, it adds a token one on the spool, but
8205 marks it "old" so that it doesn't get used or transmitted.)
8207 27. Installed eximstats 1.28 (addition of -nt option).
8209 28. There was no check for failure on the call to getsockname() in the daemon
8210 code. This can fail if there is a shortage of resources on the system, with
8211 ENOMEM, for example. A temporary error is now given on failure.
8213 29. Contrary to the C standard, it seems that in some environments, the
8214 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8215 program. Exim now does this explicitly; it affects the formatting of
8216 timestamps using strftime().
8218 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8219 complaints. I've now initialized all the variables, to avoid this.
8221 32. Header lines added by a system filter were not being "seen" during
8222 transport-time rewrites.
8224 33. The info_callback() function passed to OpenSSL is set up with type void
8225 (*)(SSL *, int, int), as described somewhere. However, when calling the
8226 function (actually a macro) that sets it up, the type void(*)() is
8227 expected. I've put in a cast to prevent warnings from picky compilers.
8229 34. If a DNS black list lookup found a CNAME record, but there were no A
8230 records associated with the domain it pointed at, Exim crashed.
8232 35. If a DNS black list lookup returned more than one A record, Exim ignored
8233 all but the first. It now scans all returned addresses if a particular IP
8234 value is being sought. In this situation, the contents of the
8235 $dnslist_value variable are a list of all the addresses, separated by a
8238 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8239 to accept a host name and all its aliases if the forward lookup for any of
8240 them yielded the IP address of the incoming connection. Now it accepts only
8241 those names whose forward lookup yields the correct IP address. Any other
8242 names are discarded. This closes a loophole whereby a rogue DNS
8243 administrator could create reverse DNS records to break through a
8244 wildcarded host restriction in an ACL.
8246 37. If a user filter or a system filter that ran in a subprocess used any of
8247 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8248 the wrong values were passed to the pipe command ($thisaddress had the
8249 value of $0, $0 had the value of $1, etc). This bug was introduced by
8250 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8252 38. Improved the line breaking for long SMTP error messages from ACLs.
8253 Previously, if there was no break point between 40 and 75 characters, Exim
8254 left the rest of the message alone. Two changes have been made: (a) I've
8255 reduced the minimum length to 35 characters; (b) if it can't find a break
8256 point between 35 and 75 characters, it looks ahead and uses the first one
8257 that it finds. This may give the occasional overlong line, but at least the
8258 remaining text gets split now.
8260 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8261 file descriptors might be low, and that setting 1000 would always raise it.
8262 It turns out that in some environments, the limit is already over 1000 and
8263 that lowering it causes trouble. So now Exim takes care not to decrease it.
8265 40. When delivering a message, the value of $return_path is set to $sender_
8266 address at the start of routing (routers may change the value). By an
8267 oversight, this default was not being set up when an address was tested by
8268 -bt or -bv, which affected the outcome if any router or filter referred to
8271 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8272 log only when "message" or "log_message" are set. However, if one of the
8273 conditions was an address verification, or a call to a nested ACL, the
8274 messages generated by the underlying test were being passed through. This
8275 no longer happens. The underlying message is available in $acl_verify_
8276 message for both "message" and "log_message" expansions, so it can be
8277 passed through if needed.
8279 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8280 new expansion $bh_ to give the encoded byte string without charset
8281 translation. Translation happens only if iconv() is available; HAVE_ICONV
8282 indicates this at build time. HEADERS_CHARSET gives the charset to
8283 translate to; headers_charset can change it in the configuration, and
8284 "headers charset" can change it in an individual filter file.
8286 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8287 that creates RFC 2047 encoded "words" labels them as that charset instead
8288 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8289 expansion operator; (ii) when Exim creates a From: line for a local
8290 message; (iii) when a header line is rewritten to include a "phrase" part.
8292 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8293 buggy, causing it to skip the first lines of messages whose message ID
8294 ended in 'D'. This would not have bitten before Exim release 4.14, because
8295 message IDs were unlikely to end in 'D' before then. The effect was to have
8296 incorrect size information for certain domains.
8298 45. #include "config.h" was missing at the start of the crypt16.c module. This
8299 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8302 46. If there was a timeout during a "random" callout check, Exim treated it as
8303 a failure of the random address, and carried on sending RSET and the real
8304 address. If the delay was just some slowness somewhere, the response to the
8305 original RCPT would be taken as a response to RSET and so on, causing
8306 mayhem of various kinds.
8308 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8309 when I implemented it. It didn't allow for the fact that some option values
8310 may legitimately be negative (e.g. size_addition), and it didn't even do
8311 the right test for positive values.
8313 48. Domain names in DNS records are case-independent. Exim always looks them up
8314 in lower case. Some resolvers return domain names in exactly the case they
8315 appear in the zone file, that is, they may contain uppercase letters. Not
8316 all resolvers do this - some return always lower case. Exim was treating a
8317 change of case by a resolver as a change of domain, similar to a widening
8318 of a domain abbreviation. This triggered its re-routing code and so it was
8319 trying to route what was effectively the same domain again. This normally
8320 caused routing to fail (because the router wouldn't handle the domain
8321 twice). Now Exim checks for this case specially, and just changes the
8322 casing of the domain that it ultimately uses when it transmits the message
8325 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8328 50. If a filter generated a file delivery with a non-absolute name (possible if
8329 no home directory exists for the router), the forbid_file option was not
8332 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8333 the existing equality matching.
8335 52. Exim was using ints instead of ino_t variables in some places where it was
8336 dealing with inode numbers.
8338 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8339 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8340 called TMPDIR, and if it finds it is different, it changes its value.
8342 54. The smtp_printf() function is now made available to local_scan() so
8343 additional output lines can be written before returning. There is also an
8344 smtp_fflush() function to enable the detection of a dropped connection.
8345 The variables smtp_input and smtp_batched_input are exported to
8348 55. Changed the default runtime configuration: the message "Unknown user"
8349 has been removed from the ACL, and instead placed on the localuser router,
8350 using the cannot_route_message feature. This means that any verification
8351 failures that generate their own messages won't get overridden. Similarly,
8352 the "Unrouteable address" message that was in the ACL for unverifiable
8353 relay addresses has also been removed.
8355 56. Added hosts_avoid_esmtp to the smtp transport.
8357 57. The exicyclog script was not checking for the esoteric option
8358 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8359 will work only if exicyclog is run under the appropriate euid.
8361 58. Following a discussion on the list, the rules by which Exim recognises line
8362 endings on incoming messages have been changed. The -dropcr and drop_cr
8363 options are now no-ops, retained only for backwards compatibility. The
8364 following line terminators are recognized: LF CRLF CR. However, special
8365 processing applies to CR:
8367 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8368 nor a local message in the state where . is a terminator.
8370 (ii) If a bare CR is encountered in a header line, an extra space is added
8371 after the line terminator so as not to end the header. The reasoning
8372 behind this is that bare CRs in header lines are most likely either
8373 to be mistakes, or people trying to play silly games.
8375 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8376 was being incorrectly given as 18 bytes larger than it should have been.
8377 This is a VOB (very old bug).
8379 60. This may never have affected anything current, but just in case it has:
8380 When the local host is found other than at the start of a list of hosts,
8381 the local host, those with the same MX, and any that follow, are discarded.
8382 When the list in question was part of a longer list of hosts, the following
8383 hosts (not currently being processed) were also being discarded. This no
8384 longer happens. I'm not sure if this situation could ever has previously
8387 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8390 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8391 header. This is something that is recommended in a new Internet Draft, and
8392 is something that is documented as being done by Sendmail. There are two
8393 possible values. For messages generated by the autoreply transport, Exim
8396 Auto-Submitted: auto-replied
8398 whereas for all other generated messages (e.g. bounces) it adds
8400 Auto-Submitted: auto-generated
8402 63. The "personal" condition in filters now includes a test for the
8403 Auto-Submitted: header. If it contains the string "auto-" the message it
8404 not considered personal.
8406 64. Added rcpt_include_affixes as a generic transport option.
8408 65. Added queue_only_override (default true).
8410 66. Added the syslog_duplication option.
8412 67. If what should have been the first header line of a message consisted of
8413 a space followed by a colon, Exim was mis-interpreting it as a header line.
8414 It isn't of course - it is syntactically invalid and should therefore be
8415 treated as the start of the message body. The misbehaviour could have
8416 caused a number of strange effects, including loss of data in subsequent
8417 header lines, and spool format errors.
8419 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8420 client host had authenticated. This control can now be exercised by an ACL
8421 for more flexibility.
8423 69. By default, callouts do not happen when testing with -bh. There is now a
8424 variant, -bhc, which does actually run the callout code, including
8425 consulting and updating the callout cache.
8427 70. Added support for saslauthd authentication, courtesy of Alexander
8430 71. If statvfs() failed on the spool or log directories while checking their
8431 size for availability, Exim confusingly gave the error "space shortage".
8432 Furthermore, in debugging mode it crashed with a floating point exception.
8433 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8434 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8435 serious problem, Exim now writes to the main and panic logs when this
8436 happens, with details of the failure. It then refuses to accept the
8437 incoming message, giving the message "spool directory problem" or "log
8438 directory problem" with a 421 code for SMTP messages.
8440 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8441 0, 1, and 2 exist, because some OS complain for execs without them (see
8442 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8443 descriptors. However, the code omitted to check that the open succeeded,
8444 causing mysterious errors if for some reason the permissions on /dev/null
8445 got screwed. Now Exim writes a message to the main and panic logs, and
8446 bombs out if it can't open /dev/null.
8448 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8449 interact so that it is all more flexible. It is supposed to remain
8450 backwards compatible. Also added extra_local_interfaces.
8452 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8453 to bomb out with an assertion failure - to the client this appears as a
8454 connection drop. This problem occurs in the part of the code that was taken
8455 from the Samba project. Fortunately, the assertion is in a very simple
8456 function, so I have fixed this by reproducing the function inline in the
8457 one place where it is called, and arranging for authentication to fail
8458 instead of killing the process with assert().
8460 75. The SPA client code was not working when the server requested OEM rather
8461 than Unicode encoding.
8463 76. Added code to make require_files with a specific uid setting more usable in
8464 the case where statting the file as root fails - usually a non-root-mounted
8465 NFS file system. When this happens and the failure is EACCES, Exim now
8466 forks a subprocess and does the per-uid checking as the relevant uid.
8468 77. Added process_log_path.
8470 78. If log_file_path was not explicitly set, a setting of check_log_space or
8471 check_log_inodes was ignored.
8473 79. If a space check for the spool or log partitions fails, the incident is now
8474 logged. Of course, in the latter case the data may get lost...
8476 80. Added the %p formatting code to string_format() so that it can be used to
8477 print addresses in debug_print(). Adjusted all the address printing in the
8478 debugging in store.c to use %p rather than %d.
8480 81. There was a concern that a line of code in smtp_in.c could overflow a
8481 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8482 initially expressed, the concern was not well-founded, because trailing
8483 spaces are removed early. However, if the trailing spaces were followed by
8484 a NULL, they did not get removed, so the overflow was possible. Two fixes
8487 (a) I re-wrote the offending code in a cleaner fashion.
8488 (b) If an incoming SMTP command contains a NULL character, it is rejected
8491 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8492 runs initgroups(), so that if the Exim user is in any additional groups,
8493 they will be used during message reception.
8499 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob