1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
21 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
22 (under Linux) we watch the dir containing the final file; previously
23 it would be the dir with the first symlink. We still do not monitor
26 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
27 dn_expand() is happy to pass them through.
29 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
30 the file with the proof had an unchanged name, the new proof(s) were
31 loaded on top of the old ones (and nover used; the old ones were stapled).
33 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
34 more than one message arrived in a single connection a reference from
35 the earlier message could be re-used. Often a sigsegv resulted.
36 These variables were introduced in Exim 4.87.
37 Debug help from Graeme Fowler.
39 JH/09 Fix ${filter } for conditions that modify $value. Previously the
40 modified version would be used in construction the result, and a memory
43 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
44 Find and fix by Jasen Betts.
46 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
47 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
48 the systemwide configuration to override the Exim config.
50 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
51 API changes in libopendmarc.
53 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
54 pid 1, in the normal "background daemon" mode, having to drop process-
55 group leadership also lost track of needing to create listener sockets.
57 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
58 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
60 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
61 a capture group which obtained no text (eg. "(abc)*" matching zero
62 occurrences) could cause a segfault if the corresponding $<n> was
65 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
66 included a close-brace character (eg. it itself used an expansion) an
69 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
70 starting TLS. Previously it was after, meaning that attackers on such
71 ports had to be screened using the host_reject_connection main config
72 option. The new sequence aligns better with the STARTTLS behaviour, and
73 permits defences against crypto-processing load attacks, even though it
74 is strictly an incompatible change.
75 Also, avoid sending any SMTP fail response for either the connect ACL
76 or host_reject_connection, for TLS-on-connect ports.
78 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
79 Previously this was not permitted, but it makes reasonable sense.
80 While there, restore a restriction on using it from a connect ACL; given
81 the change JH/16 it could only return false (and before 4.91 was not
84 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
85 was exactly sized compared to the log buffer, a crash occurred with the
86 misleading message "bad memory reference; pool not found".
87 Found and traced by Jasen Betts.
89 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
90 dns_again_means_nonexist included an element causing a DNS lookup which
91 iteslf returned DNS_AGAIN, unbounded recursion occurred. Possible results
92 included (though probably not limited to) a process crash from stack
93 memory limit, or from excessive open files. Replace this with a paniclog
94 whine (as this is likely a configuration error), and returning
97 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
98 this always failed, probably leading to the usual downgrade to in-clear
101 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
102 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
103 to not checking that list for these lookups.
105 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
108 JH/23 Fix crash in string expansions. Previously, if an empty variable was
109 immediately followed by an expansion operator, a null-indirection read
110 was done, killing the process.
112 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
113 include an SMTP response string which is longer than that supported
114 by the delivering transport. Alleviate by wrapping such lines before
117 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
118 chars (RFC limit). Previously a limit of 12 items was made, which with
119 a not-impossible References: in the message being bounced could still
120 be over-large and get stopped in the transport.
122 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
123 close. Previously a bare socket close was done.
125 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
128 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
129 message being considered for delivery was already being handled by
130 another process, and having an SMTP connection already open, the function
131 to close it tried to use an uninitialized variable. This would afftect
132 high-volume sites more, especially when running mailing-list-style loads.
133 Pollution of logs was the major effect, as the other process delivered
134 the message. Found and partly investigated by Graeme Fowler.
136 JH/29 Change format of the internal ID used for message identification. The old
137 version only supported 31 bits for a PID element; the new 64 (on systems
138 which can use Base-62 encoding, which is all currently supported ones
139 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
140 and must use Base-36). The new ID is 23 characters rather than 16, and is
141 visible in various places - notably logs, message headers, and spool file
142 names. Various of the ancillary utilities also have to know the format.
143 As well as the expanded PID portion, the sub-second part of the time
144 recorded in the ID is expanded to support finer precision. Theoretically
145 this permits a receive rate from a single comms channel of better than the
147 The major timestamp part of the ID is not changed; at 6 characters it is
148 usable until about year 3700.
149 Updating from previously releases is fully supported: old-format spool
150 files are still usable, and the utilities support both formats. New
151 message will use the new format. The one hints-DB file type which uses
152 message-IDs (the transport wait- DB) will be discarded if an old-format ID
153 is seen; new ones will be built with only new-format IDs.
154 Optionally, a utility can be used to convert spool files from old to new,
155 but this is only an efficiency measure not a requirement for operation
156 Downgrading from new to old requires running a provided utility, having
157 first stopped all operations. This will convert any spool files from new
158 back to old (losing time-precision and PID information) and remove any
159 wait- hints databases.
161 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
162 we treated them as item separators when parsing for a list item, but they
163 need to be protected by the doublequotes. While there, add handling for
166 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
167 Found and fixed by Jasen Betts. No testcase for this as my usual text
168 editor insists on emitting only valid UTF-8.
170 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
171 a null-indirection SIGSEGV for the receive process.
173 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
174 -bh use. Internal checking would spot this and take a panic.
176 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
177 In 4.96 this would expand to empty.
179 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
180 certificate. Find and fix by Andreas Metzler.
182 JH/36 Add ARC info to DMARC hostory records.
184 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
185 or fakedefer. Previously the sender could discover that the message
186 had in fact been accepted.
188 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
189 sequences. Previously the input was not noted as being tainted; notably
190 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
191 bad coding of authenticators.
193 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
194 and ${tr...}. Found and diagnosed by Heiko Schlichting.
196 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
197 could be triggered by externally-supplied input. Found by Trend Micro.
199 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
200 be triggered by externally-controlled input. Found by Trend Micro.
206 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
207 after reception to before a subsequent reception. This should
208 mean slightly faster delivery, and also confirmation of reception
211 JH/02 Move from using the pcre library to pcre2. The former is no longer
212 being developed or supported (by the original developer).
214 JH/03 Constification work in the filters module required a major version
215 bump for the local-scan API. Specifically, the "headers_charset"
216 global which is visible via the API is now const and may therefore
217 not be modified by local-scan code.
219 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
220 sendfile() didi not account for the way the ClamAV driver code called it.
222 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
223 time check was being done for every character; replace that with one
226 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
227 sent was prefixed with a length byte.
229 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
230 RFC 5321. Previously Dovecot (at least) would log errors during
233 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
234 in 4.61 without a reason logged. Binaries will be bigger, which might
235 matter on diskspace-constrained systems, but debug is easier.
237 JH/09 Fix macro-definition during "-be" expansion testing. The move to
238 write-protected store for macros had not accounted for these runtime
239 additions; fix by removing this protection for "-be" mode.
241 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
242 handing out large-numbered file descriptors, violating the usual Unix
243 assumption (and required by Posix) that the lowest possible number will be
244 allocated by the kernel when a new one is needed. In the daemon, and any
245 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
246 are not useable for FD_SET() [and hence select()] and overwrite the stack.
247 Assorted crashes happen.
249 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
250 main-section options or in a connect ACL, the value from the first ever
251 connection was never replaced for subsequent connections. Found by
254 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
255 though only once PCRE2 was introduced: the memory accounting used under
256 debug offset allocations by an int, giving a hard trap in early startup.
257 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
259 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
260 with underbars is given. The write-protection of configuration introduced
261 in 4.95 trapped when normalisation was applied to an option not needing
264 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
266 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
267 failing to close the certificates directory, every hour or any time it
270 JH/16 Debugging initiated by an ACL control now continues through into routing
271 and transport processes. Previously debugging stopped any time Exim
272 re-execs, or for processing a queued message.
274 JH/17 The "expand" debug selector now gives more detail, specifically on the
275 result of expansion operators and items.
277 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
278 bad comparison between the option value and the name of the file to
279 be included was done, and a mismatch was wrongly identified.
280 4.88 to 4.95 are affected.
282 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
284 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
285 supplied as the db file-pair basename. Previously, if a directory
286 path was given, for example via the autoreply "once" option, the DB
287 file.pag and file.dir files would be created in that directory's
290 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
291 "taint" log_selector. These were previously deprecated.
293 JH/22 Fix static address-list lookups to properly return the matched item.
294 Previously only the domain part was returned.
296 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
297 the call into OpenSSL to send a TLS Close was being repeated; this
298 resulted in the library waiting for the peer's Close. If that was never
299 sent we waited forever. Fix by tracking send calls.
301 JH/24 The ${run} expansion item now expands its command string elements after
302 splitting. Previously it was before; the new ordering makes handling
303 zero-length arguments simpler. The old ordering can be obtained by
304 appending a new option "preexpand", after a comma, to the "run".
306 JH/25 Taint-check exec arguments for transport-initiated external processes.
307 Previously, tainted values could be used. This affects "pipe", "lmtp" and
308 "queryprogram" transport, transport-filter, and ETRN commands.
309 The ${run} expansion is also affected: in "preexpand" mode no part of
310 the command line may be tainted, in default mode the executable name
313 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
314 the the facility was not passed across execs, and only the first message
315 passed over a connection could use BDAT; any further ones using DATA.
317 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
318 uses $sending_ip_address and an interface is specified.
319 Previously any use of the local address in the EHLO name disabled
320 PIPECONNECT, the common case being to use the rDNS of it.
322 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
323 resumption. Previously verify failed because no certificate status is
324 passed on the wire for the restarted session. Fix by using the recorded
325 ocsp status of the stored session for the new connection.
327 JH/29 TLS resumption: the key for session lookup in the client now includes
328 more info that a server could potentially use in configuring a TLS
329 session, avoiding oferring mismatching sessions to such a server.
330 Previously only the server IP was used.
332 JH/30 Fix string_copyn() for limit greater than actual string length.
333 Previously the copied amount was the limit, which could result in a
334 overlapping memcpy for newly allocated destination soon after a
335 source string shorter than the limit. Found/investigated by KM.
337 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
338 close; it may be needed for a subsequent connection. This caused a
339 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
341 JH/32 Fix CHUNKING for a second message on a connection when the first was
342 rejected. Previously we did not reset the chunking-offered state, and
343 erroneously rejected the BDAT command. Investigation help from
346 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
347 an empty address. Previously the expansion returned an error.
349 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
350 proxy. Previously these were misparsed, leading to paniclog entries.
356 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
357 related applications. Previously an "H" was used where available info
358 says that "M" should be, so change to match.
360 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
361 as arguments, so an implementation trying to copy these into a local
362 buffer was taking a taint-enforcement trap. Fix by using dynamically
363 created buffers. Similar fix for radius expansion condition.
365 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
366 reasonable, eg. to count headers. Fix by using dynamically created
367 buffers rather than a local. Do similar fixes for ACL actions "dcc",
368 "log_reject_target", "malware" and "spam"; the arguments are expanded
369 so could be handling tainted values.
371 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
372 broken the (no-op) support for this sendmail command. Restore it
373 to doing nothing, silently, and returning good status.
375 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
376 record path was given (or the default used) without a leading directory
377 path, an error occurred on trying to open it. Use the transport's working
380 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
381 transport. Previously it was the name on the DNS A-record; use instead
382 the head of the CNAME chain leading there (if there is one). This seems
383 to align better with RFC 6125.
385 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
386 smtp_accept_max_per_host allocated resources which were not released
387 when the limit was exceeded. This eventually crashed the daemon. Fix
388 by adding a release action in that path.
390 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
391 expanded; previously using tainted values was rejected. Fix by using
392 dynamically-created buffers.
394 JH/09 Relax restrictions on ACL verify condition needing access to message
395 headers. Previously they were only permitted in data and non-smtp ACLs;
396 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
397 not_blind, header_sender and header_names_ascii verification.
399 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
400 Previously a macro used one argument twice; when called with the
401 argument as an expression having side-effects, incorrect operation
402 resulted. Use an inlineable function.
404 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
405 held open for a verify callout. Previously this wan not accounted for
406 and a corrupt onward SMTP conversation resulted.
408 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
409 Previously they were embedded naively in the constructed address; when
410 needed, strip the quoting and quote the entire local_part.
411 Also make the inbound_srs expansion condition handle quoting.
413 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
414 excluded, not matching the documentation.
416 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
417 was given for the sqlite_dbfile a trap resulted.
419 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
420 "name" argument resulted in a trap. There is no reason to disallow such;
421 this was a coding error.
423 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
424 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
425 spent suspended, ignoring the POSIX definition. Previously we assumed
426 it did and a constant offset from real time could be used as a correction.
427 Change to using the same clock source for the start-of-message and the
428 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
429 exists, just to get a clock slightly more aligned to reality.
431 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
432 RFC says it is optional some validators care. The missing char was not
433 intended but triggered by a line-wrap alignment. Discovery and fix by
434 Guillaume Outters, hacked on by JH.
436 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
437 name being quoted was tainted a trap would be taken. Fix by using
438 dynamically created buffers. The routine could have been called by a
439 rewrite with the "h" flag, by using the "-F" command-line option, or
440 by using a "name=" option on a control=submission ACL modifier.
442 JH/19 SPF: change the Authentication-Results expansion component to give
443 smtp.helo when the sender domain is empty. Previously it gave
446 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
447 not in 127.0.0.0/8 to help in spotting list domains taken over by a
448 domain-parking registrar.
450 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
451 Previously when a whitespace character was specified it was not inserted
452 after removing the newline.
454 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
455 the domain part of the recipient address. This overrides any tls_sni
456 option set, which was previously used.
458 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
461 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
462 is_tainted() had an off-by-one error in the overenthusiastic direction.
463 Find and fix by Gavan. Although NetBSD is not a supported platform for
464 4.94 this bug could affect other platforms.
466 PP/01 Fix default prime selection to be consistent.
467 One path used ike23 still, instead of exim.dev.20160529.3; now both
468 execution flows will use the same DH primes (currently
469 exim.dev.20160529.3).
471 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
472 option in smtp transport, to match the documentation. Previously
473 verification was not being done in some cases where it should have been.
475 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
476 than one server was defined and depending on the platform memory layout
477 details, an internal consistency trap could be hit while walking the list
480 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
481 files. The value is used by the authresults expansion item. Previously
482 if this was used in a router or transport, a crash could result.
484 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
485 have a sufficient incidence of EINTR returns from select that an
486 interaction with other operations done by the main daemon loop exposed
487 a bug in the error-handling. This was benign apart from the log
490 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
491 for consistency with delivered "=>" and failed "**" lines. While we're
492 there, handle PRX and TFO.
494 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
495 applied. This resulted, if any header-line rewrite rules were configured,
496 in a panic-log triggerable by sending a message with a long address in
497 a header. Fix by increasing the arbitrary limit to larger than a single
498 (dewrapped) 5322 header line maximum size.
500 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
501 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
502 RFC 6648 which deprecates X- options in protocols as a general practice.
503 Changeover between the implementations is handled by the mechanisms
506 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
507 a remote which uses callout/hold. Previously the local was queued.
509 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
512 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
513 log line. The asterisk is now consistently appended to the remote IP
514 (and port, if given), and will also be provided on defer and fail log
515 lines. Previously it could be placed on the local IP if that was being
516 logged, and was only provided on delivery lines.
518 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
520 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
521 authenticator client_send option. Previously the next char, after a pair
522 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
523 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
524 documentation. There is still no way to get a leading ^ immediately
525 after a NUL (ie. for the password of a PLAIN method authenticator.
527 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
528 files. For bad sizes read, delete the record and whine to paniclog.
530 JH/38 When logging an AUTH failure, as server, do not include sensitive
531 information. Previously, the credentials would be included if given
532 as part of the AUTH command line and an ACL denied authentication.
534 JH/39 Bug 2691: fix $local_part_data. When the matching list element
535 referred to a file, bad data was returned. This likely also affected
538 JH/40 The gsasl authenticator now supports caching of the salted password
539 generated by the client-side implementation. This required the addition
540 of a new variable: $auth4.
542 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
543 left undeleted; the attempt to re-create it then failed - resulting in
544 the usual "SIGHUP tp have daemon reload configuration" to not work.
545 This affected any platform not supporting "abstract" Unix-domain
546 sockets (i.e. not Linux).
548 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
549 recipients" response to RCPT in a later response, with a 250. The
550 previous coding assumed this would not happen, and under PIPELINING
551 would result in both lost and duplicate recipients for a message.
553 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
554 Previously the weighting was incorrectly applied. Similar fix for socks
555 proxies. Found and fixed by Heiko Schlichting.
557 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
558 not handle sub-lists included using the +namedlist syntax. While
559 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
560 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
563 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
564 Previously the permanent pool was used, so the sore could not be freed.
565 This meant a connection with many messages would use continually-growing
568 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
569 per-pool so as not to waste too much space. Previously a constant size
570 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
571 DOS attacks harder. The cost is wasted memory use in the larger blocks.
573 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
574 same space repeatedly, and a smaller process footprint.
576 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
577 for GnuTLS. Previously the name started "%s" which, while not a bug,
578 looked as if if might be one.
580 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
581 "continued connection") make the $proxy_* variables available. Previously
582 the information was not passed across the exec() call for subsequent
583 transport executions. This also mean that the log lines for the
584 messages can show the proxy information.
586 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
587 receive time. With modern systems the difference is significant.
588 The historical behaviour can be restored by disabling (a new) log_selector
589 "queue_time_exclusive".
591 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
592 content) were specifically tested for. Now, also cover expansions
593 resulting in ACL names and inline ACL content.
595 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
596 making it unusable in complex expressions.
598 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
599 fail with a taint-check on the spoolfile name, and leave the message
602 HS/01 Enforce absolute PID file path name.
604 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
606 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
608 PP/02 Bug 2643: Correct TLS DH constants.
609 A missing NUL termination in our code-generation tool had led to some
610 incorrect Diffie-Hellman constants in the Exim source.
611 Reported by kylon94, code-gen tool fix by Simon Arlott.
613 PP/03 Impose security length checks on various command-line options.
614 Fixes CVE-2020-SPRSS reported by Qualys.
616 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
617 better. Reported by Qualys.
619 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
620 providing a particularly obnoxious sender full name.
623 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
625 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
628 PP/08 Change default for recipients_max from unlimited to 50,000.
630 PP/09 Fix security issue with too many recipients on a message (to remove a
631 known security problem if someone does set recipients_max to unlimited,
632 or if local additions add to the recipient list).
633 Fixes CVE-2020-RCPTL reported by Qualys.
635 PP/10 Fix security issue in SMTP verb option parsing
636 Fixes CVE-2020-EXOPT reported by Qualys.
638 PP/11 Fix security issue in BDAT state confusion.
639 Ensure we reset known-good where we know we need to not be reading BDAT
640 data, as a general case fix, and move the places where we switch to BDAT
641 mode until after various protocol state checks.
642 Fixes CVE-2020-BDATA reported by Qualys.
644 HS/03 Die on "/../" in msglog file names
646 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
647 the Exim runtime user are allowed to create files.
649 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
652 QS/03 When reading the output from interpreted forward files we do not
653 pass the pipe between the parent and the interpreting process to
654 executed child processes (if any).
656 QS/04 Always die if requested from internal logging, even is logging is
659 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
660 the API; compilation noo longer completes with DMARC support included.
661 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
662 been reported on other platforms.
664 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
666 JH/56 Make the majority of info read from config files readonly, for defence-in-
667 depth against exploits. Suggestion by Qualys.
668 Not supported on Solaris 10.
670 JH/57 Fix control=fakreject for a custom message containing tainted data.
671 Previously this resulted in a log complaint, due to a re-expansion present
672 since fakereject was originally introduced.
674 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
675 modifier was given, a loop resulted.
677 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
678 pipelined SMTP command followed the BDAT LAST then it would be
679 incorrectly treated as part of the message body, causing a verification
682 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
683 require looser validation rules than those for 821-level addresses,
684 which only permit IP addresses.
690 JH/01 Avoid costly startup code when not strictly needed. This reduces time
691 for some exim process initialisations. It does mean that the logging
692 of TLS configuration problems is only done for the daemon startup.
694 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
696 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
697 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
700 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
701 DKIM signing is being done. Previously a transport_filter would always
702 disable CHUNKING, falling back to traditional DATA.
704 JH/05 Regard command-line recipients as tainted.
706 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
708 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
709 PAM library frees one of the arguments given to it, despite the
710 documentation. Therefore a plain malloc must be used.
712 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
713 on-stack buffers were used, resulting in a taint trap when DSN information
714 copied from a received message was written into the buffer.
716 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
717 the ordering of its ARC headers. This caused a crash.
719 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
720 a new record was being constructed with information from the peer, a trap
723 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
724 installation would get error messages from DMARC verify, when it hit the
725 nonexistent file indicated by the default. Distros wanting DMARC enabled
726 should both provide the file and set the option.
727 Also enforce no DMARC verification for command-line sourced messages.
729 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
730 could, depending on the platform, hang at the STARTTLS response.
732 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
733 message on a connection. Previously if one message had ARC headers and
734 the following one did not, a crash could result when adding an
735 Authentication-Results: header.
737 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
738 Exim main code and Exim-related utities. The introduction of taint
739 tracking also did many adjustments to string handling. Since then, eximon
740 frequently terminated with an assert failure.
742 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
743 check for 452 responses. This slightly helps the inefficieny of doing
744 a large alias-expansion into a recipient-limited target. The max_rcpt
745 transport option still applies (and at the current default, will override
746 the new feature). The check is done for either cause of synch, and forces
747 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
748 connection. The new facility is not tunable at this time.
750 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
751 library live data was being used, so the results became garbage. Make
752 copies while it is still usable.
754 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
755 on delivery deferred (==) and failed (**) lines (if a delivery was
756 attemtped). Previously it was only on completion (=>) lines.
758 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
759 for the expansion of the server_scram_iter and server_scram_salt options.
761 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
762 are now specifically given a NO_DATA response without hitting the system
763 resolver. The library goes on to do the now-standard TXT lookup.
764 Use of dnsdb lookups is not affected.
766 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
767 only retrieve the errormessage once. Previously two calls to dlerror()
768 were used, and the second one (for mainlog/paniclog) retrieved null
771 JH/20 Taint checking: disallow use of tainted data for
772 - the appendfile transport file and directory options
773 - the pipe transport command
774 - the autoreply transport file, log and once options
775 - file names used by the redirect router (including filter files)
777 - paths used by single-key lookups
778 Previously this was permitted.
780 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
781 adjusted the size of a major service buffer; this failed because the
782 buffer was in use at the time. Change to a compile-time increase in the
783 buffer size, when this authenticator is compiled into exim.
785 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
786 previous fast-mode was untenable in the face of glibs using mmap to
787 support larger malloc requests.
789 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
790 New values supported, if defined on system where compiled:
791 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
792 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
794 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
795 running a limited number of proceses in parallel, a benefit is gained. The
796 amount varies with the platform hardware and load. The use of the option
797 queue_run_in_order means we cannot do this, as ordering becomes
800 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
801 had introduced a string-copy (for ensuring NUL-termination) which was not
802 appropriate for that case, which can include embedded NUL bytes in the
803 block of data. Investigation showed the copy to actually be needless, the
804 data being length-specified.
806 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
807 done during a receiving connection, and both used TLS, global info was
808 used rather than per-connection info for tracking the state of data
809 queued for transmission. This could result in a connection hang.
811 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
812 Previously, when delivering serveral messages down a single connection
813 only the first would provide a SIZE. This was due to the size information
814 not being properly tracked.
816 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
817 TAI (at 37 seconds currently), pretend to be in UTC for time-related
818 expansion and logging. Previously, spurious values such as a future
819 minute could be seen.
821 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
822 it could crash from a null-deref. This could also affect the
823 ${addresses: } operator and ${readsock } item.
825 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
826 message following a mime one, the variable was not reset.
828 JH/30 When an pipelined-connect fails at the first response, assume incorrect
829 cached capability (perhaps the peer reneged?) and immediately retry in
832 JH/31 Fix spurious detection of timeout while writing to transport filter.
834 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
835 an attempt to copy the string was made before checking it.
837 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
838 taint of the lookup key was maintained; we now regard the presence in the
839 filesystem as sufficient validation.
841 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
842 argument is supplied.
844 JH/35 The dsearch lookup now requires that the directory is an absolute path.
845 Previously this was not checked, and nonempty relative paths made an
846 access under Exim's current working directory.
848 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
849 Previously no event was raised.
851 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
852 parameter supplied by the sender MAIL FROM command. Previously it was
853 ignored, and only the check_spool_space option value for the required
856 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
857 the size of the signing public-key. Previously it was instead giving
858 the size of the signature hash.
860 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
861 the default. See the (new) dkim_verify_min_keysizes option.
863 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
864 and an ACL use a lookup for checking either the local_part or domain,
865 stale data could be accessed. Ensure that variable references are
866 dropped between messages.
868 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
869 by the client was not checked as pointing within response data before
870 being used. A malicious client could thus cause an out-of-bounds read and
871 possibly gain authentication. Fix by adding the check.
873 JH/42 Internationalisation: change the default for downconversion in the smtp
874 transport to be "if needed". Previously it was "as previously set" for
875 the message, which usually meant "if needed" for message-submission but
876 "no" for everything else. However, MTAs have been seen using SMTPUTF8
877 even when the envelope addresses did not need it, resulting in forwarding
878 failures to non-supporting MTAs. A downconvert in such cases will be
879 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
880 transport. The change does mean that addresses needing conversion will
881 be converted when previously a delivery failure would occur.
883 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
884 response was received it would be used unchecked in a fail-DSN, violating
885 standards on line-length limits. Truncate if needed.
887 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
888 doesn't work. (Additionally add a new main config option to configure the
895 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
896 side, to decode a TLS 1.3 packet capture.
898 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
899 Previously the default library behaviour applied, sending two, each in
902 JH/03 Debug output for ACL now gives the config file name and line number for
905 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
907 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
909 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
910 buffer overrun for (non-chunking) other transports.
912 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
913 TLS1.3, means that a server rejecting a client certificate is not visible
914 to the client until the first read of encrypted data (typically the
915 response to EHLO). Add detection for that case and treat it as a failed
916 TLS connection attempt, so that the normal retry-in-clear can work (if
917 suitably configured).
919 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
920 and/or domain. Found and fixed by Jason Betts.
922 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
923 configuration). If a CNAME target was not a wellformed name pattern, a
926 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
927 the OS reports them interleaved with other addresses.
929 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
930 used both for input and for a verify callout, both encrypted, SMTP
931 responses being sent by the server could be lost. This resulted in
932 dropped connections and sometimes bounces generated by a peer sending
935 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
936 string. Previously it was possible to cause undefined behaviour in a
937 library routine (usually a crash). Found by "zerons".
939 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
942 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
943 API was removed, so update to use the newer ones.
945 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
946 any timeout set, is taking a long time. Previously we would hang on to a
947 rotated logfile "forever" if the input was arriving with long gaps
948 (a previous attempt to fix addressed lack, for a long time, of initial
951 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
952 shared (NFS) environment. The length of the tempfile name is now
953 4 + 16 ("hdr.$message_exim_id") which might break on file
954 systems which restrict the file name length to lower values.
957 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
958 shared (NFS) environment.
960 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
961 did for all versions <4.90). Notably -M, -m, --invert, -I may be
964 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
965 on some platforms for bit 31.
967 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
968 to changes apparently associated with TLS1.3 handling some of the APIs
969 previously used were either nonfunctional or inappropriate. Strings
970 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
971 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
972 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
973 This affects log line X= elements, the $tls_{in,out}_cipher variables,
974 and the use of specific cipher names in the encrypted= ACL condition.
976 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
978 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
979 verification result was not updated unless hosts_require_ocsp applied.
981 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
982 queue_list_requires_admin set to false, non-admin users were denied the
985 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
986 directory-of-certs mode. Previously they were advertised despite the
989 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
990 A single TCP connection by a client will now hold a TLS connection open
991 for multiple message deliveries, by default. Previously the default was to
994 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
995 default. If built with the facility, DANE will be used. The facility
996 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
998 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
999 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1000 must be defined and you must still, unless you define DISABLE_TLS, manage
1001 the the include-dir and library-file requirements that go with that
1002 choice. Non-TLS builds are still supported.
1004 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1007 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1008 default. If the platform supports and has the facility enabled, it will
1009 be requested on all coneections.
1011 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1012 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1014 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1016 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1017 success-DSN messages. Previously the From: header was always the default
1018 one for these; the option was ignored.
1020 JH/28 Fix the timeout on smtp response to apply to the whole response.
1021 Previously it was reset for every read, so a teergrubing peer sending
1022 single bytes within the time limit could extend the connection for a
1023 long time. Credit to Qualsys Security Advisory Team for the discovery.
1025 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1026 delivery address, which leaked information of the results of local
1027 forwarding. Change to the original envelope recipient address, per
1030 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1031 requested. Previously not bounce was generated and a log entry of
1032 error ignored was made.
1034 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1036 JH/32 Introduce a general tainting mechanism for values read from the input
1037 channel, and values derived from them. Refuse to expand any tainted
1038 values, to catch one form of exploit.
1040 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1041 was unused and the unexpanded text used for the test. Found and
1042 fixed by Ruben Jenster.
1044 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1045 an attempt to use a TLS library read routine dereffed a nul pointer,
1048 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1049 them in our smtp response.
1051 JH/36 Have the general router option retry_use_local_part default to true when
1052 any of the restrictive preconditions are set (to anything). Previously it
1053 was only for check_local user. The change removes one item of manual
1054 configuration which is required for proper retries when a remote router
1055 handles a subset of addresses for a domain.
1057 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1058 link count into consideration.
1060 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1061 caused the extension of big_buffer, the following lines were ignored.
1063 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1064 accordance with RFC 2308. Previously there was no expiry, so a longlived
1065 receive process (eg. due to ACL delays) versus a short SOA value could
1068 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1070 JH/39 Promote DMARC support to mainline.
1072 JH/40 Bug 2452: Add a References: header to DSNs.
1074 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1075 parameters. The relevant library call is documented as "Deprecated: This
1076 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1077 3.6.0, DH parameters are negotiated following RFC7919."
1079 HS/06 Change the default of dnssec_request_domains to "*"
1081 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1082 carried on and emitted a BDAT command, even when PIPELINING was not
1085 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1086 buffer was used for the filename, resulting in a trap when tainted
1087 arguments (eg. $domain) were used.
1089 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1090 recommended to avoid a possible server-load attack. The feature can be
1091 re-enabled via the openssl_options main cofiguration option.
1093 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1094 for version 4.90 - adding a "more data" boolean to the arguments.
1095 Bumped the ABI version number also, this having been missed previously;
1096 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1097 respect of usage of smtp_printf() by either local_scan code or libraries
1098 accessed via the ${dlfunc } expansion item. Both will need coding
1099 adjustment for any calls to smtp_printf() to match the new function
1100 signature; a FALSE value for the new argument is always safe.
1102 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1103 the file-offset (which the Linux syscall does, and exim expects); this
1104 resulted in an indefinite loop.
1106 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1107 to do ARC verification. The Authentication-Results: header line added
1108 by the configuration then had no ARC item.
1110 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1111 an incorrect mode was used for reading the result, resulting in it being
1118 JH/01 Remove code calling the customisable local_scan function, unless a new
1119 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1121 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1122 non-signal-safe functions being used.
1124 JH/03 Bug 2269: When presented with a received message having a stupidly large
1125 number of DKIM-Signature headers, disable DKIM verification to avoid
1126 a resource-consumption attack. The limit is set at twenty.
1128 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1129 report of oldest_pass in ${authres } in consequence, and separate out
1130 some descriptions of reasons for verification fail.
1132 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1133 files in the spool were present and unlocked. A queue-runner could spot
1134 them, resulting in a duplicate delivery. Fix that by doing the unlock
1135 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1136 add more error-checking on spoolfile handling while that code is being
1139 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1140 No known attacks, no CVE, this is defensive hardening.
1142 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1143 a queue-runner could start a delivery while other operations were ongoing.
1144 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1145 Found and investigated by Tim Stewart. Fix by using the open message data
1146 file handle rather than opening another, and not locally closing it (which
1147 releases a lock) for that case, while creating the temporary .eml format
1148 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1150 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1151 $sender_verify_failure/$recipient_verify_failure to "random".
1153 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1156 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1157 Previously this would segfault.
1159 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1162 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1163 like zero, since the resolver should be doing this for us, But we need one
1164 as a CNAME but no MX presence gets the CNAME returned; we need to check
1165 that doesn't point to an MX to declare it "no MX returned" rather than
1166 "error, loop". A new main option is added so the older capability of
1167 following some limited number of chain links is maintained.
1169 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1171 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1172 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1173 now-deprecated earlier definitions used only specified the range up to TLS
1174 1.2 (in the older-version library docs).
1176 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1178 JH/15 Rework TLS client-side context management. Stop using a global, and
1179 explicitly pass a context around. This enables future use of TLS for
1180 connections to service-daemons (eg. malware scanning) while a client smtp
1181 connection is using TLS; with cutthrough connections this is quite likely.
1183 JH/16 Fix ARC verification to do AS checks in reverse order.
1185 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1187 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1188 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1189 Previously the "utf8" would be re-prepended for every additional message.
1191 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1192 Previously thery were accepted, resulting in issues when attempting to
1193 forward messages to a non-supporting MTA.
1195 PP/02 Let -n work with printing macros too, not just options.
1197 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1198 one parent address was copied, and bogus data was used at delivery-logging
1199 time. Either a crash (after delivery) or bogus log data could result.
1200 Discovery and analysis by Tim Stewart.
1202 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1203 Previously if the string ended mid-character, we did not insert the
1204 promised '?' replacement.
1206 PP/04 Documentation: current string operators work on bytes, not codepoints.
1208 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1209 should pack well giving a smaller memory footprint so better caching and
1210 therefore performance. Group the declarations where this can't be done so
1211 that the byte-sized flag variables are not interspersed among pointer
1212 variables, giving a better chance of good packing by the compiler.
1214 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1215 non-null, to avoid issues with sites running BATV. Previously reports were
1216 sent with an empty envelope sender so looked like bounces.
1218 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1219 The ignore_error flag wasn't being returned from the filter subprocess so
1220 was not set for later routers. Investigation and fix by Matthias Kurz.
1222 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1223 and a msg:complete for the whole, when a message is manually removed using
1224 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1226 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1227 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1228 hope that is portable enough.
1230 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1231 requiring a known-CA anchor certificate; make it now rely entirely on the
1232 TLSA as an anchor. Checking the name on the leaf cert against the name
1233 on the A-record for the host is still done for TA (but not for EE mode).
1235 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1236 included in delivery lines for non-proxied connections, when compiled with
1237 SUPPORT_SOCKS and running with proxy logging enabled.
1239 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1240 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1241 move the existing event to fire before the normal logging of message
1242 failure so that custom logging is bracketed by normal logging.
1244 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1245 msg:fail:internal event. Developement by Matthias Kurz.
1247 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1248 far too small for todays use of crypto signatures stored there. Go all
1249 the way to the max DNS message size of 64kB, even though this might be
1250 overmuch for IOT constrained device use.
1252 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1253 copy a string over itself. The library routine is documented as not
1254 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1256 JH/32 For main options check_spool_space and check_inode_space, where the
1257 platform supports 64b integers, support more than the previous 2^31 kB
1258 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1259 the previous G, M, k.
1261 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1262 $authenticated_fail_id variable on authentication failure. Previously
1265 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1266 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1267 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1268 GNUTLS_SEC_PARAM_MEDIUM.
1270 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1271 side. Previously we would continue as if no SNI had been received.
1273 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1274 "<" character, which should be a list-separator specification, we walked
1275 off past the nul-terimation.
1277 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1278 causes) even when the retry time is not yet met. Previously they were
1279 not, meaning that when (say) an account was over-quota and temp-rejecting,
1280 and multiple senders' messages were queued, only one sender would get
1281 notified on each configured delay_warning cycle.
1283 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1285 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1286 cipher-suites, an error can be left on the stack even for a succeeding
1287 accept; this results in impossible error messages when a later operation
1290 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1291 return error codes indicating retry. Under TLS1.3 this becomes required.
1293 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1294 it only wrote the new authenticators, resulting in a lack of tracking of
1295 peer changes of ESMTP extensions until the next cache flush.
1297 JH/41 Fix the loop reading a message header line to check for integer overflow,
1298 and more-often against header_maxsize. Previously a crafted message could
1299 induce a crash of the recive process; now the message is cleanly rejected.
1301 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1302 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1308 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1309 When redis_servers is set to a list of > 1 element, and the Redis servers
1310 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1311 case of MOVED into a DEFER case instead, thus moving the query onto the
1312 next server in the list. For a cluster of N elements, all N servers must
1313 be defined in redis_servers.
1315 GF/02 Catch and remove uninitialized value warning in exiqsumm
1316 Check for existence of @ARGV before looking at $ARGV[0]
1318 JH/01 Replace the store_release() internal interface with store_newblock(),
1319 which internalises the check required to safely use the old one, plus
1320 the allocate and data copy operations duplicated in both (!) of the
1321 extant use locations.
1323 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1324 modifier. This matches the restriction on the commandline.
1326 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1327 Previously only the last row was returned.
1329 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1330 we assumed that tags in the header were well-formed, and parsed the
1331 element content after inspecting only the first char of the tag.
1332 Assumptions at that stage could crash the receive process on malformed
1335 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1336 While running the DKIM ACL we operate on the Permanent memory pool so that
1337 variables created with "set" persist to the DATA ACL. Also (at any time)
1338 DNS lookups that fail create cache records using the Permanent pool. But
1339 expansions release any allocations made on the current pool - so a dnsdb
1340 lookup expansion done in the DKIM ACL releases the memory used for the
1341 DNS negative-cache, and bad things result. Solution is to switch to the
1342 Main pool for expansions.
1343 While we're in that code, add checks on the DNS cache during store_reset,
1344 active in the testsuite.
1345 Problem spotted, and debugging aided, by Wolfgang Breyha.
1347 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1348 When none of the hosts presented to a transport match an already-open
1349 connection, close it and proceed with the list. Previously we would
1350 queue the message. Spotted by Lena with Yahoo, probably involving
1353 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1354 Previously a spurious "250 OK id=" response was appended to the proper
1357 JH/08 The "support for" informational output now, which built with Content
1358 Scanning support, has a line for the malware scanner interfaces compiled
1359 in. Interface can be individually included or not at build time.
1361 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1362 by the template makefile "src/EDITME". The "STREAM" support for an older
1363 ClamAV interface method is removed.
1365 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1366 rows affected is given instead).
1368 JH/11 The runtime Berkeley DB library version is now additionally output by
1369 "exim -d -bV". Previously only the compile-time version was shown.
1371 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1372 SMTP connection. Previously, when one had more recipients than the
1373 first, an abortive onward connection was made. Move to full support for
1374 multiple onward connections in sequence, handling cutthrough connection
1375 for all multi-message initiating connections.
1377 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1378 routers. Previously, a multi-recipient message would fail to match the
1379 onward-connection opened for the first recipient, and cause its closure.
1381 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1382 a timeout on read on a GnuTLS initiating connection, resulting in the
1383 initiating connection being dropped. This mattered most when the callout
1384 was marked defer_ok. Fix to keep the two timeout-detection methods
1387 JH/15 Relax results from ACL control request to enable cutthrough, in
1388 unsupported situations, from error to silently (except under debug)
1389 ignoring. This covers use with PRDR, frozen messages, queue-only and
1392 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1394 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1395 metadata, resulting in a crash in free().
1397 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1398 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1399 Broken also in d185889f4, with init system revamp.
1401 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1402 Previously we abruptly closed the connection after reading a malware-
1403 found indication; now we go on to read the "scan ok" response line,
1406 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1407 ACL. Previously, a crash would result.
1409 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1410 macro text after a replacement (previously it was only once per line) and
1411 by skipping builtin macros when searching for an uppercase lead character.
1413 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1414 for the build is renamed.
1416 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1417 was allocated for every new TLS startup, meaning one per message. Fix
1418 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1420 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1421 reported the original. Fix to report (as far as possible) the ACL
1422 result replacing the original.
1424 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1425 OpenSSL. Certificate information is loaded for every new TLS startup,
1426 and the resources needed to be freed.
1428 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1430 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1433 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1434 DATA response info to the (existing) per-recipient response info for
1435 the "C=" log element. It can have useful tracking info from the
1436 destination system. Patch from Simon Arlott.
1438 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1439 length value. Previously this would segfault.
1441 HS/02 Support Avast multiline protoocol, this allows passing flags to
1442 newer versions of the scanner.
1444 JH/28 Ensure that variables possibly set during message acceptance are marked
1445 dead before release of memory in the daemon loop. This stops complaints
1446 about them when the debug_store option is enabled. Discovered specifically
1447 for sender_rate_period, but applies to a whole set of variables.
1448 Do the same for the queue-runner and queue-list loops, for variables set
1449 from spool message files. Do the same for the SMTP per-message loop, for
1450 certain variables indirectly set in ACL operations.
1452 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1453 as a multi-recipient message from a mailinglist manager). The coding had
1454 an arbitrary cutoff number of characters while checking for more input;
1455 enforced by writing a NUL into the buffer. This corrupted long / fast
1456 input. The problem was exposed more widely when more pipelineing of SMTP
1457 responses was introduced, and one Exim system was feeding another.
1458 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1459 receiving system, and refused recipients seen by the sending system
1460 (propating to people being dropped from mailing lists).
1461 Discovered and pinpointed by David Carter.
1463 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1464 replaced by the ${authresults } expansion.
1466 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1468 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1469 allows proper process termination in container environments.
1471 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1472 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1474 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1475 and "err_perm", deprecated since 4.83 when the RFC-defined words
1476 "temperror" and "permerror" were introduced.
1478 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1479 transport-filters or DKIM-signing. The restriction was lost in the
1480 consolidation of verify-callout and delivery SMTP handling.
1481 Extend the restriction to also cover ARC-signing.
1483 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1484 in defer=pass mode supply a 450 to the initiator. Previously the message
1487 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1488 tls_require_ciphers is used as before.
1490 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1491 "pass_unscanned". Only tmpfails from the scanner are written to
1492 the paniclog, as they may require admin intervention (permission
1493 denied, license issues). Other scanner errors (like decompression
1494 bombs) do not cause a paniclog entry.
1496 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1497 Previously it was possible to log spurious information in receive log
1500 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1501 triggered odd behaviour from Outlook Express clients.
1503 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1506 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1507 since the IETF WG has not yet settled on that versus the original
1508 "bare" representation.
1510 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1511 Previously the millisecond value corrupted the output.
1512 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1513 corrupted the output.
1519 JH/01 Rework error string handling in TLS interface so that the caller in
1520 more cases is responsible for logging. This permits library-sourced
1521 string to be attached to addresses during delivery, and collapses
1522 pairs of long lines into single ones.
1524 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1525 during configuration. Wildcards are allowed and expanded.
1527 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1528 This permits better logging.
1530 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1531 do not close it down and have the child transport start it up again on
1532 the passed-on TCP connection. Instead, proxy the child (and any
1533 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1534 affected: the continued delivery log lines do not have any DNSSEC, TLS
1535 Certificate or OCSP information. TLS cipher information is still logged.
1537 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1538 identical IP addresses on different listening ports. Will also affect
1541 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1542 add noisy ifdef guards to special-case this sillyness.
1543 Patch from Bernd Kuhls.
1545 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1546 than 255 are no longer allowed.
1548 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1549 Disable the session-cache too, which might reduce our load. Since we
1550 currrectly use a new context for every connection, both as server and
1551 client, there is no benefit for these.
1552 GnuTLS appears to not support tickets server-side by default (we don't
1553 call gnutls_session_ticket_enable_server()) but client side is enabled
1554 by default on recent versions (3.1.3 +) unless the PFS priority string
1557 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1558 <https://reproducible-builds.org/specs/source-date-epoch/>.
1560 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1561 the check for any unsuccessful recipients did not notice the limit, and
1562 erroneously found still-pending ones.
1564 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1565 MSG_MORE. Only in-clear (not on TLS connections).
1567 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1568 transport-filter is involved we can buffer the headers in memory for
1569 creating the signature, and read the spool data file once for the
1570 signature and again for transmission.
1572 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1573 4.77 as the kernel support then wasn't solid, having issues in 64bit
1574 mode. Now, it's been long enough. Add support for FreeBSD also.
1576 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1577 case where the routing stage had gathered several addresses to send to
1578 a host before calling the transport for the first, we previously failed
1579 to close down TLS in the old transport process before passing the TCP
1580 connection to the new process. The new one sent a STARTTLS command
1581 which naturally failed, giving a failed delivery and bloating the retry
1582 database. Investigation and fix prototype from Wolfgang Breyha.
1584 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1585 false-negatives in the check that the sender had not preempted a response
1586 or prompt from Exim (running as a server), due to that code's lack of
1587 awareness of the SMTP input buffering.
1589 PP/04 Add commandline_checks_require_admin option.
1590 Exim drops privileges sanely, various checks such as -be aren't a
1591 security problem, as long as you trust local users with access to their
1592 own account. When invoked by services which pass untrusted data to
1593 Exim, this might be an issue. Set this option in main configuration
1594 AND make fixes to the calling application, such as using `--` to stop
1597 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1598 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1599 responses to those, into a single TLS record each way (this usually means
1600 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1603 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1604 affects you only if you're dancing at the edge of the param size limits.
1605 If you are, and this message makes sense to you, then: raise the
1606 configured limit or use OpenSSL 1.1. Nothing we can do for older
1609 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1610 cmdline element to get the documented default one. Previously it was
1613 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1614 are unsuable for non-message-specific reasons - eg. connection timeout,
1615 banner-time rejection.
1617 JH/16 Fix logging of delivery remote port, when specified by router, under
1620 PP/06 Repair manualroute's ability to take options in any order, even if one
1621 is the name of a transport.
1624 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1626 JH/17 Change the list-building routines interface to use the expanding-string
1627 triplet model, for better allocation and copying behaviour.
1629 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1630 Previously it was constructed the first time a possibly-matching string
1631 was met in the configuration file input during startup; now it is done
1634 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1635 compatible one, to avoid the (poorly documented) possibility of a config
1636 file in the working directory redirecting the DB files, possibly correpting
1637 some existing file. CVE-2017-10140 assigned for BDB.
1639 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1640 cache-hot. Previously, although the result was properly cached, the
1641 initial verify call returned a defer.
1643 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1644 the main verify for receipient in uncached-mode.
1646 JH/22 Retire historical build files to an "unsupported" subdir. These are
1647 defined as "ones for which we have no current evidence of testing".
1649 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1650 if present. Previously it was ignored.
1652 JH/24 Start using specified-initialisers in C structure init coding. This is
1653 a C99 feature (it's 2017, so now considered safe).
1655 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1656 if was a fixed-sized field and bitmask ops via macros; it is now more
1659 PP/07 GitHub PR 56: Apply MariaDB build fix.
1660 Patch provided by Jaroslav Škarvada.
1662 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1663 during Coverity cleanups [4.87 JH/47]
1664 Diagnosis and fix provided by Michael Fischer v. Mollard.
1666 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1667 the right size to place the terminating semicolon on its own folded
1668 line, the header hash was calculated to an incorrect value thanks to
1669 the (relaxed) space the fold became.
1671 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1672 and confused the parent.
1674 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1675 which could crash as a result. This could lead to undeliverable messages.
1677 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1680 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1681 matters on fast-turnover and PID-randomising systems, which were getting
1682 out-of-order delivery.
1684 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1685 a possibly-overlapping copy. The symptom was that "Remote host closed
1686 connection in response to HELO" was logged instead of the actual 4xx
1689 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1690 Previously only that bufferd was discarded, resulting in SYMTP command
1693 JH/32 DKIM: when a message has multiple signatures matching an identity given
1694 in dkim_verify_signers, run the dkim acl once for each. Previously only
1695 one run was done. Bug 2189.
1697 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1698 "panic the current process" to "deliberately defer". The panic log is
1699 still written with the problem list name; the mail and reject logs now
1700 get a temp-reject line for the message that was being handled, saying
1701 something like "domains check lookup or other defer". The SMTP 451
1702 message is still "Temporary local problem".
1704 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1705 A crafted sequence of BDAT commands could result in in-use memory beeing
1706 freed. CVE-2017-16943.
1708 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1709 from SMTP input. Previously it was always done; now only done for DATA
1710 and not BDAT commands. CVE-2017-16944.
1712 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1713 to the message (such as an overlong header line). Previously this was
1714 not done and we did not exit BDAT mode. Followon from the previous item
1715 though a different problem.
1721 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1722 than -2003 did; needs libidn2 in addition to libidn.
1724 JH/02 The path option on a pipe transport is now expanded before use.
1726 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1727 Patch provided by "Björn", documentation fix added too.
1729 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1730 missing a wire-to-host endian conversion.
1732 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1733 close after a BDAT command line could be taken as a following command,
1734 giving a synch failure. Fix by only checking for synch immediately
1735 before acknowledging the chunk.
1737 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1738 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1739 macro. Patches provided by Josh Soref.
1741 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1742 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1743 by IANA as a service extension. However, John Klensin suggests that we
1746 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1747 the dkim code may be unix-mode line endings rather than smtp wire-format
1748 CRLF, so prepend a CR to any bare LF.
1750 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1751 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1753 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1754 the first were themselves being wrongly included in the feed into dkim
1755 processing; with most chunk sizes in use this resulted in an incorrect
1756 body hash calculated value.
1758 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1759 DKIM signature block, for verification. Although advised against by
1760 standards it is specifically not ruled illegal.
1762 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1764 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1765 missing a body hash (the bh= tag).
1767 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1768 It seems that HAProxy sends the Proxy Protocol information in clear and
1769 only then does a TLS startup, so do the same.
1771 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1772 TCP connections (such as for Spamd) unless the daemon successfully set
1773 Fast Open mode on its listening sockets. This fixes breakage seen on
1774 too-old kernels or those not configured for Fast Open, at the cost of
1775 requiring both directions being enabled for TFO, and TFO never being used
1776 by non-daemon-related Exim processes.
1778 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1779 endings, at least on the first header line. Try to canonify any that get
1780 past that check, despite the cost.
1782 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1783 now limited to an arbitrary five deep, while parsing addresses with the
1784 strip_excess_angle_brackets option enabled.
1786 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1787 instead leave the unprompted TLS handshake in socket buffer for the
1788 TLS library to consume.
1790 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1792 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1794 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1795 region in one operation, for speed, and this leaves assigned pointers
1796 dangling. Add checks run only under the testsuite which checks all
1797 variables at a store-reset and panics on a dangling pointer; add code
1798 explicitly nulling out all the variables discovered. Fixes one known
1799 bug: a transport crash, where a dangling pointer for $sending_ip_address
1800 originally assigned in a verify callout, is re-used.
1802 PP/06 Drop '.' from @INC in various Perl scripts.
1804 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1806 PP/08 Reduce a number of compilation warnings under clang; building with
1807 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1808 should be warning-free.
1810 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1812 HS/01 Fix portability problems introduced by PP/08 for platforms where
1813 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1815 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1816 chunk. This allows us to accept broken chunked messages. We need a more
1817 general solution here.
1819 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1820 already-broken messages in the queue.
1822 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1824 JH/19 Fix reference counting bug in routing-generated-address tracking.
1830 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1831 supports it and a size is available (ie. the sending peer gave us one).
1833 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1834 years of being deprecated). The replacements are the ACLs
1835 acl_smtp_mime and acl_not_smtp_mime.
1837 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1838 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1839 or even an in-clear connection were permitted. Now, if the host lookup
1840 was dnssec and dane was requested then the host is only used if the
1841 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1842 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1843 if one fails this test.
1844 This means that a poorly-configured remote DNS will make it incommunicado;
1845 but it protects against a DNS-interception attack on it.
1847 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1848 non-noisy when a race steals the message being considered.
1850 JH/05 If main configuration option tls_certificate is unset, generate a
1851 self-signed certificate for inbound TLS connections.
1853 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1854 in rewrites and routers.
1856 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1857 and logged a warning sing 4.83; now they are a configuration file error.
1859 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1860 (lacking @domain). Apply the same qualification processing as RCPT.
1862 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1864 JH/10 Support ${sha256:} applied to a string (as well as the previous
1867 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1868 a cutthrough deliver is pending, as we always want to make a connection.
1869 This also avoids re-routing the message when later placing the cutthrough
1870 connection after a verify cache hit.
1871 Do not update it with the verify result either.
1873 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1874 when routing results in more than one destination address.
1876 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1877 signing (which inhibits the cutthrough capability). Previously only
1878 the presence of an option was tested; now an expansion evaluating as
1879 empty is permissible (obviously it should depend only on data available
1880 when the cutthrough connection is made).
1882 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1883 the relevant preceding SMTP command did not note the pipelining mode.
1885 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1886 Previously they were not counted.
1888 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1889 as one having no matching records. Previously we deferred the message
1890 that needed the lookup.
1892 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1893 distinguished as "(=".
1895 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1896 for missing MX records. Previously it only worked for missing A records.
1898 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1900 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1901 after the data-go-ahead and data-ack. Patch from Jason Betts.
1903 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1904 even for a "none" policy. Patch from Tony Meyer.
1906 JH/22 Fix continued use of a connection for further deliveries. If a port was
1907 specified by a router, it must also match for the delivery to be
1910 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1911 When one of the recipients of a message was unsuitable for the connection
1912 (has no matching addresses), we lost track of needing to mark it
1913 deferred. As a result mail would be lost.
1915 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1917 JH/25 Decoding ACL controls is now done using a binary search; the source code
1918 takes up less space and should be simpler to maintain. Merge the ACL
1919 condition decode tables also, with similar effect.
1921 JH/26 Fix problem with one_time used on a redirect router which returned the
1922 parent address unchanged. A retry would see the parent address marked as
1923 delivered, so not attempt the (identical) child. As a result mail would
1926 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1927 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1928 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1931 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1932 Main config options check_{log,spool}_{inodes,space} are now
1933 100 inodes, 10MB unless set otherwise in the configuration.
1935 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1936 Previously it only applied to the main-section connection policy
1939 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1941 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1942 by me. Added RFC7919 DH primes as an alternative.
1944 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1945 are not in the system include path.
1947 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1948 GnuTLS, when a session startup failed (eg because the client disconnected)
1949 Exim did stdio operations after fclose. This was exposed by a recent
1950 change which nulled out the file handle after the fclose.
1952 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1953 signed directly by the cert-signing cert, rather than an intermediate
1954 OCSP-signing cert. This is the model used by LetsEncrypt.
1956 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1958 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1959 an incoming connection.
1961 HS/02 Bug 1802: Do not half-close the connection after sending a request
1964 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1965 fallback to "prime256v1".
1967 JH/34 SECURITY: Use proper copy of DATA command in error message.
1968 Could leak key material. Remotely exploitable. CVE-2016-9963.
1974 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1975 and 3.4.4 - once the server is enabled to respond to an OCSP request
1976 it does even when not requested, resulting in a stapling non-aware
1977 client dropping the TLS connection.
1979 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1980 support variable-length bit vectors. No functional change.
1982 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1983 The I= interface field on outgoing lines is now after the H= remote
1984 host field, same as incoming lines. There is a separate
1985 outgoing_interface log selector which allows you to disable the
1988 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1989 If not running log_selector +smtp_connection the mainlog would be held
1990 open indefinitely after a "too many connections" event, including to a
1991 deleted file after a log rotate. Leave the per net connection logging
1992 leaving it open for efficiency as that will be quickly detected by the
1993 check on the next write.
1995 HS/01 Bug 1671: Fix post transport crash.
1996 Processing the wait-<transport> messages could crash the delivery
1997 process if the message IDs didn't exist for some reason. When
1998 using 'split_spool_directory=yes' the construction of the spool
1999 file name failed already, exposing the same netto behaviour.
2001 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2002 mime_regex ACL conditions.
2004 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2005 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2006 to HELO, local diagnostic string.
2008 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2009 log line to debug. Even when configured with a tls authenticator many
2010 client connections are expected to not authenticate in this way, so
2011 an authenticate fail is not an error.
2013 HS/02 Add the Exim version string to the process info. This way exiwhat
2014 gives some more detail about the running daemon.
2016 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2017 matter for fast-change records such as DNSBLs.
2019 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2020 retry record, even if constant. There may be multiple transports with
2021 different interface settings and the retry behaviour needs to be kept
2024 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2026 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2028 JH/10 Bug 840: fix log_defer_output option of pipe transport
2030 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2031 hosts_randomize. This matters a lot when combined with mua_wrapper.
2033 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2034 ${quote_pgsql:<string>} operator.
2036 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2038 JH/14 Change header file name for freeradius-client. Relevant if compiling
2039 with Radius support; from the Gentoo tree and checked under Fedora.
2041 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2043 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2044 option transport_filter, meaning no filtering.
2046 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2048 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2049 defaults to "*" (all hosts). The variable is now available when not built
2050 with TLS, default unset, mainly to enable keeping the testsuite sane.
2051 If a server certificate is not supplied (via tls_certificate) an error is
2052 logged, and clients will find TLS connections fail on startup. Presumably
2053 they will retry in-clear.
2054 Packagers of Exim are strongly encouraged to create a server certificate
2055 at installation time.
2057 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2058 with the $config_file variable.
2060 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2061 in transport context, after the attempt, and per-recipient. The latter type
2062 is per host attempted. The event data is the error message, and the errno
2063 information encodes the lookup type (A vs. MX) used for the (first) host,
2064 and the trailing two digits of the smtp 4xx response.
2066 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2067 to write to mainlog (or rejectlog, paniclog) in the window between file
2068 creation and permissions/ownership being changed. Particularly affects
2069 installations where exicyclog is run as root, rather than exim user;
2070 result is that the running daemon panics and dies.
2072 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2074 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2075 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2076 "pri" and "weight". Note that the previous implicit priority given by the
2077 list order is no longer honoured.
2079 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2080 for DKIM processing.
2082 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2083 by defining SUPPORT_SOCKS.
2085 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2086 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2087 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2088 variables are renamed to proxy_{local,external}_{address,port}.
2090 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2091 for a build by defining SUPPORT_I18N
2093 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2094 of the query string, and make ${quote_redis:} do that quoting.
2096 JH/29 Move Events support from Experimental to mainline, enabled by default
2097 and removable for a build by defining DISABLE_EVENT.
2099 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2101 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2102 cached by the daemon.
2104 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2105 by defining LOOKUP_REDIS. The libhiredis library is required.
2107 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2108 keys are given for lookup.
2110 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2111 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2112 only supported when built with TLS support. The PolarSSL SHA routines
2113 are still used when the TLS library is too old for convenient support.
2115 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2116 openssl_options), for security. OpenSSL forces this from version 1.1.0
2117 server-side so match that on older versions.
2119 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2120 allocation for $value could be released as the expansion processing
2121 concluded, but leaving the global pointer active for it.
2123 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2124 and to use the domains and local_parts ACL conditions.
2126 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2127 incorrectly not doubled on cutthrough transmission, hence seen as a
2128 body-termination at the receiving system - resulting in truncated mails.
2129 Commonly the sender saw a TCP-level error, and retransmitted the message
2130 via the normal store-and-forward channel. This could result in duplicates
2131 received - but deduplicating mailstores were liable to retain only the
2132 initial truncated version.
2134 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2136 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2138 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2139 we're in there, support oversigning also; bug 1309.
2141 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2143 HS/04 Add support for keep_environment and add_environment options.
2145 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2146 either intentional arithmetic overflow during PRNG, or testing config-
2149 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2150 delivery resulted in actual delivery. Cancel cutthrough before DATA
2153 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2154 rejects a recipient: pass the reject to the originator.
2156 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2157 Many were false-positives and ignorable, but it's worth fixing the
2160 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2161 for the new environment-manipulation done at startup. Move the routines
2162 from being local to tls.c to being global via the os.c file.
2164 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2165 an extract embedded as result-arg for a map, the first arg for extract
2166 is unavailable so we cannot tell if this is a numbered or keyed
2167 extraction. Accept either.
2173 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2176 JH/02 The smtp transport option "multi_domain" is now expanded.
2178 JH/03 The smtp transport now requests PRDR by default, if the server offers
2181 JH/04 Certificate name checking on server certificates, when exim is a client,
2182 is now done by default. The transport option tls_verify_cert_hostnames
2183 can be used to disable this per-host. The build option
2184 EXPERIMENTAL_CERTNAMES is withdrawn.
2186 JH/05 The value of the tls_verify_certificates smtp transport and main options
2187 default to the word "system" to access the system default CA bundle.
2188 For GnuTLS, only version 3.0.20 or later.
2190 JH/06 Verification of the server certificate for a TLS connection is now tried
2191 (but not required) by default. The verification status is now logged by
2192 default, for both outbound TLS and client-certificate supplying inbound
2195 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2198 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2199 Status Notification (bounce) messages are now MIME format per RFC 3464.
2200 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2201 under the control of the dsn_advertise_hosts option, and routers may
2202 have a dsn_lasthop option.
2204 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2205 default, modifiable by a malware= option. The list separator for
2206 the options can now be changed in the usual way. Bug 68.
2208 JH/10 The smtp_receive_timeout main option is now expanded before use.
2210 JH/11 The incoming_interface log option now also enables logging of the
2211 local interface on delivery outgoing connections.
2213 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2214 if the interface and destination host and port all match.
2216 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2219 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2220 Patch from Andrew Lewis.
2222 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2223 now supports optional time-restrictions, weighting, and priority
2224 modifiers per server. Patch originally by <rommer@active.by>.
2226 JH/16 The spamd_address main option now supports a mixed list of local
2227 and remote servers. Remote servers can be IPv6 addresses, and
2228 specify a port-range.
2230 JH/17 Bug 68: The spamd_address main option now supports an optional
2231 timeout value per server.
2233 JH/18 Bug 1581: Router and transport options headers_add/remove can
2234 now have the list separator specified.
2236 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2239 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2242 JH/21 Support for the A6 type of dns record is withdrawn.
2244 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2245 rather than the verbs used.
2247 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2248 from 255 to 1024 chars.
2250 JH/24 Verification callouts now attempt to use TLS by default.
2252 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2253 are generic router options now. The defaults didn't change.
2255 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2256 Original patch from Alexander Shikoff, worked over by JH.
2258 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2259 files if ZCAT_COMMAND is not executable.
2261 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2263 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2265 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2266 Normally benign, it bites when the pair was led to by a CNAME;
2267 modern usage is to not canonicalize the domain to a CNAME target
2268 (and we were inconsistent anyway for A-only vs AAAA+A).
2270 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2272 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2273 when evaluating $sender_host_dnssec.
2275 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2276 $sender_helo_dnssec variable.
2278 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2280 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2282 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2284 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2285 documented as working, but never had. Support all but $spam_report.
2287 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2288 added for tls authenticator.
2290 HS/03 Add perl_taintmode main config option
2296 TL/01 When running the test suite, the README says that variables such as
2297 no_msglog_check are global and can be placed anywhere in a specific
2298 test's script, however it was observed that placement needed to be near
2299 the beginning for it to behave that way. Changed the runtest perl
2300 script to read through the entire script once to detect and set these
2301 variables, reset to the beginning of the script, and then run through
2302 the script parsing/test process like normal.
2304 TL/02 The BSD's have an arc4random API. One of the functions to induce
2305 adding randomness was arc4random_stir(), but it has been removed in
2306 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2307 function when detected.
2309 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2310 cause callback expansion.
2312 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2313 syntax errors in an expansion can be treated as a string instead of
2314 logging or causing an error, due to the internal use of bool_lax
2315 instead of bool when processing it.
2317 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2318 server certificates when making smtp deliveries.
2320 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2322 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2324 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2326 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2327 Merged patch from Sebastian Wiedenroth.
2329 JH/05 Fix results-pipe from transport process. Several recipients, combined
2330 with certificate use, exposed issues where response data items split
2331 over buffer boundaries were not parsed properly. This eventually
2332 resulted in duplicates being sent. This issue only became common enough
2333 to notice due to the introduction of connection certificate information,
2334 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2336 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2337 size buffer was used, resulting in syntax errors when an expansion
2340 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2341 version 3.3.6 or later.
2343 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2344 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2345 both become "event_action", the variables become $event_name, $event_data
2346 and $event_defer_errno. There is a new variable $verify_mode, usable in
2347 routers, transports and related events. The tls:cert event is now also
2348 raised for inbound connections, if the main configuration event_action
2351 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2352 early OCSP support, but no stapling (appears to be less than 1.0.0).
2354 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2355 server certificate names available under the smtp transport option
2356 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2359 JH/10 Time-related extraction expansions from certificates now use the main
2360 option "timezone" setting for output formatting, and are consistent
2361 between OpenSSL and GnuTLS compilations. Bug 1541.
2363 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2364 encoded parameter in the incoming message. Bug 1558.
2366 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2367 include certificate info, eximon was claiming there were spoolfile
2370 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2372 JH/14 Log delivery-related information more consistently, using the sequence
2373 "H=<name> [<ip>]" wherever possible.
2375 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2376 are problematic for Debian distribution, omit them from the release
2379 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2381 JH/16 Fix string representation of time values on 64bit time_t architectures.
2384 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2385 output list separator was used.
2390 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2391 checkers that were complaining about end of non-void function with no
2394 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2395 This was a regression introduced in 4.83 by another bugfix.
2397 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2399 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2400 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2406 TF/01 Correctly close the server side of TLS when forking for delivery.
2408 When a message was received over SMTP with TLS, Exim failed to clear up
2409 the incoming connection properly after forking off the child process to
2410 deliver the message. In some situations the subsequent outgoing
2411 delivery connection happened to have the same fd number as the incoming
2412 connection previously had. Exim would try to use TLS and fail, logging
2413 a "Bad file descriptor" error.
2415 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2416 utilities have not been installed.
2418 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2419 temporary space as the ACL may create new global variables.
2421 TL/01 LDAP support uses per connection or global context settings, depending
2422 upon the detected version of the libraries at build time.
2424 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2425 to extract and use the src ip:port in logging and expansions as if it
2426 were a direct connection from the outside internet. PPv2 support was
2427 updated based on HAProxy spec change in May 2014.
2429 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2431 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2432 Properly escape header and check for NULL return.
2434 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2437 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2439 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2440 characters in header names, implemented as a verify condition.
2441 Contributed by Michael Fischer v. Mollard.
2443 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2444 results permerror and temperror. Previous values are deprecated but
2445 still accepted. In a future release, err_perm and err_temp will be
2446 completely removed, which will be a backward incompatibility if the
2447 ACL tests for either of these two old results. Patch contributed by
2448 user bes-internal on the mailing list.
2450 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2452 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2453 selectors, in both main and reject logs.
2455 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2458 JH/07 Add malware type "sock" for talking to simple daemon.
2460 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2462 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2463 routers/transports under cutthrough routing.
2465 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2466 numbers. Touch up "bool" conditional to keep the same definition.
2468 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2470 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2472 JH/12 Expand items in router/transport headers_add or headers_remove lists
2473 individually rather than the list as a whole. Bug 1452.
2475 Required for reasonable handling of multiple headers_ options when
2476 they may be empty; requires that headers_remove items with embedded
2477 colons must have them doubled (or the list-separator changed).
2479 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2480 view the policy declared in the DMARC record. Currently, $dmarc_status
2481 is a combined value of both the record presence and the result of the
2484 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2486 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2487 dnslookup router and the smtp transport (applying to the forward
2490 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2491 of ldap servers used for a specific lookup. Patch provided by Heiko
2494 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2495 New variable $lookup_dnssec_authenticated for observability.
2497 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2498 Patch submitted by Lars Timman.
2500 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2502 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2503 Requires trusted mode and valid format message id, aborts otherwise.
2504 Patch contributed by Heiko Schlichting.
2506 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2507 certextract with support for various fields. Bug 1358.
2509 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2510 is requested by default, modifiable by smtp transport option
2513 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2514 operate on certificate variables to give certificate fingerprints
2515 Also new ${sha256:cert_variable}.
2517 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2519 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2520 Christian Aistleitner.
2522 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2524 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2525 file. Patch from Wolfgang Breyha.
2527 JH/25 Expand the coverage of the delivery $host and $host_address to
2528 client authenticators run in verify callout. Bug 1476.
2530 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2531 align with daemon_smtp_ports. Bug 72.
2533 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2534 support and error reporting did not work properly.
2536 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2537 and is readable. Patch from Andrew Colin Kissa.
2539 TL/14 Enhance documentation of ${run expansion and how it parses the
2540 commandline after expansion, particularly in the case when an
2541 unquoted variable expansion results in an empty value.
2543 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2545 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2546 by renaming away from it. Addresses GH issue 15, reported by
2549 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2551 TL/15 SECURITY: prevent double expansion in math comparison functions
2552 (can expand unsanitized data). Not remotely exploitable.
2559 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2561 PP/02 Make -n do something, by making it not do something.
2562 When combined with -bP, the name of an option is not output.
2564 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2567 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2568 $sender_host_name and config options to manage this, and basic check
2571 PP/05 DSCP support for outbound connections and control modifier for inbound.
2573 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2574 (Only plugin which currently uses this is kerberos4, which nobody should
2575 be using, but we should make it available and other future plugins might
2576 conceivably use it, even though it would break NAT; stuff *should* be
2577 using channel bindings instead).
2579 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2580 name; added for Sendmail compatibility; requires admin caller.
2581 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2582 just ignore it); requires trusted caller.
2583 Also parse but ignore: -Ac -Am -X<logfile>
2586 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2588 TL/02 Add +smtp_confirmation as a default logging option.
2590 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2591 Patch by Magnus Holmgren from 2007-02-20.
2593 TL/04 Bugzilla 1281 - Spec typo.
2594 Bugzilla 1283 - Spec typo.
2595 Bugzilla 1290 - Spec grammar fixes.
2597 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2599 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2601 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2602 mailing list by by Dmitry Isaikin.
2604 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2606 JH/02 Support "G" suffix to numbers in ${if comparisons.
2608 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2610 NM/01 Bugzilla 1197 - Spec typo
2611 Bugzilla 1196 - Spec examples corrections
2613 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2615 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2616 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2619 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2620 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2622 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2623 "acl {{name}{arg}...}", and optional args on acl condition
2626 JH/05 Permit multiple router/transport headers_add/remove lines.
2628 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2630 JH/07 Avoid using a waiting database for a single-message-only transport.
2631 Performance patch from Paul Fisher. Bugzilla 1262.
2633 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2636 JH/09 Add $headers_added variable, with content from use of ACL modifier
2637 add_header (but not yet added to the message). Bugzilla 199.
2639 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2640 Pulled from Bugzilla 817 by Wolfgang Breyha.
2642 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2644 (nb: this is the same fix as in Exim 4.80.1)
2646 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2649 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2652 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2654 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2655 advertise SMTP AUTH mechanism to us, instead of a generic
2656 protocol violation error. Also, make Exim more robust to bad
2657 data from the Dovecot auth socket.
2659 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2661 When a queue runner is handling a message, Exim first routes the
2662 recipient addresses, during which it prunes them based on the retry
2663 hints database. After that it attempts to deliver the message to
2664 any remaining recipients. It then updates the hints database using
2667 So if a recipient address works intermittently, it can get repeatedly
2668 deferred at routing time. The retry hints record remains fresh so the
2669 address never reaches the final cutoff time.
2671 This is a fairly common occurrence when a user is bumping up against
2672 their storage quota. Exim had some logic in its local delivery code
2673 to deal with this. However it did not apply to per-recipient defers
2674 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2676 This change adds a proper retry rule check during routing so that the
2677 final cutoff time is checked against the message's age. We only do
2678 this check if there is an address retry record and there is not a
2679 domain retry record; this implies that previous attempts to handle
2680 the address had the retry_use_local_parts option turned on. We use
2681 this as an approximation for the destination being like a local
2682 delivery, as in LMTP.
2684 I suspect this new check makes the old local delivery cutoff check
2685 redundant, but I have not verified this so I left the code in place.
2687 TF/02 Correct gecos expansion when From: is a prefix of the username.
2689 Test 0254 submits a message to Exim with the header
2693 When I ran the test suite under the user fanf2, Exim expanded
2694 the header to contain my full name, whereas it should have added
2695 a Resent-Sender: header. It erroneously treats any prefix of the
2696 username as equal to the username.
2698 This change corrects that bug.
2700 GF/01 DCC debug and logging tidyup
2701 Error conditions log to paniclog rather than rejectlog.
2702 Debug lines prefixed by "DCC: " to remove any ambiguity.
2704 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2706 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2707 Bug spotted by Jeremy Harris; was flawed since initial commit.
2708 Would have resulted in OCSP responses post-SNI triggering an Exim
2709 NULL dereference and crash.
2711 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2713 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2714 Bug detection, analysis and fix by Samuel Thibault.
2715 Bugzilla 1331, Debian bug #698092.
2717 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2719 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2720 Server implementation by Todd Lyons, client by JH.
2721 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2722 config variable "prdr_enable" controls whether the server
2723 advertises the facility. If the client requests PRDR a new
2724 acl_data_smtp_prdr ACL is called once for each recipient, after
2725 the body content is received and before the acl_smtp_data ACL.
2726 The client is controlled by both of: a hosts_try_prdr option
2727 on the smtp transport, and the server advertisement.
2728 Default client logging of deliveries and rejections involving
2729 PRDR are flagged with the string "PRDR".
2731 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2732 fclose(). Diagnosis by Todd Lyons.
2734 PP/17 Update configure.default to handle IPv6 localhost better.
2735 Patch by Alain Williams (plus minor tweaks).
2738 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2739 This is now consistent with GnuTLS, and is now documented: the
2740 previous undocumented portable approach to treating the option as
2741 unset was to force an expansion failure. That still works, and
2742 an empty string is now equivalent.
2744 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2745 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2746 not performing validation itself.
2748 PP/20 Added force_command boolean option to pipe transport.
2749 Patch from Nick Koston, of cPanel Inc.
2751 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2754 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2756 PP/21 Fix eximon continuous updating with timestamped log-files.
2757 Broken in a format-string cleanup in 4.80, missed when I repaired the
2758 other false fix of the same issue.
2759 Report and fix from Heiko Schlichting.
2762 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2763 Report from Prashanth Katuri.
2765 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2766 It's SecureTransport, so affects any MacOS clients which use the
2767 system-integrated TLS libraries, including email clients.
2769 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2770 using a MIME ACL for non-SMTP local injection.
2771 Report and assistance in diagnosis by Warren Baker.
2773 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2775 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2777 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2778 last id that failed so it may be referenced in subsequent ACL's.
2780 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2783 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2784 ldap library initialization, allowing self-signed CA's to be
2785 used. Also properly sets require_cert option later in code by
2786 using NULL (global ldap config) instead of ldap handle (per
2787 session). Bug diagnosis and testing by alxgomz.
2789 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2790 the src/util/ subdirectory.
2792 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2793 renamed to Transport Post Delivery Action by Jeremy Harris, as
2796 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2797 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2798 redis_servers = needs to be configured which will be used by the redis
2799 lookup. Patch from Warren Baker, of The Packet Hub.
2801 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2803 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2804 hostname or reverse DNS when processing a host list. Used suggestions
2805 from multiple comments on this bug.
2807 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2809 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2810 Missed a few lines, added it to make the runtest require no keyboard
2813 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2814 contains upper case chars. Make router use caseful_local_part.
2816 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2817 support when GnuTLS has been built with p11-kit.
2823 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2825 This, or similar/improved, will also be change PP/11 of 4.82.
2831 PP/01 Handle short writes when writing local log-files.
2832 In practice, only affects FreeBSD (8 onwards).
2833 Bugzilla 1053, with thanks to Dmitry Isaikin.
2835 NM/01 Bugzilla 949 - Documentation tweak
2837 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2840 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2842 PP/02 Implemented gsasl authenticator.
2844 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2846 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2847 `pkg-config foo` for cflags/libs.
2849 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2850 with rest of GSASL and with heimdal_gssapi.
2852 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2853 `pkg-config foo` for cflags/libs for the TLS implementation.
2855 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2856 properties get this fed in as external SSF. A number of robustness
2857 and debugging improvements to the cyrus_sasl authenticator.
2859 PP/08 cyrus_sasl server now expands the server_realm option.
2861 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2862 Patch by Jeremy Harris.
2864 PP/10 Added dbmjz lookup type.
2866 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2868 PP/12 MAIL args handles TAB as well as SP, for better interop with
2869 non-compliant senders.
2870 Analysis and variant patch by Todd Lyons.
2872 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2873 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2874 Patch from Dirk Mueller <dmueller@suse.com>
2876 PP/13 tls_peerdn now print-escaped for spool files.
2877 Observed some $tls_peerdn in wild which contained \n, which resulted
2878 in spool file corruption.
2880 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2881 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2882 or write after TLS renegotiation, which otherwise led to messages
2885 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2886 as a tracking header (ie: a signed header comes before the signature).
2887 Patch from Wolfgang Breyha.
2889 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2890 comma-sep list; embedded commas doubled.
2892 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2894 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2896 Report and patch from Dmitry Banschikov.
2898 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2899 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2900 failures. We appear to now support TLS1.1+ with Exim.
2902 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2903 lets Exim select keys and certificates based upon TLS SNI from client.
2904 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2905 before an outbound SMTP session. New log_selector, +tls_sni.
2907 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2908 NULL dereference. Report and patch from Alun Jones.
2910 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2911 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2912 Not seeing resolver debug output on NetBSD, but suspect this is a
2913 resolver implementation change.
2915 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2916 Left warnings. Added "eximon gdb" invocation mode.
2918 PP/21 Defaulting "accept_8bitmime" to true, not false.
2920 PP/22 Added -bw for inetd wait mode support.
2922 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2923 locate the relevant includes and libraries. Made this the default.
2925 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2926 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2928 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2929 This may cause build issues on older platforms.
2931 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2932 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2933 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2934 Added SNI support via GnuTLS too.
2935 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2937 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2939 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2940 Applied second patch from Janne, implementing suggestion to default
2941 multiple-strings-in-record handling to match SPF spec.
2943 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2945 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2946 read-only, out of scope).
2947 Patch from Wolfgang Breyha, report from Stuart Northfield.
2949 PP/29 Fix three issues highlighted by clang analyser static analysis.
2950 Only crash-plausible issue would require the Cambridge-specific
2951 iplookup router and a misconfiguration.
2952 Report from Marcin Mirosław.
2954 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2956 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2957 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2958 As part of this, removing so much warning spew let me fix some minor
2959 real issues in debug logging.
2961 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2962 assignment on my part. Fixed.
2964 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2965 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2966 Janne Snabb (who went above and beyond: thank you).
2968 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2969 string otherwise requires a connection and a bunch more work and it's
2970 relatively easy to get wrong. Should also expose TLS library linkage
2973 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2974 64-bit ${eval} (JH/03).
2976 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2977 GNU libc to support some of the 64-bit stuff, should not lead to
2978 conflicts. Defined before os.h is pulled in, so if a given platform
2979 needs to override this, it can.
2981 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2982 protection layer was required, which is not implemented.
2983 Bugzilla 1254, patch from Wolfgang Breyha.
2985 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2986 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2987 tls_dhparam take prime identifiers. Also unbreak combination of
2988 OpenSSL+DH_params+TLSSNI.
2990 PP/39 Disable SSLv2 by default in OpenSSL support.
2996 PP/01 Solaris build fix for Oracle's LDAP libraries.
2997 Bugzilla 1109, patch from Stephen Usher.
2999 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3001 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3004 TF/02 Fix a couple more cases where we did not log the error message
3005 when unlink() failed. See also change 4.74-TF/03.
3007 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3008 lock up or crash if it happened to be inside a call to libc when it
3009 got a SIGUSR1 from exiwhat.
3011 The SIGUSR1 handler appends the current process status to the process
3012 log which is later printed by exiwhat. It used to use the general
3013 purpose logging code to do this, but several functions it calls are
3014 not safe for signals.
3016 The new output code in the SIGUSR1 handler is specific to the process
3017 log, and simple enough that it's easy to inspect for signal safety.
3018 Removing some special cases also simplifies the general logging code.
3019 Removing the spurious timestamps from the process log simplifies
3022 TF/04 Improved ratelimit ACL condition.
3024 The /noupdate option has been deprecated in favour of /readonly which
3025 has clearer semantics. The /leaky, /strict, and /readonly update modes
3026 are mutually exclusive. The update mode is no longer included in the
3027 database key; it just determines when the database is updated. (This
3028 means that when you upgrade Exim will forget old rate measurements.)
3030 Exim now checks that the per_* options are used with an update mode that
3031 makes sense for the current ACL. For example, when Exim is processing a
3032 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3033 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3034 must specify per_mail/readonly. If you omit the update mode it defaults to
3035 /leaky where that makes sense (as before) or /readonly where required.
3037 The /noupdate option is now undocumented but still supported for
3038 backwards compatibility. It is equivalent to /readonly except that in
3039 ACLs where /readonly is required you may specify /leaky/noupdate or
3040 /strict/noupdate which are treated the same as /readonly.
3042 A useful new feature is the /count= option. This is a generalization
3043 of the per_byte option, so that you can measure the throughput of other
3044 aggregate values. For example, the per_byte option is now equivalent
3045 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3047 The per_rcpt option has been generalized using the /count= mechanism
3048 (though it's more complicated than the per_byte equivalence). When it is
3049 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3050 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3051 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3052 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3053 Note that using per_rcpt with a non-readonly update mode in more than
3054 one ACL will cause the recipients to be double-counted. (The per_mail
3055 and per_byte options don't have this problem.)
3057 The handling of very low rates has changed slightly. If the computed rate
3058 is less than the event's count (usually one) then this event is the first
3059 after a long gap. In this case the rate is set to the same as this event's
3060 count, so that the first message of a spam run is counted properly.
3062 The major new feature is a mechanism for counting the rate of unique
3063 events. The new per_addr option counts the number of different
3064 recipients that someone has sent messages to in the last time period. It
3065 behaves like per_rcpt if all the recipient addresses are different, but
3066 duplicate recipient addresses do not increase the measured rate. Like
3067 the /count= option this is a general mechanism, so the per_addr option
3068 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3069 example, measure the rate that a client uses different sender addresses
3070 with the options per_mail/unique=$sender_address. There are further
3071 details in the main documentation.
3073 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3075 TF/06 Removed a few PCRE remnants.
3077 TF/07 Automatically extract Exim's version number from tags in the git
3078 repository when doing development or release builds.
3080 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3081 Bugzilla 879. Patch from Paul Fisher.
3083 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3084 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3087 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3089 PP/05 Variable $av_failed, true if the AV scanner deferred.
3090 Bugzilla 1078. Patch from John Horne.
3092 PP/06 Stop make process more reliably on build failure.
3093 Bugzilla 1087. Patch from Heiko Schlittermann.
3095 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3096 Bugzilla 1089. Patch from Heiko Schlittermann.
3098 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3099 Bugzilla 1131. Patch from Holger Weiß.
3101 PP/09 Handle IPv6 addresses with SPF.
3102 Bugzilla 860. Patch from Wolfgang Breyha.
3104 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3106 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3109 PP/11 match_* no longer expand right-hand-side by default.
3110 New compile-time build option, EXPAND_LISTMATCH_RHS.
3111 New expansion conditions, "inlist", "inlisti".
3113 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3115 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3117 PP/14 fix log_write() format string regression from TF/03.
3118 Bugzilla 1152. Patch from Dmitry Isaikin.
3124 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3126 PP/02 Harmonised TLS library version reporting; only show if debugging.
3127 Layout now matches that introduced for other libraries in 4.74 PP/03.
3129 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3131 PP/04 New "dns_use_edns0" global option.
3133 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3136 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3137 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3139 TK/01 Updated PolarSSL code to 0.14.2.
3140 Bugzilla 1097. Patch from Andreas Metzler.
3142 PP/07 Catch divide-by-zero in ${eval:...}.
3143 Fixes bugzilla 1102.
3145 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3148 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3149 format-string attack -- SECURITY: remote arbitrary code execution.
3151 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3152 time unintentionally subject to list matching rules, letting the header
3153 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3154 arbitrary string expansion). This allowed for information disclosure.
3156 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3157 INT_MIN/-1 -- value coerced to INT_MAX.
3163 NM/01 Workaround for PCRE version dependency in version reporting
3166 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3167 This fixes portability to compilers other than gcc, notably
3168 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3170 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3171 makefiles for portability to HP-UX and POSIX correctness.
3173 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3174 Also via indented variable definition in the Makefile.
3175 (Debugging by Oliver Heesakkers).
3177 PP/02 Restore caching of spamd results with expanded spamd_address.
3178 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3180 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3181 Improves build reliability. Fix from: Frank Elsner
3183 NM/02 Fix wide character breakage in the rfc2047 coding
3184 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3186 NM/03 Allow underscore in dnslist lookups
3187 Fixes bug 1026. Patch from Graeme Fowler
3189 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3190 Code patches from Adam Ciarcinski of NetBSD.
3192 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3195 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3196 is logged, to avoid truncation. Patch from John Horne.
3198 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3199 Patch from Jakob Hirsch.
3201 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3202 SQL string expansion failure details.
3203 Patch from Andrey Oktyabrski.
3205 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3206 Patch from Simon Arlott.
3208 PP/09 New lookups functionality failed to compile on old gcc which rejects
3209 extern declarations in function scope.
3210 Patch from Oliver Fleischmann
3212 PP/10 Use sig_atomic_t for flags set from signal handlers.
3213 Check getgroups() return and improve debugging.
3214 Fixed developed for diagnosis in bug 927 (which turned out to be
3217 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3218 Patch from Mark Zealey.
3220 PP/12 Bugzilla 1056: Improved spamd server selection.
3221 Patch from Mark Zealey.
3223 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3224 Based on patch from Heiko Schlittermann.
3226 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3227 Patch from Uwe Doering, sign-off by Michael Haardt.
3229 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3230 variable declaration deep within a block. Bug and patch from
3233 PP/15 lookups-Makefile IRIX compatibility coercion.
3235 PP/16 Make DISABLE_DKIM build knob functional.
3237 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3238 Patch by Simon Arlott
3240 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3241 variable argument macros. Our copy now differs from upstream.
3247 TF/01 Failure to get a lock on a hints database can have serious
3248 consequences so log it to the panic log.
3250 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3251 controlled using the smtp_confirmation log selector.
3253 TF/03 Include the error message when we fail to unlink a spool file.
3255 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3256 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3257 for maintaining out-of-tree patches for some time.
3259 PP/01 Bugzilla 139: Documentation and portability issues.
3260 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3261 Handle per-OS dynamic-module compilation flags.
3263 PP/02 Let /dev/null have normal permissions.
3264 The 4.73 fixes were a little too stringent and complained about the
3265 permissions on /dev/null. Exempt it from some checks.
3266 Reported by Andreas M. Kirchwitz.
3268 PP/03 Report version information for many libraries, including
3269 Exim version information for dynamically loaded libraries. Created
3270 version.h, now support a version extension string for distributors
3271 who patch heavily. Dynamic module ABI change.
3273 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3274 privilege escalation vulnerability whereby the Exim run-time user
3275 can cause root to append content of the attacker's choosing to
3278 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3281 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3282 If dropping privileges for untrusted macros, we disabled normal logging
3283 on the basis that it would fail; for the Exim run-time user, this is not
3284 the case, and it resulted in successful deliveries going unlogged.
3285 Fixed. Reported by Andreas Metzler.
3291 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3292 only prepend for the Resent-* case. Fixes regression introduced in
3293 Exim 4.70 by NM/22 for Bugzilla 607.
3295 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3296 increasing numbers of administrators be bitten by this.
3298 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3300 PP/03 Bugzilla 994: added openssl_options main configuration option.
3302 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3304 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3306 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3308 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3309 without a peer certificate, leading to a segfault because of an
3310 assumption that peers always have certificates. Be a little more
3311 paranoid. Problem reported by Martin Tscholak.
3313 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3314 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3315 NB: ClamAV planning to remove STREAM in "middle of 2010".
3316 CL also introduces -bmalware, various -d+acl logging additions and
3317 more caution in buffer sizes.
3319 PP/09 Implemented reverse_ip expansion operator.
3321 PP/10 Bugzilla 937: provide a "debug" ACL control.
3323 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3325 PP/12 Bugzilla 973: Implement --version.
3327 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3329 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3331 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3333 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3334 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3335 ignore trailing whitespace.
3337 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3339 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3342 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3343 Notification from Dr Andrew Aitchison.
3345 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3346 ExtendedDetectionInfo response format.
3347 Notification from John Horne.
3349 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3352 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3353 XSL and documented dependency on system catalogs, with examples of how
3356 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3359 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3360 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3361 configuration file which is writeable by the Exim user or group.
3363 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3364 of configuration files to cover files specified with the -C option if
3365 they are going to be used with root privileges, not just the default
3368 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3369 option (effectively making it always true).
3371 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3372 files to be used while preserving root privileges.
3374 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3375 that rogue child processes cannot use them.
3377 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3378 run-time user, instead of root.
3380 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3381 Exim run-time user without dropping privileges.
3383 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3384 result string, instead of calling string_vformat() twice with the same
3387 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3388 for other users. Others should always drop root privileges if they use
3389 -C on the command line, even for a whitelisted configure file.
3391 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3393 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3399 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3400 $data_path, and $header_path variables; fixed documentation bugs and
3403 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3404 exipick to access non-standard spools, including the "frozen" queue
3407 NM/01 Bugzilla 965: Support mysql stored procedures.
3408 Patch from Alain Williams
3410 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3412 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3413 Patch from Andreas Metzler
3415 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3416 Patch from Kirill Miazine
3418 NM/05 Bugzilla 671: Added umask to procmail example.
3420 JJ/03 installed exipick 20100323.0, fixing doc bug
3422 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3423 directory. Notification and patch from Dan Rosenberg.
3425 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3427 TK/02 Improve log output when DKIM signing operation fails.
3429 MH/01 Treat the transport option dkim_domain as a colon separated
3430 list, not as a single string, and sign the message with each element,
3431 omitting multiple occurences of the same signer.
3433 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3434 Bugzilla 985, 986. Patch by Simon Arlott
3436 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3437 Patch by Simon Arlott
3439 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3440 MBX locking. Notification from Dan Rosenberg.
3446 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3448 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3450 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3452 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3454 NM/04 Bugzilla 915: Flag broken perl installation during build.
3460 TK/01 Added patch by Johannes Berg that expands the main option
3461 "spamd_address" if it starts with a dollar sign.
3463 TK/02 Write list of recipients to X-Envelope-Sender header when building
3464 the mbox-format spool file for content scanning (suggested by Jakob
3467 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3468 (http://www.dcc-servers.net/) support via dccifd. Activated by
3469 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3471 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3472 by Mark Daniel Reidel <mr@df.eu>.
3474 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3475 When building exim an external PCRE library is now needed -
3476 PCRE is a system library on the majority of modern systems.
3477 See entry on PCRE_LIBS in EDITME file.
3479 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3480 conversation. Added nologin parameter to request.
3481 Patch contributed by Kirill Miazine.
3483 TF/01 Do not log submission mode rewrites if they do not change the address.
3485 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3487 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3488 log files in place. Contributed by Roberto Lima.
3490 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3492 TF/03 Bugzilla 615: When checking the local_parts router precondition
3493 after a local_part_suffix or local_part_prefix option, Exim now
3494 does not use the address's named list lookup cache, since this
3495 contains cached lookups for the whole local part.
3497 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3498 Robert Millan. Documentation is in experimental-spec.txt.
3500 TF/04 Bugzilla 668: Fix parallel build (make -j).
3502 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3504 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3505 Patch provided by Jan Srzednicki.
3507 TF/05 Leading white space used to be stripped from $spam_report which
3508 wrecked the formatting. Now it is preserved.
3510 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3511 that they are available at delivery time.
3513 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3515 TF/08 TLS error reporting now respects the incoming_interface and
3516 incoming_port log selectors.
3518 TF/09 Produce a more useful error message if an SMTP transport's hosts
3519 setting expands to an empty string.
3521 NM/06 Bugzilla 744: EXPN did not work under TLS.
3522 Patch provided by Phil Pennock.
3524 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3525 Patch provided by Richard Godbee.
3527 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3528 acl_smtp_notquit, added index entry.
3530 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3531 Patch provided by Eugene Bujak.
3533 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3534 accept(). Patch provided by Maxim Dounin.
3536 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3537 Patch provided by Phil Pennock.
3539 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3541 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3542 Patch provided by Brad "anomie" Jorsch.
3544 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3545 Patch provided by Dean Brooks.
3547 TK/05 Add native DKIM support (does not depend on external libraries).
3549 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3550 Patch provided by Graeme Fowler.
3552 NM/16 Bugzilla 851: Documentation example syntax fix.
3554 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3556 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3559 NM/19 Bugzilla 745: TLS version reporting.
3560 Patch provided by Phil Pennock.
3562 NM/20 Bugzilla 167: bool: condition support.
3563 Patch provided by Phil Pennock.
3565 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3566 clients. Patch provided by Phil Pennock.
3568 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3569 Patch provided by Brad "anomie" Jorsch.
3571 NM/23 Bugzilla 687: Fix misparses in eximstats.
3572 Patch provided by Heiko Schlittermann.
3574 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3575 Patch provided by Heiko Schlittermann.
3577 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3578 plus update to original patch.
3580 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3582 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3583 Patch provided by David Brownlee.
3585 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3587 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3589 NM/30 Bugzilla 888: TLS documentation bugfixes.
3591 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3593 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3594 Unlike the original bugzilla I have changed all shell scripts in src tree.
3596 NM/33 Bugzilla 898: Transport filter timeout fix.
3597 Patch by Todd Rinaldo.
3599 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3600 Patch by Serge Demonchaux.
3602 NM/35 Bugzilla 39: Base64 decode bug fixes.
3603 Patch by Jakob Hirsch.
3605 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3607 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3609 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3611 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3617 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3618 ALT-N's libdkim that I have put here:
3619 http://duncanthrax.net/exim-experimental/
3621 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3622 were called 'true' and it seems that C99 defines that as a reserved
3623 keyword to be used with 'bool' variable types. That means you could
3624 not include C99-style headers which use bools without triggering
3625 build errors in sieve.c.
3627 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3628 as mailq or other aliases. Changed the --help handling significantly
3629 to do whats expected. exim_usage() emits usage/help information.
3631 SC/01 Added the -bylocaldomain option to eximstats.
3633 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3635 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3637 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3643 PH/01 Another patch from the Sieve maintainer.
3645 PH/02 When an IPv6 address is converted to a string for single-key lookup
3646 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3647 dots are used instead of colons so that keys in lsearch files need not
3648 contain colons. This was done some time before quoting was made available
3649 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3650 (notated using the quote facility) so as to distinguish them from IPv4
3651 keys. This meant that lookups for IP addresses in host lists did not work
3652 for iplsearch lookups.
3654 This has been fixed by arranging for IPv6 addresses to be expressed with
3655 colons if the lookup type is iplsearch. This is not incompatible, because
3656 previously such lookups could never work.
3658 The situation is now rather anomalous, since one *can* have colons in
3659 ordinary lsearch keys. However, making the change in all cases is
3660 incompatible and would probably break a number of configurations.
3662 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3665 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3666 conversion specification without a maximum field width, thereby enabling
3667 a rogue spamd server to cause a buffer overflow. While nobody in their
3668 right mind would setup Exim to query an untrusted spamd server, an
3669 attacker that gains access to a server running spamd could potentially
3670 exploit this vulnerability to run arbitrary code as the Exim user.
3672 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3673 $primary_hostname instead of what libspf2 thinks the hosts name is.
3675 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3676 a directory entry by the name of the lookup key. Previously, if a
3677 symlink pointed to a non-existing file or a file in a directory that
3678 Exim lacked permissions to read, a lookup for a key matching that
3679 symlink would fail. Now it is enough that a matching directory entry
3680 exists, symlink or not. (Bugzilla 503.)
3682 PH/03 The body_linecount and body_zerocount variables are now exported in the
3685 PH/04 Added the $dnslist_matched variable.
3687 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3688 This means they are set thereafter only if the connection becomes
3691 PH/06 Added the client_condition to authenticators so that some can be skipped
3692 by clients under certain conditions.
3694 PH/07 The error message for a badly-placed control=no_multiline_responses left
3695 "_responses" off the end of the name.
3697 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3699 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3700 (without spaces) instead of just copying the configuration text.
3702 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3704 PH/11 Added $max_received_linelength.
3706 PH/12 Added +ignore_defer and +include_defer to host lists.
3708 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3709 way in which PCRE > 7.0 is built.
3711 PH/14 Implemented queue_only_load_latch.
3713 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3714 MAIL command. The effect was to mangle the value on 64-bit systems.
3716 PH/16 Another patch from the Sieve maintainer.
3718 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3720 PH/18 If a system quota error occurred while trying to create the file for
3721 a maildir delivery, the message "Mailbox is full" was not appended to the
3722 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3723 only to a quota excession during the actual writing of the file.
3725 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3726 characters?) which causes problems in log lines. The DN values are now
3727 passed through string_printing() before being added to log lines.
3729 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3730 and InterBase are left for another time.)
3732 PH/21 Added message_body_newlines option.
3734 PH/22 Guard against possible overflow in moan_check_errorcopy().
3736 PH/23 POSIX allows open() to be a macro; guard against that.
3738 PH/24 If the recipient of an error message contained an @ in the local part
3739 (suitably quoted, of course), incorrect values were put in $domain and
3740 $local_part during the evaluation of errors_copy.
3746 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3747 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3750 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3751 issue a MAIL command.
3753 PH/02 In an ACL statement such as
3755 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3757 if a client was not listed at all, or was listed with a value other than
3758 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3759 the condition was not true (as it should be), so access was not denied.
3760 The bug was that the ! inversion was incorrectly passed on to the second
3761 item. This has been fixed.
3763 PH/03 Added additional dnslists conditions == and =& which are different from
3764 = and & when the dns lookup returns more than one IP address.
3766 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3767 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3769 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3770 FSYNC, which compiles an option called disable_fsync that allows for
3771 bypassing fsync(). The documentation is heavily laced with warnings.
3773 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3775 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3776 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3777 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3778 including adding "make clean"; (3) Added -fPIC when compiling the test
3779 dynamically loaded module, to get rid of a warning.
3781 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3782 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3783 = 0s. The bug is otherwise harmless.
3785 PH/07 There was a bug in the dovecot authenticator such that the value of
3786 $auth1 could be overwritten, and so not correctly preserved, after a
3787 successful authentication. This usually meant that the value preserved by
3788 the server_setid option was incorrect.
3790 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3792 PH/09 Installed PCRE release 7.0.
3794 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3795 run for batched SMTP input. It is now run at the start of every message
3796 in the batch. While fixing this I discovered that the process information
3797 (output by running exiwhat) was not always getting set for -bs and -bS
3798 input. This is fixed, and it now also says "batched" for BSMTP.
3800 PH/11 Added control=no_pipelining.
3802 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3803 patch, slightly modified), and move the expansion of helo_data till after
3804 the connection is made in the smtp transport (so it can use these
3807 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3809 PH/14 Added log_selector = +pid.
3811 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3813 PH/16 Add ${if forany and ${if forall.
3815 PH/17 Added dsn_from option to vary the From: line in DSNs.
3817 PH/18 Flush SMTP output before performing a callout, unless control =
3818 no_callout_flush is set.
3820 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3821 was true (the default) a successful delivery failed to delete the retry
3822 item, thus causing premature timeout of the address. The bug is now
3825 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3827 PH/21 Long custom messages for fakedefer and fakereject are now split up
3828 into multiline responses in the same way that messages for "deny" and
3829 other ACL rejections are.
3831 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3832 with slight modification.
3834 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3835 draft, changing the syntax and factoring some duplicate code".
3837 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3838 for deliveries of the second and subsequent messages over the same SMTP
3841 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3842 ${reduce, with only minor "tidies".
3844 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3846 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3847 expansion side effects.
3849 PH/27 When a message times out after an over-quota error from an Exim-imposed
3850 quota, the bounce message says "mailbox is full". This message was not
3851 being given when it was a system quota that was exceeded. It now should
3854 MH/03 Made $recipients available in local_scan(). local_scan() already has
3855 better access to the recipient list through recipients_list[], but
3856 $recipients can be useful in postmaster-provided expansion strings.
3858 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3859 in the case of a MAIL command with additional options following the
3860 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3861 were accidentally chopped off.
3863 PH/29 SMTP synchronization checks are implemented when a command is read -
3864 there is a check that no more input is waiting when there shouldn't be
3865 any. However, for some commands, a delay in an ACL can mean that it is
3866 some time before the response is written. In this time, more input might
3867 arrive, invalidly. So now there are extra checks after an ACL has run for
3868 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3869 pipelining has not been advertised.
3871 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3873 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3874 returned in the SMTP response when smtp_return_error_details was false.
3875 This has been fixed.
3877 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3878 instead of the C I/O that was originally supplied, because problems were
3879 reported on Solaris.
3881 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3882 Exim which did not show up earlier: it was assuming that a call to
3883 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3884 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3885 was a macro that became an assignment, so it seemed to work. This has
3886 changed to a proper function call with a void return, hence the compile
3887 error. Exim's code has been fixed.
3889 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3892 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3894 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3896 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3897 because it turns out to be too big or there is a timeout) there is no
3898 "Completed" line in the log. When some messages of this type were
3899 selected by exigrep, they were listed as "not completed". Others were
3900 picked up by some special patterns. I have improved the selection
3901 criteria to be more general.
3903 PH/38 The host_find_failed option in the manualroute router can now be set
3904 to "ignore", to completely ignore a host whose IP address cannot be
3905 found. If all hosts are ignored, the behaviour is controlled by the new
3906 host_all_ignored option.
3908 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3909 homing or because of multiple MX records with /mx) generated more than
3910 one IP address, and the following item turned out to be the local host,
3911 all the secondary addresses of the first item were incorrectly removed
3912 from the list, along with the local host and any following hosts (which
3913 is what is supposed to happen).
3915 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3916 whoever called Exim into the -H file. In the case of the daemon it was
3917 behaving confusingly. When first started, it used values for whoever
3918 started the daemon, but after a SIGHUP it used the Exim user (because it
3919 calls itself on a restart). I have changed the code so that it now always
3922 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3923 message are rejected with the same error (e.g. no authentication or bad
3924 sender address), and a DATA command is nevertheless sent (as can happen
3925 with PIPELINING or a stupid MUA), the error message that was given to the
3926 RCPT commands is included in the rejection of the DATA command. This is
3927 intended to be helpful for MUAs that show only the final error to their
3930 PH/42 Another patch from the Sieve maintainer.
3932 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3933 Eximstats - Fixed some broken HTML links and added missing column headers
3935 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3936 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3938 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3940 PH/43 Yet another patch from the Sieve maintainer.
3942 PH/44 I found a way to check for a TCP/IP connection going away before sending
3943 the response to the final '.' that terminates a message, but only in the
3944 case where the client has not sent further data following the '.'
3945 (unfortunately, this is allowed). However, in many cases there won't be
3946 any further data because there won't be any more messages to send. A call
3947 to select() can be used: if it shows that the input is "ready", there is
3948 either input waiting, or the socket has been closed. An attempt to read
3949 the next input character can distinguish the two cases. Previously, Exim
3950 would have sent an OK response which the client would never have see.
3951 This could lead to message repetition. This fix should cure that, at
3952 least in a lot of common cases.
3954 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3955 advertised in response to EHLO.
3961 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3962 fixed by 4.65/MH/01 (is this a record?) are fixed:
3964 (i) An empty string was always treated as zero by the numeric comparison
3965 operators. This behaviour has been restored.
3967 (ii) It is documented that the numeric comparison operators always treat
3968 their arguments as decimal numbers. This was broken in that numbers
3969 starting with 0 were being interpreted as octal.
3971 While fixing these problems I realized that there was another issue that
3972 hadn't been noticed. Values of message_size_limit (both the global option
3973 and the transport option) were treated as octal if they started with 0.
3974 The documentation was vague. These values are now always treated as
3975 decimal, and I will make that clear in the documentation.
3981 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3982 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3985 MH/01 Don't check that the operands of numeric comparison operators are
3986 integers when their expansion is in "skipping" mode (fixes bug
3987 introduced by 4.64-PH/07).
3989 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3990 child addresses, Exim now panics and dies. Previously, because the count
3991 is held in a short int, deliveries were likely to be lost. As such a
3992 large number of recipients for a single message is ridiculous
3993 (performance will be very, very poor), I have chosen to impose a limit
3994 rather than extend the field.
4000 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4001 leftover -K file (the existence of which was triggered by #402).
4002 While we were at it, introduced process PID as part of the -K
4003 filename. This should rule out race conditions when creating
4006 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4007 processing considerably. Previous code took too long for large mails,
4008 triggering a timeout which in turn triggers #401.
4010 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4011 in the DK code in transports.c. sendfile() is not really portable,
4012 hence the _LINUX specificness.
4014 TF/01 In the add_headers option to the mail command in an Exim filter,
4015 there was a bug that Exim would claim a syntax error in any
4016 header after the first one which had an odd number of characters
4019 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4020 callout verification, Exim cached a "reject" for the entire domain. This
4021 is correct for most verifications, but it is not correct for a recipient
4022 verification with use_sender or use_postmaster set, because in that case
4023 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4024 case of MAIL FROM:<> rejection from other early rejections (e.g.
4025 rejection of HELO). When verifying a recipient using a non-null MAIL
4026 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4027 Whatever the result of the callout, the value of the domain cache is
4028 left unchanged (for any other kind of callout, getting as far as trying
4029 RCPT means that the domain itself is ok).
4031 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4034 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4035 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4036 session, the daemon ignores ECONNECT errors and logs others; it now
4037 ignores EPIPE as well.
4039 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4040 (quoted-printable decoding).
4042 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4043 later the small subsequent patch to fix an introduced bug.
4045 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4047 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4049 PH/08 An error is now given if message_size_limit is specified negative.
4051 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4052 to be given (somewhat) arbitrary names.
4054 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4057 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4058 miscellaneous code fixes
4060 PH/10 Added the log_reject_target ACL modifier to specify where to log
4063 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4064 hostname. This is wrong, because it relates to the incoming message (and
4065 probably the interface on which it is arriving) and not to the outgoing
4066 callout (which could be using a different interface). This has been
4067 changed to use the value of the helo_data option from the smtp transport
4068 instead - this is what is used when a message is actually being sent. If
4069 there is no remote transport (possible with a router that sets up host
4070 addresses), $smtp_active_hostname is used.
4072 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4073 tweaks were necessary in order to get it to work (see also 21 below):
4074 (a) The code assumed that strncpy() returns a negative number on buffer
4075 overflow, which isn't the case. Replaced with Exim's string_format()
4077 (b) There were several signed/unsigned issues. I just did the minimum
4078 hacking in of casts. There is scope for a larger refactoring.
4079 (c) The code used strcasecmp() which is not a standard C function.
4080 Replaced with Exim's strcmpic() function.
4081 (d) The code set only $1; it now sets $auth1 as well.
4082 (e) A simple test gave the error "authentication client didn't specify
4083 service in request". It would seem that Dovecot has changed its
4084 interface. Fortunately there's a specification; I followed it and
4085 changed what the client sends and it appears to be working now.
4087 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4090 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4091 address A is aliased to B and C, where B exists and C does not. Without
4092 -v the output is "A verified" because verification stops after a
4093 successful redirection if more than one address is generated. However,
4094 with -v the child addresses are also verified. Exim was outputting "A
4095 failed to verify" and then showing the successful verification for C,
4096 with its parentage. It now outputs "B failed to verify", showing B's
4097 parentage before showing the successful verification of C.
4099 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4100 look up a TXT record in a specific list after matching in a combined
4103 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4104 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4105 they consult the DNS. I had assumed they would set it the way they
4106 wanted; and indeed my experiments on Linux seem to show that in some
4107 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4108 To be on the safe side, however, I have now made the interface to
4109 host_find_byname() similar to host_find_bydns(), with an argument
4110 containing the DNS resolver options. The host_find_byname() function now
4111 sets these options at its start, just as host_find_bydns() does. The smtp
4112 transport options dns_qualify_single and dns_search_parents are passed to
4113 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4114 of host_find_byname() use the default settings of RES_DEFNAMES
4115 (qualify_single) but not RES_DNSRCH (search_parents).
4117 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4118 spool_read_header() do less string testing, by means of a preliminary
4119 switch on the second character of optional "-foo" lines. (This is
4120 overdue, caused by the large number of possibilities that now exist.
4121 Originally there were few.) While I was there, I also converted the
4122 str(n)cmp tests so they don't re-test the leading "-" and the first
4123 character, in the hope this might squeeze out yet more improvement.
4125 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4126 flag allowing group syntax was set by the header_syntax check but not
4127 turned off, possible causing trouble later; (2) The flag was not being
4128 set at all for the header_verify test, causing "group"-style headers to
4129 be rejected. I have now set it in this case, and also caused header_
4130 verify to ignore an empty address taken from a group. While doing this, I
4131 came across some other cases where the code for allowing group syntax
4132 while scanning a header line wasn't quite right (mostly, not resetting
4133 the flag correctly in the right place). These bugs could have caused
4134 trouble for malformed header lines. I hope it is now all correct.
4136 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4137 with the "reply" argument non-NULL. The code, however (which originally
4138 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4139 but it didn't always do it. This confused somebody who was copying the
4140 code for some other use. I have removed all the tests.
4142 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4143 feature that was used to support insecure browsers during the U.S. crypto
4144 embargo. It requires special client support, and Exim is probably the
4145 only MTA that supported it -- and would never use it because real RSA is
4146 always available. This code has been removed, because it had the bad
4147 effect of slowing Exim down by computing (never used) parameters for the
4148 RSA_EXPORT functionality.
4150 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4151 authenticator to fail if there's a tab character in the incoming data
4152 (there should never be unless someone is messing about, as it's supposed
4153 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4154 if the connection is using TLS or if the remote IP is the same as the
4155 local IP, and the "valid-client-cert option" if a client certificate has
4158 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4159 authenticators. This can be used for authorization after authentication
4160 succeeds. (In the case of plaintext, it servers for both authentication
4163 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4164 if any retry times were supplied.
4166 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4167 connection, where there is no client IP address to check. In this
4168 situation, the verify now always succeeds.
4170 PH/25 Applied John Jetmore's -Mset patch.
4172 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4174 PH/27 In a string expansion for a processed (not raw) header when multiple
4175 headers of the same name were present, leading whitespace was being
4176 removed from all of them, but trailing whitespace was being removed only
4177 from the last one. Now trailing whitespace is removed from each header
4178 before concatenation. Completely empty headers in a concatenation (as
4179 before) are ignored.
4181 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4182 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4184 PH/29 [Removed. This was a change that I later backed out, and forgot to
4185 correct the ChangeLog entry (that I had efficiently created) before
4186 committing the later change.]
4188 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4189 address errors (4xx response to RCPT) over the same connection as other
4190 messages routed to the same hosts. Such deliveries are always "forced",
4191 so retry times are not inspected. This resulted in far too many retries
4192 for the affected addresses. The effect occurred only when there were more
4193 hosts than the hosts_max_try setting in the smtp transport when it had
4194 the 4xx errors. Those hosts that it had tried were not added to the list
4195 of hosts for which the message was waiting, so if all were tried, there
4196 was no problem. Two fixes have been applied:
4198 (i) If there are any address or message errors in an SMTP delivery, none
4199 of the hosts (tried or untried) are now added to the list of hosts
4200 for which the message is waiting, so the message should not be a
4201 candidate for sending over the same connection that was used for a
4202 successful delivery of some other message. This seems entirely
4203 reasonable: after all the message is NOT "waiting for some host".
4204 This is so "obvious" that I'm not sure why it wasn't done
4205 previously. Hope I haven't missed anything, but it can't do any
4206 harm, as the worst effect is to miss an optimization.
4208 (ii) If, despite (i), such a delivery is accidentally attempted, the
4209 routing retry time is respected, so at least it doesn't keep
4210 hammering the server.
4212 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4213 in ${readsocket because some servers need this prod.
4215 PH/32 Added some extra debug output when updating a wait-xxx database.
4217 PH/33 The hint "could be header name not terminated by colon", which has been
4218 given for certain expansion errors for a long time, was not being given
4219 for the ${if def:h_colon_omitted{... case.
4221 PH/34 The spec says: "With one important exception, whenever a domain list is
4222 being scanned, $domain contains the subject domain." There was at least
4223 one case where this was not true.
4225 PH/35 The error "getsockname() failed: connection reset by peer" was being
4226 written to the panic log as well as the main log, but it isn't really
4227 panic-worthy as it just means the connection died rather early on. I have
4228 removed the panic log writing for the ECONNRESET error when getsockname()
4231 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4232 runs only) independently of the message's sender address. This meant
4233 that, if the 4xx error was in fact related to the sender, a different
4234 message to the same recipient with a different sender could confuse
4235 things. In particular, this can happen when sending to a greylisting
4236 server, but other circumstances could also provoke similar problems.
4237 I have changed the default so that the retry time for these errors is now
4238 based a combination of the sender and recipient addresses. This change
4239 can be overridden by setting address_retry_include_sender=false in the
4242 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4243 remote server are returned as part of bounce messages. This was not
4244 happening for LMTP over a pipe (the lmtp transport), but now it is the
4245 same for both kinds of LMTP.
4247 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4248 in header lines that were in fact CNAMEs. This is no longer the case.
4250 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4251 and queue runs started by the daemon processed all messages. This has
4252 been fixed so that -R and -S can now usefully be given with -q<time>.
4254 PH/40 Import PCRE release 6.7 (fixes some bugs).
4256 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4258 PH/42 Give an error if -q is specified more than once.
4260 PH/43 Renamed the variables $interface_address and $interface_port as
4261 $received_ip_address and $received_port, to make it clear that these
4262 values apply to message reception, and not to the outgoing interface when
4263 a message is delivered. (The old names remain recognized, of course.)
4265 PH/44 There was no timeout on the connect() call when using a Unix domain
4266 socket in the ${readsocket expansion. There now is.
4268 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4269 be meaningful with "accept".
4271 SC/01 Eximstats V1.43
4272 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4274 SC/02 Eximstats V1.44
4275 Use a glob alias rather than an array ref in the generated
4276 parser. This improves both readability and performance.
4278 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4279 Collect SpamAssassin and rejection statistics.
4280 Don't display local sender or destination tables unless
4281 there is data to show.
4282 Added average volumes into the top table text output.
4284 SC/04 Eximstats V1.46
4285 Collect data on the number of addresses (recipients)
4286 as well as the number of messages.
4288 SC/05 Eximstats V1.47
4289 Added 'Message too big' to the list of mail rejection
4290 reasons (thanks to Marco Gaiarin).
4292 SC/06 Eximstats V1.48
4293 Mainlog lines which have GMT offsets and are too short to
4294 have a flag are now skipped.
4296 SC/07 Eximstats V1.49 (Alain Williams)
4297 Added the -emptyok flag.
4299 SC/08 Eximstats V1.50
4300 Fixes for obtaining the IP address from reject messages.
4302 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4303 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4304 whitespace changes from 4.64-PH/27
4306 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4309 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4310 are found, allow negative numbers in numeric criteria)
4312 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4314 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4315 to match changes made in 4.64-PH/43
4317 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4319 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4320 "domain" is actually an IP address, and force a failure. This locks out
4321 those revolvers/nameservers that support "A-for-A" lookups, in
4322 contravention of the specifications.
4324 PH/48 When a host name was looked up from an IP address, and the subsequent
4325 forward lookup of the name timed out, the host name was left in
4326 $sender_host_name, contrary to the specification.
4328 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4329 restricted to single-key lookups, Exim was not diagnosing an error if
4330 * or *@ was used with a query-style lookup.
4332 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4334 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4335 long ago, but noone interested enough thought of it. Let's just say that
4336 the "1.1" means that there are some new functions that weren't there at
4337 some point in the past.
4339 PH/51 Error processing for expansion failure of helo_data from an smtp
4340 transport during callout processing was broken.
4342 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4343 tested/used via the -bh/-bhc/-bs options.
4345 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4346 bug, fixed in subsequent PCRE releases).
4348 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4349 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4351 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4357 SC/01 Use a glob alias rather than an array ref in eximstats generated
4358 parser. This improves both readability and performance.
4360 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4361 Don't display local sender or destination tables in eximstats unless
4362 there is data to show.
4363 Added average volumes into the eximstats top table text output.
4365 SC/03 Collect data on the number of addresses (recipients) as well
4366 as the number of messages in eximstats.
4368 TF/01 Correct an error in the documentation for the redirect router. Exim
4369 does (usually) call initgroups() when daemonizing.
4371 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4372 with consistent privilege compared to when running as a daemon.
4374 TF/03 Note in the spec that $authenticated_id is not set for local
4375 submissions from trusted users.
4377 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4378 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4380 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4381 by adding some example configuration directives to the default
4382 configuration file. A little bit of work is required to uncomment the
4383 directives and define how usernames and passwords are checked, but
4384 there is now a framework to start from.
4386 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4387 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4388 without this. I don't know how relevant this is to other LDAP libraries.
4390 PH/02 Add the verb name to the "unknown ACL verb" error.
4392 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4394 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4396 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4397 directory not expanded when it should be if an expanded home directory
4398 was set for the address (which is overridden by the transport).
4400 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4403 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4404 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4405 because it is too late at that time, and has no effect.
4407 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4408 security issue with \' (bugzilla #107). I could not use the
4409 PQescapeStringConn() function, because it needs a PGconn value as one of
4412 PH/08 When testing addresses using -bt, indicate those final addresses that
4413 are duplicates that would not cause an additional delivery. At least one
4414 person was confused, thinking that -bt output corresponded to deliveries.
4415 (Suppressing duplicates isn't a good idea as you lose the information
4416 about possibly different redirections that led to the duplicates.)
4418 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4419 systems where poll() doesn't work, in particular OS X.
4421 PH/10 Added more information to debugging output for retry time not reached.
4423 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4424 operations in malware.c.
4426 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4429 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4430 syslog_duplication set false, log lines that would normally be written
4431 both the the main log and to the reject log were not written to syslog at
4434 PH/14 In the default configuration, change the use of "message" in ACL warn
4435 statements to "add_header".
4437 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4438 not followed by a command (e.g. "seen endif").
4440 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4441 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4444 PH/17 Added extra conditions to the default value of delay_warning_condition
4448 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4449 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4450 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4453 The Auto-Submitted: and various List- headers are standardised, whereas I
4454 don't think Precedence: ever was.
4456 PH/18 Refactored debugging code in route_finduser() to show more information,
4457 in particular, the error code if getpwnam() issues one.
4459 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4460 This is apparently needed in addition to the PH/07 change above to avoid
4461 any possible encoding problems.
4463 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4464 but not after initializing Perl.
4466 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4467 output them only if debugging. By default they are written stderr,
4468 apparently, which is not desirable.
4470 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4473 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4476 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4478 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4479 authenticated or an ident call has been made. Suppress the default
4480 values for $authenticated_id and $authenticated_sender (but permit -oMai
4481 and -oMas) when testing with -bh.
4483 PH/24 Re-jigged the order of the tests in the default configuration so that the
4484 tests for valid domains and recipients precede the DNS black list and CSA
4485 tests, on the grounds that those ones are more expensive.
4487 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4488 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4489 This bug exists in every version of Exim that I still have, right back to
4492 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4493 However, an attempt to turn on case-sensitivity in a regex key by
4494 including (?-i) didn't work because the subject string was already
4495 lowercased, and the effects were non-intuitive. It turns out that a
4496 one-line patch can be used to allow (?-i) to work as expected.
4502 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4503 other effects) broke the use of negated acl sub-conditions.
4505 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4508 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4509 "Deny" causes Exim to reject the incoming connection with a 554 error.
4510 Unfortunately, if there is a major crisis, such as a disk failure,
4511 tcp-wrappers gives "deny", whereas what one would like would be some
4512 kind of temporary error. A kludge has been added to help with this.
4513 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4514 554 error is used if errno is still zero or contains ENOENT (which occurs
4515 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4518 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4520 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4521 errors. Otherwise a message that provokes a temporary error (when other
4522 messages do not) can cause a whole host to time out.
4524 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4525 addresses were routed directly to files or pipes from a redirect router.
4526 File deliveries just didn't batch; pipe deliveries might have suffered
4529 PH/06 A failure to get a lock for a hints database would erroneously always say
4530 "Failed to get write lock", even when it was really a read lock.
4532 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4533 of 0600. This has been changed to use the value of the lockfile_mode
4534 option (which defaults to 0600).
4536 PH/08 Applied small patch from the Sieve maintainer.
4538 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4539 folder from quota calculations, a direct delivery into this folder messed
4540 up the contents of the maildirsize file. This was because the regex was
4541 used only to exclude .Trash (or whatever) when the size of the mailbox
4542 was calculated. There was no check that a delivery was happening into an
4543 excluded directory. This bug has been fixed by ignoring all quota
4544 processing for deliveries into excluded directories.
4546 PH/10 Added the maildirfolder_create_regex option to appendfile.
4552 PH/01 The code for finding all the local interface addresses on a FreeBSD
4553 system running IPv6 was broken. This may well have applied to all BSD
4554 systems, as well as to others that have similar system calls. The broken
4555 code found IPv4 interfaces correctly, but gave incorrect values for the
4556 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4557 that it would not match correctly against @[] and not recognize the IPv6
4560 PH/02 The ipliteral router was not recognizing addresses of the form user@
4561 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4563 PH/03 Added disable_ipv6.
4565 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4566 decoded form, because it is most often used to construct To: headers
4567 lines in autoreplies, and the decoded form may well be syntactically
4568 invalid. However, $reply_address has leading white space removed, and all
4569 newlines turned into spaces so that the autoreply transport does not
4572 PH/05 If group was specified without a user on a router, and no group or user
4573 was specified on a transport, the group from the router was ignored.
4575 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4576 for visible compile-time settings that can be used to change these
4577 numbers, for those that want even more. Backwards compatibility with old
4578 spool files has been maintained. However, going back to a previous Exim
4579 release will lost any variables that are in spool files.
4581 PH/07 Two small changes when running in the test harness: increase delay when
4582 passing a TCP/IP connection to a new process, in case the original
4583 process has to generate a bounce, and remove special handling of
4584 127.0.0.2 (sic), which is no longer necessary.
4586 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4587 be the same on different OS.
4589 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4592 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4593 whether --show-vars was specified or not
4595 JJ/02 exipick: Added support for new ACL variable spool format introduced
4598 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4599 syntactically invalid From: or Reply-to: line, and a filter used this to
4600 generate an autoreply, and therefore failed to obtain an address for the
4601 autoreply, Exim could try to deliver to a non-existent relative file
4602 name, causing unrelated and misleading errors. What now happens is that
4603 it logs this as a hard delivery error, but does not attempt to create a
4606 PH/11 The exinext utility has a -C option for testing purposes, but although
4607 the given file was scanned by exinext itself; it wasn't being passed on
4608 when Exim was called.
4610 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4611 an end-of-file indication when reading a command response.
4613 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4614 compiled. In many other places in Exim, IPv6 addresses are always
4615 recognized, so I have changed this. It also means that IPv4 domain
4616 literals of the form [IPV4:n.n.n.n] are now always recognized.
4618 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4619 used if the router is not running as root, for example, when verifying at
4620 ACL time, or when using -bh. The debugging output from this situation was
4621 non-existent - all you got was a failure to exec. I have made two
4624 (a) Failures to set uid/gid, the current directory, or a process leader
4625 in a subprocess such as that created by queryprogram now generate
4626 suitable debugging output when -d is set.
4628 (b) The queryprogram router detects when it is not running as root,
4629 outputs suitable debugging information if -d is set, and then runs
4630 the subprocess without attempting to change uid/gid.
4632 PH/15 Minor change to Makefile for building test_host (undocumented testing
4635 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4636 additional section of a DNS packet that returns MX or SRV records.
4637 Instead, it always explicitly searches for A/AAAA records. This avoids
4638 major problems that occur when a DNS server includes only records of one
4639 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4640 fixed another bug: if SRV records were looked up and the corresponding
4641 address records were *not* found in the additional section, the port
4642 values from the SRV records were lost.
4644 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4645 using the correct key (the original address) when searching the retry
4646 rules in order to find which one to use for generating the retry hint.
4648 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4649 adding the default one. Similarly, if it contains a Reply-To: header, the
4650 errors_reply_to option, if set, is not used.
4652 PH/19 When calculating a retry time, Exim used to measure the "time since
4653 failure" by looking at the "first failed" field in the retry record. Now
4654 it does not use this if it is later than than the arrival time of the
4655 message. Instead it uses the arrival time. This makes for better
4656 behaviour in cases where some deliveries succeed, thus re-setting the
4657 "first failed" field. An example is a quota failure for a huge message
4658 when small messages continue to be delivered. Without this change, the
4659 "time since failure" will always be short, possible causing more frequent
4660 delivery attempts for the huge message than are intended.
4661 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4663 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4664 $1, $2, $3) because the numerical variables can be reset during some
4665 expansion items (e.g. "match"), thereby losing the authentication data.
4667 PH/21 Make -bV show the size of off_t variables so that the test suite can
4668 decide whether to run tests for quotas > 2G.
4670 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4671 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4672 filecount value is greater than 2G or if a quota value is greater than 2G
4673 on a system where the size of off_t is not greater than 4, a panic error
4676 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4677 never match. The debug and -bh output now contains an explicit error
4678 message indicating a malformed IPv4 address or mask.
4680 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4681 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4682 PH/23 above applies.
4684 PH/25 Do not write to syslog when running in the test harness. The only
4685 occasion when this arises is a failure to open the main or panic logs
4686 (for which there is an explicit test).
4688 PH/26 Added the /no_tell option to "control=freeze".
4690 PH/27 If a host name lookup failed very early in a connection, for example, if
4691 the IP address matched host_lookup and the reverse lookup yielded a name
4692 that did not have a forward lookup, an error message of the form "no IP
4693 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4694 could be logged. Now it outputs the IP address instead of "NULL".
4696 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4697 allows the sender and the authenticated sender to be set when
4698 submitting a message from within Exim. Since child_open_exim() is
4699 documented for local_scan(), the new function should be too.
4701 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4702 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4703 results in an empty string is now treated as unset.
4705 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4707 PH/31 Added sender_verify_fail logging option.
4709 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4710 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4711 tidied the source and removed it altogether.
4713 PH/33 When a queue run was abandoned because the load average was too high, a
4714 log line was always written; now it is written only if the queue_run log
4715 selector is set. In addition, the log line for abandonment now contains
4716 information about the queue run such as the pid. This is always present
4717 in "start" and "stop" lines but was omitted from the "abandon" line.
4719 PH/34 Omit spaces between a header name and the colon in the error message that
4720 is given when verify = headers_syntax fails (if there are lots of them,
4721 the message gets confusing).
4723 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4724 names, as there are now some PTR records that contain slashes. This check
4725 is only to protect against broken name servers that fall over on strange
4726 characters, so the fact that it applies to all lookups doesn't matter.
4728 PH/36 Now that the new test suite is complete, we can remove some of the
4729 special code in Exim that was needed for the old test suite. For example,
4730 sorting DNS records because real resolvers return them in an arbitrary
4731 order. The new test suite's fake resolver always returns records in the
4734 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4735 bounces) except when queue_only is set, to avoid logging races between
4736 the different processes.
4738 PH/38 Panic-die if .include specifies a non-absolute path.
4740 PH/39 A tweak to the "H" retry rule from its user.
4742 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4743 a label. They prevented compilation on older perls.
4745 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4746 a warning to be raised on newish perls.
4748 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4749 on queue. Changes to match documented behaviour of showing count of
4750 messages matching specified criteria.
4752 PH/40 Changed the default ident timeout from 30s to 5s.
4754 PH/41 Added support for the use of login_cap features, on those BSD systems
4755 that have them, for controlling the resources used by pipe deliveries.
4757 PH/42 The content-scanning code uses fopen() to create files in which to put
4758 message data. Previously it was not paying any attention to the mode of
4759 the files. Exim runs with umask(0) because the rest of the code creates
4760 files with open(), and sets the required mode explicitly. Thus, these
4761 files were ending up world-writeable. This was not a big issue, because,
4762 being within the spool directory, they were not world-accessible. I have
4763 created a function called modefopen, which takes an additional mode
4764 argument. It sets umask(777), creates the file, chmods it to the required
4765 mode, then resets the umask. All the relevant calls to fopen() in the
4766 content scanning code have been changed to use this function.
4768 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4769 to 24 hours. This avoids potential overflow problems when processing G
4770 and H retry rules. I suspect nobody ever tinkers with this value.
4772 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4774 PH/45 When the plaintext authenticator is running as a client, the server's
4775 challenges are checked to ensure they are valid base64 strings. By
4776 default, the authentication attempt is cancelled if an invalid string is
4777 received. Setting client_ignore_invalid_base64 true ignores these errors.
4778 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4779 they are received. Thus, the responses can be made to depend on the
4780 challenges. If an invalid string is ignored, an empty string is placed in
4783 PH/46 Messages that are created by the autoreply transport now contains a
4784 References: header, in accordance with RFCs 2822 and 3834.
4786 PH/47 Added authenticated_sender_force to the smtp transport.
4788 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4790 PH/49 Installed latest patch from the Sieve maintainer.
4792 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4793 was also set, the appendfile transport was unnecessarily scanning a
4794 directory of message files (e.g. for maildir delivery) to find the count
4795 of files (along with the size), even though it did not need this
4796 information. It now does the scan only if it needs to find either the
4797 size of the count of files.
4799 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4801 PH/52 Two bugs concerned with error handling when the smtp transport is
4804 (i) Exim was not creating retry information for temporary errors given
4805 for individual recipients after the DATA command when the smtp transport
4806 was used in LMTP mode. This meant that they could be retried too
4807 frequently, and not timed out correctly.
4809 (ii) Exim was setting the flag that allows error details to be returned
4810 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4811 recipients that were returned after the DATA command.
4813 PH/53 This is related to PH/52, but is more general: for any failing address,
4814 when detailed error information was permitted to be returned to the
4815 sender, but the error was temporary, then after the final timeout, only
4816 "retry timeout exceeded" was returned. Now it returns the full error as
4817 well as "retry timeout exceeded".
4819 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4820 do this, and (what is worse) MTAs that accept it.
4822 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4823 will now be deprecated.
4825 PH/56 New os.c-cygwin from the Cygwin maintainer.
4827 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4828 formats (previously only available in exim formats via -bpr, -bpru,
4829 and -bpra. Now also available in native and exiqgrep formats)
4831 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4832 with very large, slow to parse queues
4834 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4836 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4838 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4839 responses to authentication challenges, though it was showing the
4840 challenges; (ii) I've removed the CR characters from the debug output for
4843 PH/58 Allow for the insertion of a newline as well as a space when a string
4844 is turned into more than one encoded-word during RFC 2047 encoding. The
4845 Sieve code now uses this.
4847 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4848 data_4xx, lost_connection, tls_required.
4850 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4851 message was being sent as an SMTP response.
4853 PH/61 Add -l and -k options to exicyclog.
4855 PH/62 When verifying, if an address was redirected to one new address, so that
4856 verification continued, and the new address failed or deferred after
4857 having set something in $address_data, the value of $address_data was not
4858 passed back to the ACL. This was different to the case when no
4859 redirection occurred. The value is now passed back in both cases.
4861 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4862 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4863 use login_cap.h, so on its own it isn't the distinguishing feature. The
4864 new name refers directly to the setclassresources() function.
4866 PH/65 Added configuration files for NetBSD3.
4868 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4870 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4871 is preferred over IPv4.
4873 PH/68 The bounce_return_message and bounce_return_body options were not being
4874 honoured for bounces generated during the reception of non-SMTP messages.
4875 In particular, this applied to messages rejected by the ACL. This bug has
4876 been fixed. However, if bounce_return_message is true and bounce_return_
4877 body is false, the headers that are returned for a non-SMTP message
4878 include only those that have been read before the error was detected.
4879 (In the case of an ACL rejection, they have all been read.)
4881 PH/69 The HTML version of the specification is now built in a directory called
4882 spec_html instead of spec.html, because the latter looks like a path with
4883 a MIME-type, and this confuses some software.
4885 PH/70 Catch two compiler warnings in sieve.c.
4887 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4888 function verify_get_ident() calls ip_connect() to connect a socket, but
4889 if the "connect()" function timed out, ip_connect() used to close the
4890 socket. However, verify_get_ident() also closes the socket later, and in
4891 between Exim writes to the log, which may get opened at this point. When
4892 the socket was closed in ip_connect(), the log could get the same file
4893 descriptor number as the socket. This naturally causes chaos. The fix is
4894 not to close the socket in ip_connect(); the socket should be closed by
4895 the function that creates it. There was only one place in the code where
4896 this was missing, in the iplookup router, which I don't think anybody now
4897 uses, but I've fixed it anyway.
4899 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4900 well as to direct DNS lookups. Otherwise the handling of names in host
4901 lists is inconsistent and therefore confusing.
4907 PH/01 Two changes to the default runtime configuration:
4909 (1) Move the checks for relay_from_hosts and authenticated clients from
4910 after to before the (commented out) DNS black list checks.
4912 (2) Add control=submission to the relay_from_hosts and authenticated
4913 clients checks, on the grounds that messages accepted by these
4914 statements are most likely to be submissions.
4916 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4918 (1) Generate an error if the third argument for the ${prvs expansion is
4921 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4924 (3) Reset the variables that are obtained from the first argument of
4925 ${prvscheck and used in the second argument before leaving the code,
4926 because their memory is reclaimed, so using them afterwards may do
4929 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4930 one (it's much easier than Tom thought :-).
4932 (5) Because of (4), we can now allow for the use of $prvscheck_result
4933 inside the third argument.
4935 PH/03 For some reason, the default setting of PATH when running a command from
4936 a pipe transport was just "/usr/bin". I have changed it to
4939 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4940 anything to be listed in the output from -bV.
4942 PH/05 When a filter generated an autoreply, the entire To: header line was
4943 quoted in the delivery log line, like this:
4945 => >A.N.Other <ano@some.domain> <original@ddress> ...
4947 This has been changed so that it extracts the operative address. There
4948 may be more than one such address. If so, they are comma-separated, like
4951 => >ano@some.domain,ona@other.domain <original@ddress> ...
4953 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4954 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4955 not being looked up in the rDNS to get a host name, Exim was showing the
4956 IP address twice in Received: lines, even though the IP addresses were
4957 identical. For example:
4959 Received: from [1.2.3.4] (helo=[1.2.3.4])
4961 However, if the real host name was known, it was omitting the HELO data
4962 if it matched the actual IP address. This has been tidied up so that it
4963 doesn't show the same IP address twice.
4965 PH/07 When both +timestamp and +memory debugging was on, the value given by
4966 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4967 called by the debug printing, thereby overwriting the timestamp buffer.
4968 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4970 PH/08 When the original message was included in an autoreply transport, it
4971 always said "this is a copy of the message, including all the headers",
4972 even if body_only or headers_only was set. It now gives an appropriate
4975 PH/09 Applied a patch from the Sieve maintainer which:
4977 o fixes some comments
4978 o adds the (disabled) notify extension core
4979 o adds some debug output for the result of if/elsif tests
4980 o points to the current vacation draft in the documentation
4981 and documents the missing references header update
4985 o fixes a bug in processing the envelope test (when testing
4986 multiple envelope elements, the last element determined the
4989 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4990 Electronic Mail") by including:
4992 Auto-submitted: auto-generated
4994 in the messages that it generates (bounce messages and others, such as
4995 warnings). In the case of bounce messages for non-SMTP messages, there was
4996 also a typo: it was using "Auto_submitted" (underscore instead of
4997 hyphen). Since every message generated by Exim is necessarily in response
4998 to another message, thes have all been changed to:
5000 Auto-Submitted: auto-replied
5002 in accordance with these statements in the RFC:
5004 The auto-replied keyword:
5006 - SHOULD be used on messages sent in direct response to another
5007 message by an automatic process,
5009 - MUST NOT be used on manually-generated messages,
5011 - MAY be used on Delivery Status Notifications (DSNs) and Message
5012 Disposition Notifications (MDNs),
5014 - MUST NOT be used on messages generated by automatic or periodic
5015 processes, except for messages which are automatic responses to
5018 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5019 to the default Received: header definition.
5021 PH/12 Added log selector acl_warn_skipped (default on).
5023 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5024 variables because (a) they are in the wrong storage pool and (b) even if
5025 they were copied, it wouldn't work properly because of the caching.
5027 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5028 checking when decoding. Apparently there are clients that generate
5029 overlong encoded strings. Why am I not surprised?
5031 PH/15 If the first argument of "${if match_address" was not empty, but did not
5032 contain an "@" character, Exim crashed. Now it writes a panic log message
5033 and treats the condition as false.
5035 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5037 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5038 extension "envelope-auth". The code is finished and in agreement with
5039 other implementations, but there is no documentation so far and in fact,
5040 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5041 not changing the active code.
5043 Print executed "if" and "elsif" statements when debugging is used. This
5044 helps a great deal to understand what a filter does.
5046 Document more things not specified clearly in RFC3028. I had all this
5047 sorted out, when out of a sudden new issues came to my mind. Oops."
5049 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5052 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5053 canonical form (as documented). However, after a host name lookup from
5054 the IP address, check_host() was doing a simple string comparison with
5055 addresses acquired from the DNS when checking that the found name did
5056 have the original IP as one of its addresses. Since any found IPv6
5057 addresses are likely to be in abbreviated form, the comparison could
5058 fail. Luckily, there already exists a function for doing the comparison
5059 by converting both addresses to binary, so now that is used instead of
5060 the text comparison.
5062 PH/20 There was another similar case to PH/19, when a complete host name was
5063 given in a host list; looking up its IP address could give an abbreviated
5064 form, whereas the current host's name might or might not be abbreviated.
5065 The same fix has been applied.
5071 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5072 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5075 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5076 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5078 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5080 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5081 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5082 Also, for "Auto-Submitted", treat anything other than "no" as
5083 non-personal, in accordance with RFC 3834. (Previously it treated
5084 anything starting "auto-" as non-personal.)
5086 TF/01 The control=submission/name=... option had a problem with syntax
5087 errors if the name included a slash character. The /name= option
5088 now slurps the rest of the string, so it can include any characters
5089 but it must come last in the list of options (after /sender_retain
5092 PH/05 Some modifications to the interface to the fake nameserver for the new
5100 TK/01 Added the "success_on_redirect" address verification option. See
5101 NewStuff for rationale and an example.
5103 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5105 PH/02 Patch to exigrep to allow it to work on syslog lines.
5107 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5108 fread() to skip over the body file's header line, because in Cygwin the
5109 header line is locked and is inaccessible.
5111 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5112 co-exist for some time) to make it clear that it is the Exim ID that is
5113 referenced, not the Message-ID: header line.
5115 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5116 string_format() function, because snprintf() does not exist on all
5119 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5120 ${stat: expansion item.
5122 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5123 protocol synchronization error", to keep the pedants happy.
5125 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5126 well as for IRIX systems, when gcc is being used. See the host.c source
5129 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5131 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5134 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5135 transport and to the smtp transport in LMTP mode.
5137 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5139 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5140 run a filter in a subprocess. This could lead to confusion in subsequent
5141 lookups in the parent process. There should also be a search_tidyup() at
5142 the end of the subprocess.
5144 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5145 only if the host matched helo_try_verify_hosts, which caused the
5146 verification to occur when the EHLO/HELO command was issued. The ACL just
5147 tested the remembered result. Now, if a previous verification attempt has
5148 not happened, "verify = helo" does it there and then.
5150 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5152 TK/03 Fix log output including CR from clamd.
5154 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5155 exist provoked a memory error which could cause a segfault.
5157 PH/15 Installed PCRE 6.2
5159 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5161 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5162 of the problem. Specifically, suggested +O2 rather than +O1 for the
5165 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5167 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5168 record, the port was not correctly logged when the outgoing_port log
5169 selector was set (it logged the transort's default port).
5171 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5172 fallback_hosts, and "hosts" in the smtp transport.
5174 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5175 host errors such as "Connection refused".
5177 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5178 authentication with radiusclient 0.4.9:
5180 - Error returned from rc_read_config was caught wrongly
5181 - Username/password not passed on to radius server due to wrong length.
5183 The presumption is that some radiusclient API changes for 4.51/PH/17
5184 were not taken care of correctly. The code is still untested by me (my
5185 Linux distribution still has 0.3.2 of radiusclient), but it was
5186 contributed by a Radius user.
5188 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5189 expanding the "port" option of the smtp transport.
5191 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5192 while reading a MIME header. Thanks to Tom Hughes for a patch.
5194 PH/24 Include config.h inside local_scan.h so that configuration settings are
5197 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5198 that in an ACL for RCPT (for example), you can examine exactly what was
5201 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5202 commands, but it was not correctly comparing the address with the actual
5203 client host address. Thus, it would show the EHLO address in Received:
5204 header lines when this was not necessary.
5206 PH/27 Added the % operator to ${eval:}.
5208 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5209 it should be ignoring failures (because with -C, for example, it has lost
5210 privilege). It wasn't ignoring creation failures other than "already
5213 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5216 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5217 because an input line was too long, either on its own, or by virtue of
5218 too many continuations, the temporary file was not being removed, and the
5219 return code was incorrect.
5221 PH/31 Missing "BOOL" in function definition in filtertest.c.
5223 PH/32 Applied Sieve patches from the maintainer.
5225 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5227 PH/33 Added "verify = not_blind".
5229 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5230 Local/Makefile (with some defaults set). These are used in built scripts
5231 such as exicyclog, but they have never been used in the exim_install
5232 script (though there are many overriding facilities there). I have
5233 arranged that the exim_install script now takes note of these two
5236 PH/35 Installed configuration files for Dragonfly.
5238 PH/36 When a locally submitted message by a trusted user did not contain a
5239 From: header, and the sender address was obtained from -f or from an SMTP
5240 MAIL command, and the trusted user did not use -F to supply a sender
5241 name, $originator_name was incorrectly used when constructing a From:
5242 header. Furthermore, $originator_name was used for submission mode
5243 messages from external hosts without From: headers in a similar way,
5244 which is clearly wrong.
5246 PH/37 Added control=suppress_local_fixups.
5248 PH/38 When log_selector = +received_sender was set, and the addition of the
5249 sender made the log line's construction buffer exactly full, or one byte
5250 less than full, an overflow happened when the terminating "\n" was
5253 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5254 when the result of a list match is failure because a DNS lookup failed.
5256 PH/40 RM_COMMAND is now used in the building process.
5258 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5259 the "build-* directories that it finds.
5261 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5262 address in a domain literal was a prefix of an interface address.
5264 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5265 when verifying a sender address, unless rewrite_headers is false.
5267 PH/44 Wrote a long comment about why errors_to addresses are verified as
5268 recipients, not senders.
5270 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5271 the ratelimit ACL was added.
5273 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5275 PH/46 Added extra information about PostgreSQL errors to the error string.
5277 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5278 avoiding the need to install special zones in a real server. This is
5279 backwards compatible; if it can't find the fake resolver, it drops back.
5280 Thus, both old and new test suites can be run.
5282 TF/02 Added util/ratelimit.pl
5284 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5285 clock is set back in time.
5287 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5288 Candler <B.Candler@pobox.com>.
5290 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5291 OK for addresses that are the result of redirections.
5293 PH/48 A number of further additions for the benefit of the new test suite,
5294 including a fake gethostbyname() that interfaces to the fake DNS resolver
5297 TF/06 The fix for widen_domains has also been applied to qualify_single and
5298 search_parents which are the other dnslookup options that can cause
5301 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5304 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5306 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5307 service" if it encountered an MX record with an empty target hostname.
5308 The message is now "an MX or SRV record indicated no SMTP service".
5310 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5311 if the DNS of the sending site is misconfigured. This is quite a
5312 common situation. This change restores the behaviour of treating a
5313 helo verification defer as a failure.
5315 PH/51 If self=fail was set on a router, the bounce message did not include the
5316 actual error message.
5322 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5324 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5325 command itself ended in error, the underlying message about the transport
5326 filter timeout was being overwritten with the pipe command error. Now the
5327 underlying error message should be appended to the second error message.
5329 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5331 PH/02 Reduce the amount of output that "make" produces by default. Full output
5332 can still be requested.
5334 PH/03 The warning log line about a condition test deferring for a "warn" verb
5335 was being output only once per connection, rather than after each
5336 occurrence (because it was using the same function as for successful
5337 "warn" verbs). This seems wrong, so I have changed it.
5339 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5340 it should not have, which might have caused a crash in the right
5341 circumstances, but probably never did.
5343 PH/04 Installed a modified version of Tony Finch's patch to make submission
5344 mode fix the return path as well as the Sender: header line, and to
5345 add a /name= option so that you can make the user's friendly name appear
5348 TF/03 Added the control = fakedefer ACL modifier.
5350 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5351 Mark Lowes for thorough testing.
5353 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5355 TK/03 Merged latest SRS patch from Miles Wilton.
5357 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5358 with the definition in sysexits.h (which is #included earlier).
5359 Fortunately, Exim does not actually use EX_OK. The code used to try to
5360 preserve the sysexits.h value, by assuming that macro definitions were
5361 scanned for macro replacements. I have been disabused of this notion,
5362 so now the code just undefines EX_OK before #including unistd.h.
5364 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5365 in the smtp transport. When a block could not be written in a single
5366 write() function, the timeout was being re-applied to each part-write.
5367 This seems wrong - if the receiver was accepting one byte at a time it
5368 would take for ever. The timeout is now adjusted when this happens. It
5369 doesn't have to be particularly precise.
5371 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5372 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5374 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5375 without a domain if the check to <postmaster@domain> fails.
5377 SC/01 Eximstats: added -xls and the ability to specify output files
5378 (patch written by Frank Heydlauf).
5380 SC/02 Eximstats: use FileHandles for outputting results.
5382 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5384 SC/04 Eximstats: fixed display of large numbers with -nvr option
5386 SC/05 Eximstats: fixed merging of reports with empty tables.
5388 SC/06 Eximstats: added the -include_original_destination flag
5390 SC/07 Eximstats: removed tabs and trailing whitespace.
5392 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5394 TK/06 MBOX spool code: Add real "From " MBOX separator line
5395 so the .eml file is really in mbox format (even though
5396 most programs do not really care). Patch from Alex Miller.
5398 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5399 The latter is generated from $received_to and is only set if the
5400 message has one envelope recipient. SA can use these headers,
5401 obviously out-of-the-box. Patch from Alex Miller.
5403 PH/08 The ${def test on a variable was returning false if the variable's
5404 value was "0", contrary to what the specification has always said!
5405 The result should be true unless the variable is empty.
5407 PH/09 The syntax error of a character other than { following "${if
5408 def:variable_name" (after optional whitespace) was not being diagnosed.
5409 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5410 accidental colon was present, for example, could give incorrect results.
5412 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5413 result is used (not including appendfile, where other changes are about
5416 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5417 This involved changing a lot of size variables from int to off_t. It
5418 should work with maildirs and everything.
5420 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5421 spamd dying while we are connected to it.
5423 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5426 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5427 function for BDB 4.3.
5429 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5431 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5432 the sources for PCRE have been re-arranged and more files are now
5435 PH/15 The code I had for printing potentially long long variables in PH/11
5436 above was not the best (it lost precision). The length of off_t variables
5437 is now inspected at build time, and an appropriate printing format (%ld
5438 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5439 to be "long long int" or "long int". This is needed for the internal
5440 formatting function string_vformat().
5442 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5443 the configuration file to be ":syslog", then the script "guesses" where
5444 the logs files are, rather than using the compiled in default. In our
5445 case the guess is not the same as the compiled default, so the script
5446 suddenly stopped working when I started to use syslog. The patch checks
5447 to see if log_file_path is "". If so, it attempts to read it from exim
5448 with no configuration file to get the compiled in version, before it
5449 falls back to the previous guessing code."
5451 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5452 implementing BATV in an Exim configuration. See NewStuff for the gory
5455 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5456 Makefile that are specific to HP-UX.
5458 PH/18 If the "use_postmaster" option was set for a recipient callout together
5459 with the "random" option, the postmaster address was used as the MAIL
5460 FROM address for the random test, but not for the subsequent recipient
5461 test. It is now used for both.
5463 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5464 patch removes a few documentation additions to RFC 3028, because the
5465 latest draft now contains them. It adds the new en;ascii-case comparator
5466 and a new error check for 8bit text in MIME parts. Comparator and
5467 require names are now matched exactly. I enabled the subaddress
5468 extension, but it is not well tested yet (read: it works for me)."
5470 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5471 rework some of the code of TK/09 above to avoid the hardwired use of
5472 "%lld" and "long long". Replaced the call to snprintf() with a call to
5475 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5476 records point to non-existent hosts", "retry timeout exceeded", and
5477 "retry time not reached for any host after a long failure period".
5479 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5480 experimental DomainKeys support:
5482 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5483 (2) On an error such as an illegally used "control", the wrong name for
5484 the control was given.
5486 These problems did NOT occur unless DomainKeys support was compiled.
5488 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5490 PH/24 Added ${if match_ip condition.
5492 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5493 databases so that it will be absolutely obvious if a crash occurs in the
5494 DB library. This is a regular occurrence (often caused by mis-matched
5497 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5498 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5499 picked up on a user's system that detects such things. There doesn't seem
5500 to be a gcc warning option for this - only an attribute that has to be
5501 put on the function's prototype. It seems that in Fedora Core 4 they have
5502 set this on a number of new functions. No doubt there will be more in due
5505 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5506 specify a transport. However, if an address that was verified by such a
5507 router was the subject of a callout, Exim crashed because it tried to
5508 read the rcpt_include_affixes from the non-existent transport. Now it
5509 just assumes that the setting of that option is false. This bug was
5510 introduced by 4.51/PH/31.
5512 PH/28 Changed -d+all to exclude +memory, because that information is very
5513 rarely of interest, but it makes the output a lot bigger. People tend to
5514 do -d+all out of habit.
5516 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5517 code in os-type was giving problems when libc.so lives in lib64, like on
5520 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5521 aren't the modern standard, and it seems that some systems' include files
5522 don't always have them. Exim was already checking for some of the newer
5523 ones like T_AAAA, and defining it itself. I've added checks for all the
5524 record types that Exim uses.
5526 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5527 not automatically generating a new one, as it is supposed to. This
5528 prevented TLS from working. If the file did exist, but contained invalid
5529 data, a new version was generated, as expected. It was only the case of a
5530 non-existent file that was broken.
5532 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5533 with a change in libdomainkeys > 0.64.
5535 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5536 from DNS. If the selector record carries the flag, it now has
5537 precedence over the domain-wide flag.
5539 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5541 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5542 the use of a port name that isn't defined in /etc/services) occurred, the
5543 message was deferred as in a normal delivery, and thus remained on the
5544 spool, instead of being failed because of the mua_wrapper setting. This
5545 is now fixed, and I tidied up some of the mua_wrapper messages at the
5548 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5549 the messages in a hash of arrays rather than using individual hashes.
5550 This is a bit cleaner and results in dramatic memory savings, albeit
5551 at a slight CPU cost.
5553 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5554 as requested by Marc Sherman.
5556 SC/10 Eximstats: added histograms for user specified patterns as requested
5559 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5561 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5562 fopen() in the content-scanning modules that did not already have it.
5568 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5569 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5571 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5573 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5575 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5576 to test Sieve filters that use "vacation".
5578 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5579 that changes the way the GnuTLS parameters are stored in the cache file.
5580 The new format can be generated externally. For backward compatibility,
5581 if the data in the cache doesn't make sense, Exim assumes it has read an
5582 old-format file, and it generates new data and writes a new file. This
5583 means that you can't go back to an older release without removing the
5586 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5587 work if there are any delivery delays because "one_time" forces the
5588 parent to be marked "delivered", so its unseen clone is never tried
5589 again. For this reason, Exim now forbids the simultaneous setting of
5592 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5593 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5594 entry if you want to know the details. The fix, however, neglected to
5595 consider the case when local delivery batching is involved. The test for
5596 "previously delivered" was not happening when checking to see if an
5597 address could be batched with a previous (undelivered) one; under
5598 certain circumstances this could lead to multiple deliveries to the same
5601 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5602 in its include files, and this causes problems building Exim.
5604 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5605 header_syntax) but Exim was just ignoring anything given after a slash.
5606 In particular, this caused confusion with an attempt to use "verify =
5607 reverse_host_lookup/defer_ok". An error is now given when options are
5608 supplied for verify items that do not have them. (Maybe reverse_host_
5609 lookup should have a defer_ok option, but that's a different point.)
5611 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5612 defined by RFC 821) to 2048, because there were problems with some AUTH
5613 commands, and RFC 1869 says the size should be increased for extended
5614 SMTP commands that take arguments.
5616 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5619 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5620 "unknown" error; now it says that the functionality isn't in the binary.
5622 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5623 an address' error message when a string expansion fails (syntax or
5624 whatever). Otherwise the password may appear in the log. Following change
5625 PH/42 below, there is no longer a chance of it appearing in a bounce
5628 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5630 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5631 of its addresses was ever tried. (Bugzilla bug #2.)
5633 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5634 the result incorrectly in the debug output. (It correctly added a newline
5635 to what was transported.)
5637 TF/01 Added $received_time.
5639 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5640 commented out examples of how to interface to a virus scanner and to
5641 SpamAssassin. Also added commented examples of av_scanner and
5642 spamd_address settings.
5644 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5645 and controls are allowed in which ACLs. There were a couple of minor
5646 errors. Some of the entries in the conditions table (which is a table of
5647 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5648 negation of where the condition IS allowed.
5650 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5652 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5653 header file does not have a version number, so I've had to invent a new
5654 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5655 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5656 radiusclient), but it was contributed by a Radius user.
5658 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5659 files or directories, for OpenSSL.
5661 PH/19 When an Exim process that is running as root has to create an Exim log
5662 file, it does so in a subprocess that runs as exim:exim so as to get the
5663 ownership right at creation (otherwise, other Exim processes might see
5664 the file with the wrong ownership). There was no test for failure of this
5665 fork() call, which would lead to the process getting stuck as it waited
5666 for a non-existent subprocess. Forks do occasionally fail when resources
5667 run out. I reviewed all the other calls to fork(); they all seem to check
5670 PH/20 When checking for unexpected SMTP input at connect time (before writing
5671 the banner), Exim was not dealing correctly with a non-positive return
5672 from the read() function. If the client had disconnected by this time,
5673 the result was a log entry for a synchronization error with an empty
5674 string after "input=" when read() returned zero. If read() returned -1
5675 (an event I could not check), uninitialized data bytes were printed.
5676 There were reports of junk text (parts of files, etc) appearing after
5679 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5681 PH/22 Added support for macro redefinition, and (re)definition in between
5682 driver and ACL definitions.
5684 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5685 forgetting to use the resulting value; it was using the unexpanded value.
5687 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5688 hadn't been configured. The fix is from Juergen Kreileder, who
5689 understands it better than I do:
5691 "Here's what I see happening with three configured cyrus_sasl
5692 authenticators configured (plain, login, cram-md5):
5694 On startup auth_cyrus_sasl_init() gets called for each of these.
5695 This means three calls to sasl_listmech() without a specified mech_list.
5696 => SASL tests which mechs of all available mechs actually work
5697 => three warnings about OTP not working
5698 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5700 With the patch, sasl_listmech() also gets called three times. But now
5701 SASL's mech_list option is set to the server_mech specified in the the
5702 authenticator. Or in other words, the answer from sasl_listmech()
5703 gets limited to just the mech you're testing for (which is different
5705 => the return list contains just 'plain' or 'login', 'cram-md5' or
5706 nothing depending on the value of ob->server_mech.
5708 I've just tested the patch: Authentication still works fine,
5709 unavailable mechs specified in the exim configuration are still
5710 caught, and the auth.log warnings about OTP are gone."
5712 PH/25 When debugging is enabled, the contents of the command line are added
5713 to the debugging output, even when log_selector=+arguments is not
5716 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5717 answer is "GNU", and only if the return is "GNU/something" is the answer
5720 PH/27 $acl_verify_message is now set immediately after the failure of a
5721 verification in an ACL, and so is available in subsequent modifiers. In
5722 particular, the message can be preserved by coding like this:
5724 warn !verify = sender
5725 set acl_m0 = $acl_verify_message
5727 Previously, $acl_verify_message was set only while expanding "message"
5728 and "log_message" when a very denied access.
5730 PH/28 Modified OS/os.c-Linux with
5732 -#ifndef OS_LOAD_AVERAGE
5733 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5735 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5736 nomenclature these days.)
5738 PH/29 Installed patch from the Sieve maintainer that adds the options
5739 sieve_useraddress and sieve_subaddress to the redirect router.
5741 PH/30 In these circumstances:
5742 . Two addresses routed to the same list of hosts;
5743 . First host does not offer TLS;
5744 . First host accepts first address;
5745 . First host gives temporary error to second address;
5746 . Second host offers TLS and a TLS session is established;
5747 . Second host accepts second address.
5748 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5749 and peerdn, if requested) that were in fact used only for the second
5752 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5753 attention to any local part prefix or suffix that was matched by the
5754 router that accepted the address. It now behaves in the same way as it
5755 does for delivery: the affixes are removed from the local part unless
5756 rcpt_include_affixes is set on the transport.
5758 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5759 timeout during the DATA phase of an incoming message.
5761 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5762 applied a patch sanctioned by the Sieve maintainer.
5764 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5765 the uid or gid is negative. A case of a negative gid caused this to be
5766 noticed. The fix allows for either to be negative.
5768 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5769 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5772 PH/36 The change PH/12 above was broken. Fixed it.
5774 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5775 the grounds that routing the same address twice would always produce the
5776 same answer. This might have been true once, but it is certainly no
5777 longer true now. Routing a child address may depend on the previous
5778 routing that produced that child. Some complicated redirection strategies
5779 went wrong when messages had multiple recipients, and made Exim's
5780 behaviour dependent on the order in which the addresses were given.
5782 I have moved the duplicate checking until after the routing is complete.
5783 Exim scans the addresses that are assigned to local and remote
5784 transports, and removes any duplicates. This means that more work will be
5785 done, as duplicates will always all be routed, but duplicates are
5786 presumably rare, so I don't expect this is of any significance.
5788 For deliveries to pipes, files, and autoreplies, the duplicate checking
5789 still happens during the routing process, since they are not going to be
5792 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5793 It corrects a timeout issue with spamd. This is Ian's comment: "The
5794 background is that sometimes spamd either never reads data from a
5795 connection it has accepted, or it never writes response data. The exiscan
5796 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5797 blindly assumes that writes won't block so it may never time out."
5799 PH/39 Allow G after quota size as well as K and M.
5801 PH/40 The value set for $authenticated_id in an authenticator may not contain
5802 binary zeroes or newlines because the value is written to log lines and
5803 to spool files. There was no check on this. Now the value is run through
5804 the string_printing() function so that such characters are converted to
5805 printable escape sequences.
5807 PH/41 $message_linecount is a new variable that contains the total number of
5808 lines in the message. Compare $body_linecount, which is the count for the
5811 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5812 bounce and delay warning messages, except in certain special cases, which
5815 (a) An SMTP error message from a remote host;
5816 (b) A message specified in a :fail: redirection;
5817 (c) A message specified in a "fail" command in a system filter;
5818 (d) A message specified in a FAIL return from the queryprogram router;
5819 (e) A message specified by the cannot_route_message router option.
5821 In these cases only, Exim does include the error details in bounce and
5822 warning messages. There are also a few cases where bland messages such
5823 as "unrouteable address" or "local delivery error" are given.
5825 PH/43 $value is now also set for the "else" part of a ${run expansion.
5827 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5828 being worked on, but at least Exim now implements the latest version to
5831 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5832 process to complete was treated as a delivery failure, a timeout while
5833 writing the message to the pipe was logged, but erroneously treated as a
5834 successful delivery. Such timeouts include transport filter timeouts. For
5835 consistency with the overall process timeout, these timeouts are now
5836 treated as errors, giving rise to delivery failures by default. However,
5837 there is now a new Boolean option for the pipe transport called
5838 timeout_defer, which, if set TRUE, converts the failures into defers for
5839 both kinds of timeout. A transport filter timeout is now identified in
5842 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5843 systems where "make" and "gmake" are different, calling "gmake" at top
5844 level broke things. I've arranged for the value of $(MAKE) to be passed
5845 from the Makefile to this script so that it can call the same version of
5849 A note about Exim versions 4.44 and 4.50
5850 ----------------------------------------
5852 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5853 changes of various kinds. As a consequence, a big documentation update was
5854 needed. This delayed the release for rather longer than seemed good, especially
5855 in the light of a couple of (minor) security issues. Therefore, the changes
5856 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5857 release. So 4.44 and 4.50 are in effect two different branches that both start
5860 I have left the 4.50 change log unchanged; it contains all the changes since
5861 4.43. The change log for 4.44 is below; many of its items are identical to
5862 those for 4.50. This seems to be the most sensible way to preserve the
5863 historical information.
5869 1. Minor wording change to the doc/README.SIEVE file.
5871 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5872 computation of the current number of files was incorrect.
5874 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5875 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5878 4. Give more explanation in the error message when the command for a transport
5879 filter fails to execute.
5881 5. There are several places where Exim runs a non-Exim command in a
5882 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5883 was being done only for the command run by the queryprogram router. It is
5884 now done for all such subprocesses. The other cases are: ${run, transport
5885 filters, and the commands run by the lmtp and pipe transports.
5887 6. Added CONFIGURE_GROUP build-time option.
5889 7. Some older OS have a limit of 256 on the maximum number of file
5890 descriptors. Exim was using setrlimit() to set 1000 as a large value
5891 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5892 systems. I've change it so that if it can't get 1000, it tries for 256.
5894 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5895 was an oversight, and furthermore, ever since the addition of extra
5896 controls (e.g. 4.43/32), the checks on when to allow different forms of
5897 "control" were broken. There should now be diagnostics for all cases when a
5898 control that does not make sense is encountered.
5900 9. Added the /retain_sender option to "control=submission".
5902 10. $recipients is now available in the predata ACL (oversight).
5904 11. Tidy the search cache before the fork to do a delivery from a message
5905 received from the command line. Otherwise the child will trigger a lookup
5906 failure and thereby defer the delivery if it tries to use (for example) a
5907 cached ldap connection that the parent has called unbind on.
5909 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5910 of $address_data from the recipient verification was clobbered by the
5911 sender verification.
5913 13. The value of address_data from a sender verification is now available in
5914 $sender_address_data in subsequent conditions in the ACL statement.
5916 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5918 15. Added a new option "connect=<time>" to callout options, to set a different
5921 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5922 was its contents. (It was OK if the option was not defined at all.)
5924 17. A "Completed" log line is now written for messages that are removed from
5925 the spool by the -Mrm option.
5927 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5928 information about exactly what failed.
5930 19. Added -dd to debug only the daemon process.
5932 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5933 handles timeouts, both on the server side and network timeouts. Renamed the
5934 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5936 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5937 It is now set to "smtps".
5939 22. $host_address is now set to the target address during the checking of
5940 ignore_target_hosts.
5942 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5943 being passed; this would have caused $sender_host_name to have been used if
5944 matching the list had actually called for a host name (not very likely,
5945 since this list is usually IP addresses). A host name is now passed as
5948 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5949 code that specifies a non-restarting handler (typically sigaction() in
5950 modern systems) in an attempt to fix a rare and obscure crash bug.
5952 25. Narrowed the window for a race in the daemon that could cause it to ignore
5953 SIGCHLD signals. This is not a major problem, because they are used only to
5954 wake it up if nothing else does.
5956 26. A malformed maildirsize file could cause Exim to calculate negative values
5957 for the mailbox size or file count. Odd effects could occur as a result.
5958 The maildirsize information is now recalculated if the size or filecount
5961 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5962 support for a long time. Removed HAVE_SYS_VFS_H.
5964 28. Installed the latest version of exipick from John Jetmore.
5966 29. In an address list, if the pattern was not a regular expression, an empty
5967 subject address (from a bounce message) matched only if the pattern was an
5968 empty string. Non-empty patterns were not even tested. This was the wrong
5969 because it is perfectly reasonable to use an empty address as part of a
5970 database query. An empty address is now tested by patterns that are
5971 lookups. However, all the other forms of pattern expect the subject to
5972 contain a local part and a domain, and therefore, for them, an empty
5973 address still always fails if the pattern is not itself empty.
5975 30. Exim went into a mad DNS loop when attempting to do a callout where the
5976 host was specified on an smtp transport, and looking it up yielded more
5977 than one IP address.
5979 31. Re-factored the code for checking spool and log partition space into a
5980 function that finds that data and another that does the check. The former
5981 is then used to implement four new variables: $spool_space, $log_space,
5982 $spool_inodes, and $log_inodes.
5984 32. The RFC2047 encoding function was originally intended for short strings
5985 such as real names; it was not keeping to the 75-character limit for
5986 encoded words that the RFC imposes. It now respects the limit, and
5987 generates multiple encoded words if necessary. To be on the safe side, I
5988 have increased the buffer size for the ${rfc2047: expansion operator from
5991 33. It is now permitted to omit both strings after an "if" condition; if the
5992 condition is true, the result is "true". As before, when the second string
5993 is omitted, a false condition yields an empty string. This makes it less
5994 cumbersome to write custom ACL and router conditions.
5996 34. Failure to deliver a bounce message always caused it to be frozen, even if
5997 there was an errors_to setting on the router. The errors_to setting is now
6000 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6001 canonical form (fully expanded) before being placed in
6002 $sender_host_address.
6004 36. The table in the code that translates DNS record types into text (T_A to
6005 "A" for instance) was missing entries for NS and CNAME. It is just possible
6006 that this could have caused confusion if both these types were looked up
6007 for the same domain, because the text type is used as part of Exim's
6008 per-process caching. But the chance of anyone hitting this buglet seems
6011 37. The dnsdb lookup has been extended in a number of ways.
6013 (1) There is a new type, "zns", which walks up the domain tree until it
6014 finds some nameserver records. It should be used with care.
6016 (2) There is a new type, "mxh", which is like "mx" except that it returns
6017 just the host names, not the priorities.
6019 (3) It is now possible to give a list of domains (or IP addresses) to be
6020 looked up. The behaviour when one of the lookups defers can be
6021 controlled by a keyword.
6023 (4) It is now possible to specify the separator character for use when
6024 multiple records are returned.
6026 38. The dnslists ACL condition has been extended: it is now possible to supply
6027 a list of IP addresses and/or domains to be looked up in a particular DNS
6030 39. Added log_selector=+queue_time_overall.
6032 40. When running the queue in the test harness, wait just a tad after forking a
6033 delivery process, to get repeatability of debugging output.
6035 41. Include certificate and key file names in error message when GnuTLS fails
6036 to set them up, because the GnuTLS error message doesn't include the name
6037 of the failing file when there is a problem reading it.
6039 42. Allow both -bf and -bF in the same test run.
6041 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6043 44. The "Exiscan patch" is now merged into the mainline Exim source.
6045 45. Sometimes the final signoff response after QUIT could fail to get
6046 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6047 < 0 before doing a fflush(). This bug looks as though it goes back to the
6048 introduction of TLS in release 3.20, but "sometimes" must have been rare
6049 because the tests only now provoked it.
6051 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6052 (this can affect the format of dates).
6054 47. exim_tidydb, when checking for the continued existence of a message for
6055 which it has found a message-specific retry record, was not finding
6056 messages that were in split spool directories. Consequently, it was
6057 deleting retry records that should have stayed in existence.
6059 48. Steve fixed some bugs in eximstats.
6061 49. The SPA authentication driver was not abandoning authentication and moving
6062 on to the next authenticator when an expansion was forced to fail,
6063 contradicting the general specification for all authenticators. Instead it
6064 was generating a temporary error. It now behaves as specified.
6066 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6067 (the order specifies the preference for clients). The order is now AES256,
6068 AES128, 3DES, ARCFOUR128.
6070 51. Small patch to Sieve code - explicitly set From: when generating an
6073 52. Exim crashed if a remote delivery caused a very long error message to be
6074 recorded - for instance if somebody sent an entire SpamAssassin report back
6075 as a large number of 550 error lines. This bug was coincidentally fixed by
6076 increasing the size of one of Exim's internal buffers (big_buffer) that
6077 happened as part of the Exiscan merge. However, to be on the safe side, I
6078 have made the code more robust (and fixed the comments that describe what
6081 53. Now that there can be additional text after "Completed" in log lines (if
6082 the queue_time_overall log selector is set), a one-byte patch to exigrep
6083 was needed to allow it to recognize "Completed" as not the last thing in
6086 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6087 patch that reportedly fixes this has been added. I am not expert enough to
6088 create a test for it. This is what the patch creator wrote:
6090 "I found a little strange behaviour of ldap code when working with
6091 Windows 2003 AD Domain, where users was placed in more than one
6092 Organization Units. When I tried to give exim partial DN, the exit code
6093 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6094 But simultaneously result of request was absolutely normal ldap result,
6095 so I produce this patch..."
6097 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6098 REFERENCE, so I have modified the code to exclude the patch when that macro
6101 55. Some experimental protocols are using DNS PTR records for new purposes. The
6102 keys for these records are domain names, not reversed IP addresses. The
6103 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6104 leaves it alone. Component reversal etc. now happens only for IP addresses.
6107 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6109 57. Double the size of the debug message buffer (to 2048) so that more of very
6110 long debug lines gets shown.
6112 58. The exicyclog utility now does better if the number of log files to keep
6113 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6115 59. Two changes related to the smtp_active_hostname option:
6117 (1) $smtp_active_hostname is now available as a variable.
6118 (2) The default for smtp_banner uses $smtp_active_hostname instead
6119 of $primary_hostname.
6121 60. The host_aton() function is supposed to be passed a string that is known
6122 to be a valid IP address. However, in the case of IPv6 addresses, it was
6123 not checking this. This is a hostage to fortune. Exim now panics and dies
6124 if the condition is not met. A case was found where this could be provoked
6125 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6126 components; fortuitously, this particular loophole had already been fixed
6127 by change 4.50/55 above.
6129 If there are any other similar loopholes, the new check in host_aton()
6130 itself should stop them being exploited. The report I received stated that
6131 data on the command line could provoke the exploit when Exim was running as
6132 exim, but did not say which command line option was involved. All I could
6133 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6134 running as the user.
6137 61. There was a buffer overflow vulnerability in the SPA authentication code
6138 (which came originally from the Samba project). I have added a test to the
6139 spa_base64_to_bits() function which I hope fixes it.
6142 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6143 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6145 63. The daemon start-up calls getloadavg() while still root for those OS that
6146 need the first call to be done as root, but it missed one case: when
6147 deliver_queue_load_max is set with deliver_drop_privilege. This is
6148 necessary for the benefit of the queue runner, because there is no re-exec
6149 when deliver_drop_privilege is set.
6151 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6152 This has been fixed.
6154 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6155 was in use, was not putting the data itself into the right store pool;
6156 consequently, it could be overwritten for a subsequent message in the same
6157 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6160 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6162 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6163 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6164 the function were treating the return as a boolean value, which happened to
6165 work because 0=false and not-0=true, but is not correct code.
6167 68. The host_aton() function was not handling scoped IPv6 addresses (those
6168 with, for example, "%eth0" on the end) correctly.
6170 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6171 negated items (that is, ~something) in unsigned ints. Some compilers
6172 apparently mutter when there is no cast.
6174 70. If an address verification called from an ACL failed, and did not produce a
6175 user-specific message (i.e. there was only a "system" message), nothing was
6176 put in $acl_verify_message. In this situation, it now puts the system
6179 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6180 session; change 4.31/43 added the unwanted input to the log line - except
6181 that it did not do this in the start of session case. It now does.
6183 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6184 This is wrong and can cause the other end to generate a synchronization
6185 error if it is another Exim or anything else that does the synchronization
6186 check. A QUIT command is no longer sent after a timeout.
6188 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6189 during host lookups.
6191 74. The defer_ok option of callout verification was not working if it was used
6192 when verifying addresses in header lines, that is, for this case:
6194 verify = header_sender/callout=defer_ok
6196 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6197 those file descriptors could be used for SMTP connections. If anything
6198 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6199 could be sent to the SMTP client, causing chaos. The daemon now opens
6200 stdin, stdout, and stderr to /dev/null when it puts itself into the
6203 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6204 log by default. The user can override this with suitable Perl magic.
6206 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6207 the log message when discard triggers, was not working for the DATA ACL or
6208 for the non-SMTP ACL.
6210 78. Error message wording change in sieve.c.
6212 79. If smtp_accept_max_per_host was set, the number of connections could be
6213 restricted to fewer than expected, because the daemon was trying to set up
6214 a new connection before checking whether the processes handling previous
6215 connections had finished. The check for completed processes is now done
6216 earlier. On busy systems, this bug wouldn't be noticed because something
6217 else would have woken the daemon, and it would have reaped the completed
6220 80. If a message was submitted locally by a user whose login name contained one
6221 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6222 It caused a spool format error. I have fixed the spool reading code. A
6223 related problem was that the "from" clause in the Received: line became
6224 illegal because of the space(s). It is now covered by ${quote_local_part.
6226 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6229 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6231 83. Patch from Sieve maintainer to fix unterminated string problem in
6232 "vacation" handling.
6234 84. Some minor changes to the Linux configuration files to help with other
6235 OS variants using glibc.
6237 85. One more patch for Sieve to update vacation handling to latest spec.
6240 ----------------------------------------------------
6241 See the note above about the 4.44 and 4.50 releases.
6242 ----------------------------------------------------
6248 1. Change 4.43/35 introduced a bug that caused file counts to be
6249 incorrectly computed when quota_filecount was set in an appendfile
6252 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6253 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6256 3. Give more explanation in the error message when the command for a transport
6257 filter fails to execute.
6259 4. There are several places where Exim runs a non-Exim command in a
6260 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6261 was being done only for the command run by the queryprogram router. It is
6262 now done for all such subprocesses. The other cases are: ${run, transport
6263 filters, and the commands run by the lmtp and pipe transports.
6265 5. Some older OS have a limit of 256 on the maximum number of file
6266 descriptors. Exim was using setrlimit() to set 1000 as a large value
6267 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6268 systems. I've change it so that if it can't get 1000, it tries for 256.
6270 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6271 was an oversight, and furthermore, ever since the addition of extra
6272 controls (e.g. 4.43/32), the checks on when to allow different forms of
6273 "control" were broken. There should now be diagnostics for all cases when a
6274 control that does not make sense is encountered.
6276 7. $recipients is now available in the predata ACL (oversight).
6278 8. Tidy the search cache before the fork to do a delivery from a message
6279 received from the command line. Otherwise the child will trigger a lookup
6280 failure and thereby defer the delivery if it tries to use (for example) a
6281 cached ldap connection that the parent has called unbind on.
6283 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6284 of $address_data from the recipient verification was clobbered by the
6285 sender verification.
6287 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6288 was its contents. (It was OK if the option was not defined at all.)
6290 11. A "Completed" log line is now written for messages that are removed from
6291 the spool by the -Mrm option.
6293 12. $host_address is now set to the target address during the checking of
6294 ignore_target_hosts.
6296 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6297 being passed; this would have caused $sender_host_name to have been used if
6298 matching the list had actually called for a host name (not very likely,
6299 since this list is usually IP addresses). A host name is now passed as
6302 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6303 code that specifies a non-restarting handler (typically sigaction() in
6304 modern systems) in an attempt to fix a rare and obscure crash bug.
6306 15. Narrowed the window for a race in the daemon that could cause it to ignore
6307 SIGCHLD signals. This is not a major problem, because they are used only to
6308 wake it up if nothing else does.
6310 16. A malformed maildirsize file could cause Exim to calculate negative values
6311 for the mailbox size or file count. Odd effects could occur as a result.
6312 The maildirsize information is now recalculated if the size or filecount
6315 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6316 support for a long time. Removed HAVE_SYS_VFS_H.
6318 18. Updated exipick to current release from John Jetmore.
6320 19. Allow an empty sender to be matched against a lookup in an address list.
6321 Previously the only cases considered were a regular expression, or an
6324 20. Exim went into a mad DNS lookup loop when doing a callout where the
6325 host was specified on the transport, if the DNS lookup yielded more than
6328 21. The RFC2047 encoding function was originally intended for short strings
6329 such as real names; it was not keeping to the 75-character limit for
6330 encoded words that the RFC imposes. It now respects the limit, and
6331 generates multiple encoded words if necessary. To be on the safe side, I
6332 have increased the buffer size for the ${rfc2047: expansion operator from
6335 22. Failure to deliver a bounce message always caused it to be frozen, even if
6336 there was an errors_to setting on the router. The errors_to setting is now
6339 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6340 canonical form (fully expanded) before being placed in
6341 $sender_host_address.
6343 24. Updated eximstats to version 1.33
6345 25. Include certificate and key file names in error message when GnuTLS fails
6346 to set them up, because the GnuTLS error message doesn't include the name
6347 of the failing file when there is a problem reading it.
6349 26. Expand error message when OpenSSL has problems setting up cert/key files.
6352 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6353 (this can affect the format of dates).
6355 28. exim_tidydb, when checking for the continued existence of a message for
6356 which it has found a message-specific retry record, was not finding
6357 messages that were in split spool directories. Consequently, it was
6358 deleting retry records that should have stayed in existence.
6360 29. eximstats updated to version 1.35
6361 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6362 1.35 - bugfix such that pie charts by volume are generated correctly
6364 30. The SPA authentication driver was not abandoning authentication and moving
6365 on to the next authenticator when an expansion was forced to fail,
6366 contradicting the general specification for all authenticators. Instead it
6367 was generating a temporary error. It now behaves as specified.
6369 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6370 (the order specifies the preference for clients). The order is now AES256,
6371 AES128, 3DES, ARCFOUR128.
6373 31. Small patch to Sieve code - explicitly set From: when generating an
6376 32. Exim crashed if a remote delivery caused a very long error message to be
6377 recorded - for instance if somebody sent an entire SpamAssassin report back
6378 as a large number of 550 error lines. This bug was coincidentally fixed by
6379 increasing the size of one of Exim's internal buffers (big_buffer) that
6380 happened as part of the Exiscan merge. However, to be on the safe side, I
6381 have made the code more robust (and fixed the comments that describe what
6384 33. Some experimental protocols are using DNS PTR records for new purposes. The
6385 keys for these records are domain names, not reversed IP addresses. The
6386 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6387 leaves it alone. Component reversal etc. now happens only for IP addresses.
6390 34. The host_aton() function is supposed to be passed a string that is known
6391 to be a valid IP address. However, in the case of IPv6 addresses, it was
6392 not checking this. This is a hostage to fortune. Exim now panics and dies
6393 if the condition is not met. A case was found where this could be provoked
6394 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6395 components; fortuitously, this particular loophole had already been fixed
6396 by change 4.50/55 or 4.44/33 above.
6398 If there are any other similar loopholes, the new check in host_aton()
6399 itself should stop them being exploited. The report I received stated that
6400 data on the command line could provoke the exploit when Exim was running as
6401 exim, but did not say which command line option was involved. All I could
6402 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6403 running as the user.
6406 35. There was a buffer overflow vulnerability in the SPA authentication code
6407 (which came originally from the Samba project). I have added a test to the
6408 spa_base64_to_bits() function which I hope fixes it.
6411 36. The daemon start-up calls getloadavg() while still root for those OS that
6412 need the first call to be done as root, but it missed one case: when
6413 deliver_queue_load_max is set with deliver_drop_privilege. This is
6414 necessary for the benefit of the queue runner, because there is no re-exec
6415 when deliver_drop_privilege is set.
6417 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6418 was in use, was not putting the data itself into the right store pool;
6419 consequently, it could be overwritten for a subsequent message in the same
6420 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6423 38. Sometimes the final signoff response after QUIT could fail to get
6424 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6425 < 0 before doing a fflush(). This bug looks as though it goes back to the
6426 introduction of TLS in release 3.20, but "sometimes" must have been rare
6427 because the tests only now provoked it.
6433 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6434 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6435 Now it returns an integer. A number of calls were still expecting a T/F
6436 return. Fortuitously, in all cases, the tests worked in OK situations,
6437 which is the norm. However, things would have gone wrong on any write
6438 failures on the smtp file descriptor. This function is used when sending
6439 messages over SMTP and also when doing verify callouts.
6441 2. When Exim is called to do synchronous delivery of a locally submitted
6442 message (the -odf or -odi options), it no longer closes stderr before doing
6445 3. Implemented the mua_wrapper option.
6447 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6449 5. Implemented the functions header_remove(), header_testname(),
6450 header_add_at_position(), and receive_remove_recipient(), and exported them
6453 6. If an ACL "warn" statement specified the addition of headers, Exim already
6454 inserted X-ACL-Warn: at the start if there was no header name. However, it
6455 was not making this test for the second and subsequent header lines if
6456 there were newlines in the string. This meant that an invalid header could
6457 be inserted if Exim was badly configured.
6459 7. Allow an ACL "warn" statement to add header lines at the start or after all
6460 the Received: headers, as well as at the end.
6462 8. Added the rcpt_4xx retry error code.
6464 9. Added postmaster_mailfrom=xxx to callout verification option.
6466 10. Added mailfrom=xxxx to the callout verification option, for verify=
6469 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6470 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6472 12. Inserted some casts to stop certain compilers warning when using pointer
6473 differences as field lengths or precisions in printf-type calls (mostly
6474 affecting debugging statements).
6476 13. Added optional readline() support for -be (dynamically loaded).
6478 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6479 same clock tick as a message's arrival, so that its received time was the
6480 same as the "first fail" time on the retry record, and that message
6481 remained on the queue past the ultimate address timeout, every queue runner
6482 would try a delivery (because it was past the ultimate address timeout) but
6483 after another failure, the ultimate address timeout, which should have then
6484 bounced the address, did not kick in. This was a "< instead of <=" error;
6485 in most cases the first failure would have been in the next clock tick
6486 after the received time, and all would be well.
6488 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6489 being recognized when the domain list was tested by the match_domain
6490 condition in an expansion string.
6492 16. Added the ${str2b64: operator.
6494 17. Exim was always calling setrlimit() to set a large limit for the number of
6495 processes, without checking whether the existing limit was already
6496 adequate. (It did check for the limit on file descriptors.) Furthermore,
6497 errors from getrlimit() and setrlimit() were being ignored. Now they are
6498 logged to the main and panic logs, but Exim does carry on, to try to do its
6499 job under whatever limits there are.
6501 18. Imported PCRE 5.0.
6503 19. Trivial typo in log message " temporarily refused connection" (the leading
6506 20. If the log selector return_path_on_delivery was set and an address was
6507 redirected to /dev/null, the delivery process crashed because it assumed
6508 that a return path would always be set for a "successful" delivery. In this
6509 case, the whole delivery is bypassed as an optimization, and therefore no
6512 21. Internal re-arrangement: the function for sending a challenge and reading
6513 a response while authentication was assuming a zero-terminated challenge
6514 string. It's now changed to take a pointer and a length, to allow for
6515 binary data in such strings.
6517 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6519 23. Exim was not respecting finduser_retries when seeking the login of the
6520 uid under which it was called; it was always trying 10 times. (The default
6521 setting of finduser_retries is zero.) Also, it was sleeping after the final
6522 failure, which is pointless.
6524 24. Implemented tls_on_connect_ports.
6526 25. Implemented acl_smtp_predata.
6528 26. If the domain in control=submission is set empty, Exim assumes that the
6529 authenticated id is a complete email address when it generates From: or
6530 Sender: header lines.
6532 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6533 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6534 chown and chgrp in /bin and hostname in /usr/bin.
6536 28. Exim was keeping the "process log" file open after each use, just as it
6537 does for the main log. This opens the possibility of it remaining open for
6538 long periods when the USR1 signal hits a daemon. Occasional processlog
6539 errors were reported, that could have been caused by this. Anyway, it seems
6540 much more sensible not to leave this file open at all, so that is what now
6543 29. The long-running daemon process does not normally write to the log once it
6544 has entered its main loop, and it closes the log before doing so. This is
6545 so that log files can straightforwardly be renamed and moved. However,
6546 there are a couple of unusual error situations where the daemon does write
6547 log entries, and I had neglected to close the log afterwards.
6549 30. The text of an SMTP error response that was received during a remote
6550 delivery was being truncated at 512 bytes. This is too short for some of
6551 the long messages that one sometimes sees. I've increased the limit to
6554 31. It is now possible to make retry rules that apply only when a message has a
6555 specific sender, in particular, an empty sender.
6557 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6558 it possible to be selective about when SMTP synchronization is enforced.
6560 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6562 32. Implemented hosts_connection_nolog.
6564 33. Added an ACL for QUIT.
6566 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6569 35. Added mailbox_size and mailbox_filecount to appendfile.
6571 36. Added control = no_multiline_responses to ACLs.
6573 37. There was a bug in the logic of the code that waits for the clock to tick
6574 in the case where the clock went backwards by a substantial amount such
6575 that the microsecond fraction of "now" was more than the microsecond
6576 fraction of "then" (but the whole seconds number was less).
6578 38. Added support for the libradius Radius client library this is found on
6579 FreeBSD (previously only the radiusclient library was supported).
6585 1. When certain lookups returned multiple values in the form name=value, the
6586 quoting of the values was not always being done properly. Specifically:
6587 (a) If the value started with a double quote, but contained no whitespace,
6589 (b) If the value contained whitespace other than a space character (i.e.
6590 tabs or newlines or carriage returns) it was not quoted.
6591 This fix has been applied to the mysql and pgsql lookups by writing a
6592 separate quoting function and calling it from the lookup code. The fix
6593 should probably also be applied to nisplus, ibase and oracle lookups, but
6594 since I cannot test any of those, I have not disturbed their existing code.
6596 2. A hit in the callout cache for a specific address caused a log line with no
6597 reason for rejecting RCPT. Now it says "Previous (cached) callout
6598 verification failure".
6600 3. There was an off-by-one bug in the queryprogram router. An over-long
6601 return line was truncated at 256 instead of 255 characters, thereby
6602 overflowing its buffer with the terminating zero. As well as fixing this, I
6603 have increased the buffer size to 1024 (and made a note to document this).
6605 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6606 when Exim is waiting for an SMTP response from a remote server, Exim
6607 restarts its select() call on the socket, thereby resetting its timeout.
6608 This is not a problem when such interrupts are rare. Somebody set up a cron
6609 job to run exiwhat every 2 minutes, which is less than the normal select()
6610 timeout (5 or 10 minutes). This meant that the select() timeout never
6611 kicked in because it was always reset. I have fixed this by comparing the
6612 time when an interrupt arrives with the time at the start of the first call
6613 to select(). If more time than the timeout has elapsed, the interrupt is
6614 treated as a timeout.
6616 5. Some internal re-factoring in preparation for the addition of Sieve
6617 extensions (by MH). In particular, the "personal" test is moved to a
6618 separate function, and given an option for scanning Cc: and Bcc: (which is
6619 not set for Exim filters).
6621 6. When Exim created an email address using the login of the caller as the
6622 local part (e.g. when creating a From: or Sender: header line), it was not
6623 quoting the local part when it contained special characters such as @.
6625 7. Installed new OpenBSD configuration files.
6627 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6628 try to make them clearer.
6630 9. Callout options, other than the timeout value, were being ignored when
6631 verifying sender addresses in header lines. For example, when using
6633 verify = header_sender/callout=no_cache
6635 the cache was (incorrectly) being used.
6637 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6638 only the Cygwin environment.
6640 11. When return_path_on_delivery was set as a log selector, if different remote
6641 addresses in the same message used different return paths and parallel
6642 remote delivery occurred, the wrong values would sometimes be logged.
6643 (Whenever a remote delivery process finished, the return path value from
6644 the most recently started remote delivery process was logged.)
6646 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6647 lines when AUTH and/or TLS are in use. This is the "received protocol"
6648 field. Exim used to use "asmtp" for authenticated SMTP, without any
6649 indication (in the protocol name) for TLS use. Now it follows the RFC and
6650 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6651 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6652 names appear in log lines as well as in Received: header lines.
6654 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6655 extensions, and comparison tests, and to fix some bugs.
6657 14. Changes to the "personal" filter test:
6659 (1) The test was buggy in that it was just doing the equivalent of
6660 "contains" tests on header lines. For example, if a user's address was
6661 anne@some.where, the "personal" test would incorrectly be true for
6663 To: susanne@some.where
6665 This test is now done by extracting each address from the header in turn,
6666 and checking the entire address. Other tests that are part of "personal"
6667 are now done using regular expressions (for example, to check local parts
6668 of addresses in From: header lines).
6670 (2) The list of non-personal local parts in From: addresses has been
6671 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6672 taken from the Sieve specification recommendations.
6674 (3) If the message contains any header line starting with "List-" it is
6675 treated as non-personal.
6677 (4) The test for "circular" in the Subject: header line has been removed
6678 because it now seems ill-conceived.
6680 15. Minor typos in src/EDITME comments corrected.
6682 16. Installed latest exipick from John Jetmore.
6684 17. If headers_add on a router specified a text string that was too long for
6685 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6686 of string_sprintf() is now avoided.
6688 18. $message_body_size was not set (it was always zero) when running the DATA
6689 ACL and the local_scan() function.
6691 19. For the "mail" command in an Exim filter, no default was being set for
6692 the once_repeat time, causing a random time value to be used if "once" was
6693 specified. (If the value happened to be <= 0, no repeat happened.) The
6694 default is now 0s, meaning "never repeat". The "vacation" command was OK
6695 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6696 (I found it when inspecting the code).
6698 20. There is now an overall timeout for performing a callout verification. It
6699 defaults to 4 times the callout timeout, which applies to individual SMTP
6700 commands during the callout. The overall timeout applies when there is more
6701 than one host that can be tried. The timeout is checked before trying the
6702 next host. This prevents very long delays if there are a large number of
6703 hosts and all are timing out (e.g. when the network connections are timing
6704 out). The value of the overall timeout can be changed by specifying an
6705 additional sub-option for "callout", called "maxwait". For example:
6707 verify = sender/callout=5s,maxwait=20s
6709 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6710 to the end before writing, but this should make it even safer).
6712 22. Exim was forgetting that it had advertised PIPELINING for the second and
6713 subsequent messages on an SMTP connection. It was also not resetting its
6714 memory on STARTTLS and an internal HELO.
6716 23. When Exim logs an SMTP synchronization error within a session, it now
6717 records whether PIPELINING has been advertised or not.
6719 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6720 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6721 rather than long int.
6723 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6725 26. Added the never_mail option to autoreply.
6731 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6732 crash if the getsockname() call failed; this can happen if a connection is
6733 closed very soon after it is established. The problem was simply in the
6734 order in which certain operations were done, causing Exim to try to write
6735 to the SMTP stream before it had set up the file descriptor. The bug has
6736 been fixed by making things happen in the correct order.
6742 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6743 before the connection was closed, thus losing the rejection response.
6745 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6746 some early Solaris releases, but causes trouble in current releases where
6747 socklen_t is defined.
6749 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6752 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6755 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6756 that releasing the top of it at the end releases what was used for sub-
6757 expansions (unless the block got too big). However, discard this block if
6758 the first thing is a variable or header, so that we can use its block when
6759 it is dynamic (useful for very large $message_headers, for example).
6761 6. Lookups now cache *every* query, not just the most recent. A new, separate
6762 store pool is used for this. It can be recovered when all lookup caches are
6763 flushed. Lookups now release memory at the end of their result strings.
6764 This has involved some general refactoring of the lookup sources.
6766 7. Some code has been added to the store_xxx() functions to reduce the amount
6767 of flapping under certain conditions.
6769 8. log_incoming_interface used to affect only the <= reception log lines. Now
6770 it causes the local interface and port to be added to several more SMTP log
6771 lines, for example "SMTP connection from", and rejection lines.
6773 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6775 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6777 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6778 could be overwritten at the end of the current message (or the start of a
6779 new message if it was set in a HELO ACL). The value is now preserved for
6780 the duration of the SMTP connection.
6782 12. If a transport had a headers_rewrite setting, and a matching header line
6783 contained an unqualified address, that address was qualified, even if it
6784 did not match any rewriting rules. The underlying bug was that the values
6785 of the flags that permit the existence of unqualified sender and recipient
6786 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6787 non-local messages, and by -bnq for local messages) were not being
6788 preserved with the message after it was received.
6790 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6791 "next input=" as part of the text comprising the host identity instead of
6792 the correct text. The code was using the same buffer for two different
6793 strings. However, depending on which order the printing function evaluated
6794 its arguments, the bug did not always show up. Under Linux, for example, my
6795 test suite worked just fine.
6797 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6798 doesn't work with some older versions of Perl. It has been changed to "my",
6799 which in any case is probably the better facility to use.
6801 15. A really picky compiler found some instances of statements for creating
6802 error messages that either had too many or two few arguments for the format
6805 16. The size of the buffer for calls to the DNS resolver has been increased
6806 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6807 for addresses that have a lot of PTR records. This alleviates a problem; it
6808 does not fully solve it.
6810 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6811 buffer now truncates the list and logs the incident, which is the same
6812 action as happens when Exim is looking up a host name and its aliases.
6813 Previously in this situation something unpredictable would happen;
6814 sometimes it was "internal error: store_reset failed".
6816 18. If a server dropped the connection unexpectedly when an Exim client was
6817 using GnuTLS and trying to read a response, the client delivery process
6818 crashed while trying to generate an error log message.
6820 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6821 string, for example:
6823 warn message = H1: something\nH2: something
6825 the text was added as a single header line from Exim's point of view
6826 though it ended up OK in the delivered message. However, searching for the
6827 second and subsequent header lines using $h_h2: did not work. This has been
6828 fixed. Similarly, if a system filter added multiple headers in this way,
6829 the routers could not see them.
6831 20. Expanded the error message when iplsearch is called with an invalid key to
6832 suggest using net-iplsearch in a host list.
6834 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6835 ACLs are no longer actually imposed (and a message to that effect is
6838 22. If a "gecos" field in a passwd entry contained escaped characters, in
6839 particular, if it contained a \" sequence, Exim got it wrong when building
6840 a From: or a Sender: header from that name. A second bug also caused
6841 incorrect handling when an unquoted " was present following a character
6842 that needed quoting.
6844 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6845 was not being matched caselessly.
6847 24. Arranged for all hyphens in the exim.8 source to be escaped with
6850 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6851 properly. Recipient callout cache records were still being keyed to include
6852 the sender, even when use_sender was set false. This led to far more
6853 callouts that were necessary. The sender is no longer included in the key
6854 when use_sender is false.
6856 26. Added "control = submission" modifier to ACLs.
6858 27. Added the ${base62d: operator to decode base 62 numbers.
6860 28. dnsdb lookups can now access SRV records.
6862 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6863 the configuration file.
6865 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6866 (-v) mode. This makes the output for a verbose queue run more intelligible.
6868 31. Added a use_postmaster feature to recipient callouts.
6870 32. Added the $body_zerocount variable, containing the number of binary zero
6871 bytes in the message body.
6873 33. The time of last modification of the "new" subdirectory is now used as the
6874 "mailbox time last read" when there is a quota error for a maildir
6877 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6879 35. Added +ignore_unknown as a special item in host lists.
6881 36. Code for decoding IPv6 addresses in host lists is now included, even if
6882 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6883 address was recognized as an IP address, but was then not correctly decoded
6884 into binary, causing unexpected and incorrect effects when compared with
6891 1. Very minor rewording of debugging text in manualroute to say "list of
6892 hosts" instead of "hostlist".
6894 2. If verify=header_syntax was set, and a header line with an unqualified
6895 address (no domain) and a large number of spaces between the end of the
6896 name and the colon was received, the reception process suffered a buffer
6897 overflow, and (when I tested it) crashed. This was caused by some obsolete
6898 code that should have been removed. The fix is to remove it!
6900 3. When running in the test harness, delay a bit after writing a bounce
6901 message to get a bit more predictability in the log output.
6903 4. Added a call to search_tidyup() just before forking a reception process. In
6904 theory, someone could use a lookup in the expansion of smtp_accept_max_
6905 per_host which, without the tidyup, could leave open a database connection.
6907 5. Added the variables $recipient_data and $sender_data which get set from a
6908 lookup success in an ACL "recipients" or "senders" condition, or a router
6909 "senders" option, similar to $domain_data and $local_part_data.
6911 6. Moved the writing of debug_print from before to after the "senders" test
6914 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6915 problems for message scanning, either using a data ACL, or using
6916 local_scan() because the Received: header was not generated till after they
6917 were called (in order to set the time as the time of reception completion).
6918 I have revised the way this works. The header is now generated after the
6919 body is received, but before the ACL or local_scan() are called. After they
6920 are run, the timestamp in the header is updated.
6926 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6927 before starting a queue runner without re-exec. This happened only when
6928 deliver_drop_privilege was set or when the Exim user was set to root. The
6929 effect of the bug was that timeouts during subsequent deliveries caused
6930 crashes instead of being properly handled. The handler is now left at its
6931 default (and expected) setting.
6933 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6934 message, again when deliver_drop_privilege is set or Exim is run as root.
6935 The bug described in (1) was not present in this case, but the tidying up
6936 of the other signals was missing. I have made the two cases consistent.
6938 3. The ignore_target_hosts setting on a manualroute router was being ignored
6939 for hosts that were looked up using the /MX notation.
6941 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6944 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6945 operated on the sender address. After changing the $sender_address to <>
6946 for the sender address verify, Exim was re-instated it as the original
6947 (before rewriting) address, but remembering that it had rewritten it, so it
6948 wasn't rewriting it again. This bug also had the effect of breaking the
6949 sender address verification caching when the sender address was rewritten.
6951 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6952 This has been changed so that if the ip literal address matches
6953 ignore_target_hosts, the router declines.
6955 7. Added expansion conditions match_domain, match_address, and match_local_
6956 part (NOT match_host).
6958 8. The placeholder for the Received: header didn't have a length field set.
6960 9. Added code to Exim itself and to exim_lock to test for a specific race
6961 condition that could lead to file corruption when using MBX delivery. The
6962 issue is with the lockfile that is created in /tmp. If this file is removed
6963 after a process has opened it but before that process has acquired a lock,
6964 there is the potential for a second process to recreate the file and also
6965 acquire a lock. This could lead to two Exim processes writing to the file
6966 at the same time. The added code performs the same test as UW imapd; it
6967 checks after acquiring the lock that its file descriptor still refers to
6968 the same named file.
6970 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6971 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6974 11. Added the smtp_active_hostname option. If used, this will typically be made
6975 to depend on the incoming interface address. Because $interface_address is
6976 not set up until the daemon has forked a reception process, error responses
6977 that can happen earlier (such as "too many connections") no longer contain
6980 12. If an expansion in a condition on a "warn" statement fails because a lookup
6981 defers, the "warn" statement is abandoned, and the next ACL statement is
6982 processed. Previously this caused the whole ACL to be aborted.
6984 13. Added the iplsearch lookup type.
6986 14. Added ident_timeout as a log selector.
6988 15. Added tls_certificate_verified as a log selector.
6990 16. Added a global option tls_require_ciphers (compare the smtp transport
6991 option of the same name). This controls incoming TLS connections.
6993 17. I finally figured out how to make tls_require_ciphers do a similar thing
6994 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6995 before starting the TLS session.
6997 18. Tabs are now shown as \t in -bP output.
6999 19. If the log selector return_path_on_delivery was set, Exim crashed when
7000 bouncing a message because it had too many Received: header lines.
7002 20. If two routers both had headers_remove settings, and the first one included
7003 a superfluous trailing colon, the final name in the first list and the
7004 first name in the second list were incorrectly joined into one item (with a
7005 colon in the middle).
7011 1. Added -C and -D options to the exinext utility, mainly to make it easier
7012 to include in the automated testing, but these could be helpful when
7013 multiple configurations are in use.
7015 2. The exinext utility was not formatting the output nicely when there was
7016 an alternate port involved in the retry record key, nor when there was a
7017 message id as well (for retries that were specific to a specific message
7018 and a specific host). It was also confused by IPv6 addresses, because of
7019 the additional colons they contain. I have fixed the IPv4 problem, and
7020 patched it up to do a reasonable job for IPv6.
7022 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7023 delivery, the log line now contains "pipelined" if PIPELINING was used.
7025 4. An SMTP transport process used to panic and die if the bind() call to set
7026 an explicit outgoing interface failed. This has been changed; it is now
7027 treated in the same way as a connect() failure.
7029 5. A reference to $sender_host_name in the part of a conditional expansion
7030 that was being skipped was still causing a DNS lookup. This no longer
7033 6. The def: expansion condition was not recognizing references to header lines
7034 that used bh_ and bheader_.
7036 7. Added the _cache feature to named lists.
7038 8. The code for checking quota_filecount in the appendfile transport was
7039 allowing one more file than it should have been.
7041 9. For compatibility with Sendmail, the command line option
7049 and sets the incoming protocol and host name (for trusted callers). The
7050 host name and its colon can be omitted when only the protocol is to be set.
7051 Note the Exim already has two private options, -pd and -ps, that refer to
7052 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7053 "s", but I don't think that's a major issue.
7055 10. A number of refactoring changes to the code, none of which should affect
7058 (a) The number of logging options was getting close to filling up the
7059 32-bit word that was used as a bit map. I have split them into two classes:
7060 those that are passed in the argument to log_write(), and those that are
7061 only ever tested independently outside of that function. These are now in
7062 separate 32-bit words, so there is plenty of room for expansion again.
7063 There is no change in the user interface or the logging behaviour.
7065 (b) When building, for example, log lines, the code previously used a
7066 macro that called string_cat() twice, in order to add two strings. This is
7067 not really sufficiently general. Furthermore, there was one instance where
7068 it was actually wrong because one of the argument was used twice, and in
7069 one call a function was used. (As it happened, calling the function twice
7070 did not affect the overall behaviour.) The macro has been replaced by a
7071 function that can join an arbitrary number of extra strings onto a growing
7074 (c) The code for expansion conditions now uses a table and a binary chop
7075 instead of a serial search (which was left over from when there were very
7076 few conditions). Also, it now recognizes conditions like "pam" even when
7077 the relevant support is not compiled in: a suitably worded error message is
7078 given if an attempt is made to use such a condition.
7080 11. Added ${time_interval:xxxxx}.
7082 12. A bug was causing one of the ddress fields not to be passed back correctly
7083 from remote delivery subprocesses. The field in question was not being
7084 subsequently used, so this caused to problems in practice.
7086 13. Added new log selectors queue_time and deliver_time.
7088 14. Might have fixed a bug in maildirsizefile handling that threw up
7089 "unexpected character" debug warnings, and recalculated the data
7090 unnecessarily. In any case, I expanded the warning message to give more
7093 15. Added the message "Restricted characters in address" to the statements in
7094 the default ACL that block characters like @ and % in local parts.
7096 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7097 Three changes have been made:
7099 (a) There was a serious bug; a negative response to MAIL caused the whole
7100 recipient domain to be cached as invalid, thereby blocking all messages
7101 to all local parts at the same domain, from all senders. This bug has
7102 been fixed. The domain is no longer cached after a negative response to
7103 MAIL if the sender used is not empty.
7105 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7108 (c) A new callout option, "use_sender" has been added for people who want
7109 the modified behaviour.
7115 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7118 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7119 indeed breaks things for older releases.
7121 3. Added additional logging to the case where there is a problem reading data
7122 from a filter that is running in a subprocess using a pipe, in order to
7123 try to track down a specific problem.
7125 4. Testing facility fudge: when running in the test harness and attempting
7126 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7127 getting "No route to host". Convert this to a timeout.
7129 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7132 6. Some OS don't have socklen_t but use size_t instead. This affects the
7133 fifth argument of getsockopt() amongst other things. This is now
7134 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7135 can be set for individual OS. I have set it for SunOS5, OSF1, and
7136 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7137 some earlier ones do not.
7139 7. Change 4.30/15 was not doing the test caselessly.
7141 8. The standard form for an IPv6 address literal was being rejected by address
7142 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7143 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7144 this, as well as the form without the "IPv6" on the front (but only when
7145 address literals are enabled, of course).
7147 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7149 10. Exim crashed if a message with an empty sender address specified by -f
7150 encountered a router with an errors_to setting. This could be provoked only
7151 by a command such as
7155 where an empty string was supplied; "<>" did not hit this bug.
7157 11. Installed PCRE release 4.5.
7159 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7160 remained set. It is now erased.
7162 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7163 times from message ids (which are base 36 rather than the normal 62).
7165 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7166 were being counted as actual protocol errors, and logged if the log
7167 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7168 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7169 and DATA following a set of rejected RCPTs do not count as protocol errors.
7170 In other words, Exim assumes they were pipelined, though this may not
7171 actually be the case. Of course, in all cases the client gets an
7172 appropriate error code.
7174 15. If a lookup fails in an ACL condition, a message about the failure may
7175 be available; it is used if testing the ACL cannot continue, because most
7176 such messages specify what the cause of the deferral is. However, some
7177 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7178 that caused an old message to be retained and used if a later statement
7179 caused a defer, replacing the real cause of the deferral.
7181 16. If an IP address had so many PTR records that the DNS lookup buffer
7182 was not large enough to hold them, Exim could crash while trying to process
7183 the truncated data. It now detects and logs this case.
7185 17. Further to 4.21/58, another change has been made: if (and only if) the
7186 first line of a message (the first header line) ends with CRLF, a bare LF
7187 in a subsequent header line has a space inserted after it, so as not to
7188 terminate the header.
7190 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7191 unnecessarily, used atoi() instead of strtol(), and didn't check the
7192 termination when getting file sizes from file names by regex.
7194 19. Completely re-implemented the support for maildirsize files, in the light
7195 of a number of problems with the previous contributed implementation
7196 (4.30/29). In particular:
7198 . If the quota is zero, the maildirsize file is maintained, but no quota is
7201 . If the maildir directory does not exist, it is created before any attempt
7202 to write a maildirsize file.
7204 . The quota value in the file is just a cache; if the quota is changed in
7205 the transport, the new value overrides.
7207 . A regular expression is available for excluding directories from the
7210 20. The autoreply transport checks the characters in options that define the
7211 message's headers; it allows continued headers, but it was checking with
7212 isspace() after an embedded newline instead of explicitly looking for a
7215 21. If all the "regular" hosts to which an address was routed had passed their
7216 expiry times, and had not reached their retry times, the address was
7217 bounced, even if fallback hosts were defined. Now Exim should go on to try
7220 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7221 equivalent code in the SMTP transport. Some hosts send humungous responses
7222 to HELO/EHLO, more than 1024 it seems.
7224 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7225 gives compiler warnings in some environments. I've now done it "properly",
7228 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7229 (because of problems with the built-in one) was declared to return uschar *
7230 instead of char *, causing compiler failure.
7232 25. Fixed a file descriptor leak when processing alias/forward files.
7234 26. Fixed a minor format string issue in dbfn.c.
7236 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7238 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7239 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7242 29. When a sender address is verified, it is cached, to save repeating the test
7243 when there is more than one recipient in a message. However, when the
7244 verification involves a callout, it is possible for different callout
7245 options to be set for different recipients. It is too complicated to keep
7246 track of this in the cache, so now Exim always runs a verification when a
7247 callout is required, relying on the callout cache for the optimization.
7248 The overhead is duplication of the address routing, but this should not be
7251 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7252 to be verified with callout=postmaster, and the main callout worked but the
7253 postmaster check failed, the verification correctly failed. However, if a
7254 subsequent RCPT command asked for sender verification *without* the
7255 postmaster check, incorrect caching caused this verification also to fail,
7258 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7259 it was not caching the DNS options (qualify_single, search_parents) that
7260 were used when the lookup failed. A subsequent lookup with different
7261 options therefore always gave the same answer, though there were cases
7262 where it should not have. (Example: a "domains = !$mx_any" option on a
7263 dnslookup router: the "domains" option is always processed without any
7264 widening, but the router might have qualify_single set.) Now Exim uses the
7265 cached value only when the same options are set.
7267 32. Added John Jetmore's "exipick" utility to the distribution.
7269 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7270 than a timeout (e.g. a certificate is required, and is not provided), an
7271 Exim server now closes the connection immediately. Previously it waited for
7272 the client to close - but if the client is SSL, it seems that they each
7273 wait for each other, leading to a delay before one of them times out.
7275 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7276 maintained 0.8.x compatibility because I don't think many are using it, and
7277 it is clearly obsolete.
7279 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7282 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7283 client certificate was expired. A simple patch fixes this, though I don't
7284 understand the full logic of why the verify callback is called multiple
7287 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7288 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7289 which causes problems with some clients (such as the Certicom SSL Plus
7290 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7291 disables the coutermeasure allowing Eudora to connect."
7293 38. Exim was not checking that a write() to a log file succeeded. This could
7294 lead to Bad Things if a log got too big, in particular if it hit a file
7295 size limit. Exim now panics and dies if it cannot write to a log file, just
7296 as it does if it cannot open a log file.
7298 39. Modified OS/Makefile-Linux so that it now contains
7300 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7302 The two -D definitions ensure that Exim is compiled with large file
7303 support, which makes it possible to handle log files that are bigger than
7306 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7307 instance) a domain was checked against a named list that involved a lookup,
7308 causing $domain_data to be set, then another domain was checked against the
7309 same list, then the first domain was re-checked, the value of $domain_data
7310 after the final check could be wrong. In particular, if the second check
7311 failed, it could be set empty. This bug probably also applied to
7314 41. The strip_trailing_dot option was not being applied to the address given
7315 with the -f command-line option.
7317 42. The code for reading a message's header from the spool was incrementing
7318 $received_count, but never initializing it. This meant that the value was
7319 incorrect (doubled) while delivering a message in the same process in which
7320 it was received. In the most common configuration of Exim, this never
7321 happens - a fresh exec is done - but it can happen when
7322 deliver_drop_privilege is set.
7324 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7325 it now includes up to 150 characters of the unexpected data in the log
7328 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7329 and building data strings. The size of both of these buffers was 10 000
7330 bytes - far larger than anybody would *ever* want, thought I. Needless to
7331 say, somebody hit the limit. I have increased the maximum line length to
7332 20 000 and the maximum data length of concatenated lines to 100 000. I have
7333 also fixed two bugs, because there was no checking on these buffers. Tsk,
7334 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7335 buffer is too small.
7337 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7338 lsearch lookups. Now it does.
7340 46. When parsing a route_list item in a manualroute router, a fixed-length
7341 buffer was used for the list of hosts. I made this 1024 bytes long,
7342 thinking that nobody would ever have a list of hosts that long. Wrong.
7343 Somebody had a whole pile of complicated expansion conditions, and the
7344 string was silently truncated, leading to an expansion error. It turns out
7345 that it is easier to change to an unlimited length (owing to other changes
7346 that have happened since this code was originally written) than to build
7347 structure for giving a limitation error. The length of the item that
7348 expands into the list of hosts is now unlimited.
7350 47. The lsearch lookup could not handle data where the length of text line was
7351 more than 4095 characters. Such lines were truncated, leading to shortened
7352 data being returned. It should now handle lines of any length.
7354 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7355 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7358 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7359 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7360 no longer happens in comment lines. A list of replacements is now placed
7361 at the head of all of the source files, except those whose only change is
7362 to replace PERL_COMMAND in the very first #! line.
7364 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7365 messages on the queue, with a bottom-up merge sort, using code contributed
7366 by Michael Haardt. This should make operations like -bp somewhat faster on
7367 large queues. It won't affect queue runners, except when queue_run_in_order
7370 51. Installed eximstats 1.31 in the distribution.
7372 52. Added support for SRV lookups to the dnslookup router.
7374 53. If an ACL referred to $message_body or $message_body_end, the value was not
7375 reset for any messages that followed in the same SMTP session.
7377 54. The store-handling optimization for building very long strings was not
7378 differentiating between the different store pools. I don't think this
7379 actually made any difference in practice, but I've tidied it.
7381 55. While running the routers to verify a sender address, $sender_address
7382 was still set to the sender address. This is wrong, because when routing to
7383 send a bounce to the sender, it would be empty. Therefore, I have changed
7384 it so that, while verifying a sender address, $sender_address is set to <>.
7385 (There is no change to what happens when verifying a recipient address.)
7387 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7388 target A or AAAA records (if not already returned) without resetting the
7389 qualify_single or search_parents options of the DNS resolver. These are
7390 inappropriate in this case because the targets of MX and SRV records must
7391 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7392 target that, when qualified, matched something in the local domain. These
7393 two options are now turned off when doing these lookups.
7395 57. It seems that at least some releases of Reiserfs (which does not have the
7396 concept of a fixed number of inodes) returns zero and not -1 for the
7397 number of available inodes. This interacted badly with check_spool_inodes,
7398 which assumed that -1 was the "no such thing" setting. What I have done is
7399 to check that the total number of inodes is greater than zero before doing
7400 the test of how many are available.
7402 58. When a "warn" ACL statement has a log_message modifier, the message is
7403 remembered, and not repeated. This is to avoid a lot of repetition when a
7404 message has many recipients that cause the same warning to be written.
7405 However, Exim was preserving the list of already written lines for an
7406 entire SMTP session, which doesn't seem right. The memory is now reset if a
7407 new message is started.
7409 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7410 debugging output unless log_rewrite was also set.
7412 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7413 of (int)(handle) when we know that handle contains (void *)(-1).
7415 61. The Exim daemon panic-logs an error return when it closes the incoming
7416 connection. However "connection reset by peer" seems to be common, and
7417 isn't really an error worthy of noting specially, so that particular error
7420 62. When Exim is trying to find all the local interfaces, it used to panic and
7421 die if the ioctl to get the interface flags failed. However, it seems that
7422 on at least one OS (Solaris 9) it is possible to have an interface that is
7423 included in the list of interfaces, but for which you get a failure error
7424 for this call. This happens when the interface is not "plumbed" into a
7425 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7426 failure of the "get flags" call assumes that the interface is down.
7428 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7429 makes life easier for people who are doing arithmetic on fields extracted
7430 from dates, where you often get leading zeros that should not be
7431 interpreted as octal.
7433 64. Added qualify_domain to the redirect router, to override the global
7436 65. If a pathologically long header line contained very many addresses (the
7437 report of this problem mentioned 10 000) and each of them was rewritten,
7438 Exim could use up a very large amount of memory. (It kept on making new
7439 copies of the header line as it rewrote, and never released the old ones.)
7440 At the expense of a bit more processing, the header rewriting function has
7441 been changed so that it no longer eats memory in this way.
7443 66. The generation of the Received: header has been moved from the time that a
7444 message starts to be received, to the time that it finishes. The timestamp
7445 in the Received: header should now be very close to that of the <= log
7446 line. There are two side-effects of this change:
7448 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7449 logged header lines no longer include the local Received: line, because
7450 it has not yet been created. The same applies to a copy of the message
7451 that is returned to a non-SMTP sender when a message is rejected.
7453 (b) When a filter file is tested using -bf, no additional Received: header
7454 is added to the test message. After some thought, I decided that this
7457 This change does not affect the value of $received_for. It is still set
7458 after address rewriting, but before local_scan() is called.
7460 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7462 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7463 gave an unhelpful panic error message, and a defer error. I have managed to
7464 change this behaviour so that it now rejects any supplied certificate,
7465 which seems right, as the list of acceptable certificates is empty.
7467 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7468 gave an unhelpful defer error. I have not managed to make this reject any
7469 supplied certificates, but the error message it gives is "no certificate
7470 supplied", which is not helpful.
7472 70. exigrep's output now also includes lines that are not associated with any
7473 message, but which match the given pattern. Implemented by a patch from
7474 Martin Sluka, which also tidied up the Perl a bit.
7476 71. Recipient callout verification, like sender verification, was using <> in
7477 the MAIL FROM command. This isn't really the right thing, since the actual
7478 sender may affect whether the remote host accepts the recipient or not. I
7479 have changed it to use the actual sender in the callout; this means that
7480 the cache record is now keyed on a recipient/sender pair, not just the
7481 recipient address. There doesn't seem to be a real danger of callout loops,
7482 since a callout by the remote host to check the sender would use <>.
7483 [SEE ABOVE: changed after hitting problems.]
7485 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7486 temporary errors. However, in the case of such a code being given after
7487 the end of a data transmission (i.e. after ".") Exim was failing to write
7488 a retry record for the message. (Yes, there was some broken host that was
7489 actually sending 8xx at this point.)
7491 73. An unknown lookup type in a host list could cause Exim to panic-die when
7492 the list was checked. (An example that provoked this was putting <; in the
7493 middle of a list instead of at the start.) If this happened during a DATA
7494 ACL check, a -D file could be left lying around. This kind of configuration
7495 error no longer causes Exim to die; instead it causes a defer error. The
7496 incident is still logged to the main and panic logs.
7498 74. Buglet left over from Exim 3 conversion. The message "too many messages
7499 in one connection" was written to the rejectlog but not the mainlog, except
7500 when address rewriting (yes!) was being logged.
7502 75. Added write_rejectlog option.
7504 76. When a system filter was run not as root (that is, when system_filter_user
7505 was set), the values of the $n variables were not being returned to the
7506 main process; thus, they were not subsequently available in the $sn
7509 77. Added +return_path_on_delivery log selector.
7511 78. A connection timeout was being treated differently from recipients deferred
7512 when testing hosts_max_try with a message that was older than the host's
7513 retry timeout. (The host should not be counted, thus allowing all hosts to
7514 be tried at least once before bouncing.) This may have been the cause of an
7515 occasionally reported bug whereby a message would remain on the queue
7516 longer than the retry timeout, but would be bounced if a delivery was
7517 forced. I say "may" because I never totally pinned down the problem;
7518 setting up timeout/retry tests is difficult. See also the next item.
7520 79. The ultimate address timeout was not being applied to errors that involved
7521 a combination of host plus message (for example, a timeout on a MAIL
7522 command). When an address resolved to a number of possible hosts, and they
7523 were not all tried for each delivery (e.g. because of hosts_max_try), a
7524 message could remain on the queue longer than the retry timeout.
7526 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7529 81. Fixed an obscure SMTP outgoing bug which required at least the following
7530 conditions: (a) there was another message waiting for the same server;
7531 (b) the server returned 5xx to all RCPT commands in the first message so
7532 that the message was not completed; (c) the server dropped the connection
7533 or gave a negative response to the RSET that Exim sends to abort the
7534 transaction. The observed case was a dropped connection after DATA that had
7535 been sent in pipelining mode. That is, the server had advertised PIPELINING
7536 but was not implementing it correctly. The effect of the bug was incorrect
7537 behaviour, such as trying another host, and this could lead to a crash.
7543 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7544 and daemon.c were passed as pointers to ints; they should have been
7545 pointers to socklen_t variables (which are typically unsigned ints).
7547 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7550 3. Fixed a really odd bug that affected only the testing scheme; patching a
7551 certain fixed string in the binary changed the value of another string that
7552 happened to be identical to the end of the original first string.
7554 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7555 name", it returns that address as the IP address. On some operating
7556 systems (e.g. Solaris), it also passes back the IP address string as the
7557 "host name". However, on others (e.g. Linux), it passes back an empty
7558 string. Exim wasn't checking for this, and was changing the host name to an
7559 empty string, assuming it had been canonicalized.
7561 5. Although rare, it is permitted to have more than one PTR record for a given
7562 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7563 all the names associated with an address, because they do in Solaris.
7564 However, it seems that they do not in Linux for data that comes from the
7565 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7566 I found this out when I moved to a new Linux workstation and tried to run
7567 the Exim test suite.
7569 To get round this problem I have changed the code so that it now does its
7570 own call to the DNS to look up PTR records when searching for a host name.
7571 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7572 addresses that are only in /etc/hosts are still found.
7574 This behaviour is, however, controlled by an option called host_lookup_
7575 order, which defaults to "bydns:byaddr". If people want to use the other
7576 order, or indeed, just use one or the other means of lookup, they can
7577 specify it in this variable.
7579 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7580 some operating systems, this comes back from gethostbyaddr() as an empty
7581 string, and this is what Exim used to test for. However, it seems that in
7582 other systems, "." is yielded. Exim now tests for this case too.
7584 7. The values of check_spool_space and check_log_space are now held internally
7585 as a number of kilobytes instead of an absolute number of bytes. If a
7586 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7587 kilobyte. This means that much larger values can be stored.
7589 8. Exim monitor: an attempt to get the action menu when not actually pointing
7590 at a message produces an empty menu entitled "No message selected". This
7591 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7592 no entries in it ("Shell widget menu has zero width and/or height"). So I
7593 have added a single, blank menu entry in this case.
7595 9. Added ${quote_local_part.
7597 10. MIME decoding is now applied to the contents of Subject: header lines when
7600 11. Now that a reference to $sender_host_address automatically causes a reverse
7601 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7602 host lookup before query-style lookups in lists that might use this
7603 variable. This has therefore been abolished, and the "net-" prefix is no
7604 longer necessary for query-style lookups.
7606 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7607 have been a typo for LFLAGS, so it has been changed.
7609 13. The install script calls Exim with "-C /dev/null" in order to find the
7610 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7611 to be output. However, since Exim outputs its version number before the
7612 error, it didn't break the script. It just looked ugly. I fixed this by
7613 always allowing "-C /dev/null" if the caller is root.
7615 14. Ignore overlarge ACL variable number when reading spool file - insurance
7616 against a later release with more variables having written the file.
7618 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7619 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7620 this, as well as the form without the "IPv6" on the front.
7622 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7623 OS/Makefile-Darwin file.
7625 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7626 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7628 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7629 variable. However, some OS use ints for inodes. I've added cast to long int
7630 to get rid of the compiler warning.
7632 19. I had forgotten to lock out "/../" in configuration file names when
7633 ALT_CONFIG_PREFIX was set.
7635 20. Routers used for verification do not need to specify transports. However,
7636 if such a router generated a host list, and callout was configured, Exim
7637 crashed, because it could not find a port number from the (non-existent)
7638 transport. It now assumes port 25 in this circumstance.
7640 21. Added the -t option to exigrep.
7642 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7643 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7644 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7645 them is set, LOOKUP_LSEARCH is forced.
7647 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7649 24. Added sender and host information to the "rejected by local_scan()" log
7650 line; previously there was no indication of these.
7652 25. Added .include_if_exists.
7654 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7655 Linux. It turns out that not all file systems support this. Apparently some
7656 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7657 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7658 not supported on the file descriptor, is now ignored when Exim is trying to
7659 sync a directory. This applies only to Linux.
7661 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7663 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7664 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7667 29. Added support for maildirsize files from supplied patch (modified a bit).
7669 30. The use of :fail: followed by an empty string could lead Exim to respond to
7670 sender verification failures with (e.g.):
7672 550 Verification failed for <xxx>
7673 550 Sender verify failed
7675 where the first response line was missing the '-' that indicates it is not
7676 the final line of the response.
7678 31. The loop for finding the name of the user that called Exim had a hardwired
7679 limit of 10; it now uses the value of finduser_retries, which is used for
7680 all other user lookups.
7682 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7685 33. Exim was neglecting to zero errno before one call of strtol() when
7686 expanding a string and expecting an integer value. On some systems this
7687 resulted in spurious "integer overflow" errors. Also, it was casting the
7688 result into an int without checking.
7690 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7691 did not work. The code looks as if it has *never* worked, though it appears
7692 to have been documented since at least release 1.62. I have made it work.
7694 35. The "timeout_DNS" error in retry rules, also documented since at least
7695 1.62, also never worked. As it isn't clear exactly what this means, and
7696 clearly it isn't a major issue, I have abolished the feature by treating it
7697 as "timeout", and writing a warning to the main and panic logs.
7699 36. The display of retry rules for -brt wasn't always showing the error code
7702 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7703 timeout_connect_A, timeout_connect_MX.
7705 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7706 to the empty sender.
7708 39. The daemon was not analysing the content of -oX till after it had closed
7709 stderr and disconnected from the controlling terminal. This meant that any
7710 syntax errors were only noted on the panic log, and the return code from
7711 the command was 0. By re-arranging the code a little, I've made the
7712 decoding happen first, so such errors now appear on stderr, and the return
7713 code is 1. However, the actual setting up of the sockets still happens in
7714 the disconnected process, so errors there are still only recorded on the
7717 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7718 connections (as happens on some IP stacks) was logged at start up time as
7719 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7720 it from "IPv6 and IPv4", which means that two separate sockets are being
7723 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7724 says whether AF_INET or AF_INET6 was passed as an argument.
7726 42. Exiwhat output was messed up when time zones were included in log
7729 43. Exiwhat now gives more information about the daemon's listening ports,
7730 and whether -tls-on-connect was used.
7732 44. The "port" option of the smtp transport is now expanded.
7734 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7735 silently ignored. Now an error message is written to the main and panic
7738 46. There's a new ACL modifier called "logwrite" which writes to a log file
7739 as soon as it is encountered.
7741 47. Added $local_user_uid and $local_user_gid at routing time.
7743 48. Exim crashed when trying to verify a sender address that was being
7746 49. Exim was recognizing only a space character after ".include". It now also
7747 recognizes a tab character.
7749 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7750 extracting the relevant information from the specification. The man page no
7751 longer contains scrambled data for the -d option, and I've added a section
7752 at the front about calling Exim under different names.
7754 51. Added "extra_headers" argument to the "mail" command in filter files.
7756 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7759 53. Installed eximstats 1.29.
7761 54. Added transport_filter_timeout as a generic transport option.
7763 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7764 Cc: header lines. This was required by RFC 822, but it not required by RFC
7767 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7768 incoming messages that did not have them. Now it does so only if the
7769 message originates locally, that is, if there is no associated remote host
7770 address. When Resent- header lines are present, this applies to the Resent-
7771 lines rather than the non-Resent- lines.
7773 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7774 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7776 58. Messages for configuration errors now include the name of the main
7777 configuration file - useful now that there may be more than one file in a
7778 list (.included file names were always shown).
7780 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7781 for those rare installations that do not start the daemon as root or run it
7782 setuid root. I've cut out the call to initgroups() if the daemon is not
7785 60. The Exim user and group can now be bound into the binary as text strings
7786 that are looked up at the start of Exim's processing.
7788 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7790 62. Added $mailstore_basename variable.
7792 63. Installed patch to sieve.c from Michael Haardt.
7794 64. When Exim failed to open the panic log after failing to open the main log,
7795 the original message it was trying to log was written to stderr and debug
7796 output, but if they were not available (the usual case in production), it
7797 was lost. Now it is written to syslog before the two lines that record the
7798 failures to open the logs.
7800 65. Users' Exim filters run in subprocesses under the user's uid. It is
7801 possible for a "deliver" command or an alias in a "personal" command to
7802 provoke an address rewrite. If logging of address rewriting is configured,
7803 this fails because the process is not running as root or exim. There may be
7804 a better way of dealing with this, but for the moment (because 4.30 needs
7805 to be released), I have disabled address rewrite logging when running a
7806 filter in a non-root, non-exim process.
7812 1. The buildconfig auxiliary program wasn't quoting the value set for
7813 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7814 not defined. This bug was masked in 4.22 by the effect that was fixed in
7817 2. Some messages that were rejected after a message id was allocated were
7818 shown as "incomplete" by exigrep. It no longer does this for messages that
7819 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7821 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7822 have allow_domain_literals set, the ID did not get logged in the <= line.
7823 Domain literals are now always recognized in Message-ID: header lines.
7825 4. The first argument for a ${extract expansion item is the key name or field
7826 number. Leading and trailing spaces in this item were not being ignored,
7827 causing some misleading effects.
7829 5. When deliver_drop_privilege was set, single queue runner processes started
7830 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7831 same command in the process it spins off) were not dropping privilege.
7833 6. When the daemon running as "exim" started a queue runner, it always
7834 re-executed Exim in the spun-off process. This is a waste of effort when
7835 deliver_drop_privilege is set. The new process now just calls the
7836 queue-runner function directly.
7842 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7845 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7846 ignored. Though the use of -oP was forcing the writing of a pid file, it
7847 was always written to the default place.
7849 3. If the message "no IP address found for host xxxx" is generated during
7850 incoming verification, it is now followed by identification of the incoming
7851 connection (so you can more easily find what provoked it).
7853 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7855 5. Added some features to "harden" Exim a bit more against certain attacks:
7857 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7858 be put in Local/Makefile. This is like the never_users runtime option,
7859 but it cannot be overridden. The default setting is "root".
7861 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7862 prefix string with which any file named in a -C command line option
7865 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7866 is retained for -C and -D only if the caller of Exim is root. Without
7867 it, the exim user may also use -C and -D and retain privilege.
7869 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7870 command line option is disabled.
7872 6. Macro names set by the -D option must start with an upper case letter, just
7873 like macro names defined in the configuration file.
7875 7. Added "dereference=" facility to LDAP.
7877 8. Two instances of the typo "uknown" in the source files are fixed.
7879 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7880 the Configure-Makefile script screwed up while processing it.
7882 10. Incorporated PCRE 4.4.
7884 11. The SMTP synchronization check was not operating right at the start of an
7885 SMTP session. For example, it could not catch a HELO sent before the client
7886 waited for the greeting. There is now a check for outstanding input at the
7887 point when the greeting is written. Because of the duplex, asynchronous
7888 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7889 way, but not yet received, when the check is performed.
7891 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7892 on TCP/IP sockets, because this apparently causes some broken clients to
7895 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7896 unchanged) from the Cygwin maintainer.
7898 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7899 was supported instead of testing for mbx. Effectively a typo.
7901 15. The spa authenticator server code was not checking that the input it
7902 received was valid base64.
7904 16. The debug output line for the "set" modifier in ACLs was not showing the
7905 name of the variable that was being set.
7907 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7909 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7910 lookup on its own. Something else was needed to trigger the lookup. For
7911 example, a match in host_lookup or the need for a host name in a host list.
7912 Now, if $sender_host_name is referenced and the host name has not yet been
7913 looked up, a lookup is performed. If the lookup fails, the variable remains
7914 empty, and $host_lookup_failed is set to "1".
7916 19. Added "eqi" as a case-independent comparison operator.
7918 20. The saslauthd authentication condition could segfault if neither service
7919 nor realm was specified.
7921 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7922 error message that was logged was misleading, and incoming SMTP
7923 connections were dropped. The message is now more accurate, and temporary
7924 errors are given to SMTP connections.
7926 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7927 (except RSET) with a 421 error, until QUIT is received. However, it was
7928 failing to send a response to QUIT.
7930 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7931 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7932 ACL is now run after the helo_try_verify_hosts code.
7934 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7935 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7936 case-independent, so other case variants are also recognized). Apparently
7937 some systems use these upper case variants.
7939 25. If more than two messages were waiting for the same host, and a transport
7940 filter was specified for the transport, Exim sent two messages over the
7941 same TCP/IP connection, and then failed with "socket operation on non-
7942 socket" when it tried to send the third.
7944 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7946 27. The extern definition of crypt16() in expand.c was not being excluded when
7947 the OS had its own crypt16() function.
7949 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7950 as a preferred synonym for return_size_limit, both as an option and as an
7953 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7955 30. Changed the default configuration ACL to relax the local part checking rule
7956 for addresses that are not in any local domains. For these addresses,
7957 slashes and pipe symbols are allowed within local parts, but the sequence
7958 /../ is explicitly forbidden.
7960 31. SPA server authentication was not clearing the challenge buffer before
7963 32. log_message in a "warn" ACL statement was writing to the reject log as
7964 well as to the main log, which contradicts the documentation and doesn't
7965 seem right (because no rejection is happening). So I have stopped it.
7967 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7968 However, I am unable to do any testing of this.
7970 34. Fixed an infelicity in the appendfile transport. When checking directories
7971 for a mailbox, to see if any needed to be created, it was accidentally
7972 using path names with one or more superfluous leading slashes; tracing
7973 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7975 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7976 contents are added to the log line that is written for every discarded
7977 recipient. (Previously a log_message setting was ignored.)
7979 36. The ${quote: operator now quotes the string if it is empty.
7981 37. The install script runs exim in order to find its version number. If for
7982 some reason other than non-existence or emptiness, which it checks, it
7983 could not run './exim', it was installing it with an empty version number,
7984 i.e. as "exim-". This error state is now caught, and the installation is
7987 38. An argument was missing from the function that creates an error message
7988 when Exim fails to connect to the socket for saslauthd authentication.
7989 This could cause Exim to crash, or give a corrupted message.
7991 39. Added isip, isip4, and isip6 to ${if conditions.
7993 40. The ACL variables $acl_xx are now saved with the message, and can be
7994 accessed later in routers, transports, and filters.
7996 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7997 strings in the file are not string-expanded.
7999 42. If a MAIL command specified a SIZE value that was too large to fit into an
8000 int variable, the check against message_size_limit failed. Such values are
8001 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8002 day this will have to be increased, but I don't think I want to be around
8003 when emails are that large.
8010 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8011 iconv() is not standard in FreeBSD.
8013 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8014 IPv6 enabled. The observed symptom was a segmentation fault on return from
8015 the function os_common_find_running_interfaces() in src/os.c.
8017 3. In the check_special_case() function in daemon.c I had used "errno" as an
8018 argument name, which causes warnings on some systems. This was basically a
8019 typo, since it was named "eno" in the comments!
8021 4. The code that waits for the clock to tick (at a resolution of some fraction
8022 of a second) so as to ensure message-id uniqueness was always waiting for
8023 at least one whole tick, when it could have waited for less. [This is
8024 almost certainly not relevant at current processor speeds, where it is
8025 unlikely to ever wait at all. But we try to future-proof.]
8027 5. The function that sleeps for a time interval that includes fractions of a
8028 second contained a race. It did not block SIGALRM between setting the
8029 timer, and suspending (a couple of lines later). If the interval was short
8030 and the sigsuspend() was delayed until after it had expired, the suspension
8031 never ended. On busy systems this could lead to processes getting stuck for
8034 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8035 process, before it forks any delivery processes. The open lookup caching
8036 mechanism meant that the open file or database connection was passed into
8037 the delivery process. The problem was that delivery processes always tidy
8038 up cached lookup data. This could cause a problem for the next delivery
8039 process started by the queue runner, because the external queue runner
8040 process does not know about the closure. So the next delivery process
8041 still has data in the lookup cache. In the case of a file lookup, there was
8042 no problem because closing a file descriptor in a subprocess doesn't affect
8043 the parent. However, if the lookup was caching a connection to a database,
8044 the connection was closed, and the second delivery process was likely to
8045 see errors such as "PGSQL: query failed: server closed the connection
8046 unexpectedly". The problem has been fixed by closing all cached lookups
8047 in a queue runner before running a delivery process.
8049 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8050 seem to have the "const" qualifier which it has on other OS. I've
8053 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8054 *of the same type* (client or server) with the same public name that an
8055 error should be diagnosed.
8057 9. When Exim looked up a host name for an IP address, but failed to find the
8058 original IP address when looking up the host name (a safety check), it
8059 output the message "<ip address> does not match any IP for NULL", which was
8060 confusing, to say the least. The bug was that the host name should have
8061 appeared instead of "NULL".
8063 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8064 user that is built into the binary, and the -C or -D options is used, root
8065 privilege is dropped before the configuration file is read. In addition,
8066 logging is switched to stderr instead of the normal log files. If the
8067 configuration then re-defines the Exim user, the unprivileged environment
8068 is probably not what is expected, so Exim logs a panic warning message (but
8071 However, if deliver_drop_privilege is set, the unprivileged state may well
8072 be exactly what is intended, so the warning has been cut out in that case,
8073 and Exim is allowed to try to write to its normal log files.
8079 1. smtp_return_error_details was not giving details for temporary sender
8080 or receiver verification errors.
8082 2. Diagnose a configuration error if two authenticators have the same public
8085 3. Exim used not to create the message log file for a message until the first
8086 delivery attempt. This could be confusing when incoming messages were held
8087 for policy or load reasons. The message log file is now created at the time
8088 the message is received, and an initial "Received" line is written to it.
8090 4. The automatically generated man page for command line options had a minor
8091 bug that caused no ill effects; however, a more serious problem was that
8092 the procedure for building the man page automatically didn't always
8093 operate. Consequently, release 4.20 contains an out-of-date version. This
8094 shouldn't happen again.
8096 5. When building Exim with embedded Perl support, the script that builds the
8097 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8098 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8100 6. The freeze_tell option was not being used for messages that were frozen on
8101 arrival, either by an ACL or by local_scan().
8103 7. Added the smtp_incomplete_transaction log selector.
8105 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8106 was accepting AUTH without a new EHLO.
8108 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8109 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8112 10. Logging of TCP/IP connections (when configured) now happens in the main
8113 daemon process instead of the child process, so that the TCP/IP connection
8114 count is more accurate (but it can never be perfect).
8116 11. The use of "drop" in a nested ACL was not being handled correctly in the
8117 outer ACL. Now, if condition failure induced by the nested "drop" causes
8118 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8119 or "require"), the connection is dropped.
8121 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8122 that yield "discard" can now be used with an "accept" or a "discard" verb,
8123 but an error is generated for any others (because I can't see a useful way
8124 to define what should happen).
8126 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8127 are now processed in the same way as lines in the Exim configuration file.
8128 In particular, continuation lines are supported.
8130 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8132 15. Added -ti meaning -t -i.
8134 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8135 domains, and warn by logging if others are found.
8137 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8138 returned from GIFCONF when Exim is trying to find the list of interfaces on
8139 a host. The code in os.c has been modified to copy each ifreq to an aligned
8140 structure in all cases.
8142 Also, in some cases, the returned ifreq's were being copied to a 'struct
8143 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8144 means the last couple of bytes of an IPv6 address could be chopped if the
8145 ifreq contained only a normal sockaddr (14 bytes storage).
8147 18. Named domain lists were not supported in the hosts_treat_as_local option.
8148 An entry such as +xxxx was not recognized, and was treated as a literal
8151 19. Ensure that header lines added by a DATA ACL are included in the reject log
8152 if the ACL subsequently rejects the message.
8154 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8155 MD5 (which is deprecated).
8157 21. When testing a filter file using -bf, Exim was writing a message when it
8158 took the sender from a "From " line in the message, but it was not doing so
8159 when it took $return_path from a Return-Path: header line. It now does.
8161 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8162 with a valid header line field name (a series of printing characters
8163 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8165 23. Changed "disc" in the source to "disk" to conform to the documentation and
8166 the book and for uniformity.
8168 24. Ignore Sendmail's -Ooption=value command line item.
8170 25. When execve() failed while trying to run a command in a pipe transport,
8171 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8172 could be confused with a return value of 69 from the command itself. This
8173 has been changed to 127, the value the shell returns if it is asked to run
8174 a non-existent command. The wording for the related log line suggests a
8175 non-existent command as the problem.
8177 26. If received_header_text expands to an empty string, do not add a Received:
8178 header line to the message. (Well, it adds a token one on the spool, but
8179 marks it "old" so that it doesn't get used or transmitted.)
8181 27. Installed eximstats 1.28 (addition of -nt option).
8183 28. There was no check for failure on the call to getsockname() in the daemon
8184 code. This can fail if there is a shortage of resources on the system, with
8185 ENOMEM, for example. A temporary error is now given on failure.
8187 29. Contrary to the C standard, it seems that in some environments, the
8188 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8189 program. Exim now does this explicitly; it affects the formatting of
8190 timestamps using strftime().
8192 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8193 complaints. I've now initialized all the variables, to avoid this.
8195 32. Header lines added by a system filter were not being "seen" during
8196 transport-time rewrites.
8198 33. The info_callback() function passed to OpenSSL is set up with type void
8199 (*)(SSL *, int, int), as described somewhere. However, when calling the
8200 function (actually a macro) that sets it up, the type void(*)() is
8201 expected. I've put in a cast to prevent warnings from picky compilers.
8203 34. If a DNS black list lookup found a CNAME record, but there were no A
8204 records associated with the domain it pointed at, Exim crashed.
8206 35. If a DNS black list lookup returned more than one A record, Exim ignored
8207 all but the first. It now scans all returned addresses if a particular IP
8208 value is being sought. In this situation, the contents of the
8209 $dnslist_value variable are a list of all the addresses, separated by a
8212 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8213 to accept a host name and all its aliases if the forward lookup for any of
8214 them yielded the IP address of the incoming connection. Now it accepts only
8215 those names whose forward lookup yields the correct IP address. Any other
8216 names are discarded. This closes a loophole whereby a rogue DNS
8217 administrator could create reverse DNS records to break through a
8218 wildcarded host restriction in an ACL.
8220 37. If a user filter or a system filter that ran in a subprocess used any of
8221 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8222 the wrong values were passed to the pipe command ($thisaddress had the
8223 value of $0, $0 had the value of $1, etc). This bug was introduced by
8224 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8226 38. Improved the line breaking for long SMTP error messages from ACLs.
8227 Previously, if there was no break point between 40 and 75 characters, Exim
8228 left the rest of the message alone. Two changes have been made: (a) I've
8229 reduced the minimum length to 35 characters; (b) if it can't find a break
8230 point between 35 and 75 characters, it looks ahead and uses the first one
8231 that it finds. This may give the occasional overlong line, but at least the
8232 remaining text gets split now.
8234 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8235 file descriptors might be low, and that setting 1000 would always raise it.
8236 It turns out that in some environments, the limit is already over 1000 and
8237 that lowering it causes trouble. So now Exim takes care not to decrease it.
8239 40. When delivering a message, the value of $return_path is set to $sender_
8240 address at the start of routing (routers may change the value). By an
8241 oversight, this default was not being set up when an address was tested by
8242 -bt or -bv, which affected the outcome if any router or filter referred to
8245 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8246 log only when "message" or "log_message" are set. However, if one of the
8247 conditions was an address verification, or a call to a nested ACL, the
8248 messages generated by the underlying test were being passed through. This
8249 no longer happens. The underlying message is available in $acl_verify_
8250 message for both "message" and "log_message" expansions, so it can be
8251 passed through if needed.
8253 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8254 new expansion $bh_ to give the encoded byte string without charset
8255 translation. Translation happens only if iconv() is available; HAVE_ICONV
8256 indicates this at build time. HEADERS_CHARSET gives the charset to
8257 translate to; headers_charset can change it in the configuration, and
8258 "headers charset" can change it in an individual filter file.
8260 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8261 that creates RFC 2047 encoded "words" labels them as that charset instead
8262 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8263 expansion operator; (ii) when Exim creates a From: line for a local
8264 message; (iii) when a header line is rewritten to include a "phrase" part.
8266 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8267 buggy, causing it to skip the first lines of messages whose message ID
8268 ended in 'D'. This would not have bitten before Exim release 4.14, because
8269 message IDs were unlikely to end in 'D' before then. The effect was to have
8270 incorrect size information for certain domains.
8272 45. #include "config.h" was missing at the start of the crypt16.c module. This
8273 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8276 46. If there was a timeout during a "random" callout check, Exim treated it as
8277 a failure of the random address, and carried on sending RSET and the real
8278 address. If the delay was just some slowness somewhere, the response to the
8279 original RCPT would be taken as a response to RSET and so on, causing
8280 mayhem of various kinds.
8282 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8283 when I implemented it. It didn't allow for the fact that some option values
8284 may legitimately be negative (e.g. size_addition), and it didn't even do
8285 the right test for positive values.
8287 48. Domain names in DNS records are case-independent. Exim always looks them up
8288 in lower case. Some resolvers return domain names in exactly the case they
8289 appear in the zone file, that is, they may contain uppercase letters. Not
8290 all resolvers do this - some return always lower case. Exim was treating a
8291 change of case by a resolver as a change of domain, similar to a widening
8292 of a domain abbreviation. This triggered its re-routing code and so it was
8293 trying to route what was effectively the same domain again. This normally
8294 caused routing to fail (because the router wouldn't handle the domain
8295 twice). Now Exim checks for this case specially, and just changes the
8296 casing of the domain that it ultimately uses when it transmits the message
8299 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8302 50. If a filter generated a file delivery with a non-absolute name (possible if
8303 no home directory exists for the router), the forbid_file option was not
8306 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8307 the existing equality matching.
8309 52. Exim was using ints instead of ino_t variables in some places where it was
8310 dealing with inode numbers.
8312 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8313 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8314 called TMPDIR, and if it finds it is different, it changes its value.
8316 54. The smtp_printf() function is now made available to local_scan() so
8317 additional output lines can be written before returning. There is also an
8318 smtp_fflush() function to enable the detection of a dropped connection.
8319 The variables smtp_input and smtp_batched_input are exported to
8322 55. Changed the default runtime configuration: the message "Unknown user"
8323 has been removed from the ACL, and instead placed on the localuser router,
8324 using the cannot_route_message feature. This means that any verification
8325 failures that generate their own messages won't get overridden. Similarly,
8326 the "Unrouteable address" message that was in the ACL for unverifiable
8327 relay addresses has also been removed.
8329 56. Added hosts_avoid_esmtp to the smtp transport.
8331 57. The exicyclog script was not checking for the esoteric option
8332 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8333 will work only if exicyclog is run under the appropriate euid.
8335 58. Following a discussion on the list, the rules by which Exim recognises line
8336 endings on incoming messages have been changed. The -dropcr and drop_cr
8337 options are now no-ops, retained only for backwards compatibility. The
8338 following line terminators are recognized: LF CRLF CR. However, special
8339 processing applies to CR:
8341 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8342 nor a local message in the state where . is a terminator.
8344 (ii) If a bare CR is encountered in a header line, an extra space is added
8345 after the line terminator so as not to end the header. The reasoning
8346 behind this is that bare CRs in header lines are most likely either
8347 to be mistakes, or people trying to play silly games.
8349 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8350 was being incorrectly given as 18 bytes larger than it should have been.
8351 This is a VOB (very old bug).
8353 60. This may never have affected anything current, but just in case it has:
8354 When the local host is found other than at the start of a list of hosts,
8355 the local host, those with the same MX, and any that follow, are discarded.
8356 When the list in question was part of a longer list of hosts, the following
8357 hosts (not currently being processed) were also being discarded. This no
8358 longer happens. I'm not sure if this situation could ever has previously
8361 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8364 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8365 header. This is something that is recommended in a new Internet Draft, and
8366 is something that is documented as being done by Sendmail. There are two
8367 possible values. For messages generated by the autoreply transport, Exim
8370 Auto-Submitted: auto-replied
8372 whereas for all other generated messages (e.g. bounces) it adds
8374 Auto-Submitted: auto-generated
8376 63. The "personal" condition in filters now includes a test for the
8377 Auto-Submitted: header. If it contains the string "auto-" the message it
8378 not considered personal.
8380 64. Added rcpt_include_affixes as a generic transport option.
8382 65. Added queue_only_override (default true).
8384 66. Added the syslog_duplication option.
8386 67. If what should have been the first header line of a message consisted of
8387 a space followed by a colon, Exim was mis-interpreting it as a header line.
8388 It isn't of course - it is syntactically invalid and should therefore be
8389 treated as the start of the message body. The misbehaviour could have
8390 caused a number of strange effects, including loss of data in subsequent
8391 header lines, and spool format errors.
8393 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8394 client host had authenticated. This control can now be exercised by an ACL
8395 for more flexibility.
8397 69. By default, callouts do not happen when testing with -bh. There is now a
8398 variant, -bhc, which does actually run the callout code, including
8399 consulting and updating the callout cache.
8401 70. Added support for saslauthd authentication, courtesy of Alexander
8404 71. If statvfs() failed on the spool or log directories while checking their
8405 size for availability, Exim confusingly gave the error "space shortage".
8406 Furthermore, in debugging mode it crashed with a floating point exception.
8407 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8408 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8409 serious problem, Exim now writes to the main and panic logs when this
8410 happens, with details of the failure. It then refuses to accept the
8411 incoming message, giving the message "spool directory problem" or "log
8412 directory problem" with a 421 code for SMTP messages.
8414 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8415 0, 1, and 2 exist, because some OS complain for execs without them (see
8416 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8417 descriptors. However, the code omitted to check that the open succeeded,
8418 causing mysterious errors if for some reason the permissions on /dev/null
8419 got screwed. Now Exim writes a message to the main and panic logs, and
8420 bombs out if it can't open /dev/null.
8422 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8423 interact so that it is all more flexible. It is supposed to remain
8424 backwards compatible. Also added extra_local_interfaces.
8426 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8427 to bomb out with an assertion failure - to the client this appears as a
8428 connection drop. This problem occurs in the part of the code that was taken
8429 from the Samba project. Fortunately, the assertion is in a very simple
8430 function, so I have fixed this by reproducing the function inline in the
8431 one place where it is called, and arranging for authentication to fail
8432 instead of killing the process with assert().
8434 75. The SPA client code was not working when the server requested OEM rather
8435 than Unicode encoding.
8437 76. Added code to make require_files with a specific uid setting more usable in
8438 the case where statting the file as root fails - usually a non-root-mounted
8439 NFS file system. When this happens and the failure is EACCES, Exim now
8440 forks a subprocess and does the per-uid checking as the relevant uid.
8442 77. Added process_log_path.
8444 78. If log_file_path was not explicitly set, a setting of check_log_space or
8445 check_log_inodes was ignored.
8447 79. If a space check for the spool or log partitions fails, the incident is now
8448 logged. Of course, in the latter case the data may get lost...
8450 80. Added the %p formatting code to string_format() so that it can be used to
8451 print addresses in debug_print(). Adjusted all the address printing in the
8452 debugging in store.c to use %p rather than %d.
8454 81. There was a concern that a line of code in smtp_in.c could overflow a
8455 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8456 initially expressed, the concern was not well-founded, because trailing
8457 spaces are removed early. However, if the trailing spaces were followed by
8458 a NULL, they did not get removed, so the overflow was possible. Two fixes
8461 (a) I re-wrote the offending code in a cleaner fashion.
8462 (b) If an incoming SMTP command contains a NULL character, it is rejected
8465 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8466 runs initgroups(), so that if the Exim user is in any additional groups,
8467 they will be used during message reception.
8473 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob