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.
203 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
204 after reception to before a subsequent reception. This should
205 mean slightly faster delivery, and also confirmation of reception
208 JH/02 Move from using the pcre library to pcre2. The former is no longer
209 being developed or supported (by the original developer).
211 JH/03 Constification work in the filters module required a major version
212 bump for the local-scan API. Specifically, the "headers_charset"
213 global which is visible via the API is now const and may therefore
214 not be modified by local-scan code.
216 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
217 sendfile() didi not account for the way the ClamAV driver code called it.
219 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
220 time check was being done for every character; replace that with one
223 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
224 sent was prefixed with a length byte.
226 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
227 RFC 5321. Previously Dovecot (at least) would log errors during
230 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
231 in 4.61 without a reason logged. Binaries will be bigger, which might
232 matter on diskspace-constrained systems, but debug is easier.
234 JH/09 Fix macro-definition during "-be" expansion testing. The move to
235 write-protected store for macros had not accounted for these runtime
236 additions; fix by removing this protection for "-be" mode.
238 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
239 handing out large-numbered file descriptors, violating the usual Unix
240 assumption (and required by Posix) that the lowest possible number will be
241 allocated by the kernel when a new one is needed. In the daemon, and any
242 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
243 are not useable for FD_SET() [and hence select()] and overwrite the stack.
244 Assorted crashes happen.
246 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
247 main-section options or in a connect ACL, the value from the first ever
248 connection was never replaced for subsequent connections. Found by
251 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
252 though only once PCRE2 was introduced: the memory accounting used under
253 debug offset allocations by an int, giving a hard trap in early startup.
254 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
256 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
257 with underbars is given. The write-protection of configuration introduced
258 in 4.95 trapped when normalisation was applied to an option not needing
261 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
263 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
264 failing to close the certificates directory, every hour or any time it
267 JH/16 Debugging initiated by an ACL control now continues through into routing
268 and transport processes. Previously debugging stopped any time Exim
269 re-execs, or for processing a queued message.
271 JH/17 The "expand" debug selector now gives more detail, specifically on the
272 result of expansion operators and items.
274 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
275 bad comparison between the option value and the name of the file to
276 be included was done, and a mismatch was wrongly identified.
277 4.88 to 4.95 are affected.
279 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
281 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
282 supplied as the db file-pair basename. Previously, if a directory
283 path was given, for example via the autoreply "once" option, the DB
284 file.pag and file.dir files would be created in that directory's
287 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
288 "taint" log_selector. These were previously deprecated.
290 JH/22 Fix static address-list lookups to properly return the matched item.
291 Previously only the domain part was returned.
293 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
294 the call into OpenSSL to send a TLS Close was being repeated; this
295 resulted in the library waiting for the peer's Close. If that was never
296 sent we waited forever. Fix by tracking send calls.
298 JH/24 The ${run} expansion item now expands its command string elements after
299 splitting. Previously it was before; the new ordering makes handling
300 zero-length arguments simpler. The old ordering can be obtained by
301 appending a new option "preexpand", after a comma, to the "run".
303 JH/25 Taint-check exec arguments for transport-initiated external processes.
304 Previously, tainted values could be used. This affects "pipe", "lmtp" and
305 "queryprogram" transport, transport-filter, and ETRN commands.
306 The ${run} expansion is also affected: in "preexpand" mode no part of
307 the command line may be tainted, in default mode the executable name
310 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
311 the the facility was not passed across execs, and only the first message
312 passed over a connection could use BDAT; any further ones using DATA.
314 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
315 uses $sending_ip_address and an interface is specified.
316 Previously any use of the local address in the EHLO name disabled
317 PIPECONNECT, the common case being to use the rDNS of it.
319 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
320 resumption. Previously verify failed because no certificate status is
321 passed on the wire for the restarted session. Fix by using the recorded
322 ocsp status of the stored session for the new connection.
324 JH/29 TLS resumption: the key for session lookup in the client now includes
325 more info that a server could potentially use in configuring a TLS
326 session, avoiding oferring mismatching sessions to such a server.
327 Previously only the server IP was used.
329 JH/30 Fix string_copyn() for limit greater than actual string length.
330 Previously the copied amount was the limit, which could result in a
331 overlapping memcpy for newly allocated destination soon after a
332 source string shorter than the limit. Found/investigated by KM.
334 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
335 close; it may be needed for a subsequent connection. This caused a
336 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
338 JH/32 Fix CHUNKING for a second message on a connection when the first was
339 rejected. Previously we did not reset the chunking-offered state, and
340 erroneously rejected the BDAT command. Investigation help from
343 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
344 an empty address. Previously the expansion returned an error.
346 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
347 proxy. Previously these were misparsed, leading to paniclog entries.
353 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
354 related applications. Previously an "H" was used where available info
355 says that "M" should be, so change to match.
357 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
358 as arguments, so an implementation trying to copy these into a local
359 buffer was taking a taint-enforcement trap. Fix by using dynamically
360 created buffers. Similar fix for radius expansion condition.
362 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
363 reasonable, eg. to count headers. Fix by using dynamically created
364 buffers rather than a local. Do similar fixes for ACL actions "dcc",
365 "log_reject_target", "malware" and "spam"; the arguments are expanded
366 so could be handling tainted values.
368 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
369 broken the (no-op) support for this sendmail command. Restore it
370 to doing nothing, silently, and returning good status.
372 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
373 record path was given (or the default used) without a leading directory
374 path, an error occurred on trying to open it. Use the transport's working
377 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
378 transport. Previously it was the name on the DNS A-record; use instead
379 the head of the CNAME chain leading there (if there is one). This seems
380 to align better with RFC 6125.
382 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
383 smtp_accept_max_per_host allocated resources which were not released
384 when the limit was exceeded. This eventually crashed the daemon. Fix
385 by adding a release action in that path.
387 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
388 expanded; previously using tainted values was rejected. Fix by using
389 dynamically-created buffers.
391 JH/09 Relax restrictions on ACL verify condition needing access to message
392 headers. Previously they were only permitted in data and non-smtp ACLs;
393 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
394 not_blind, header_sender and header_names_ascii verification.
396 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
397 Previously a macro used one argument twice; when called with the
398 argument as an expression having side-effects, incorrect operation
399 resulted. Use an inlineable function.
401 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
402 held open for a verify callout. Previously this wan not accounted for
403 and a corrupt onward SMTP conversation resulted.
405 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
406 Previously they were embedded naively in the constructed address; when
407 needed, strip the quoting and quote the entire local_part.
408 Also make the inbound_srs expansion condition handle quoting.
410 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
411 excluded, not matching the documentation.
413 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
414 was given for the sqlite_dbfile a trap resulted.
416 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
417 "name" argument resulted in a trap. There is no reason to disallow such;
418 this was a coding error.
420 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
421 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
422 spent suspended, ignoring the POSIX definition. Previously we assumed
423 it did and a constant offset from real time could be used as a correction.
424 Change to using the same clock source for the start-of-message and the
425 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
426 exists, just to get a clock slightly more aligned to reality.
428 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
429 RFC says it is optional some validators care. The missing char was not
430 intended but triggered by a line-wrap alignment. Discovery and fix by
431 Guillaume Outters, hacked on by JH.
433 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
434 name being quoted was tainted a trap would be taken. Fix by using
435 dynamically created buffers. The routine could have been called by a
436 rewrite with the "h" flag, by using the "-F" command-line option, or
437 by using a "name=" option on a control=submission ACL modifier.
439 JH/19 SPF: change the Authentication-Results expansion component to give
440 smtp.helo when the sender domain is empty. Previously it gave
443 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
444 not in 127.0.0.0/8 to help in spotting list domains taken over by a
445 domain-parking registrar.
447 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
448 Previously when a whitespace character was specified it was not inserted
449 after removing the newline.
451 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
452 the domain part of the recipient address. This overrides any tls_sni
453 option set, which was previously used.
455 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
458 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
459 is_tainted() had an off-by-one error in the overenthusiastic direction.
460 Find and fix by Gavan. Although NetBSD is not a supported platform for
461 4.94 this bug could affect other platforms.
463 PP/01 Fix default prime selection to be consistent.
464 One path used ike23 still, instead of exim.dev.20160529.3; now both
465 execution flows will use the same DH primes (currently
466 exim.dev.20160529.3).
468 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
469 option in smtp transport, to match the documentation. Previously
470 verification was not being done in some cases where it should have been.
472 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
473 than one server was defined and depending on the platform memory layout
474 details, an internal consistency trap could be hit while walking the list
477 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
478 files. The value is used by the authresults expansion item. Previously
479 if this was used in a router or transport, a crash could result.
481 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
482 have a sufficient incidence of EINTR returns from select that an
483 interaction with other operations done by the main daemon loop exposed
484 a bug in the error-handling. This was benign apart from the log
487 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
488 for consistency with delivered "=>" and failed "**" lines. While we're
489 there, handle PRX and TFO.
491 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
492 applied. This resulted, if any header-line rewrite rules were configured,
493 in a panic-log triggerable by sending a message with a long address in
494 a header. Fix by increasing the arbitrary limit to larger than a single
495 (dewrapped) 5322 header line maximum size.
497 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
498 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
499 RFC 6648 which deprecates X- options in protocols as a general practice.
500 Changeover between the implementations is handled by the mechanisms
503 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
504 a remote which uses callout/hold. Previously the local was queued.
506 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
509 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
510 log line. The asterisk is now consistently appended to the remote IP
511 (and port, if given), and will also be provided on defer and fail log
512 lines. Previously it could be placed on the local IP if that was being
513 logged, and was only provided on delivery lines.
515 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
517 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
518 authenticator client_send option. Previously the next char, after a pair
519 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
520 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
521 documentation. There is still no way to get a leading ^ immediately
522 after a NUL (ie. for the password of a PLAIN method authenticator.
524 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
525 files. For bad sizes read, delete the record and whine to paniclog.
527 JH/38 When logging an AUTH failure, as server, do not include sensitive
528 information. Previously, the credentials would be included if given
529 as part of the AUTH command line and an ACL denied authentication.
531 JH/39 Bug 2691: fix $local_part_data. When the matching list element
532 referred to a file, bad data was returned. This likely also affected
535 JH/40 The gsasl authenticator now supports caching of the salted password
536 generated by the client-side implementation. This required the addition
537 of a new variable: $auth4.
539 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
540 left undeleted; the attempt to re-create it then failed - resulting in
541 the usual "SIGHUP tp have daemon reload configuration" to not work.
542 This affected any platform not supporting "abstract" Unix-domain
543 sockets (i.e. not Linux).
545 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
546 recipients" response to RCPT in a later response, with a 250. The
547 previous coding assumed this would not happen, and under PIPELINING
548 would result in both lost and duplicate recipients for a message.
550 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
551 Previously the weighting was incorrectly applied. Similar fix for socks
552 proxies. Found and fixed by Heiko Schlichting.
554 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
555 not handle sub-lists included using the +namedlist syntax. While
556 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
557 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
560 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
561 Previously the permanent pool was used, so the sore could not be freed.
562 This meant a connection with many messages would use continually-growing
565 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
566 per-pool so as not to waste too much space. Previously a constant size
567 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
568 DOS attacks harder. The cost is wasted memory use in the larger blocks.
570 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
571 same space repeatedly, and a smaller process footprint.
573 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
574 for GnuTLS. Previously the name started "%s" which, while not a bug,
575 looked as if if might be one.
577 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
578 "continued connection") make the $proxy_* variables available. Previously
579 the information was not passed across the exec() call for subsequent
580 transport executions. This also mean that the log lines for the
581 messages can show the proxy information.
583 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
584 receive time. With modern systems the difference is significant.
585 The historical behaviour can be restored by disabling (a new) log_selector
586 "queue_time_exclusive".
588 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
589 content) were specifically tested for. Now, also cover expansions
590 resulting in ACL names and inline ACL content.
592 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
593 making it unusable in complex expressions.
595 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
596 fail with a taint-check on the spoolfile name, and leave the message
599 HS/01 Enforce absolute PID file path name.
601 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
603 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
605 PP/02 Bug 2643: Correct TLS DH constants.
606 A missing NUL termination in our code-generation tool had led to some
607 incorrect Diffie-Hellman constants in the Exim source.
608 Reported by kylon94, code-gen tool fix by Simon Arlott.
610 PP/03 Impose security length checks on various command-line options.
611 Fixes CVE-2020-SPRSS reported by Qualys.
613 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
614 better. Reported by Qualys.
616 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
617 providing a particularly obnoxious sender full name.
620 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
622 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
625 PP/08 Change default for recipients_max from unlimited to 50,000.
627 PP/09 Fix security issue with too many recipients on a message (to remove a
628 known security problem if someone does set recipients_max to unlimited,
629 or if local additions add to the recipient list).
630 Fixes CVE-2020-RCPTL reported by Qualys.
632 PP/10 Fix security issue in SMTP verb option parsing
633 Fixes CVE-2020-EXOPT reported by Qualys.
635 PP/11 Fix security issue in BDAT state confusion.
636 Ensure we reset known-good where we know we need to not be reading BDAT
637 data, as a general case fix, and move the places where we switch to BDAT
638 mode until after various protocol state checks.
639 Fixes CVE-2020-BDATA reported by Qualys.
641 HS/03 Die on "/../" in msglog file names
643 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
644 the Exim runtime user are allowed to create files.
646 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
649 QS/03 When reading the output from interpreted forward files we do not
650 pass the pipe between the parent and the interpreting process to
651 executed child processes (if any).
653 QS/04 Always die if requested from internal logging, even is logging is
656 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
657 the API; compilation noo longer completes with DMARC support included.
658 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
659 been reported on other platforms.
661 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
663 JH/56 Make the majority of info read from config files readonly, for defence-in-
664 depth against exploits. Suggestion by Qualys.
665 Not supported on Solaris 10.
667 JH/57 Fix control=fakreject for a custom message containing tainted data.
668 Previously this resulted in a log complaint, due to a re-expansion present
669 since fakereject was originally introduced.
671 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
672 modifier was given, a loop resulted.
674 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
675 pipelined SMTP command followed the BDAT LAST then it would be
676 incorrectly treated as part of the message body, causing a verification
679 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
680 require looser validation rules than those for 821-level addresses,
681 which only permit IP addresses.
687 JH/01 Avoid costly startup code when not strictly needed. This reduces time
688 for some exim process initialisations. It does mean that the logging
689 of TLS configuration problems is only done for the daemon startup.
691 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
693 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
694 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
697 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
698 DKIM signing is being done. Previously a transport_filter would always
699 disable CHUNKING, falling back to traditional DATA.
701 JH/05 Regard command-line recipients as tainted.
703 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
705 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
706 PAM library frees one of the arguments given to it, despite the
707 documentation. Therefore a plain malloc must be used.
709 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
710 on-stack buffers were used, resulting in a taint trap when DSN information
711 copied from a received message was written into the buffer.
713 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
714 the ordering of its ARC headers. This caused a crash.
716 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
717 a new record was being constructed with information from the peer, a trap
720 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
721 installation would get error messages from DMARC verify, when it hit the
722 nonexistent file indicated by the default. Distros wanting DMARC enabled
723 should both provide the file and set the option.
724 Also enforce no DMARC verification for command-line sourced messages.
726 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
727 could, depending on the platform, hang at the STARTTLS response.
729 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
730 message on a connection. Previously if one message had ARC headers and
731 the following one did not, a crash could result when adding an
732 Authentication-Results: header.
734 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
735 Exim main code and Exim-related utities. The introduction of taint
736 tracking also did many adjustments to string handling. Since then, eximon
737 frequently terminated with an assert failure.
739 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
740 check for 452 responses. This slightly helps the inefficieny of doing
741 a large alias-expansion into a recipient-limited target. The max_rcpt
742 transport option still applies (and at the current default, will override
743 the new feature). The check is done for either cause of synch, and forces
744 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
745 connection. The new facility is not tunable at this time.
747 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
748 library live data was being used, so the results became garbage. Make
749 copies while it is still usable.
751 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
752 on delivery deferred (==) and failed (**) lines (if a delivery was
753 attemtped). Previously it was only on completion (=>) lines.
755 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
756 for the expansion of the server_scram_iter and server_scram_salt options.
758 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
759 are now specifically given a NO_DATA response without hitting the system
760 resolver. The library goes on to do the now-standard TXT lookup.
761 Use of dnsdb lookups is not affected.
763 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
764 only retrieve the errormessage once. Previously two calls to dlerror()
765 were used, and the second one (for mainlog/paniclog) retrieved null
768 JH/20 Taint checking: disallow use of tainted data for
769 - the appendfile transport file and directory options
770 - the pipe transport command
771 - the autoreply transport file, log and once options
772 - file names used by the redirect router (including filter files)
774 - paths used by single-key lookups
775 Previously this was permitted.
777 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
778 adjusted the size of a major service buffer; this failed because the
779 buffer was in use at the time. Change to a compile-time increase in the
780 buffer size, when this authenticator is compiled into exim.
782 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
783 previous fast-mode was untenable in the face of glibs using mmap to
784 support larger malloc requests.
786 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
787 New values supported, if defined on system where compiled:
788 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
789 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
791 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
792 running a limited number of proceses in parallel, a benefit is gained. The
793 amount varies with the platform hardware and load. The use of the option
794 queue_run_in_order means we cannot do this, as ordering becomes
797 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
798 had introduced a string-copy (for ensuring NUL-termination) which was not
799 appropriate for that case, which can include embedded NUL bytes in the
800 block of data. Investigation showed the copy to actually be needless, the
801 data being length-specified.
803 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
804 done during a receiving connection, and both used TLS, global info was
805 used rather than per-connection info for tracking the state of data
806 queued for transmission. This could result in a connection hang.
808 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
809 Previously, when delivering serveral messages down a single connection
810 only the first would provide a SIZE. This was due to the size information
811 not being properly tracked.
813 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
814 TAI (at 37 seconds currently), pretend to be in UTC for time-related
815 expansion and logging. Previously, spurious values such as a future
816 minute could be seen.
818 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
819 it could crash from a null-deref. This could also affect the
820 ${addresses: } operator and ${readsock } item.
822 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
823 message following a mime one, the variable was not reset.
825 JH/30 When an pipelined-connect fails at the first response, assume incorrect
826 cached capability (perhaps the peer reneged?) and immediately retry in
829 JH/31 Fix spurious detection of timeout while writing to transport filter.
831 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
832 an attempt to copy the string was made before checking it.
834 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
835 taint of the lookup key was maintained; we now regard the presence in the
836 filesystem as sufficient validation.
838 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
839 argument is supplied.
841 JH/35 The dsearch lookup now requires that the directory is an absolute path.
842 Previously this was not checked, and nonempty relative paths made an
843 access under Exim's current working directory.
845 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
846 Previously no event was raised.
848 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
849 parameter supplied by the sender MAIL FROM command. Previously it was
850 ignored, and only the check_spool_space option value for the required
853 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
854 the size of the signing public-key. Previously it was instead giving
855 the size of the signature hash.
857 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
858 the default. See the (new) dkim_verify_min_keysizes option.
860 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
861 and an ACL use a lookup for checking either the local_part or domain,
862 stale data could be accessed. Ensure that variable references are
863 dropped between messages.
865 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
866 by the client was not checked as pointing within response data before
867 being used. A malicious client could thus cause an out-of-bounds read and
868 possibly gain authentication. Fix by adding the check.
870 JH/42 Internationalisation: change the default for downconversion in the smtp
871 transport to be "if needed". Previously it was "as previously set" for
872 the message, which usually meant "if needed" for message-submission but
873 "no" for everything else. However, MTAs have been seen using SMTPUTF8
874 even when the envelope addresses did not need it, resulting in forwarding
875 failures to non-supporting MTAs. A downconvert in such cases will be
876 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
877 transport. The change does mean that addresses needing conversion will
878 be converted when previously a delivery failure would occur.
880 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
881 response was received it would be used unchecked in a fail-DSN, violating
882 standards on line-length limits. Truncate if needed.
884 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
885 doesn't work. (Additionally add a new main config option to configure the
892 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
893 side, to decode a TLS 1.3 packet capture.
895 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
896 Previously the default library behaviour applied, sending two, each in
899 JH/03 Debug output for ACL now gives the config file name and line number for
902 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
904 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
906 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
907 buffer overrun for (non-chunking) other transports.
909 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
910 TLS1.3, means that a server rejecting a client certificate is not visible
911 to the client until the first read of encrypted data (typically the
912 response to EHLO). Add detection for that case and treat it as a failed
913 TLS connection attempt, so that the normal retry-in-clear can work (if
914 suitably configured).
916 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
917 and/or domain. Found and fixed by Jason Betts.
919 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
920 configuration). If a CNAME target was not a wellformed name pattern, a
923 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
924 the OS reports them interleaved with other addresses.
926 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
927 used both for input and for a verify callout, both encrypted, SMTP
928 responses being sent by the server could be lost. This resulted in
929 dropped connections and sometimes bounces generated by a peer sending
932 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
933 string. Previously it was possible to cause undefined behaviour in a
934 library routine (usually a crash). Found by "zerons".
936 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
939 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
940 API was removed, so update to use the newer ones.
942 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
943 any timeout set, is taking a long time. Previously we would hang on to a
944 rotated logfile "forever" if the input was arriving with long gaps
945 (a previous attempt to fix addressed lack, for a long time, of initial
948 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
949 shared (NFS) environment. The length of the tempfile name is now
950 4 + 16 ("hdr.$message_exim_id") which might break on file
951 systems which restrict the file name length to lower values.
954 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
955 shared (NFS) environment.
957 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
958 did for all versions <4.90). Notably -M, -m, --invert, -I may be
961 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
962 on some platforms for bit 31.
964 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
965 to changes apparently associated with TLS1.3 handling some of the APIs
966 previously used were either nonfunctional or inappropriate. Strings
967 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
968 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
969 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
970 This affects log line X= elements, the $tls_{in,out}_cipher variables,
971 and the use of specific cipher names in the encrypted= ACL condition.
973 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
975 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
976 verification result was not updated unless hosts_require_ocsp applied.
978 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
979 queue_list_requires_admin set to false, non-admin users were denied the
982 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
983 directory-of-certs mode. Previously they were advertised despite the
986 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
987 A single TCP connection by a client will now hold a TLS connection open
988 for multiple message deliveries, by default. Previously the default was to
991 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
992 default. If built with the facility, DANE will be used. The facility
993 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
995 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
996 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
997 must be defined and you must still, unless you define DISABLE_TLS, manage
998 the the include-dir and library-file requirements that go with that
999 choice. Non-TLS builds are still supported.
1001 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1004 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1005 default. If the platform supports and has the facility enabled, it will
1006 be requested on all coneections.
1008 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1009 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1011 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1013 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1014 success-DSN messages. Previously the From: header was always the default
1015 one for these; the option was ignored.
1017 JH/28 Fix the timeout on smtp response to apply to the whole response.
1018 Previously it was reset for every read, so a teergrubing peer sending
1019 single bytes within the time limit could extend the connection for a
1020 long time. Credit to Qualsys Security Advisory Team for the discovery.
1022 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1023 delivery address, which leaked information of the results of local
1024 forwarding. Change to the original envelope recipient address, per
1027 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1028 requested. Previously not bounce was generated and a log entry of
1029 error ignored was made.
1031 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1033 JH/32 Introduce a general tainting mechanism for values read from the input
1034 channel, and values derived from them. Refuse to expand any tainted
1035 values, to catch one form of exploit.
1037 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1038 was unused and the unexpanded text used for the test. Found and
1039 fixed by Ruben Jenster.
1041 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1042 an attempt to use a TLS library read routine dereffed a nul pointer,
1045 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1046 them in our smtp response.
1048 JH/36 Have the general router option retry_use_local_part default to true when
1049 any of the restrictive preconditions are set (to anything). Previously it
1050 was only for check_local user. The change removes one item of manual
1051 configuration which is required for proper retries when a remote router
1052 handles a subset of addresses for a domain.
1054 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1055 link count into consideration.
1057 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1058 caused the extension of big_buffer, the following lines were ignored.
1060 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1061 accordance with RFC 2308. Previously there was no expiry, so a longlived
1062 receive process (eg. due to ACL delays) versus a short SOA value could
1065 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1067 JH/39 Promote DMARC support to mainline.
1069 JH/40 Bug 2452: Add a References: header to DSNs.
1071 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1072 parameters. The relevant library call is documented as "Deprecated: This
1073 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1074 3.6.0, DH parameters are negotiated following RFC7919."
1076 HS/06 Change the default of dnssec_request_domains to "*"
1078 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1079 carried on and emitted a BDAT command, even when PIPELINING was not
1082 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1083 buffer was used for the filename, resulting in a trap when tainted
1084 arguments (eg. $domain) were used.
1086 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1087 recommended to avoid a possible server-load attack. The feature can be
1088 re-enabled via the openssl_options main cofiguration option.
1090 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1091 for version 4.90 - adding a "more data" boolean to the arguments.
1092 Bumped the ABI version number also, this having been missed previously;
1093 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1094 respect of usage of smtp_printf() by either local_scan code or libraries
1095 accessed via the ${dlfunc } expansion item. Both will need coding
1096 adjustment for any calls to smtp_printf() to match the new function
1097 signature; a FALSE value for the new argument is always safe.
1099 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1100 the file-offset (which the Linux syscall does, and exim expects); this
1101 resulted in an indefinite loop.
1103 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1104 to do ARC verification. The Authentication-Results: header line added
1105 by the configuration then had no ARC item.
1107 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1108 an incorrect mode was used for reading the result, resulting in it being
1115 JH/01 Remove code calling the customisable local_scan function, unless a new
1116 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1118 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1119 non-signal-safe functions being used.
1121 JH/03 Bug 2269: When presented with a received message having a stupidly large
1122 number of DKIM-Signature headers, disable DKIM verification to avoid
1123 a resource-consumption attack. The limit is set at twenty.
1125 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1126 report of oldest_pass in ${authres } in consequence, and separate out
1127 some descriptions of reasons for verification fail.
1129 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1130 files in the spool were present and unlocked. A queue-runner could spot
1131 them, resulting in a duplicate delivery. Fix that by doing the unlock
1132 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1133 add more error-checking on spoolfile handling while that code is being
1136 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1137 No known attacks, no CVE, this is defensive hardening.
1139 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1140 a queue-runner could start a delivery while other operations were ongoing.
1141 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1142 Found and investigated by Tim Stewart. Fix by using the open message data
1143 file handle rather than opening another, and not locally closing it (which
1144 releases a lock) for that case, while creating the temporary .eml format
1145 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1147 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1148 $sender_verify_failure/$recipient_verify_failure to "random".
1150 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1153 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1154 Previously this would segfault.
1156 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1159 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1160 like zero, since the resolver should be doing this for us, But we need one
1161 as a CNAME but no MX presence gets the CNAME returned; we need to check
1162 that doesn't point to an MX to declare it "no MX returned" rather than
1163 "error, loop". A new main option is added so the older capability of
1164 following some limited number of chain links is maintained.
1166 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1168 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1169 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1170 now-deprecated earlier definitions used only specified the range up to TLS
1171 1.2 (in the older-version library docs).
1173 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1175 JH/15 Rework TLS client-side context management. Stop using a global, and
1176 explicitly pass a context around. This enables future use of TLS for
1177 connections to service-daemons (eg. malware scanning) while a client smtp
1178 connection is using TLS; with cutthrough connections this is quite likely.
1180 JH/16 Fix ARC verification to do AS checks in reverse order.
1182 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1184 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1185 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1186 Previously the "utf8" would be re-prepended for every additional message.
1188 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1189 Previously thery were accepted, resulting in issues when attempting to
1190 forward messages to a non-supporting MTA.
1192 PP/02 Let -n work with printing macros too, not just options.
1194 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1195 one parent address was copied, and bogus data was used at delivery-logging
1196 time. Either a crash (after delivery) or bogus log data could result.
1197 Discovery and analysis by Tim Stewart.
1199 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1200 Previously if the string ended mid-character, we did not insert the
1201 promised '?' replacement.
1203 PP/04 Documentation: current string operators work on bytes, not codepoints.
1205 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1206 should pack well giving a smaller memory footprint so better caching and
1207 therefore performance. Group the declarations where this can't be done so
1208 that the byte-sized flag variables are not interspersed among pointer
1209 variables, giving a better chance of good packing by the compiler.
1211 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1212 non-null, to avoid issues with sites running BATV. Previously reports were
1213 sent with an empty envelope sender so looked like bounces.
1215 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1216 The ignore_error flag wasn't being returned from the filter subprocess so
1217 was not set for later routers. Investigation and fix by Matthias Kurz.
1219 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1220 and a msg:complete for the whole, when a message is manually removed using
1221 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1223 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1224 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1225 hope that is portable enough.
1227 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1228 requiring a known-CA anchor certificate; make it now rely entirely on the
1229 TLSA as an anchor. Checking the name on the leaf cert against the name
1230 on the A-record for the host is still done for TA (but not for EE mode).
1232 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1233 included in delivery lines for non-proxied connections, when compiled with
1234 SUPPORT_SOCKS and running with proxy logging enabled.
1236 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1237 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1238 move the existing event to fire before the normal logging of message
1239 failure so that custom logging is bracketed by normal logging.
1241 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1242 msg:fail:internal event. Developement by Matthias Kurz.
1244 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1245 far too small for todays use of crypto signatures stored there. Go all
1246 the way to the max DNS message size of 64kB, even though this might be
1247 overmuch for IOT constrained device use.
1249 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1250 copy a string over itself. The library routine is documented as not
1251 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1253 JH/32 For main options check_spool_space and check_inode_space, where the
1254 platform supports 64b integers, support more than the previous 2^31 kB
1255 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1256 the previous G, M, k.
1258 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1259 $authenticated_fail_id variable on authentication failure. Previously
1262 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1263 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1264 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1265 GNUTLS_SEC_PARAM_MEDIUM.
1267 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1268 side. Previously we would continue as if no SNI had been received.
1270 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1271 "<" character, which should be a list-separator specification, we walked
1272 off past the nul-terimation.
1274 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1275 causes) even when the retry time is not yet met. Previously they were
1276 not, meaning that when (say) an account was over-quota and temp-rejecting,
1277 and multiple senders' messages were queued, only one sender would get
1278 notified on each configured delay_warning cycle.
1280 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1282 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1283 cipher-suites, an error can be left on the stack even for a succeeding
1284 accept; this results in impossible error messages when a later operation
1287 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1288 return error codes indicating retry. Under TLS1.3 this becomes required.
1290 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1291 it only wrote the new authenticators, resulting in a lack of tracking of
1292 peer changes of ESMTP extensions until the next cache flush.
1294 JH/41 Fix the loop reading a message header line to check for integer overflow,
1295 and more-often against header_maxsize. Previously a crafted message could
1296 induce a crash of the recive process; now the message is cleanly rejected.
1298 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1299 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1305 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1306 When redis_servers is set to a list of > 1 element, and the Redis servers
1307 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1308 case of MOVED into a DEFER case instead, thus moving the query onto the
1309 next server in the list. For a cluster of N elements, all N servers must
1310 be defined in redis_servers.
1312 GF/02 Catch and remove uninitialized value warning in exiqsumm
1313 Check for existence of @ARGV before looking at $ARGV[0]
1315 JH/01 Replace the store_release() internal interface with store_newblock(),
1316 which internalises the check required to safely use the old one, plus
1317 the allocate and data copy operations duplicated in both (!) of the
1318 extant use locations.
1320 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1321 modifier. This matches the restriction on the commandline.
1323 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1324 Previously only the last row was returned.
1326 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1327 we assumed that tags in the header were well-formed, and parsed the
1328 element content after inspecting only the first char of the tag.
1329 Assumptions at that stage could crash the receive process on malformed
1332 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1333 While running the DKIM ACL we operate on the Permanent memory pool so that
1334 variables created with "set" persist to the DATA ACL. Also (at any time)
1335 DNS lookups that fail create cache records using the Permanent pool. But
1336 expansions release any allocations made on the current pool - so a dnsdb
1337 lookup expansion done in the DKIM ACL releases the memory used for the
1338 DNS negative-cache, and bad things result. Solution is to switch to the
1339 Main pool for expansions.
1340 While we're in that code, add checks on the DNS cache during store_reset,
1341 active in the testsuite.
1342 Problem spotted, and debugging aided, by Wolfgang Breyha.
1344 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1345 When none of the hosts presented to a transport match an already-open
1346 connection, close it and proceed with the list. Previously we would
1347 queue the message. Spotted by Lena with Yahoo, probably involving
1350 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1351 Previously a spurious "250 OK id=" response was appended to the proper
1354 JH/08 The "support for" informational output now, which built with Content
1355 Scanning support, has a line for the malware scanner interfaces compiled
1356 in. Interface can be individually included or not at build time.
1358 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1359 by the template makefile "src/EDITME". The "STREAM" support for an older
1360 ClamAV interface method is removed.
1362 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1363 rows affected is given instead).
1365 JH/11 The runtime Berkeley DB library version is now additionally output by
1366 "exim -d -bV". Previously only the compile-time version was shown.
1368 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1369 SMTP connection. Previously, when one had more recipients than the
1370 first, an abortive onward connection was made. Move to full support for
1371 multiple onward connections in sequence, handling cutthrough connection
1372 for all multi-message initiating connections.
1374 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1375 routers. Previously, a multi-recipient message would fail to match the
1376 onward-connection opened for the first recipient, and cause its closure.
1378 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1379 a timeout on read on a GnuTLS initiating connection, resulting in the
1380 initiating connection being dropped. This mattered most when the callout
1381 was marked defer_ok. Fix to keep the two timeout-detection methods
1384 JH/15 Relax results from ACL control request to enable cutthrough, in
1385 unsupported situations, from error to silently (except under debug)
1386 ignoring. This covers use with PRDR, frozen messages, queue-only and
1389 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1391 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1392 metadata, resulting in a crash in free().
1394 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1395 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1396 Broken also in d185889f4, with init system revamp.
1398 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1399 Previously we abruptly closed the connection after reading a malware-
1400 found indication; now we go on to read the "scan ok" response line,
1403 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1404 ACL. Previously, a crash would result.
1406 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1407 macro text after a replacement (previously it was only once per line) and
1408 by skipping builtin macros when searching for an uppercase lead character.
1410 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1411 for the build is renamed.
1413 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1414 was allocated for every new TLS startup, meaning one per message. Fix
1415 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1417 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1418 reported the original. Fix to report (as far as possible) the ACL
1419 result replacing the original.
1421 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1422 OpenSSL. Certificate information is loaded for every new TLS startup,
1423 and the resources needed to be freed.
1425 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1427 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1430 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1431 DATA response info to the (existing) per-recipient response info for
1432 the "C=" log element. It can have useful tracking info from the
1433 destination system. Patch from Simon Arlott.
1435 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1436 length value. Previously this would segfault.
1438 HS/02 Support Avast multiline protoocol, this allows passing flags to
1439 newer versions of the scanner.
1441 JH/28 Ensure that variables possibly set during message acceptance are marked
1442 dead before release of memory in the daemon loop. This stops complaints
1443 about them when the debug_store option is enabled. Discovered specifically
1444 for sender_rate_period, but applies to a whole set of variables.
1445 Do the same for the queue-runner and queue-list loops, for variables set
1446 from spool message files. Do the same for the SMTP per-message loop, for
1447 certain variables indirectly set in ACL operations.
1449 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1450 as a multi-recipient message from a mailinglist manager). The coding had
1451 an arbitrary cutoff number of characters while checking for more input;
1452 enforced by writing a NUL into the buffer. This corrupted long / fast
1453 input. The problem was exposed more widely when more pipelineing of SMTP
1454 responses was introduced, and one Exim system was feeding another.
1455 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1456 receiving system, and refused recipients seen by the sending system
1457 (propating to people being dropped from mailing lists).
1458 Discovered and pinpointed by David Carter.
1460 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1461 replaced by the ${authresults } expansion.
1463 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1465 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1466 allows proper process termination in container environments.
1468 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1469 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1471 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1472 and "err_perm", deprecated since 4.83 when the RFC-defined words
1473 "temperror" and "permerror" were introduced.
1475 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1476 transport-filters or DKIM-signing. The restriction was lost in the
1477 consolidation of verify-callout and delivery SMTP handling.
1478 Extend the restriction to also cover ARC-signing.
1480 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1481 in defer=pass mode supply a 450 to the initiator. Previously the message
1484 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1485 tls_require_ciphers is used as before.
1487 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1488 "pass_unscanned". Only tmpfails from the scanner are written to
1489 the paniclog, as they may require admin intervention (permission
1490 denied, license issues). Other scanner errors (like decompression
1491 bombs) do not cause a paniclog entry.
1493 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1494 Previously it was possible to log spurious information in receive log
1497 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1498 triggered odd behaviour from Outlook Express clients.
1500 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1503 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1504 since the IETF WG has not yet settled on that versus the original
1505 "bare" representation.
1507 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1508 Previously the millisecond value corrupted the output.
1509 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1510 corrupted the output.
1516 JH/01 Rework error string handling in TLS interface so that the caller in
1517 more cases is responsible for logging. This permits library-sourced
1518 string to be attached to addresses during delivery, and collapses
1519 pairs of long lines into single ones.
1521 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1522 during configuration. Wildcards are allowed and expanded.
1524 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1525 This permits better logging.
1527 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1528 do not close it down and have the child transport start it up again on
1529 the passed-on TCP connection. Instead, proxy the child (and any
1530 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1531 affected: the continued delivery log lines do not have any DNSSEC, TLS
1532 Certificate or OCSP information. TLS cipher information is still logged.
1534 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1535 identical IP addresses on different listening ports. Will also affect
1538 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1539 add noisy ifdef guards to special-case this sillyness.
1540 Patch from Bernd Kuhls.
1542 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1543 than 255 are no longer allowed.
1545 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1546 Disable the session-cache too, which might reduce our load. Since we
1547 currrectly use a new context for every connection, both as server and
1548 client, there is no benefit for these.
1549 GnuTLS appears to not support tickets server-side by default (we don't
1550 call gnutls_session_ticket_enable_server()) but client side is enabled
1551 by default on recent versions (3.1.3 +) unless the PFS priority string
1554 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1555 <https://reproducible-builds.org/specs/source-date-epoch/>.
1557 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1558 the check for any unsuccessful recipients did not notice the limit, and
1559 erroneously found still-pending ones.
1561 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1562 MSG_MORE. Only in-clear (not on TLS connections).
1564 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1565 transport-filter is involved we can buffer the headers in memory for
1566 creating the signature, and read the spool data file once for the
1567 signature and again for transmission.
1569 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1570 4.77 as the kernel support then wasn't solid, having issues in 64bit
1571 mode. Now, it's been long enough. Add support for FreeBSD also.
1573 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1574 case where the routing stage had gathered several addresses to send to
1575 a host before calling the transport for the first, we previously failed
1576 to close down TLS in the old transport process before passing the TCP
1577 connection to the new process. The new one sent a STARTTLS command
1578 which naturally failed, giving a failed delivery and bloating the retry
1579 database. Investigation and fix prototype from Wolfgang Breyha.
1581 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1582 false-negatives in the check that the sender had not preempted a response
1583 or prompt from Exim (running as a server), due to that code's lack of
1584 awareness of the SMTP input buffering.
1586 PP/04 Add commandline_checks_require_admin option.
1587 Exim drops privileges sanely, various checks such as -be aren't a
1588 security problem, as long as you trust local users with access to their
1589 own account. When invoked by services which pass untrusted data to
1590 Exim, this might be an issue. Set this option in main configuration
1591 AND make fixes to the calling application, such as using `--` to stop
1594 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1595 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1596 responses to those, into a single TLS record each way (this usually means
1597 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1600 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1601 affects you only if you're dancing at the edge of the param size limits.
1602 If you are, and this message makes sense to you, then: raise the
1603 configured limit or use OpenSSL 1.1. Nothing we can do for older
1606 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1607 cmdline element to get the documented default one. Previously it was
1610 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1611 are unsuable for non-message-specific reasons - eg. connection timeout,
1612 banner-time rejection.
1614 JH/16 Fix logging of delivery remote port, when specified by router, under
1617 PP/06 Repair manualroute's ability to take options in any order, even if one
1618 is the name of a transport.
1621 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1623 JH/17 Change the list-building routines interface to use the expanding-string
1624 triplet model, for better allocation and copying behaviour.
1626 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1627 Previously it was constructed the first time a possibly-matching string
1628 was met in the configuration file input during startup; now it is done
1631 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1632 compatible one, to avoid the (poorly documented) possibility of a config
1633 file in the working directory redirecting the DB files, possibly correpting
1634 some existing file. CVE-2017-10140 assigned for BDB.
1636 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1637 cache-hot. Previously, although the result was properly cached, the
1638 initial verify call returned a defer.
1640 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1641 the main verify for receipient in uncached-mode.
1643 JH/22 Retire historical build files to an "unsupported" subdir. These are
1644 defined as "ones for which we have no current evidence of testing".
1646 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1647 if present. Previously it was ignored.
1649 JH/24 Start using specified-initialisers in C structure init coding. This is
1650 a C99 feature (it's 2017, so now considered safe).
1652 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1653 if was a fixed-sized field and bitmask ops via macros; it is now more
1656 PP/07 GitHub PR 56: Apply MariaDB build fix.
1657 Patch provided by Jaroslav Škarvada.
1659 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1660 during Coverity cleanups [4.87 JH/47]
1661 Diagnosis and fix provided by Michael Fischer v. Mollard.
1663 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1664 the right size to place the terminating semicolon on its own folded
1665 line, the header hash was calculated to an incorrect value thanks to
1666 the (relaxed) space the fold became.
1668 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1669 and confused the parent.
1671 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1672 which could crash as a result. This could lead to undeliverable messages.
1674 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1677 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1678 matters on fast-turnover and PID-randomising systems, which were getting
1679 out-of-order delivery.
1681 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1682 a possibly-overlapping copy. The symptom was that "Remote host closed
1683 connection in response to HELO" was logged instead of the actual 4xx
1686 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1687 Previously only that bufferd was discarded, resulting in SYMTP command
1690 JH/32 DKIM: when a message has multiple signatures matching an identity given
1691 in dkim_verify_signers, run the dkim acl once for each. Previously only
1692 one run was done. Bug 2189.
1694 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1695 "panic the current process" to "deliberately defer". The panic log is
1696 still written with the problem list name; the mail and reject logs now
1697 get a temp-reject line for the message that was being handled, saying
1698 something like "domains check lookup or other defer". The SMTP 451
1699 message is still "Temporary local problem".
1701 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1702 A crafted sequence of BDAT commands could result in in-use memory beeing
1703 freed. CVE-2017-16943.
1705 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1706 from SMTP input. Previously it was always done; now only done for DATA
1707 and not BDAT commands. CVE-2017-16944.
1709 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1710 to the message (such as an overlong header line). Previously this was
1711 not done and we did not exit BDAT mode. Followon from the previous item
1712 though a different problem.
1718 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1719 than -2003 did; needs libidn2 in addition to libidn.
1721 JH/02 The path option on a pipe transport is now expanded before use.
1723 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1724 Patch provided by "Björn", documentation fix added too.
1726 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1727 missing a wire-to-host endian conversion.
1729 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1730 close after a BDAT command line could be taken as a following command,
1731 giving a synch failure. Fix by only checking for synch immediately
1732 before acknowledging the chunk.
1734 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1735 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1736 macro. Patches provided by Josh Soref.
1738 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1739 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1740 by IANA as a service extension. However, John Klensin suggests that we
1743 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1744 the dkim code may be unix-mode line endings rather than smtp wire-format
1745 CRLF, so prepend a CR to any bare LF.
1747 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1748 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1750 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1751 the first were themselves being wrongly included in the feed into dkim
1752 processing; with most chunk sizes in use this resulted in an incorrect
1753 body hash calculated value.
1755 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1756 DKIM signature block, for verification. Although advised against by
1757 standards it is specifically not ruled illegal.
1759 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1761 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1762 missing a body hash (the bh= tag).
1764 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1765 It seems that HAProxy sends the Proxy Protocol information in clear and
1766 only then does a TLS startup, so do the same.
1768 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1769 TCP connections (such as for Spamd) unless the daemon successfully set
1770 Fast Open mode on its listening sockets. This fixes breakage seen on
1771 too-old kernels or those not configured for Fast Open, at the cost of
1772 requiring both directions being enabled for TFO, and TFO never being used
1773 by non-daemon-related Exim processes.
1775 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1776 endings, at least on the first header line. Try to canonify any that get
1777 past that check, despite the cost.
1779 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1780 now limited to an arbitrary five deep, while parsing addresses with the
1781 strip_excess_angle_brackets option enabled.
1783 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1784 instead leave the unprompted TLS handshake in socket buffer for the
1785 TLS library to consume.
1787 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1789 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1791 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1792 region in one operation, for speed, and this leaves assigned pointers
1793 dangling. Add checks run only under the testsuite which checks all
1794 variables at a store-reset and panics on a dangling pointer; add code
1795 explicitly nulling out all the variables discovered. Fixes one known
1796 bug: a transport crash, where a dangling pointer for $sending_ip_address
1797 originally assigned in a verify callout, is re-used.
1799 PP/06 Drop '.' from @INC in various Perl scripts.
1801 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1803 PP/08 Reduce a number of compilation warnings under clang; building with
1804 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1805 should be warning-free.
1807 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1809 HS/01 Fix portability problems introduced by PP/08 for platforms where
1810 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1812 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1813 chunk. This allows us to accept broken chunked messages. We need a more
1814 general solution here.
1816 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1817 already-broken messages in the queue.
1819 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1821 JH/19 Fix reference counting bug in routing-generated-address tracking.
1827 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1828 supports it and a size is available (ie. the sending peer gave us one).
1830 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1831 years of being deprecated). The replacements are the ACLs
1832 acl_smtp_mime and acl_not_smtp_mime.
1834 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1835 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1836 or even an in-clear connection were permitted. Now, if the host lookup
1837 was dnssec and dane was requested then the host is only used if the
1838 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1839 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1840 if one fails this test.
1841 This means that a poorly-configured remote DNS will make it incommunicado;
1842 but it protects against a DNS-interception attack on it.
1844 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1845 non-noisy when a race steals the message being considered.
1847 JH/05 If main configuration option tls_certificate is unset, generate a
1848 self-signed certificate for inbound TLS connections.
1850 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1851 in rewrites and routers.
1853 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1854 and logged a warning sing 4.83; now they are a configuration file error.
1856 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1857 (lacking @domain). Apply the same qualification processing as RCPT.
1859 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1861 JH/10 Support ${sha256:} applied to a string (as well as the previous
1864 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1865 a cutthrough deliver is pending, as we always want to make a connection.
1866 This also avoids re-routing the message when later placing the cutthrough
1867 connection after a verify cache hit.
1868 Do not update it with the verify result either.
1870 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1871 when routing results in more than one destination address.
1873 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1874 signing (which inhibits the cutthrough capability). Previously only
1875 the presence of an option was tested; now an expansion evaluating as
1876 empty is permissible (obviously it should depend only on data available
1877 when the cutthrough connection is made).
1879 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1880 the relevant preceding SMTP command did not note the pipelining mode.
1882 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1883 Previously they were not counted.
1885 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1886 as one having no matching records. Previously we deferred the message
1887 that needed the lookup.
1889 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1890 distinguished as "(=".
1892 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1893 for missing MX records. Previously it only worked for missing A records.
1895 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1897 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1898 after the data-go-ahead and data-ack. Patch from Jason Betts.
1900 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1901 even for a "none" policy. Patch from Tony Meyer.
1903 JH/22 Fix continued use of a connection for further deliveries. If a port was
1904 specified by a router, it must also match for the delivery to be
1907 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1908 When one of the recipients of a message was unsuitable for the connection
1909 (has no matching addresses), we lost track of needing to mark it
1910 deferred. As a result mail would be lost.
1912 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1914 JH/25 Decoding ACL controls is now done using a binary search; the source code
1915 takes up less space and should be simpler to maintain. Merge the ACL
1916 condition decode tables also, with similar effect.
1918 JH/26 Fix problem with one_time used on a redirect router which returned the
1919 parent address unchanged. A retry would see the parent address marked as
1920 delivered, so not attempt the (identical) child. As a result mail would
1923 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1924 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1925 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1928 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1929 Main config options check_{log,spool}_{inodes,space} are now
1930 100 inodes, 10MB unless set otherwise in the configuration.
1932 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1933 Previously it only applied to the main-section connection policy
1936 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1938 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1939 by me. Added RFC7919 DH primes as an alternative.
1941 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1942 are not in the system include path.
1944 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1945 GnuTLS, when a session startup failed (eg because the client disconnected)
1946 Exim did stdio operations after fclose. This was exposed by a recent
1947 change which nulled out the file handle after the fclose.
1949 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1950 signed directly by the cert-signing cert, rather than an intermediate
1951 OCSP-signing cert. This is the model used by LetsEncrypt.
1953 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1955 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1956 an incoming connection.
1958 HS/02 Bug 1802: Do not half-close the connection after sending a request
1961 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1962 fallback to "prime256v1".
1964 JH/34 SECURITY: Use proper copy of DATA command in error message.
1965 Could leak key material. Remotely exploitable. CVE-2016-9963.
1971 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1972 and 3.4.4 - once the server is enabled to respond to an OCSP request
1973 it does even when not requested, resulting in a stapling non-aware
1974 client dropping the TLS connection.
1976 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1977 support variable-length bit vectors. No functional change.
1979 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1980 The I= interface field on outgoing lines is now after the H= remote
1981 host field, same as incoming lines. There is a separate
1982 outgoing_interface log selector which allows you to disable the
1985 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1986 If not running log_selector +smtp_connection the mainlog would be held
1987 open indefinitely after a "too many connections" event, including to a
1988 deleted file after a log rotate. Leave the per net connection logging
1989 leaving it open for efficiency as that will be quickly detected by the
1990 check on the next write.
1992 HS/01 Bug 1671: Fix post transport crash.
1993 Processing the wait-<transport> messages could crash the delivery
1994 process if the message IDs didn't exist for some reason. When
1995 using 'split_spool_directory=yes' the construction of the spool
1996 file name failed already, exposing the same netto behaviour.
1998 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1999 mime_regex ACL conditions.
2001 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2002 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2003 to HELO, local diagnostic string.
2005 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2006 log line to debug. Even when configured with a tls authenticator many
2007 client connections are expected to not authenticate in this way, so
2008 an authenticate fail is not an error.
2010 HS/02 Add the Exim version string to the process info. This way exiwhat
2011 gives some more detail about the running daemon.
2013 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2014 matter for fast-change records such as DNSBLs.
2016 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2017 retry record, even if constant. There may be multiple transports with
2018 different interface settings and the retry behaviour needs to be kept
2021 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2023 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2025 JH/10 Bug 840: fix log_defer_output option of pipe transport
2027 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2028 hosts_randomize. This matters a lot when combined with mua_wrapper.
2030 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2031 ${quote_pgsql:<string>} operator.
2033 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2035 JH/14 Change header file name for freeradius-client. Relevant if compiling
2036 with Radius support; from the Gentoo tree and checked under Fedora.
2038 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2040 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2041 option transport_filter, meaning no filtering.
2043 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2045 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2046 defaults to "*" (all hosts). The variable is now available when not built
2047 with TLS, default unset, mainly to enable keeping the testsuite sane.
2048 If a server certificate is not supplied (via tls_certificate) an error is
2049 logged, and clients will find TLS connections fail on startup. Presumably
2050 they will retry in-clear.
2051 Packagers of Exim are strongly encouraged to create a server certificate
2052 at installation time.
2054 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2055 with the $config_file variable.
2057 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2058 in transport context, after the attempt, and per-recipient. The latter type
2059 is per host attempted. The event data is the error message, and the errno
2060 information encodes the lookup type (A vs. MX) used for the (first) host,
2061 and the trailing two digits of the smtp 4xx response.
2063 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2064 to write to mainlog (or rejectlog, paniclog) in the window between file
2065 creation and permissions/ownership being changed. Particularly affects
2066 installations where exicyclog is run as root, rather than exim user;
2067 result is that the running daemon panics and dies.
2069 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2071 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2072 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2073 "pri" and "weight". Note that the previous implicit priority given by the
2074 list order is no longer honoured.
2076 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2077 for DKIM processing.
2079 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2080 by defining SUPPORT_SOCKS.
2082 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2083 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2084 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2085 variables are renamed to proxy_{local,external}_{address,port}.
2087 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2088 for a build by defining SUPPORT_I18N
2090 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2091 of the query string, and make ${quote_redis:} do that quoting.
2093 JH/29 Move Events support from Experimental to mainline, enabled by default
2094 and removable for a build by defining DISABLE_EVENT.
2096 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2098 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2099 cached by the daemon.
2101 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2102 by defining LOOKUP_REDIS. The libhiredis library is required.
2104 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2105 keys are given for lookup.
2107 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2108 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2109 only supported when built with TLS support. The PolarSSL SHA routines
2110 are still used when the TLS library is too old for convenient support.
2112 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2113 openssl_options), for security. OpenSSL forces this from version 1.1.0
2114 server-side so match that on older versions.
2116 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2117 allocation for $value could be released as the expansion processing
2118 concluded, but leaving the global pointer active for it.
2120 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2121 and to use the domains and local_parts ACL conditions.
2123 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2124 incorrectly not doubled on cutthrough transmission, hence seen as a
2125 body-termination at the receiving system - resulting in truncated mails.
2126 Commonly the sender saw a TCP-level error, and retransmitted the message
2127 via the normal store-and-forward channel. This could result in duplicates
2128 received - but deduplicating mailstores were liable to retain only the
2129 initial truncated version.
2131 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2133 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2135 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2136 we're in there, support oversigning also; bug 1309.
2138 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2140 HS/04 Add support for keep_environment and add_environment options.
2142 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2143 either intentional arithmetic overflow during PRNG, or testing config-
2146 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2147 delivery resulted in actual delivery. Cancel cutthrough before DATA
2150 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2151 rejects a recipient: pass the reject to the originator.
2153 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2154 Many were false-positives and ignorable, but it's worth fixing the
2157 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2158 for the new environment-manipulation done at startup. Move the routines
2159 from being local to tls.c to being global via the os.c file.
2161 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2162 an extract embedded as result-arg for a map, the first arg for extract
2163 is unavailable so we cannot tell if this is a numbered or keyed
2164 extraction. Accept either.
2170 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2173 JH/02 The smtp transport option "multi_domain" is now expanded.
2175 JH/03 The smtp transport now requests PRDR by default, if the server offers
2178 JH/04 Certificate name checking on server certificates, when exim is a client,
2179 is now done by default. The transport option tls_verify_cert_hostnames
2180 can be used to disable this per-host. The build option
2181 EXPERIMENTAL_CERTNAMES is withdrawn.
2183 JH/05 The value of the tls_verify_certificates smtp transport and main options
2184 default to the word "system" to access the system default CA bundle.
2185 For GnuTLS, only version 3.0.20 or later.
2187 JH/06 Verification of the server certificate for a TLS connection is now tried
2188 (but not required) by default. The verification status is now logged by
2189 default, for both outbound TLS and client-certificate supplying inbound
2192 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2195 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2196 Status Notification (bounce) messages are now MIME format per RFC 3464.
2197 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2198 under the control of the dsn_advertise_hosts option, and routers may
2199 have a dsn_lasthop option.
2201 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2202 default, modifiable by a malware= option. The list separator for
2203 the options can now be changed in the usual way. Bug 68.
2205 JH/10 The smtp_receive_timeout main option is now expanded before use.
2207 JH/11 The incoming_interface log option now also enables logging of the
2208 local interface on delivery outgoing connections.
2210 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2211 if the interface and destination host and port all match.
2213 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2216 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2217 Patch from Andrew Lewis.
2219 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2220 now supports optional time-restrictions, weighting, and priority
2221 modifiers per server. Patch originally by <rommer@active.by>.
2223 JH/16 The spamd_address main option now supports a mixed list of local
2224 and remote servers. Remote servers can be IPv6 addresses, and
2225 specify a port-range.
2227 JH/17 Bug 68: The spamd_address main option now supports an optional
2228 timeout value per server.
2230 JH/18 Bug 1581: Router and transport options headers_add/remove can
2231 now have the list separator specified.
2233 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2236 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2239 JH/21 Support for the A6 type of dns record is withdrawn.
2241 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2242 rather than the verbs used.
2244 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2245 from 255 to 1024 chars.
2247 JH/24 Verification callouts now attempt to use TLS by default.
2249 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2250 are generic router options now. The defaults didn't change.
2252 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2253 Original patch from Alexander Shikoff, worked over by JH.
2255 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2256 files if ZCAT_COMMAND is not executable.
2258 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2260 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2262 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2263 Normally benign, it bites when the pair was led to by a CNAME;
2264 modern usage is to not canonicalize the domain to a CNAME target
2265 (and we were inconsistent anyway for A-only vs AAAA+A).
2267 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2269 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2270 when evaluating $sender_host_dnssec.
2272 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2273 $sender_helo_dnssec variable.
2275 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2277 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2279 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2281 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2282 documented as working, but never had. Support all but $spam_report.
2284 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2285 added for tls authenticator.
2287 HS/03 Add perl_taintmode main config option
2293 TL/01 When running the test suite, the README says that variables such as
2294 no_msglog_check are global and can be placed anywhere in a specific
2295 test's script, however it was observed that placement needed to be near
2296 the beginning for it to behave that way. Changed the runtest perl
2297 script to read through the entire script once to detect and set these
2298 variables, reset to the beginning of the script, and then run through
2299 the script parsing/test process like normal.
2301 TL/02 The BSD's have an arc4random API. One of the functions to induce
2302 adding randomness was arc4random_stir(), but it has been removed in
2303 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2304 function when detected.
2306 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2307 cause callback expansion.
2309 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2310 syntax errors in an expansion can be treated as a string instead of
2311 logging or causing an error, due to the internal use of bool_lax
2312 instead of bool when processing it.
2314 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2315 server certificates when making smtp deliveries.
2317 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2319 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2321 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2323 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2324 Merged patch from Sebastian Wiedenroth.
2326 JH/05 Fix results-pipe from transport process. Several recipients, combined
2327 with certificate use, exposed issues where response data items split
2328 over buffer boundaries were not parsed properly. This eventually
2329 resulted in duplicates being sent. This issue only became common enough
2330 to notice due to the introduction of connection certificate information,
2331 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2333 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2334 size buffer was used, resulting in syntax errors when an expansion
2337 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2338 version 3.3.6 or later.
2340 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2341 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2342 both become "event_action", the variables become $event_name, $event_data
2343 and $event_defer_errno. There is a new variable $verify_mode, usable in
2344 routers, transports and related events. The tls:cert event is now also
2345 raised for inbound connections, if the main configuration event_action
2348 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2349 early OCSP support, but no stapling (appears to be less than 1.0.0).
2351 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2352 server certificate names available under the smtp transport option
2353 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2356 JH/10 Time-related extraction expansions from certificates now use the main
2357 option "timezone" setting for output formatting, and are consistent
2358 between OpenSSL and GnuTLS compilations. Bug 1541.
2360 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2361 encoded parameter in the incoming message. Bug 1558.
2363 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2364 include certificate info, eximon was claiming there were spoolfile
2367 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2369 JH/14 Log delivery-related information more consistently, using the sequence
2370 "H=<name> [<ip>]" wherever possible.
2372 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2373 are problematic for Debian distribution, omit them from the release
2376 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2378 JH/16 Fix string representation of time values on 64bit time_t architectures.
2381 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2382 output list separator was used.
2387 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2388 checkers that were complaining about end of non-void function with no
2391 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2392 This was a regression introduced in 4.83 by another bugfix.
2394 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2396 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2397 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2403 TF/01 Correctly close the server side of TLS when forking for delivery.
2405 When a message was received over SMTP with TLS, Exim failed to clear up
2406 the incoming connection properly after forking off the child process to
2407 deliver the message. In some situations the subsequent outgoing
2408 delivery connection happened to have the same fd number as the incoming
2409 connection previously had. Exim would try to use TLS and fail, logging
2410 a "Bad file descriptor" error.
2412 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2413 utilities have not been installed.
2415 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2416 temporary space as the ACL may create new global variables.
2418 TL/01 LDAP support uses per connection or global context settings, depending
2419 upon the detected version of the libraries at build time.
2421 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2422 to extract and use the src ip:port in logging and expansions as if it
2423 were a direct connection from the outside internet. PPv2 support was
2424 updated based on HAProxy spec change in May 2014.
2426 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2428 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2429 Properly escape header and check for NULL return.
2431 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2434 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2436 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2437 characters in header names, implemented as a verify condition.
2438 Contributed by Michael Fischer v. Mollard.
2440 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2441 results permerror and temperror. Previous values are deprecated but
2442 still accepted. In a future release, err_perm and err_temp will be
2443 completely removed, which will be a backward incompatibility if the
2444 ACL tests for either of these two old results. Patch contributed by
2445 user bes-internal on the mailing list.
2447 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2449 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2450 selectors, in both main and reject logs.
2452 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2455 JH/07 Add malware type "sock" for talking to simple daemon.
2457 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2459 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2460 routers/transports under cutthrough routing.
2462 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2463 numbers. Touch up "bool" conditional to keep the same definition.
2465 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2467 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2469 JH/12 Expand items in router/transport headers_add or headers_remove lists
2470 individually rather than the list as a whole. Bug 1452.
2472 Required for reasonable handling of multiple headers_ options when
2473 they may be empty; requires that headers_remove items with embedded
2474 colons must have them doubled (or the list-separator changed).
2476 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2477 view the policy declared in the DMARC record. Currently, $dmarc_status
2478 is a combined value of both the record presence and the result of the
2481 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2483 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2484 dnslookup router and the smtp transport (applying to the forward
2487 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2488 of ldap servers used for a specific lookup. Patch provided by Heiko
2491 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2492 New variable $lookup_dnssec_authenticated for observability.
2494 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2495 Patch submitted by Lars Timman.
2497 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2499 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2500 Requires trusted mode and valid format message id, aborts otherwise.
2501 Patch contributed by Heiko Schlichting.
2503 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2504 certextract with support for various fields. Bug 1358.
2506 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2507 is requested by default, modifiable by smtp transport option
2510 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2511 operate on certificate variables to give certificate fingerprints
2512 Also new ${sha256:cert_variable}.
2514 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2516 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2517 Christian Aistleitner.
2519 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2521 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2522 file. Patch from Wolfgang Breyha.
2524 JH/25 Expand the coverage of the delivery $host and $host_address to
2525 client authenticators run in verify callout. Bug 1476.
2527 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2528 align with daemon_smtp_ports. Bug 72.
2530 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2531 support and error reporting did not work properly.
2533 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2534 and is readable. Patch from Andrew Colin Kissa.
2536 TL/14 Enhance documentation of ${run expansion and how it parses the
2537 commandline after expansion, particularly in the case when an
2538 unquoted variable expansion results in an empty value.
2540 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2542 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2543 by renaming away from it. Addresses GH issue 15, reported by
2546 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2548 TL/15 SECURITY: prevent double expansion in math comparison functions
2549 (can expand unsanitized data). Not remotely exploitable.
2556 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2558 PP/02 Make -n do something, by making it not do something.
2559 When combined with -bP, the name of an option is not output.
2561 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2564 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2565 $sender_host_name and config options to manage this, and basic check
2568 PP/05 DSCP support for outbound connections and control modifier for inbound.
2570 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2571 (Only plugin which currently uses this is kerberos4, which nobody should
2572 be using, but we should make it available and other future plugins might
2573 conceivably use it, even though it would break NAT; stuff *should* be
2574 using channel bindings instead).
2576 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2577 name; added for Sendmail compatibility; requires admin caller.
2578 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2579 just ignore it); requires trusted caller.
2580 Also parse but ignore: -Ac -Am -X<logfile>
2583 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2585 TL/02 Add +smtp_confirmation as a default logging option.
2587 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2588 Patch by Magnus Holmgren from 2007-02-20.
2590 TL/04 Bugzilla 1281 - Spec typo.
2591 Bugzilla 1283 - Spec typo.
2592 Bugzilla 1290 - Spec grammar fixes.
2594 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2596 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2598 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2599 mailing list by by Dmitry Isaikin.
2601 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2603 JH/02 Support "G" suffix to numbers in ${if comparisons.
2605 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2607 NM/01 Bugzilla 1197 - Spec typo
2608 Bugzilla 1196 - Spec examples corrections
2610 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2612 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2613 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2616 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2617 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2619 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2620 "acl {{name}{arg}...}", and optional args on acl condition
2623 JH/05 Permit multiple router/transport headers_add/remove lines.
2625 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2627 JH/07 Avoid using a waiting database for a single-message-only transport.
2628 Performance patch from Paul Fisher. Bugzilla 1262.
2630 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2633 JH/09 Add $headers_added variable, with content from use of ACL modifier
2634 add_header (but not yet added to the message). Bugzilla 199.
2636 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2637 Pulled from Bugzilla 817 by Wolfgang Breyha.
2639 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2641 (nb: this is the same fix as in Exim 4.80.1)
2643 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2646 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2649 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2651 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2652 advertise SMTP AUTH mechanism to us, instead of a generic
2653 protocol violation error. Also, make Exim more robust to bad
2654 data from the Dovecot auth socket.
2656 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2658 When a queue runner is handling a message, Exim first routes the
2659 recipient addresses, during which it prunes them based on the retry
2660 hints database. After that it attempts to deliver the message to
2661 any remaining recipients. It then updates the hints database using
2664 So if a recipient address works intermittently, it can get repeatedly
2665 deferred at routing time. The retry hints record remains fresh so the
2666 address never reaches the final cutoff time.
2668 This is a fairly common occurrence when a user is bumping up against
2669 their storage quota. Exim had some logic in its local delivery code
2670 to deal with this. However it did not apply to per-recipient defers
2671 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2673 This change adds a proper retry rule check during routing so that the
2674 final cutoff time is checked against the message's age. We only do
2675 this check if there is an address retry record and there is not a
2676 domain retry record; this implies that previous attempts to handle
2677 the address had the retry_use_local_parts option turned on. We use
2678 this as an approximation for the destination being like a local
2679 delivery, as in LMTP.
2681 I suspect this new check makes the old local delivery cutoff check
2682 redundant, but I have not verified this so I left the code in place.
2684 TF/02 Correct gecos expansion when From: is a prefix of the username.
2686 Test 0254 submits a message to Exim with the header
2690 When I ran the test suite under the user fanf2, Exim expanded
2691 the header to contain my full name, whereas it should have added
2692 a Resent-Sender: header. It erroneously treats any prefix of the
2693 username as equal to the username.
2695 This change corrects that bug.
2697 GF/01 DCC debug and logging tidyup
2698 Error conditions log to paniclog rather than rejectlog.
2699 Debug lines prefixed by "DCC: " to remove any ambiguity.
2701 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2703 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2704 Bug spotted by Jeremy Harris; was flawed since initial commit.
2705 Would have resulted in OCSP responses post-SNI triggering an Exim
2706 NULL dereference and crash.
2708 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2710 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2711 Bug detection, analysis and fix by Samuel Thibault.
2712 Bugzilla 1331, Debian bug #698092.
2714 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2716 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2717 Server implementation by Todd Lyons, client by JH.
2718 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2719 config variable "prdr_enable" controls whether the server
2720 advertises the facility. If the client requests PRDR a new
2721 acl_data_smtp_prdr ACL is called once for each recipient, after
2722 the body content is received and before the acl_smtp_data ACL.
2723 The client is controlled by both of: a hosts_try_prdr option
2724 on the smtp transport, and the server advertisement.
2725 Default client logging of deliveries and rejections involving
2726 PRDR are flagged with the string "PRDR".
2728 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2729 fclose(). Diagnosis by Todd Lyons.
2731 PP/17 Update configure.default to handle IPv6 localhost better.
2732 Patch by Alain Williams (plus minor tweaks).
2735 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2736 This is now consistent with GnuTLS, and is now documented: the
2737 previous undocumented portable approach to treating the option as
2738 unset was to force an expansion failure. That still works, and
2739 an empty string is now equivalent.
2741 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2742 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2743 not performing validation itself.
2745 PP/20 Added force_command boolean option to pipe transport.
2746 Patch from Nick Koston, of cPanel Inc.
2748 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2751 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2753 PP/21 Fix eximon continuous updating with timestamped log-files.
2754 Broken in a format-string cleanup in 4.80, missed when I repaired the
2755 other false fix of the same issue.
2756 Report and fix from Heiko Schlichting.
2759 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2760 Report from Prashanth Katuri.
2762 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2763 It's SecureTransport, so affects any MacOS clients which use the
2764 system-integrated TLS libraries, including email clients.
2766 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2767 using a MIME ACL for non-SMTP local injection.
2768 Report and assistance in diagnosis by Warren Baker.
2770 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2772 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2774 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2775 last id that failed so it may be referenced in subsequent ACL's.
2777 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2780 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2781 ldap library initialization, allowing self-signed CA's to be
2782 used. Also properly sets require_cert option later in code by
2783 using NULL (global ldap config) instead of ldap handle (per
2784 session). Bug diagnosis and testing by alxgomz.
2786 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2787 the src/util/ subdirectory.
2789 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2790 renamed to Transport Post Delivery Action by Jeremy Harris, as
2793 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2794 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2795 redis_servers = needs to be configured which will be used by the redis
2796 lookup. Patch from Warren Baker, of The Packet Hub.
2798 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2800 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2801 hostname or reverse DNS when processing a host list. Used suggestions
2802 from multiple comments on this bug.
2804 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2806 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2807 Missed a few lines, added it to make the runtest require no keyboard
2810 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2811 contains upper case chars. Make router use caseful_local_part.
2813 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2814 support when GnuTLS has been built with p11-kit.
2820 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2822 This, or similar/improved, will also be change PP/11 of 4.82.
2828 PP/01 Handle short writes when writing local log-files.
2829 In practice, only affects FreeBSD (8 onwards).
2830 Bugzilla 1053, with thanks to Dmitry Isaikin.
2832 NM/01 Bugzilla 949 - Documentation tweak
2834 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2837 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2839 PP/02 Implemented gsasl authenticator.
2841 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2843 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2844 `pkg-config foo` for cflags/libs.
2846 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2847 with rest of GSASL and with heimdal_gssapi.
2849 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2850 `pkg-config foo` for cflags/libs for the TLS implementation.
2852 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2853 properties get this fed in as external SSF. A number of robustness
2854 and debugging improvements to the cyrus_sasl authenticator.
2856 PP/08 cyrus_sasl server now expands the server_realm option.
2858 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2859 Patch by Jeremy Harris.
2861 PP/10 Added dbmjz lookup type.
2863 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2865 PP/12 MAIL args handles TAB as well as SP, for better interop with
2866 non-compliant senders.
2867 Analysis and variant patch by Todd Lyons.
2869 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2870 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2871 Patch from Dirk Mueller <dmueller@suse.com>
2873 PP/13 tls_peerdn now print-escaped for spool files.
2874 Observed some $tls_peerdn in wild which contained \n, which resulted
2875 in spool file corruption.
2877 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2878 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2879 or write after TLS renegotiation, which otherwise led to messages
2882 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2883 as a tracking header (ie: a signed header comes before the signature).
2884 Patch from Wolfgang Breyha.
2886 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2887 comma-sep list; embedded commas doubled.
2889 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2891 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2893 Report and patch from Dmitry Banschikov.
2895 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2896 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2897 failures. We appear to now support TLS1.1+ with Exim.
2899 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2900 lets Exim select keys and certificates based upon TLS SNI from client.
2901 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2902 before an outbound SMTP session. New log_selector, +tls_sni.
2904 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2905 NULL dereference. Report and patch from Alun Jones.
2907 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2908 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2909 Not seeing resolver debug output on NetBSD, but suspect this is a
2910 resolver implementation change.
2912 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2913 Left warnings. Added "eximon gdb" invocation mode.
2915 PP/21 Defaulting "accept_8bitmime" to true, not false.
2917 PP/22 Added -bw for inetd wait mode support.
2919 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2920 locate the relevant includes and libraries. Made this the default.
2922 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2923 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2925 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2926 This may cause build issues on older platforms.
2928 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2929 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2930 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2931 Added SNI support via GnuTLS too.
2932 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2934 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2936 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2937 Applied second patch from Janne, implementing suggestion to default
2938 multiple-strings-in-record handling to match SPF spec.
2940 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2942 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2943 read-only, out of scope).
2944 Patch from Wolfgang Breyha, report from Stuart Northfield.
2946 PP/29 Fix three issues highlighted by clang analyser static analysis.
2947 Only crash-plausible issue would require the Cambridge-specific
2948 iplookup router and a misconfiguration.
2949 Report from Marcin Mirosław.
2951 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2953 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2954 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2955 As part of this, removing so much warning spew let me fix some minor
2956 real issues in debug logging.
2958 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2959 assignment on my part. Fixed.
2961 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2962 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2963 Janne Snabb (who went above and beyond: thank you).
2965 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2966 string otherwise requires a connection and a bunch more work and it's
2967 relatively easy to get wrong. Should also expose TLS library linkage
2970 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2971 64-bit ${eval} (JH/03).
2973 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2974 GNU libc to support some of the 64-bit stuff, should not lead to
2975 conflicts. Defined before os.h is pulled in, so if a given platform
2976 needs to override this, it can.
2978 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2979 protection layer was required, which is not implemented.
2980 Bugzilla 1254, patch from Wolfgang Breyha.
2982 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2983 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2984 tls_dhparam take prime identifiers. Also unbreak combination of
2985 OpenSSL+DH_params+TLSSNI.
2987 PP/39 Disable SSLv2 by default in OpenSSL support.
2993 PP/01 Solaris build fix for Oracle's LDAP libraries.
2994 Bugzilla 1109, patch from Stephen Usher.
2996 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2998 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3001 TF/02 Fix a couple more cases where we did not log the error message
3002 when unlink() failed. See also change 4.74-TF/03.
3004 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3005 lock up or crash if it happened to be inside a call to libc when it
3006 got a SIGUSR1 from exiwhat.
3008 The SIGUSR1 handler appends the current process status to the process
3009 log which is later printed by exiwhat. It used to use the general
3010 purpose logging code to do this, but several functions it calls are
3011 not safe for signals.
3013 The new output code in the SIGUSR1 handler is specific to the process
3014 log, and simple enough that it's easy to inspect for signal safety.
3015 Removing some special cases also simplifies the general logging code.
3016 Removing the spurious timestamps from the process log simplifies
3019 TF/04 Improved ratelimit ACL condition.
3021 The /noupdate option has been deprecated in favour of /readonly which
3022 has clearer semantics. The /leaky, /strict, and /readonly update modes
3023 are mutually exclusive. The update mode is no longer included in the
3024 database key; it just determines when the database is updated. (This
3025 means that when you upgrade Exim will forget old rate measurements.)
3027 Exim now checks that the per_* options are used with an update mode that
3028 makes sense for the current ACL. For example, when Exim is processing a
3029 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3030 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3031 must specify per_mail/readonly. If you omit the update mode it defaults to
3032 /leaky where that makes sense (as before) or /readonly where required.
3034 The /noupdate option is now undocumented but still supported for
3035 backwards compatibility. It is equivalent to /readonly except that in
3036 ACLs where /readonly is required you may specify /leaky/noupdate or
3037 /strict/noupdate which are treated the same as /readonly.
3039 A useful new feature is the /count= option. This is a generalization
3040 of the per_byte option, so that you can measure the throughput of other
3041 aggregate values. For example, the per_byte option is now equivalent
3042 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3044 The per_rcpt option has been generalized using the /count= mechanism
3045 (though it's more complicated than the per_byte equivalence). When it is
3046 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3047 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3048 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3049 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3050 Note that using per_rcpt with a non-readonly update mode in more than
3051 one ACL will cause the recipients to be double-counted. (The per_mail
3052 and per_byte options don't have this problem.)
3054 The handling of very low rates has changed slightly. If the computed rate
3055 is less than the event's count (usually one) then this event is the first
3056 after a long gap. In this case the rate is set to the same as this event's
3057 count, so that the first message of a spam run is counted properly.
3059 The major new feature is a mechanism for counting the rate of unique
3060 events. The new per_addr option counts the number of different
3061 recipients that someone has sent messages to in the last time period. It
3062 behaves like per_rcpt if all the recipient addresses are different, but
3063 duplicate recipient addresses do not increase the measured rate. Like
3064 the /count= option this is a general mechanism, so the per_addr option
3065 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3066 example, measure the rate that a client uses different sender addresses
3067 with the options per_mail/unique=$sender_address. There are further
3068 details in the main documentation.
3070 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3072 TF/06 Removed a few PCRE remnants.
3074 TF/07 Automatically extract Exim's version number from tags in the git
3075 repository when doing development or release builds.
3077 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3078 Bugzilla 879. Patch from Paul Fisher.
3080 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3081 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3084 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3086 PP/05 Variable $av_failed, true if the AV scanner deferred.
3087 Bugzilla 1078. Patch from John Horne.
3089 PP/06 Stop make process more reliably on build failure.
3090 Bugzilla 1087. Patch from Heiko Schlittermann.
3092 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3093 Bugzilla 1089. Patch from Heiko Schlittermann.
3095 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3096 Bugzilla 1131. Patch from Holger Weiß.
3098 PP/09 Handle IPv6 addresses with SPF.
3099 Bugzilla 860. Patch from Wolfgang Breyha.
3101 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3103 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3106 PP/11 match_* no longer expand right-hand-side by default.
3107 New compile-time build option, EXPAND_LISTMATCH_RHS.
3108 New expansion conditions, "inlist", "inlisti".
3110 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3112 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3114 PP/14 fix log_write() format string regression from TF/03.
3115 Bugzilla 1152. Patch from Dmitry Isaikin.
3121 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3123 PP/02 Harmonised TLS library version reporting; only show if debugging.
3124 Layout now matches that introduced for other libraries in 4.74 PP/03.
3126 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3128 PP/04 New "dns_use_edns0" global option.
3130 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3133 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3134 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3136 TK/01 Updated PolarSSL code to 0.14.2.
3137 Bugzilla 1097. Patch from Andreas Metzler.
3139 PP/07 Catch divide-by-zero in ${eval:...}.
3140 Fixes bugzilla 1102.
3142 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3145 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3146 format-string attack -- SECURITY: remote arbitrary code execution.
3148 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3149 time unintentionally subject to list matching rules, letting the header
3150 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3151 arbitrary string expansion). This allowed for information disclosure.
3153 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3154 INT_MIN/-1 -- value coerced to INT_MAX.
3160 NM/01 Workaround for PCRE version dependency in version reporting
3163 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3164 This fixes portability to compilers other than gcc, notably
3165 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3167 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3168 makefiles for portability to HP-UX and POSIX correctness.
3170 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3171 Also via indented variable definition in the Makefile.
3172 (Debugging by Oliver Heesakkers).
3174 PP/02 Restore caching of spamd results with expanded spamd_address.
3175 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3177 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3178 Improves build reliability. Fix from: Frank Elsner
3180 NM/02 Fix wide character breakage in the rfc2047 coding
3181 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3183 NM/03 Allow underscore in dnslist lookups
3184 Fixes bug 1026. Patch from Graeme Fowler
3186 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3187 Code patches from Adam Ciarcinski of NetBSD.
3189 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3192 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3193 is logged, to avoid truncation. Patch from John Horne.
3195 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3196 Patch from Jakob Hirsch.
3198 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3199 SQL string expansion failure details.
3200 Patch from Andrey Oktyabrski.
3202 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3203 Patch from Simon Arlott.
3205 PP/09 New lookups functionality failed to compile on old gcc which rejects
3206 extern declarations in function scope.
3207 Patch from Oliver Fleischmann
3209 PP/10 Use sig_atomic_t for flags set from signal handlers.
3210 Check getgroups() return and improve debugging.
3211 Fixed developed for diagnosis in bug 927 (which turned out to be
3214 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3215 Patch from Mark Zealey.
3217 PP/12 Bugzilla 1056: Improved spamd server selection.
3218 Patch from Mark Zealey.
3220 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3221 Based on patch from Heiko Schlittermann.
3223 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3224 Patch from Uwe Doering, sign-off by Michael Haardt.
3226 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3227 variable declaration deep within a block. Bug and patch from
3230 PP/15 lookups-Makefile IRIX compatibility coercion.
3232 PP/16 Make DISABLE_DKIM build knob functional.
3234 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3235 Patch by Simon Arlott
3237 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3238 variable argument macros. Our copy now differs from upstream.
3244 TF/01 Failure to get a lock on a hints database can have serious
3245 consequences so log it to the panic log.
3247 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3248 controlled using the smtp_confirmation log selector.
3250 TF/03 Include the error message when we fail to unlink a spool file.
3252 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3253 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3254 for maintaining out-of-tree patches for some time.
3256 PP/01 Bugzilla 139: Documentation and portability issues.
3257 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3258 Handle per-OS dynamic-module compilation flags.
3260 PP/02 Let /dev/null have normal permissions.
3261 The 4.73 fixes were a little too stringent and complained about the
3262 permissions on /dev/null. Exempt it from some checks.
3263 Reported by Andreas M. Kirchwitz.
3265 PP/03 Report version information for many libraries, including
3266 Exim version information for dynamically loaded libraries. Created
3267 version.h, now support a version extension string for distributors
3268 who patch heavily. Dynamic module ABI change.
3270 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3271 privilege escalation vulnerability whereby the Exim run-time user
3272 can cause root to append content of the attacker's choosing to
3275 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3278 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3279 If dropping privileges for untrusted macros, we disabled normal logging
3280 on the basis that it would fail; for the Exim run-time user, this is not
3281 the case, and it resulted in successful deliveries going unlogged.
3282 Fixed. Reported by Andreas Metzler.
3288 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3289 only prepend for the Resent-* case. Fixes regression introduced in
3290 Exim 4.70 by NM/22 for Bugzilla 607.
3292 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3293 increasing numbers of administrators be bitten by this.
3295 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3297 PP/03 Bugzilla 994: added openssl_options main configuration option.
3299 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3301 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3303 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3305 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3306 without a peer certificate, leading to a segfault because of an
3307 assumption that peers always have certificates. Be a little more
3308 paranoid. Problem reported by Martin Tscholak.
3310 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3311 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3312 NB: ClamAV planning to remove STREAM in "middle of 2010".
3313 CL also introduces -bmalware, various -d+acl logging additions and
3314 more caution in buffer sizes.
3316 PP/09 Implemented reverse_ip expansion operator.
3318 PP/10 Bugzilla 937: provide a "debug" ACL control.
3320 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3322 PP/12 Bugzilla 973: Implement --version.
3324 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3326 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3328 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3330 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3331 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3332 ignore trailing whitespace.
3334 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3336 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3339 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3340 Notification from Dr Andrew Aitchison.
3342 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3343 ExtendedDetectionInfo response format.
3344 Notification from John Horne.
3346 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3349 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3350 XSL and documented dependency on system catalogs, with examples of how
3353 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3356 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3357 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3358 configuration file which is writeable by the Exim user or group.
3360 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3361 of configuration files to cover files specified with the -C option if
3362 they are going to be used with root privileges, not just the default
3365 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3366 option (effectively making it always true).
3368 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3369 files to be used while preserving root privileges.
3371 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3372 that rogue child processes cannot use them.
3374 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3375 run-time user, instead of root.
3377 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3378 Exim run-time user without dropping privileges.
3380 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3381 result string, instead of calling string_vformat() twice with the same
3384 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3385 for other users. Others should always drop root privileges if they use
3386 -C on the command line, even for a whitelisted configure file.
3388 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3390 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3396 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3397 $data_path, and $header_path variables; fixed documentation bugs and
3400 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3401 exipick to access non-standard spools, including the "frozen" queue
3404 NM/01 Bugzilla 965: Support mysql stored procedures.
3405 Patch from Alain Williams
3407 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3409 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3410 Patch from Andreas Metzler
3412 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3413 Patch from Kirill Miazine
3415 NM/05 Bugzilla 671: Added umask to procmail example.
3417 JJ/03 installed exipick 20100323.0, fixing doc bug
3419 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3420 directory. Notification and patch from Dan Rosenberg.
3422 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3424 TK/02 Improve log output when DKIM signing operation fails.
3426 MH/01 Treat the transport option dkim_domain as a colon separated
3427 list, not as a single string, and sign the message with each element,
3428 omitting multiple occurences of the same signer.
3430 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3431 Bugzilla 985, 986. Patch by Simon Arlott
3433 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3434 Patch by Simon Arlott
3436 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3437 MBX locking. Notification from Dan Rosenberg.
3443 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3445 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3447 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3449 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3451 NM/04 Bugzilla 915: Flag broken perl installation during build.
3457 TK/01 Added patch by Johannes Berg that expands the main option
3458 "spamd_address" if it starts with a dollar sign.
3460 TK/02 Write list of recipients to X-Envelope-Sender header when building
3461 the mbox-format spool file for content scanning (suggested by Jakob
3464 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3465 (http://www.dcc-servers.net/) support via dccifd. Activated by
3466 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3468 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3469 by Mark Daniel Reidel <mr@df.eu>.
3471 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3472 When building exim an external PCRE library is now needed -
3473 PCRE is a system library on the majority of modern systems.
3474 See entry on PCRE_LIBS in EDITME file.
3476 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3477 conversation. Added nologin parameter to request.
3478 Patch contributed by Kirill Miazine.
3480 TF/01 Do not log submission mode rewrites if they do not change the address.
3482 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3484 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3485 log files in place. Contributed by Roberto Lima.
3487 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3489 TF/03 Bugzilla 615: When checking the local_parts router precondition
3490 after a local_part_suffix or local_part_prefix option, Exim now
3491 does not use the address's named list lookup cache, since this
3492 contains cached lookups for the whole local part.
3494 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3495 Robert Millan. Documentation is in experimental-spec.txt.
3497 TF/04 Bugzilla 668: Fix parallel build (make -j).
3499 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3501 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3502 Patch provided by Jan Srzednicki.
3504 TF/05 Leading white space used to be stripped from $spam_report which
3505 wrecked the formatting. Now it is preserved.
3507 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3508 that they are available at delivery time.
3510 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3512 TF/08 TLS error reporting now respects the incoming_interface and
3513 incoming_port log selectors.
3515 TF/09 Produce a more useful error message if an SMTP transport's hosts
3516 setting expands to an empty string.
3518 NM/06 Bugzilla 744: EXPN did not work under TLS.
3519 Patch provided by Phil Pennock.
3521 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3522 Patch provided by Richard Godbee.
3524 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3525 acl_smtp_notquit, added index entry.
3527 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3528 Patch provided by Eugene Bujak.
3530 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3531 accept(). Patch provided by Maxim Dounin.
3533 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3534 Patch provided by Phil Pennock.
3536 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3538 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3539 Patch provided by Brad "anomie" Jorsch.
3541 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3542 Patch provided by Dean Brooks.
3544 TK/05 Add native DKIM support (does not depend on external libraries).
3546 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3547 Patch provided by Graeme Fowler.
3549 NM/16 Bugzilla 851: Documentation example syntax fix.
3551 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3553 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3556 NM/19 Bugzilla 745: TLS version reporting.
3557 Patch provided by Phil Pennock.
3559 NM/20 Bugzilla 167: bool: condition support.
3560 Patch provided by Phil Pennock.
3562 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3563 clients. Patch provided by Phil Pennock.
3565 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3566 Patch provided by Brad "anomie" Jorsch.
3568 NM/23 Bugzilla 687: Fix misparses in eximstats.
3569 Patch provided by Heiko Schlittermann.
3571 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3572 Patch provided by Heiko Schlittermann.
3574 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3575 plus update to original patch.
3577 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3579 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3580 Patch provided by David Brownlee.
3582 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3584 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3586 NM/30 Bugzilla 888: TLS documentation bugfixes.
3588 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3590 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3591 Unlike the original bugzilla I have changed all shell scripts in src tree.
3593 NM/33 Bugzilla 898: Transport filter timeout fix.
3594 Patch by Todd Rinaldo.
3596 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3597 Patch by Serge Demonchaux.
3599 NM/35 Bugzilla 39: Base64 decode bug fixes.
3600 Patch by Jakob Hirsch.
3602 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3604 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3606 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3608 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3614 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3615 ALT-N's libdkim that I have put here:
3616 http://duncanthrax.net/exim-experimental/
3618 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3619 were called 'true' and it seems that C99 defines that as a reserved
3620 keyword to be used with 'bool' variable types. That means you could
3621 not include C99-style headers which use bools without triggering
3622 build errors in sieve.c.
3624 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3625 as mailq or other aliases. Changed the --help handling significantly
3626 to do whats expected. exim_usage() emits usage/help information.
3628 SC/01 Added the -bylocaldomain option to eximstats.
3630 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3632 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3634 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3640 PH/01 Another patch from the Sieve maintainer.
3642 PH/02 When an IPv6 address is converted to a string for single-key lookup
3643 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3644 dots are used instead of colons so that keys in lsearch files need not
3645 contain colons. This was done some time before quoting was made available
3646 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3647 (notated using the quote facility) so as to distinguish them from IPv4
3648 keys. This meant that lookups for IP addresses in host lists did not work
3649 for iplsearch lookups.
3651 This has been fixed by arranging for IPv6 addresses to be expressed with
3652 colons if the lookup type is iplsearch. This is not incompatible, because
3653 previously such lookups could never work.
3655 The situation is now rather anomalous, since one *can* have colons in
3656 ordinary lsearch keys. However, making the change in all cases is
3657 incompatible and would probably break a number of configurations.
3659 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3662 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3663 conversion specification without a maximum field width, thereby enabling
3664 a rogue spamd server to cause a buffer overflow. While nobody in their
3665 right mind would setup Exim to query an untrusted spamd server, an
3666 attacker that gains access to a server running spamd could potentially
3667 exploit this vulnerability to run arbitrary code as the Exim user.
3669 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3670 $primary_hostname instead of what libspf2 thinks the hosts name is.
3672 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3673 a directory entry by the name of the lookup key. Previously, if a
3674 symlink pointed to a non-existing file or a file in a directory that
3675 Exim lacked permissions to read, a lookup for a key matching that
3676 symlink would fail. Now it is enough that a matching directory entry
3677 exists, symlink or not. (Bugzilla 503.)
3679 PH/03 The body_linecount and body_zerocount variables are now exported in the
3682 PH/04 Added the $dnslist_matched variable.
3684 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3685 This means they are set thereafter only if the connection becomes
3688 PH/06 Added the client_condition to authenticators so that some can be skipped
3689 by clients under certain conditions.
3691 PH/07 The error message for a badly-placed control=no_multiline_responses left
3692 "_responses" off the end of the name.
3694 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3696 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3697 (without spaces) instead of just copying the configuration text.
3699 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3701 PH/11 Added $max_received_linelength.
3703 PH/12 Added +ignore_defer and +include_defer to host lists.
3705 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3706 way in which PCRE > 7.0 is built.
3708 PH/14 Implemented queue_only_load_latch.
3710 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3711 MAIL command. The effect was to mangle the value on 64-bit systems.
3713 PH/16 Another patch from the Sieve maintainer.
3715 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3717 PH/18 If a system quota error occurred while trying to create the file for
3718 a maildir delivery, the message "Mailbox is full" was not appended to the
3719 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3720 only to a quota excession during the actual writing of the file.
3722 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3723 characters?) which causes problems in log lines. The DN values are now
3724 passed through string_printing() before being added to log lines.
3726 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3727 and InterBase are left for another time.)
3729 PH/21 Added message_body_newlines option.
3731 PH/22 Guard against possible overflow in moan_check_errorcopy().
3733 PH/23 POSIX allows open() to be a macro; guard against that.
3735 PH/24 If the recipient of an error message contained an @ in the local part
3736 (suitably quoted, of course), incorrect values were put in $domain and
3737 $local_part during the evaluation of errors_copy.
3743 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3744 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3747 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3748 issue a MAIL command.
3750 PH/02 In an ACL statement such as
3752 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3754 if a client was not listed at all, or was listed with a value other than
3755 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3756 the condition was not true (as it should be), so access was not denied.
3757 The bug was that the ! inversion was incorrectly passed on to the second
3758 item. This has been fixed.
3760 PH/03 Added additional dnslists conditions == and =& which are different from
3761 = and & when the dns lookup returns more than one IP address.
3763 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3764 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3766 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3767 FSYNC, which compiles an option called disable_fsync that allows for
3768 bypassing fsync(). The documentation is heavily laced with warnings.
3770 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3772 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3773 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3774 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3775 including adding "make clean"; (3) Added -fPIC when compiling the test
3776 dynamically loaded module, to get rid of a warning.
3778 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3779 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3780 = 0s. The bug is otherwise harmless.
3782 PH/07 There was a bug in the dovecot authenticator such that the value of
3783 $auth1 could be overwritten, and so not correctly preserved, after a
3784 successful authentication. This usually meant that the value preserved by
3785 the server_setid option was incorrect.
3787 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3789 PH/09 Installed PCRE release 7.0.
3791 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3792 run for batched SMTP input. It is now run at the start of every message
3793 in the batch. While fixing this I discovered that the process information
3794 (output by running exiwhat) was not always getting set for -bs and -bS
3795 input. This is fixed, and it now also says "batched" for BSMTP.
3797 PH/11 Added control=no_pipelining.
3799 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3800 patch, slightly modified), and move the expansion of helo_data till after
3801 the connection is made in the smtp transport (so it can use these
3804 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3806 PH/14 Added log_selector = +pid.
3808 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3810 PH/16 Add ${if forany and ${if forall.
3812 PH/17 Added dsn_from option to vary the From: line in DSNs.
3814 PH/18 Flush SMTP output before performing a callout, unless control =
3815 no_callout_flush is set.
3817 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3818 was true (the default) a successful delivery failed to delete the retry
3819 item, thus causing premature timeout of the address. The bug is now
3822 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3824 PH/21 Long custom messages for fakedefer and fakereject are now split up
3825 into multiline responses in the same way that messages for "deny" and
3826 other ACL rejections are.
3828 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3829 with slight modification.
3831 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3832 draft, changing the syntax and factoring some duplicate code".
3834 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3835 for deliveries of the second and subsequent messages over the same SMTP
3838 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3839 ${reduce, with only minor "tidies".
3841 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3843 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3844 expansion side effects.
3846 PH/27 When a message times out after an over-quota error from an Exim-imposed
3847 quota, the bounce message says "mailbox is full". This message was not
3848 being given when it was a system quota that was exceeded. It now should
3851 MH/03 Made $recipients available in local_scan(). local_scan() already has
3852 better access to the recipient list through recipients_list[], but
3853 $recipients can be useful in postmaster-provided expansion strings.
3855 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3856 in the case of a MAIL command with additional options following the
3857 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3858 were accidentally chopped off.
3860 PH/29 SMTP synchronization checks are implemented when a command is read -
3861 there is a check that no more input is waiting when there shouldn't be
3862 any. However, for some commands, a delay in an ACL can mean that it is
3863 some time before the response is written. In this time, more input might
3864 arrive, invalidly. So now there are extra checks after an ACL has run for
3865 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3866 pipelining has not been advertised.
3868 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3870 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3871 returned in the SMTP response when smtp_return_error_details was false.
3872 This has been fixed.
3874 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3875 instead of the C I/O that was originally supplied, because problems were
3876 reported on Solaris.
3878 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3879 Exim which did not show up earlier: it was assuming that a call to
3880 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3881 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3882 was a macro that became an assignment, so it seemed to work. This has
3883 changed to a proper function call with a void return, hence the compile
3884 error. Exim's code has been fixed.
3886 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3889 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3891 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3893 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3894 because it turns out to be too big or there is a timeout) there is no
3895 "Completed" line in the log. When some messages of this type were
3896 selected by exigrep, they were listed as "not completed". Others were
3897 picked up by some special patterns. I have improved the selection
3898 criteria to be more general.
3900 PH/38 The host_find_failed option in the manualroute router can now be set
3901 to "ignore", to completely ignore a host whose IP address cannot be
3902 found. If all hosts are ignored, the behaviour is controlled by the new
3903 host_all_ignored option.
3905 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3906 homing or because of multiple MX records with /mx) generated more than
3907 one IP address, and the following item turned out to be the local host,
3908 all the secondary addresses of the first item were incorrectly removed
3909 from the list, along with the local host and any following hosts (which
3910 is what is supposed to happen).
3912 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3913 whoever called Exim into the -H file. In the case of the daemon it was
3914 behaving confusingly. When first started, it used values for whoever
3915 started the daemon, but after a SIGHUP it used the Exim user (because it
3916 calls itself on a restart). I have changed the code so that it now always
3919 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3920 message are rejected with the same error (e.g. no authentication or bad
3921 sender address), and a DATA command is nevertheless sent (as can happen
3922 with PIPELINING or a stupid MUA), the error message that was given to the
3923 RCPT commands is included in the rejection of the DATA command. This is
3924 intended to be helpful for MUAs that show only the final error to their
3927 PH/42 Another patch from the Sieve maintainer.
3929 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3930 Eximstats - Fixed some broken HTML links and added missing column headers
3932 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3933 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3935 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3937 PH/43 Yet another patch from the Sieve maintainer.
3939 PH/44 I found a way to check for a TCP/IP connection going away before sending
3940 the response to the final '.' that terminates a message, but only in the
3941 case where the client has not sent further data following the '.'
3942 (unfortunately, this is allowed). However, in many cases there won't be
3943 any further data because there won't be any more messages to send. A call
3944 to select() can be used: if it shows that the input is "ready", there is
3945 either input waiting, or the socket has been closed. An attempt to read
3946 the next input character can distinguish the two cases. Previously, Exim
3947 would have sent an OK response which the client would never have see.
3948 This could lead to message repetition. This fix should cure that, at
3949 least in a lot of common cases.
3951 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3952 advertised in response to EHLO.
3958 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3959 fixed by 4.65/MH/01 (is this a record?) are fixed:
3961 (i) An empty string was always treated as zero by the numeric comparison
3962 operators. This behaviour has been restored.
3964 (ii) It is documented that the numeric comparison operators always treat
3965 their arguments as decimal numbers. This was broken in that numbers
3966 starting with 0 were being interpreted as octal.
3968 While fixing these problems I realized that there was another issue that
3969 hadn't been noticed. Values of message_size_limit (both the global option
3970 and the transport option) were treated as octal if they started with 0.
3971 The documentation was vague. These values are now always treated as
3972 decimal, and I will make that clear in the documentation.
3978 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3979 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3982 MH/01 Don't check that the operands of numeric comparison operators are
3983 integers when their expansion is in "skipping" mode (fixes bug
3984 introduced by 4.64-PH/07).
3986 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3987 child addresses, Exim now panics and dies. Previously, because the count
3988 is held in a short int, deliveries were likely to be lost. As such a
3989 large number of recipients for a single message is ridiculous
3990 (performance will be very, very poor), I have chosen to impose a limit
3991 rather than extend the field.
3997 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3998 leftover -K file (the existence of which was triggered by #402).
3999 While we were at it, introduced process PID as part of the -K
4000 filename. This should rule out race conditions when creating
4003 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4004 processing considerably. Previous code took too long for large mails,
4005 triggering a timeout which in turn triggers #401.
4007 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4008 in the DK code in transports.c. sendfile() is not really portable,
4009 hence the _LINUX specificness.
4011 TF/01 In the add_headers option to the mail command in an Exim filter,
4012 there was a bug that Exim would claim a syntax error in any
4013 header after the first one which had an odd number of characters
4016 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4017 callout verification, Exim cached a "reject" for the entire domain. This
4018 is correct for most verifications, but it is not correct for a recipient
4019 verification with use_sender or use_postmaster set, because in that case
4020 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4021 case of MAIL FROM:<> rejection from other early rejections (e.g.
4022 rejection of HELO). When verifying a recipient using a non-null MAIL
4023 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4024 Whatever the result of the callout, the value of the domain cache is
4025 left unchanged (for any other kind of callout, getting as far as trying
4026 RCPT means that the domain itself is ok).
4028 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4031 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4032 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4033 session, the daemon ignores ECONNECT errors and logs others; it now
4034 ignores EPIPE as well.
4036 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4037 (quoted-printable decoding).
4039 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4040 later the small subsequent patch to fix an introduced bug.
4042 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4044 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4046 PH/08 An error is now given if message_size_limit is specified negative.
4048 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4049 to be given (somewhat) arbitrary names.
4051 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4054 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4055 miscellaneous code fixes
4057 PH/10 Added the log_reject_target ACL modifier to specify where to log
4060 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4061 hostname. This is wrong, because it relates to the incoming message (and
4062 probably the interface on which it is arriving) and not to the outgoing
4063 callout (which could be using a different interface). This has been
4064 changed to use the value of the helo_data option from the smtp transport
4065 instead - this is what is used when a message is actually being sent. If
4066 there is no remote transport (possible with a router that sets up host
4067 addresses), $smtp_active_hostname is used.
4069 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4070 tweaks were necessary in order to get it to work (see also 21 below):
4071 (a) The code assumed that strncpy() returns a negative number on buffer
4072 overflow, which isn't the case. Replaced with Exim's string_format()
4074 (b) There were several signed/unsigned issues. I just did the minimum
4075 hacking in of casts. There is scope for a larger refactoring.
4076 (c) The code used strcasecmp() which is not a standard C function.
4077 Replaced with Exim's strcmpic() function.
4078 (d) The code set only $1; it now sets $auth1 as well.
4079 (e) A simple test gave the error "authentication client didn't specify
4080 service in request". It would seem that Dovecot has changed its
4081 interface. Fortunately there's a specification; I followed it and
4082 changed what the client sends and it appears to be working now.
4084 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4087 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4088 address A is aliased to B and C, where B exists and C does not. Without
4089 -v the output is "A verified" because verification stops after a
4090 successful redirection if more than one address is generated. However,
4091 with -v the child addresses are also verified. Exim was outputting "A
4092 failed to verify" and then showing the successful verification for C,
4093 with its parentage. It now outputs "B failed to verify", showing B's
4094 parentage before showing the successful verification of C.
4096 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4097 look up a TXT record in a specific list after matching in a combined
4100 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4101 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4102 they consult the DNS. I had assumed they would set it the way they
4103 wanted; and indeed my experiments on Linux seem to show that in some
4104 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4105 To be on the safe side, however, I have now made the interface to
4106 host_find_byname() similar to host_find_bydns(), with an argument
4107 containing the DNS resolver options. The host_find_byname() function now
4108 sets these options at its start, just as host_find_bydns() does. The smtp
4109 transport options dns_qualify_single and dns_search_parents are passed to
4110 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4111 of host_find_byname() use the default settings of RES_DEFNAMES
4112 (qualify_single) but not RES_DNSRCH (search_parents).
4114 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4115 spool_read_header() do less string testing, by means of a preliminary
4116 switch on the second character of optional "-foo" lines. (This is
4117 overdue, caused by the large number of possibilities that now exist.
4118 Originally there were few.) While I was there, I also converted the
4119 str(n)cmp tests so they don't re-test the leading "-" and the first
4120 character, in the hope this might squeeze out yet more improvement.
4122 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4123 flag allowing group syntax was set by the header_syntax check but not
4124 turned off, possible causing trouble later; (2) The flag was not being
4125 set at all for the header_verify test, causing "group"-style headers to
4126 be rejected. I have now set it in this case, and also caused header_
4127 verify to ignore an empty address taken from a group. While doing this, I
4128 came across some other cases where the code for allowing group syntax
4129 while scanning a header line wasn't quite right (mostly, not resetting
4130 the flag correctly in the right place). These bugs could have caused
4131 trouble for malformed header lines. I hope it is now all correct.
4133 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4134 with the "reply" argument non-NULL. The code, however (which originally
4135 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4136 but it didn't always do it. This confused somebody who was copying the
4137 code for some other use. I have removed all the tests.
4139 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4140 feature that was used to support insecure browsers during the U.S. crypto
4141 embargo. It requires special client support, and Exim is probably the
4142 only MTA that supported it -- and would never use it because real RSA is
4143 always available. This code has been removed, because it had the bad
4144 effect of slowing Exim down by computing (never used) parameters for the
4145 RSA_EXPORT functionality.
4147 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4148 authenticator to fail if there's a tab character in the incoming data
4149 (there should never be unless someone is messing about, as it's supposed
4150 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4151 if the connection is using TLS or if the remote IP is the same as the
4152 local IP, and the "valid-client-cert option" if a client certificate has
4155 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4156 authenticators. This can be used for authorization after authentication
4157 succeeds. (In the case of plaintext, it servers for both authentication
4160 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4161 if any retry times were supplied.
4163 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4164 connection, where there is no client IP address to check. In this
4165 situation, the verify now always succeeds.
4167 PH/25 Applied John Jetmore's -Mset patch.
4169 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4171 PH/27 In a string expansion for a processed (not raw) header when multiple
4172 headers of the same name were present, leading whitespace was being
4173 removed from all of them, but trailing whitespace was being removed only
4174 from the last one. Now trailing whitespace is removed from each header
4175 before concatenation. Completely empty headers in a concatenation (as
4176 before) are ignored.
4178 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4179 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4181 PH/29 [Removed. This was a change that I later backed out, and forgot to
4182 correct the ChangeLog entry (that I had efficiently created) before
4183 committing the later change.]
4185 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4186 address errors (4xx response to RCPT) over the same connection as other
4187 messages routed to the same hosts. Such deliveries are always "forced",
4188 so retry times are not inspected. This resulted in far too many retries
4189 for the affected addresses. The effect occurred only when there were more
4190 hosts than the hosts_max_try setting in the smtp transport when it had
4191 the 4xx errors. Those hosts that it had tried were not added to the list
4192 of hosts for which the message was waiting, so if all were tried, there
4193 was no problem. Two fixes have been applied:
4195 (i) If there are any address or message errors in an SMTP delivery, none
4196 of the hosts (tried or untried) are now added to the list of hosts
4197 for which the message is waiting, so the message should not be a
4198 candidate for sending over the same connection that was used for a
4199 successful delivery of some other message. This seems entirely
4200 reasonable: after all the message is NOT "waiting for some host".
4201 This is so "obvious" that I'm not sure why it wasn't done
4202 previously. Hope I haven't missed anything, but it can't do any
4203 harm, as the worst effect is to miss an optimization.
4205 (ii) If, despite (i), such a delivery is accidentally attempted, the
4206 routing retry time is respected, so at least it doesn't keep
4207 hammering the server.
4209 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4210 in ${readsocket because some servers need this prod.
4212 PH/32 Added some extra debug output when updating a wait-xxx database.
4214 PH/33 The hint "could be header name not terminated by colon", which has been
4215 given for certain expansion errors for a long time, was not being given
4216 for the ${if def:h_colon_omitted{... case.
4218 PH/34 The spec says: "With one important exception, whenever a domain list is
4219 being scanned, $domain contains the subject domain." There was at least
4220 one case where this was not true.
4222 PH/35 The error "getsockname() failed: connection reset by peer" was being
4223 written to the panic log as well as the main log, but it isn't really
4224 panic-worthy as it just means the connection died rather early on. I have
4225 removed the panic log writing for the ECONNRESET error when getsockname()
4228 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4229 runs only) independently of the message's sender address. This meant
4230 that, if the 4xx error was in fact related to the sender, a different
4231 message to the same recipient with a different sender could confuse
4232 things. In particular, this can happen when sending to a greylisting
4233 server, but other circumstances could also provoke similar problems.
4234 I have changed the default so that the retry time for these errors is now
4235 based a combination of the sender and recipient addresses. This change
4236 can be overridden by setting address_retry_include_sender=false in the
4239 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4240 remote server are returned as part of bounce messages. This was not
4241 happening for LMTP over a pipe (the lmtp transport), but now it is the
4242 same for both kinds of LMTP.
4244 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4245 in header lines that were in fact CNAMEs. This is no longer the case.
4247 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4248 and queue runs started by the daemon processed all messages. This has
4249 been fixed so that -R and -S can now usefully be given with -q<time>.
4251 PH/40 Import PCRE release 6.7 (fixes some bugs).
4253 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4255 PH/42 Give an error if -q is specified more than once.
4257 PH/43 Renamed the variables $interface_address and $interface_port as
4258 $received_ip_address and $received_port, to make it clear that these
4259 values apply to message reception, and not to the outgoing interface when
4260 a message is delivered. (The old names remain recognized, of course.)
4262 PH/44 There was no timeout on the connect() call when using a Unix domain
4263 socket in the ${readsocket expansion. There now is.
4265 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4266 be meaningful with "accept".
4268 SC/01 Eximstats V1.43
4269 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4271 SC/02 Eximstats V1.44
4272 Use a glob alias rather than an array ref in the generated
4273 parser. This improves both readability and performance.
4275 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4276 Collect SpamAssassin and rejection statistics.
4277 Don't display local sender or destination tables unless
4278 there is data to show.
4279 Added average volumes into the top table text output.
4281 SC/04 Eximstats V1.46
4282 Collect data on the number of addresses (recipients)
4283 as well as the number of messages.
4285 SC/05 Eximstats V1.47
4286 Added 'Message too big' to the list of mail rejection
4287 reasons (thanks to Marco Gaiarin).
4289 SC/06 Eximstats V1.48
4290 Mainlog lines which have GMT offsets and are too short to
4291 have a flag are now skipped.
4293 SC/07 Eximstats V1.49 (Alain Williams)
4294 Added the -emptyok flag.
4296 SC/08 Eximstats V1.50
4297 Fixes for obtaining the IP address from reject messages.
4299 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4300 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4301 whitespace changes from 4.64-PH/27
4303 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4306 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4307 are found, allow negative numbers in numeric criteria)
4309 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4311 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4312 to match changes made in 4.64-PH/43
4314 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4316 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4317 "domain" is actually an IP address, and force a failure. This locks out
4318 those revolvers/nameservers that support "A-for-A" lookups, in
4319 contravention of the specifications.
4321 PH/48 When a host name was looked up from an IP address, and the subsequent
4322 forward lookup of the name timed out, the host name was left in
4323 $sender_host_name, contrary to the specification.
4325 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4326 restricted to single-key lookups, Exim was not diagnosing an error if
4327 * or *@ was used with a query-style lookup.
4329 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4331 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4332 long ago, but noone interested enough thought of it. Let's just say that
4333 the "1.1" means that there are some new functions that weren't there at
4334 some point in the past.
4336 PH/51 Error processing for expansion failure of helo_data from an smtp
4337 transport during callout processing was broken.
4339 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4340 tested/used via the -bh/-bhc/-bs options.
4342 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4343 bug, fixed in subsequent PCRE releases).
4345 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4346 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4348 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4354 SC/01 Use a glob alias rather than an array ref in eximstats generated
4355 parser. This improves both readability and performance.
4357 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4358 Don't display local sender or destination tables in eximstats unless
4359 there is data to show.
4360 Added average volumes into the eximstats top table text output.
4362 SC/03 Collect data on the number of addresses (recipients) as well
4363 as the number of messages in eximstats.
4365 TF/01 Correct an error in the documentation for the redirect router. Exim
4366 does (usually) call initgroups() when daemonizing.
4368 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4369 with consistent privilege compared to when running as a daemon.
4371 TF/03 Note in the spec that $authenticated_id is not set for local
4372 submissions from trusted users.
4374 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4375 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4377 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4378 by adding some example configuration directives to the default
4379 configuration file. A little bit of work is required to uncomment the
4380 directives and define how usernames and passwords are checked, but
4381 there is now a framework to start from.
4383 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4384 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4385 without this. I don't know how relevant this is to other LDAP libraries.
4387 PH/02 Add the verb name to the "unknown ACL verb" error.
4389 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4391 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4393 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4394 directory not expanded when it should be if an expanded home directory
4395 was set for the address (which is overridden by the transport).
4397 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4400 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4401 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4402 because it is too late at that time, and has no effect.
4404 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4405 security issue with \' (bugzilla #107). I could not use the
4406 PQescapeStringConn() function, because it needs a PGconn value as one of
4409 PH/08 When testing addresses using -bt, indicate those final addresses that
4410 are duplicates that would not cause an additional delivery. At least one
4411 person was confused, thinking that -bt output corresponded to deliveries.
4412 (Suppressing duplicates isn't a good idea as you lose the information
4413 about possibly different redirections that led to the duplicates.)
4415 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4416 systems where poll() doesn't work, in particular OS X.
4418 PH/10 Added more information to debugging output for retry time not reached.
4420 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4421 operations in malware.c.
4423 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4426 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4427 syslog_duplication set false, log lines that would normally be written
4428 both the the main log and to the reject log were not written to syslog at
4431 PH/14 In the default configuration, change the use of "message" in ACL warn
4432 statements to "add_header".
4434 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4435 not followed by a command (e.g. "seen endif").
4437 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4438 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4441 PH/17 Added extra conditions to the default value of delay_warning_condition
4445 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4446 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4447 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4450 The Auto-Submitted: and various List- headers are standardised, whereas I
4451 don't think Precedence: ever was.
4453 PH/18 Refactored debugging code in route_finduser() to show more information,
4454 in particular, the error code if getpwnam() issues one.
4456 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4457 This is apparently needed in addition to the PH/07 change above to avoid
4458 any possible encoding problems.
4460 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4461 but not after initializing Perl.
4463 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4464 output them only if debugging. By default they are written stderr,
4465 apparently, which is not desirable.
4467 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4470 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4473 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4475 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4476 authenticated or an ident call has been made. Suppress the default
4477 values for $authenticated_id and $authenticated_sender (but permit -oMai
4478 and -oMas) when testing with -bh.
4480 PH/24 Re-jigged the order of the tests in the default configuration so that the
4481 tests for valid domains and recipients precede the DNS black list and CSA
4482 tests, on the grounds that those ones are more expensive.
4484 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4485 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4486 This bug exists in every version of Exim that I still have, right back to
4489 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4490 However, an attempt to turn on case-sensitivity in a regex key by
4491 including (?-i) didn't work because the subject string was already
4492 lowercased, and the effects were non-intuitive. It turns out that a
4493 one-line patch can be used to allow (?-i) to work as expected.
4499 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4500 other effects) broke the use of negated acl sub-conditions.
4502 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4505 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4506 "Deny" causes Exim to reject the incoming connection with a 554 error.
4507 Unfortunately, if there is a major crisis, such as a disk failure,
4508 tcp-wrappers gives "deny", whereas what one would like would be some
4509 kind of temporary error. A kludge has been added to help with this.
4510 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4511 554 error is used if errno is still zero or contains ENOENT (which occurs
4512 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4515 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4517 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4518 errors. Otherwise a message that provokes a temporary error (when other
4519 messages do not) can cause a whole host to time out.
4521 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4522 addresses were routed directly to files or pipes from a redirect router.
4523 File deliveries just didn't batch; pipe deliveries might have suffered
4526 PH/06 A failure to get a lock for a hints database would erroneously always say
4527 "Failed to get write lock", even when it was really a read lock.
4529 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4530 of 0600. This has been changed to use the value of the lockfile_mode
4531 option (which defaults to 0600).
4533 PH/08 Applied small patch from the Sieve maintainer.
4535 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4536 folder from quota calculations, a direct delivery into this folder messed
4537 up the contents of the maildirsize file. This was because the regex was
4538 used only to exclude .Trash (or whatever) when the size of the mailbox
4539 was calculated. There was no check that a delivery was happening into an
4540 excluded directory. This bug has been fixed by ignoring all quota
4541 processing for deliveries into excluded directories.
4543 PH/10 Added the maildirfolder_create_regex option to appendfile.
4549 PH/01 The code for finding all the local interface addresses on a FreeBSD
4550 system running IPv6 was broken. This may well have applied to all BSD
4551 systems, as well as to others that have similar system calls. The broken
4552 code found IPv4 interfaces correctly, but gave incorrect values for the
4553 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4554 that it would not match correctly against @[] and not recognize the IPv6
4557 PH/02 The ipliteral router was not recognizing addresses of the form user@
4558 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4560 PH/03 Added disable_ipv6.
4562 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4563 decoded form, because it is most often used to construct To: headers
4564 lines in autoreplies, and the decoded form may well be syntactically
4565 invalid. However, $reply_address has leading white space removed, and all
4566 newlines turned into spaces so that the autoreply transport does not
4569 PH/05 If group was specified without a user on a router, and no group or user
4570 was specified on a transport, the group from the router was ignored.
4572 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4573 for visible compile-time settings that can be used to change these
4574 numbers, for those that want even more. Backwards compatibility with old
4575 spool files has been maintained. However, going back to a previous Exim
4576 release will lost any variables that are in spool files.
4578 PH/07 Two small changes when running in the test harness: increase delay when
4579 passing a TCP/IP connection to a new process, in case the original
4580 process has to generate a bounce, and remove special handling of
4581 127.0.0.2 (sic), which is no longer necessary.
4583 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4584 be the same on different OS.
4586 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4589 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4590 whether --show-vars was specified or not
4592 JJ/02 exipick: Added support for new ACL variable spool format introduced
4595 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4596 syntactically invalid From: or Reply-to: line, and a filter used this to
4597 generate an autoreply, and therefore failed to obtain an address for the
4598 autoreply, Exim could try to deliver to a non-existent relative file
4599 name, causing unrelated and misleading errors. What now happens is that
4600 it logs this as a hard delivery error, but does not attempt to create a
4603 PH/11 The exinext utility has a -C option for testing purposes, but although
4604 the given file was scanned by exinext itself; it wasn't being passed on
4605 when Exim was called.
4607 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4608 an end-of-file indication when reading a command response.
4610 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4611 compiled. In many other places in Exim, IPv6 addresses are always
4612 recognized, so I have changed this. It also means that IPv4 domain
4613 literals of the form [IPV4:n.n.n.n] are now always recognized.
4615 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4616 used if the router is not running as root, for example, when verifying at
4617 ACL time, or when using -bh. The debugging output from this situation was
4618 non-existent - all you got was a failure to exec. I have made two
4621 (a) Failures to set uid/gid, the current directory, or a process leader
4622 in a subprocess such as that created by queryprogram now generate
4623 suitable debugging output when -d is set.
4625 (b) The queryprogram router detects when it is not running as root,
4626 outputs suitable debugging information if -d is set, and then runs
4627 the subprocess without attempting to change uid/gid.
4629 PH/15 Minor change to Makefile for building test_host (undocumented testing
4632 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4633 additional section of a DNS packet that returns MX or SRV records.
4634 Instead, it always explicitly searches for A/AAAA records. This avoids
4635 major problems that occur when a DNS server includes only records of one
4636 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4637 fixed another bug: if SRV records were looked up and the corresponding
4638 address records were *not* found in the additional section, the port
4639 values from the SRV records were lost.
4641 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4642 using the correct key (the original address) when searching the retry
4643 rules in order to find which one to use for generating the retry hint.
4645 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4646 adding the default one. Similarly, if it contains a Reply-To: header, the
4647 errors_reply_to option, if set, is not used.
4649 PH/19 When calculating a retry time, Exim used to measure the "time since
4650 failure" by looking at the "first failed" field in the retry record. Now
4651 it does not use this if it is later than than the arrival time of the
4652 message. Instead it uses the arrival time. This makes for better
4653 behaviour in cases where some deliveries succeed, thus re-setting the
4654 "first failed" field. An example is a quota failure for a huge message
4655 when small messages continue to be delivered. Without this change, the
4656 "time since failure" will always be short, possible causing more frequent
4657 delivery attempts for the huge message than are intended.
4658 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4660 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4661 $1, $2, $3) because the numerical variables can be reset during some
4662 expansion items (e.g. "match"), thereby losing the authentication data.
4664 PH/21 Make -bV show the size of off_t variables so that the test suite can
4665 decide whether to run tests for quotas > 2G.
4667 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4668 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4669 filecount value is greater than 2G or if a quota value is greater than 2G
4670 on a system where the size of off_t is not greater than 4, a panic error
4673 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4674 never match. The debug and -bh output now contains an explicit error
4675 message indicating a malformed IPv4 address or mask.
4677 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4678 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4679 PH/23 above applies.
4681 PH/25 Do not write to syslog when running in the test harness. The only
4682 occasion when this arises is a failure to open the main or panic logs
4683 (for which there is an explicit test).
4685 PH/26 Added the /no_tell option to "control=freeze".
4687 PH/27 If a host name lookup failed very early in a connection, for example, if
4688 the IP address matched host_lookup and the reverse lookup yielded a name
4689 that did not have a forward lookup, an error message of the form "no IP
4690 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4691 could be logged. Now it outputs the IP address instead of "NULL".
4693 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4694 allows the sender and the authenticated sender to be set when
4695 submitting a message from within Exim. Since child_open_exim() is
4696 documented for local_scan(), the new function should be too.
4698 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4699 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4700 results in an empty string is now treated as unset.
4702 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4704 PH/31 Added sender_verify_fail logging option.
4706 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4707 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4708 tidied the source and removed it altogether.
4710 PH/33 When a queue run was abandoned because the load average was too high, a
4711 log line was always written; now it is written only if the queue_run log
4712 selector is set. In addition, the log line for abandonment now contains
4713 information about the queue run such as the pid. This is always present
4714 in "start" and "stop" lines but was omitted from the "abandon" line.
4716 PH/34 Omit spaces between a header name and the colon in the error message that
4717 is given when verify = headers_syntax fails (if there are lots of them,
4718 the message gets confusing).
4720 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4721 names, as there are now some PTR records that contain slashes. This check
4722 is only to protect against broken name servers that fall over on strange
4723 characters, so the fact that it applies to all lookups doesn't matter.
4725 PH/36 Now that the new test suite is complete, we can remove some of the
4726 special code in Exim that was needed for the old test suite. For example,
4727 sorting DNS records because real resolvers return them in an arbitrary
4728 order. The new test suite's fake resolver always returns records in the
4731 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4732 bounces) except when queue_only is set, to avoid logging races between
4733 the different processes.
4735 PH/38 Panic-die if .include specifies a non-absolute path.
4737 PH/39 A tweak to the "H" retry rule from its user.
4739 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4740 a label. They prevented compilation on older perls.
4742 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4743 a warning to be raised on newish perls.
4745 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4746 on queue. Changes to match documented behaviour of showing count of
4747 messages matching specified criteria.
4749 PH/40 Changed the default ident timeout from 30s to 5s.
4751 PH/41 Added support for the use of login_cap features, on those BSD systems
4752 that have them, for controlling the resources used by pipe deliveries.
4754 PH/42 The content-scanning code uses fopen() to create files in which to put
4755 message data. Previously it was not paying any attention to the mode of
4756 the files. Exim runs with umask(0) because the rest of the code creates
4757 files with open(), and sets the required mode explicitly. Thus, these
4758 files were ending up world-writeable. This was not a big issue, because,
4759 being within the spool directory, they were not world-accessible. I have
4760 created a function called modefopen, which takes an additional mode
4761 argument. It sets umask(777), creates the file, chmods it to the required
4762 mode, then resets the umask. All the relevant calls to fopen() in the
4763 content scanning code have been changed to use this function.
4765 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4766 to 24 hours. This avoids potential overflow problems when processing G
4767 and H retry rules. I suspect nobody ever tinkers with this value.
4769 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4771 PH/45 When the plaintext authenticator is running as a client, the server's
4772 challenges are checked to ensure they are valid base64 strings. By
4773 default, the authentication attempt is cancelled if an invalid string is
4774 received. Setting client_ignore_invalid_base64 true ignores these errors.
4775 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4776 they are received. Thus, the responses can be made to depend on the
4777 challenges. If an invalid string is ignored, an empty string is placed in
4780 PH/46 Messages that are created by the autoreply transport now contains a
4781 References: header, in accordance with RFCs 2822 and 3834.
4783 PH/47 Added authenticated_sender_force to the smtp transport.
4785 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4787 PH/49 Installed latest patch from the Sieve maintainer.
4789 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4790 was also set, the appendfile transport was unnecessarily scanning a
4791 directory of message files (e.g. for maildir delivery) to find the count
4792 of files (along with the size), even though it did not need this
4793 information. It now does the scan only if it needs to find either the
4794 size of the count of files.
4796 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4798 PH/52 Two bugs concerned with error handling when the smtp transport is
4801 (i) Exim was not creating retry information for temporary errors given
4802 for individual recipients after the DATA command when the smtp transport
4803 was used in LMTP mode. This meant that they could be retried too
4804 frequently, and not timed out correctly.
4806 (ii) Exim was setting the flag that allows error details to be returned
4807 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4808 recipients that were returned after the DATA command.
4810 PH/53 This is related to PH/52, but is more general: for any failing address,
4811 when detailed error information was permitted to be returned to the
4812 sender, but the error was temporary, then after the final timeout, only
4813 "retry timeout exceeded" was returned. Now it returns the full error as
4814 well as "retry timeout exceeded".
4816 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4817 do this, and (what is worse) MTAs that accept it.
4819 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4820 will now be deprecated.
4822 PH/56 New os.c-cygwin from the Cygwin maintainer.
4824 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4825 formats (previously only available in exim formats via -bpr, -bpru,
4826 and -bpra. Now also available in native and exiqgrep formats)
4828 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4829 with very large, slow to parse queues
4831 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4833 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4835 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4836 responses to authentication challenges, though it was showing the
4837 challenges; (ii) I've removed the CR characters from the debug output for
4840 PH/58 Allow for the insertion of a newline as well as a space when a string
4841 is turned into more than one encoded-word during RFC 2047 encoding. The
4842 Sieve code now uses this.
4844 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4845 data_4xx, lost_connection, tls_required.
4847 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4848 message was being sent as an SMTP response.
4850 PH/61 Add -l and -k options to exicyclog.
4852 PH/62 When verifying, if an address was redirected to one new address, so that
4853 verification continued, and the new address failed or deferred after
4854 having set something in $address_data, the value of $address_data was not
4855 passed back to the ACL. This was different to the case when no
4856 redirection occurred. The value is now passed back in both cases.
4858 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4859 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4860 use login_cap.h, so on its own it isn't the distinguishing feature. The
4861 new name refers directly to the setclassresources() function.
4863 PH/65 Added configuration files for NetBSD3.
4865 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4867 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4868 is preferred over IPv4.
4870 PH/68 The bounce_return_message and bounce_return_body options were not being
4871 honoured for bounces generated during the reception of non-SMTP messages.
4872 In particular, this applied to messages rejected by the ACL. This bug has
4873 been fixed. However, if bounce_return_message is true and bounce_return_
4874 body is false, the headers that are returned for a non-SMTP message
4875 include only those that have been read before the error was detected.
4876 (In the case of an ACL rejection, they have all been read.)
4878 PH/69 The HTML version of the specification is now built in a directory called
4879 spec_html instead of spec.html, because the latter looks like a path with
4880 a MIME-type, and this confuses some software.
4882 PH/70 Catch two compiler warnings in sieve.c.
4884 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4885 function verify_get_ident() calls ip_connect() to connect a socket, but
4886 if the "connect()" function timed out, ip_connect() used to close the
4887 socket. However, verify_get_ident() also closes the socket later, and in
4888 between Exim writes to the log, which may get opened at this point. When
4889 the socket was closed in ip_connect(), the log could get the same file
4890 descriptor number as the socket. This naturally causes chaos. The fix is
4891 not to close the socket in ip_connect(); the socket should be closed by
4892 the function that creates it. There was only one place in the code where
4893 this was missing, in the iplookup router, which I don't think anybody now
4894 uses, but I've fixed it anyway.
4896 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4897 well as to direct DNS lookups. Otherwise the handling of names in host
4898 lists is inconsistent and therefore confusing.
4904 PH/01 Two changes to the default runtime configuration:
4906 (1) Move the checks for relay_from_hosts and authenticated clients from
4907 after to before the (commented out) DNS black list checks.
4909 (2) Add control=submission to the relay_from_hosts and authenticated
4910 clients checks, on the grounds that messages accepted by these
4911 statements are most likely to be submissions.
4913 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4915 (1) Generate an error if the third argument for the ${prvs expansion is
4918 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4921 (3) Reset the variables that are obtained from the first argument of
4922 ${prvscheck and used in the second argument before leaving the code,
4923 because their memory is reclaimed, so using them afterwards may do
4926 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4927 one (it's much easier than Tom thought :-).
4929 (5) Because of (4), we can now allow for the use of $prvscheck_result
4930 inside the third argument.
4932 PH/03 For some reason, the default setting of PATH when running a command from
4933 a pipe transport was just "/usr/bin". I have changed it to
4936 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4937 anything to be listed in the output from -bV.
4939 PH/05 When a filter generated an autoreply, the entire To: header line was
4940 quoted in the delivery log line, like this:
4942 => >A.N.Other <ano@some.domain> <original@ddress> ...
4944 This has been changed so that it extracts the operative address. There
4945 may be more than one such address. If so, they are comma-separated, like
4948 => >ano@some.domain,ona@other.domain <original@ddress> ...
4950 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4951 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4952 not being looked up in the rDNS to get a host name, Exim was showing the
4953 IP address twice in Received: lines, even though the IP addresses were
4954 identical. For example:
4956 Received: from [1.2.3.4] (helo=[1.2.3.4])
4958 However, if the real host name was known, it was omitting the HELO data
4959 if it matched the actual IP address. This has been tidied up so that it
4960 doesn't show the same IP address twice.
4962 PH/07 When both +timestamp and +memory debugging was on, the value given by
4963 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4964 called by the debug printing, thereby overwriting the timestamp buffer.
4965 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4967 PH/08 When the original message was included in an autoreply transport, it
4968 always said "this is a copy of the message, including all the headers",
4969 even if body_only or headers_only was set. It now gives an appropriate
4972 PH/09 Applied a patch from the Sieve maintainer which:
4974 o fixes some comments
4975 o adds the (disabled) notify extension core
4976 o adds some debug output for the result of if/elsif tests
4977 o points to the current vacation draft in the documentation
4978 and documents the missing references header update
4982 o fixes a bug in processing the envelope test (when testing
4983 multiple envelope elements, the last element determined the
4986 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4987 Electronic Mail") by including:
4989 Auto-submitted: auto-generated
4991 in the messages that it generates (bounce messages and others, such as
4992 warnings). In the case of bounce messages for non-SMTP messages, there was
4993 also a typo: it was using "Auto_submitted" (underscore instead of
4994 hyphen). Since every message generated by Exim is necessarily in response
4995 to another message, thes have all been changed to:
4997 Auto-Submitted: auto-replied
4999 in accordance with these statements in the RFC:
5001 The auto-replied keyword:
5003 - SHOULD be used on messages sent in direct response to another
5004 message by an automatic process,
5006 - MUST NOT be used on manually-generated messages,
5008 - MAY be used on Delivery Status Notifications (DSNs) and Message
5009 Disposition Notifications (MDNs),
5011 - MUST NOT be used on messages generated by automatic or periodic
5012 processes, except for messages which are automatic responses to
5015 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5016 to the default Received: header definition.
5018 PH/12 Added log selector acl_warn_skipped (default on).
5020 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5021 variables because (a) they are in the wrong storage pool and (b) even if
5022 they were copied, it wouldn't work properly because of the caching.
5024 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5025 checking when decoding. Apparently there are clients that generate
5026 overlong encoded strings. Why am I not surprised?
5028 PH/15 If the first argument of "${if match_address" was not empty, but did not
5029 contain an "@" character, Exim crashed. Now it writes a panic log message
5030 and treats the condition as false.
5032 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5034 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5035 extension "envelope-auth". The code is finished and in agreement with
5036 other implementations, but there is no documentation so far and in fact,
5037 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5038 not changing the active code.
5040 Print executed "if" and "elsif" statements when debugging is used. This
5041 helps a great deal to understand what a filter does.
5043 Document more things not specified clearly in RFC3028. I had all this
5044 sorted out, when out of a sudden new issues came to my mind. Oops."
5046 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5049 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5050 canonical form (as documented). However, after a host name lookup from
5051 the IP address, check_host() was doing a simple string comparison with
5052 addresses acquired from the DNS when checking that the found name did
5053 have the original IP as one of its addresses. Since any found IPv6
5054 addresses are likely to be in abbreviated form, the comparison could
5055 fail. Luckily, there already exists a function for doing the comparison
5056 by converting both addresses to binary, so now that is used instead of
5057 the text comparison.
5059 PH/20 There was another similar case to PH/19, when a complete host name was
5060 given in a host list; looking up its IP address could give an abbreviated
5061 form, whereas the current host's name might or might not be abbreviated.
5062 The same fix has been applied.
5068 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5069 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5072 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5073 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5075 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5077 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5078 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5079 Also, for "Auto-Submitted", treat anything other than "no" as
5080 non-personal, in accordance with RFC 3834. (Previously it treated
5081 anything starting "auto-" as non-personal.)
5083 TF/01 The control=submission/name=... option had a problem with syntax
5084 errors if the name included a slash character. The /name= option
5085 now slurps the rest of the string, so it can include any characters
5086 but it must come last in the list of options (after /sender_retain
5089 PH/05 Some modifications to the interface to the fake nameserver for the new
5097 TK/01 Added the "success_on_redirect" address verification option. See
5098 NewStuff for rationale and an example.
5100 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5102 PH/02 Patch to exigrep to allow it to work on syslog lines.
5104 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5105 fread() to skip over the body file's header line, because in Cygwin the
5106 header line is locked and is inaccessible.
5108 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5109 co-exist for some time) to make it clear that it is the Exim ID that is
5110 referenced, not the Message-ID: header line.
5112 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5113 string_format() function, because snprintf() does not exist on all
5116 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5117 ${stat: expansion item.
5119 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5120 protocol synchronization error", to keep the pedants happy.
5122 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5123 well as for IRIX systems, when gcc is being used. See the host.c source
5126 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5128 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5131 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5132 transport and to the smtp transport in LMTP mode.
5134 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5136 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5137 run a filter in a subprocess. This could lead to confusion in subsequent
5138 lookups in the parent process. There should also be a search_tidyup() at
5139 the end of the subprocess.
5141 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5142 only if the host matched helo_try_verify_hosts, which caused the
5143 verification to occur when the EHLO/HELO command was issued. The ACL just
5144 tested the remembered result. Now, if a previous verification attempt has
5145 not happened, "verify = helo" does it there and then.
5147 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5149 TK/03 Fix log output including CR from clamd.
5151 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5152 exist provoked a memory error which could cause a segfault.
5154 PH/15 Installed PCRE 6.2
5156 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5158 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5159 of the problem. Specifically, suggested +O2 rather than +O1 for the
5162 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5164 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5165 record, the port was not correctly logged when the outgoing_port log
5166 selector was set (it logged the transort's default port).
5168 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5169 fallback_hosts, and "hosts" in the smtp transport.
5171 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5172 host errors such as "Connection refused".
5174 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5175 authentication with radiusclient 0.4.9:
5177 - Error returned from rc_read_config was caught wrongly
5178 - Username/password not passed on to radius server due to wrong length.
5180 The presumption is that some radiusclient API changes for 4.51/PH/17
5181 were not taken care of correctly. The code is still untested by me (my
5182 Linux distribution still has 0.3.2 of radiusclient), but it was
5183 contributed by a Radius user.
5185 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5186 expanding the "port" option of the smtp transport.
5188 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5189 while reading a MIME header. Thanks to Tom Hughes for a patch.
5191 PH/24 Include config.h inside local_scan.h so that configuration settings are
5194 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5195 that in an ACL for RCPT (for example), you can examine exactly what was
5198 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5199 commands, but it was not correctly comparing the address with the actual
5200 client host address. Thus, it would show the EHLO address in Received:
5201 header lines when this was not necessary.
5203 PH/27 Added the % operator to ${eval:}.
5205 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5206 it should be ignoring failures (because with -C, for example, it has lost
5207 privilege). It wasn't ignoring creation failures other than "already
5210 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5213 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5214 because an input line was too long, either on its own, or by virtue of
5215 too many continuations, the temporary file was not being removed, and the
5216 return code was incorrect.
5218 PH/31 Missing "BOOL" in function definition in filtertest.c.
5220 PH/32 Applied Sieve patches from the maintainer.
5222 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5224 PH/33 Added "verify = not_blind".
5226 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5227 Local/Makefile (with some defaults set). These are used in built scripts
5228 such as exicyclog, but they have never been used in the exim_install
5229 script (though there are many overriding facilities there). I have
5230 arranged that the exim_install script now takes note of these two
5233 PH/35 Installed configuration files for Dragonfly.
5235 PH/36 When a locally submitted message by a trusted user did not contain a
5236 From: header, and the sender address was obtained from -f or from an SMTP
5237 MAIL command, and the trusted user did not use -F to supply a sender
5238 name, $originator_name was incorrectly used when constructing a From:
5239 header. Furthermore, $originator_name was used for submission mode
5240 messages from external hosts without From: headers in a similar way,
5241 which is clearly wrong.
5243 PH/37 Added control=suppress_local_fixups.
5245 PH/38 When log_selector = +received_sender was set, and the addition of the
5246 sender made the log line's construction buffer exactly full, or one byte
5247 less than full, an overflow happened when the terminating "\n" was
5250 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5251 when the result of a list match is failure because a DNS lookup failed.
5253 PH/40 RM_COMMAND is now used in the building process.
5255 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5256 the "build-* directories that it finds.
5258 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5259 address in a domain literal was a prefix of an interface address.
5261 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5262 when verifying a sender address, unless rewrite_headers is false.
5264 PH/44 Wrote a long comment about why errors_to addresses are verified as
5265 recipients, not senders.
5267 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5268 the ratelimit ACL was added.
5270 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5272 PH/46 Added extra information about PostgreSQL errors to the error string.
5274 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5275 avoiding the need to install special zones in a real server. This is
5276 backwards compatible; if it can't find the fake resolver, it drops back.
5277 Thus, both old and new test suites can be run.
5279 TF/02 Added util/ratelimit.pl
5281 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5282 clock is set back in time.
5284 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5285 Candler <B.Candler@pobox.com>.
5287 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5288 OK for addresses that are the result of redirections.
5290 PH/48 A number of further additions for the benefit of the new test suite,
5291 including a fake gethostbyname() that interfaces to the fake DNS resolver
5294 TF/06 The fix for widen_domains has also been applied to qualify_single and
5295 search_parents which are the other dnslookup options that can cause
5298 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5301 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5303 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5304 service" if it encountered an MX record with an empty target hostname.
5305 The message is now "an MX or SRV record indicated no SMTP service".
5307 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5308 if the DNS of the sending site is misconfigured. This is quite a
5309 common situation. This change restores the behaviour of treating a
5310 helo verification defer as a failure.
5312 PH/51 If self=fail was set on a router, the bounce message did not include the
5313 actual error message.
5319 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5321 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5322 command itself ended in error, the underlying message about the transport
5323 filter timeout was being overwritten with the pipe command error. Now the
5324 underlying error message should be appended to the second error message.
5326 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5328 PH/02 Reduce the amount of output that "make" produces by default. Full output
5329 can still be requested.
5331 PH/03 The warning log line about a condition test deferring for a "warn" verb
5332 was being output only once per connection, rather than after each
5333 occurrence (because it was using the same function as for successful
5334 "warn" verbs). This seems wrong, so I have changed it.
5336 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5337 it should not have, which might have caused a crash in the right
5338 circumstances, but probably never did.
5340 PH/04 Installed a modified version of Tony Finch's patch to make submission
5341 mode fix the return path as well as the Sender: header line, and to
5342 add a /name= option so that you can make the user's friendly name appear
5345 TF/03 Added the control = fakedefer ACL modifier.
5347 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5348 Mark Lowes for thorough testing.
5350 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5352 TK/03 Merged latest SRS patch from Miles Wilton.
5354 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5355 with the definition in sysexits.h (which is #included earlier).
5356 Fortunately, Exim does not actually use EX_OK. The code used to try to
5357 preserve the sysexits.h value, by assuming that macro definitions were
5358 scanned for macro replacements. I have been disabused of this notion,
5359 so now the code just undefines EX_OK before #including unistd.h.
5361 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5362 in the smtp transport. When a block could not be written in a single
5363 write() function, the timeout was being re-applied to each part-write.
5364 This seems wrong - if the receiver was accepting one byte at a time it
5365 would take for ever. The timeout is now adjusted when this happens. It
5366 doesn't have to be particularly precise.
5368 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5369 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5371 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5372 without a domain if the check to <postmaster@domain> fails.
5374 SC/01 Eximstats: added -xls and the ability to specify output files
5375 (patch written by Frank Heydlauf).
5377 SC/02 Eximstats: use FileHandles for outputting results.
5379 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5381 SC/04 Eximstats: fixed display of large numbers with -nvr option
5383 SC/05 Eximstats: fixed merging of reports with empty tables.
5385 SC/06 Eximstats: added the -include_original_destination flag
5387 SC/07 Eximstats: removed tabs and trailing whitespace.
5389 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5391 TK/06 MBOX spool code: Add real "From " MBOX separator line
5392 so the .eml file is really in mbox format (even though
5393 most programs do not really care). Patch from Alex Miller.
5395 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5396 The latter is generated from $received_to and is only set if the
5397 message has one envelope recipient. SA can use these headers,
5398 obviously out-of-the-box. Patch from Alex Miller.
5400 PH/08 The ${def test on a variable was returning false if the variable's
5401 value was "0", contrary to what the specification has always said!
5402 The result should be true unless the variable is empty.
5404 PH/09 The syntax error of a character other than { following "${if
5405 def:variable_name" (after optional whitespace) was not being diagnosed.
5406 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5407 accidental colon was present, for example, could give incorrect results.
5409 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5410 result is used (not including appendfile, where other changes are about
5413 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5414 This involved changing a lot of size variables from int to off_t. It
5415 should work with maildirs and everything.
5417 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5418 spamd dying while we are connected to it.
5420 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5423 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5424 function for BDB 4.3.
5426 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5428 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5429 the sources for PCRE have been re-arranged and more files are now
5432 PH/15 The code I had for printing potentially long long variables in PH/11
5433 above was not the best (it lost precision). The length of off_t variables
5434 is now inspected at build time, and an appropriate printing format (%ld
5435 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5436 to be "long long int" or "long int". This is needed for the internal
5437 formatting function string_vformat().
5439 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5440 the configuration file to be ":syslog", then the script "guesses" where
5441 the logs files are, rather than using the compiled in default. In our
5442 case the guess is not the same as the compiled default, so the script
5443 suddenly stopped working when I started to use syslog. The patch checks
5444 to see if log_file_path is "". If so, it attempts to read it from exim
5445 with no configuration file to get the compiled in version, before it
5446 falls back to the previous guessing code."
5448 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5449 implementing BATV in an Exim configuration. See NewStuff for the gory
5452 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5453 Makefile that are specific to HP-UX.
5455 PH/18 If the "use_postmaster" option was set for a recipient callout together
5456 with the "random" option, the postmaster address was used as the MAIL
5457 FROM address for the random test, but not for the subsequent recipient
5458 test. It is now used for both.
5460 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5461 patch removes a few documentation additions to RFC 3028, because the
5462 latest draft now contains them. It adds the new en;ascii-case comparator
5463 and a new error check for 8bit text in MIME parts. Comparator and
5464 require names are now matched exactly. I enabled the subaddress
5465 extension, but it is not well tested yet (read: it works for me)."
5467 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5468 rework some of the code of TK/09 above to avoid the hardwired use of
5469 "%lld" and "long long". Replaced the call to snprintf() with a call to
5472 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5473 records point to non-existent hosts", "retry timeout exceeded", and
5474 "retry time not reached for any host after a long failure period".
5476 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5477 experimental DomainKeys support:
5479 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5480 (2) On an error such as an illegally used "control", the wrong name for
5481 the control was given.
5483 These problems did NOT occur unless DomainKeys support was compiled.
5485 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5487 PH/24 Added ${if match_ip condition.
5489 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5490 databases so that it will be absolutely obvious if a crash occurs in the
5491 DB library. This is a regular occurrence (often caused by mis-matched
5494 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5495 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5496 picked up on a user's system that detects such things. There doesn't seem
5497 to be a gcc warning option for this - only an attribute that has to be
5498 put on the function's prototype. It seems that in Fedora Core 4 they have
5499 set this on a number of new functions. No doubt there will be more in due
5502 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5503 specify a transport. However, if an address that was verified by such a
5504 router was the subject of a callout, Exim crashed because it tried to
5505 read the rcpt_include_affixes from the non-existent transport. Now it
5506 just assumes that the setting of that option is false. This bug was
5507 introduced by 4.51/PH/31.
5509 PH/28 Changed -d+all to exclude +memory, because that information is very
5510 rarely of interest, but it makes the output a lot bigger. People tend to
5511 do -d+all out of habit.
5513 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5514 code in os-type was giving problems when libc.so lives in lib64, like on
5517 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5518 aren't the modern standard, and it seems that some systems' include files
5519 don't always have them. Exim was already checking for some of the newer
5520 ones like T_AAAA, and defining it itself. I've added checks for all the
5521 record types that Exim uses.
5523 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5524 not automatically generating a new one, as it is supposed to. This
5525 prevented TLS from working. If the file did exist, but contained invalid
5526 data, a new version was generated, as expected. It was only the case of a
5527 non-existent file that was broken.
5529 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5530 with a change in libdomainkeys > 0.64.
5532 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5533 from DNS. If the selector record carries the flag, it now has
5534 precedence over the domain-wide flag.
5536 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5538 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5539 the use of a port name that isn't defined in /etc/services) occurred, the
5540 message was deferred as in a normal delivery, and thus remained on the
5541 spool, instead of being failed because of the mua_wrapper setting. This
5542 is now fixed, and I tidied up some of the mua_wrapper messages at the
5545 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5546 the messages in a hash of arrays rather than using individual hashes.
5547 This is a bit cleaner and results in dramatic memory savings, albeit
5548 at a slight CPU cost.
5550 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5551 as requested by Marc Sherman.
5553 SC/10 Eximstats: added histograms for user specified patterns as requested
5556 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5558 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5559 fopen() in the content-scanning modules that did not already have it.
5565 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5566 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5568 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5570 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5572 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5573 to test Sieve filters that use "vacation".
5575 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5576 that changes the way the GnuTLS parameters are stored in the cache file.
5577 The new format can be generated externally. For backward compatibility,
5578 if the data in the cache doesn't make sense, Exim assumes it has read an
5579 old-format file, and it generates new data and writes a new file. This
5580 means that you can't go back to an older release without removing the
5583 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5584 work if there are any delivery delays because "one_time" forces the
5585 parent to be marked "delivered", so its unseen clone is never tried
5586 again. For this reason, Exim now forbids the simultaneous setting of
5589 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5590 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5591 entry if you want to know the details. The fix, however, neglected to
5592 consider the case when local delivery batching is involved. The test for
5593 "previously delivered" was not happening when checking to see if an
5594 address could be batched with a previous (undelivered) one; under
5595 certain circumstances this could lead to multiple deliveries to the same
5598 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5599 in its include files, and this causes problems building Exim.
5601 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5602 header_syntax) but Exim was just ignoring anything given after a slash.
5603 In particular, this caused confusion with an attempt to use "verify =
5604 reverse_host_lookup/defer_ok". An error is now given when options are
5605 supplied for verify items that do not have them. (Maybe reverse_host_
5606 lookup should have a defer_ok option, but that's a different point.)
5608 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5609 defined by RFC 821) to 2048, because there were problems with some AUTH
5610 commands, and RFC 1869 says the size should be increased for extended
5611 SMTP commands that take arguments.
5613 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5616 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5617 "unknown" error; now it says that the functionality isn't in the binary.
5619 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5620 an address' error message when a string expansion fails (syntax or
5621 whatever). Otherwise the password may appear in the log. Following change
5622 PH/42 below, there is no longer a chance of it appearing in a bounce
5625 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5627 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5628 of its addresses was ever tried. (Bugzilla bug #2.)
5630 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5631 the result incorrectly in the debug output. (It correctly added a newline
5632 to what was transported.)
5634 TF/01 Added $received_time.
5636 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5637 commented out examples of how to interface to a virus scanner and to
5638 SpamAssassin. Also added commented examples of av_scanner and
5639 spamd_address settings.
5641 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5642 and controls are allowed in which ACLs. There were a couple of minor
5643 errors. Some of the entries in the conditions table (which is a table of
5644 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5645 negation of where the condition IS allowed.
5647 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5649 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5650 header file does not have a version number, so I've had to invent a new
5651 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5652 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5653 radiusclient), but it was contributed by a Radius user.
5655 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5656 files or directories, for OpenSSL.
5658 PH/19 When an Exim process that is running as root has to create an Exim log
5659 file, it does so in a subprocess that runs as exim:exim so as to get the
5660 ownership right at creation (otherwise, other Exim processes might see
5661 the file with the wrong ownership). There was no test for failure of this
5662 fork() call, which would lead to the process getting stuck as it waited
5663 for a non-existent subprocess. Forks do occasionally fail when resources
5664 run out. I reviewed all the other calls to fork(); they all seem to check
5667 PH/20 When checking for unexpected SMTP input at connect time (before writing
5668 the banner), Exim was not dealing correctly with a non-positive return
5669 from the read() function. If the client had disconnected by this time,
5670 the result was a log entry for a synchronization error with an empty
5671 string after "input=" when read() returned zero. If read() returned -1
5672 (an event I could not check), uninitialized data bytes were printed.
5673 There were reports of junk text (parts of files, etc) appearing after
5676 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5678 PH/22 Added support for macro redefinition, and (re)definition in between
5679 driver and ACL definitions.
5681 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5682 forgetting to use the resulting value; it was using the unexpanded value.
5684 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5685 hadn't been configured. The fix is from Juergen Kreileder, who
5686 understands it better than I do:
5688 "Here's what I see happening with three configured cyrus_sasl
5689 authenticators configured (plain, login, cram-md5):
5691 On startup auth_cyrus_sasl_init() gets called for each of these.
5692 This means three calls to sasl_listmech() without a specified mech_list.
5693 => SASL tests which mechs of all available mechs actually work
5694 => three warnings about OTP not working
5695 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5697 With the patch, sasl_listmech() also gets called three times. But now
5698 SASL's mech_list option is set to the server_mech specified in the the
5699 authenticator. Or in other words, the answer from sasl_listmech()
5700 gets limited to just the mech you're testing for (which is different
5702 => the return list contains just 'plain' or 'login', 'cram-md5' or
5703 nothing depending on the value of ob->server_mech.
5705 I've just tested the patch: Authentication still works fine,
5706 unavailable mechs specified in the exim configuration are still
5707 caught, and the auth.log warnings about OTP are gone."
5709 PH/25 When debugging is enabled, the contents of the command line are added
5710 to the debugging output, even when log_selector=+arguments is not
5713 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5714 answer is "GNU", and only if the return is "GNU/something" is the answer
5717 PH/27 $acl_verify_message is now set immediately after the failure of a
5718 verification in an ACL, and so is available in subsequent modifiers. In
5719 particular, the message can be preserved by coding like this:
5721 warn !verify = sender
5722 set acl_m0 = $acl_verify_message
5724 Previously, $acl_verify_message was set only while expanding "message"
5725 and "log_message" when a very denied access.
5727 PH/28 Modified OS/os.c-Linux with
5729 -#ifndef OS_LOAD_AVERAGE
5730 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5732 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5733 nomenclature these days.)
5735 PH/29 Installed patch from the Sieve maintainer that adds the options
5736 sieve_useraddress and sieve_subaddress to the redirect router.
5738 PH/30 In these circumstances:
5739 . Two addresses routed to the same list of hosts;
5740 . First host does not offer TLS;
5741 . First host accepts first address;
5742 . First host gives temporary error to second address;
5743 . Second host offers TLS and a TLS session is established;
5744 . Second host accepts second address.
5745 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5746 and peerdn, if requested) that were in fact used only for the second
5749 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5750 attention to any local part prefix or suffix that was matched by the
5751 router that accepted the address. It now behaves in the same way as it
5752 does for delivery: the affixes are removed from the local part unless
5753 rcpt_include_affixes is set on the transport.
5755 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5756 timeout during the DATA phase of an incoming message.
5758 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5759 applied a patch sanctioned by the Sieve maintainer.
5761 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5762 the uid or gid is negative. A case of a negative gid caused this to be
5763 noticed. The fix allows for either to be negative.
5765 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5766 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5769 PH/36 The change PH/12 above was broken. Fixed it.
5771 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5772 the grounds that routing the same address twice would always produce the
5773 same answer. This might have been true once, but it is certainly no
5774 longer true now. Routing a child address may depend on the previous
5775 routing that produced that child. Some complicated redirection strategies
5776 went wrong when messages had multiple recipients, and made Exim's
5777 behaviour dependent on the order in which the addresses were given.
5779 I have moved the duplicate checking until after the routing is complete.
5780 Exim scans the addresses that are assigned to local and remote
5781 transports, and removes any duplicates. This means that more work will be
5782 done, as duplicates will always all be routed, but duplicates are
5783 presumably rare, so I don't expect this is of any significance.
5785 For deliveries to pipes, files, and autoreplies, the duplicate checking
5786 still happens during the routing process, since they are not going to be
5789 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5790 It corrects a timeout issue with spamd. This is Ian's comment: "The
5791 background is that sometimes spamd either never reads data from a
5792 connection it has accepted, or it never writes response data. The exiscan
5793 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5794 blindly assumes that writes won't block so it may never time out."
5796 PH/39 Allow G after quota size as well as K and M.
5798 PH/40 The value set for $authenticated_id in an authenticator may not contain
5799 binary zeroes or newlines because the value is written to log lines and
5800 to spool files. There was no check on this. Now the value is run through
5801 the string_printing() function so that such characters are converted to
5802 printable escape sequences.
5804 PH/41 $message_linecount is a new variable that contains the total number of
5805 lines in the message. Compare $body_linecount, which is the count for the
5808 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5809 bounce and delay warning messages, except in certain special cases, which
5812 (a) An SMTP error message from a remote host;
5813 (b) A message specified in a :fail: redirection;
5814 (c) A message specified in a "fail" command in a system filter;
5815 (d) A message specified in a FAIL return from the queryprogram router;
5816 (e) A message specified by the cannot_route_message router option.
5818 In these cases only, Exim does include the error details in bounce and
5819 warning messages. There are also a few cases where bland messages such
5820 as "unrouteable address" or "local delivery error" are given.
5822 PH/43 $value is now also set for the "else" part of a ${run expansion.
5824 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5825 being worked on, but at least Exim now implements the latest version to
5828 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5829 process to complete was treated as a delivery failure, a timeout while
5830 writing the message to the pipe was logged, but erroneously treated as a
5831 successful delivery. Such timeouts include transport filter timeouts. For
5832 consistency with the overall process timeout, these timeouts are now
5833 treated as errors, giving rise to delivery failures by default. However,
5834 there is now a new Boolean option for the pipe transport called
5835 timeout_defer, which, if set TRUE, converts the failures into defers for
5836 both kinds of timeout. A transport filter timeout is now identified in
5839 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5840 systems where "make" and "gmake" are different, calling "gmake" at top
5841 level broke things. I've arranged for the value of $(MAKE) to be passed
5842 from the Makefile to this script so that it can call the same version of
5846 A note about Exim versions 4.44 and 4.50
5847 ----------------------------------------
5849 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5850 changes of various kinds. As a consequence, a big documentation update was
5851 needed. This delayed the release for rather longer than seemed good, especially
5852 in the light of a couple of (minor) security issues. Therefore, the changes
5853 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5854 release. So 4.44 and 4.50 are in effect two different branches that both start
5857 I have left the 4.50 change log unchanged; it contains all the changes since
5858 4.43. The change log for 4.44 is below; many of its items are identical to
5859 those for 4.50. This seems to be the most sensible way to preserve the
5860 historical information.
5866 1. Minor wording change to the doc/README.SIEVE file.
5868 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5869 computation of the current number of files was incorrect.
5871 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5872 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5875 4. Give more explanation in the error message when the command for a transport
5876 filter fails to execute.
5878 5. There are several places where Exim runs a non-Exim command in a
5879 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5880 was being done only for the command run by the queryprogram router. It is
5881 now done for all such subprocesses. The other cases are: ${run, transport
5882 filters, and the commands run by the lmtp and pipe transports.
5884 6. Added CONFIGURE_GROUP build-time option.
5886 7. Some older OS have a limit of 256 on the maximum number of file
5887 descriptors. Exim was using setrlimit() to set 1000 as a large value
5888 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5889 systems. I've change it so that if it can't get 1000, it tries for 256.
5891 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5892 was an oversight, and furthermore, ever since the addition of extra
5893 controls (e.g. 4.43/32), the checks on when to allow different forms of
5894 "control" were broken. There should now be diagnostics for all cases when a
5895 control that does not make sense is encountered.
5897 9. Added the /retain_sender option to "control=submission".
5899 10. $recipients is now available in the predata ACL (oversight).
5901 11. Tidy the search cache before the fork to do a delivery from a message
5902 received from the command line. Otherwise the child will trigger a lookup
5903 failure and thereby defer the delivery if it tries to use (for example) a
5904 cached ldap connection that the parent has called unbind on.
5906 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5907 of $address_data from the recipient verification was clobbered by the
5908 sender verification.
5910 13. The value of address_data from a sender verification is now available in
5911 $sender_address_data in subsequent conditions in the ACL statement.
5913 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5915 15. Added a new option "connect=<time>" to callout options, to set a different
5918 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5919 was its contents. (It was OK if the option was not defined at all.)
5921 17. A "Completed" log line is now written for messages that are removed from
5922 the spool by the -Mrm option.
5924 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5925 information about exactly what failed.
5927 19. Added -dd to debug only the daemon process.
5929 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5930 handles timeouts, both on the server side and network timeouts. Renamed the
5931 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5933 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5934 It is now set to "smtps".
5936 22. $host_address is now set to the target address during the checking of
5937 ignore_target_hosts.
5939 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5940 being passed; this would have caused $sender_host_name to have been used if
5941 matching the list had actually called for a host name (not very likely,
5942 since this list is usually IP addresses). A host name is now passed as
5945 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5946 code that specifies a non-restarting handler (typically sigaction() in
5947 modern systems) in an attempt to fix a rare and obscure crash bug.
5949 25. Narrowed the window for a race in the daemon that could cause it to ignore
5950 SIGCHLD signals. This is not a major problem, because they are used only to
5951 wake it up if nothing else does.
5953 26. A malformed maildirsize file could cause Exim to calculate negative values
5954 for the mailbox size or file count. Odd effects could occur as a result.
5955 The maildirsize information is now recalculated if the size or filecount
5958 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5959 support for a long time. Removed HAVE_SYS_VFS_H.
5961 28. Installed the latest version of exipick from John Jetmore.
5963 29. In an address list, if the pattern was not a regular expression, an empty
5964 subject address (from a bounce message) matched only if the pattern was an
5965 empty string. Non-empty patterns were not even tested. This was the wrong
5966 because it is perfectly reasonable to use an empty address as part of a
5967 database query. An empty address is now tested by patterns that are
5968 lookups. However, all the other forms of pattern expect the subject to
5969 contain a local part and a domain, and therefore, for them, an empty
5970 address still always fails if the pattern is not itself empty.
5972 30. Exim went into a mad DNS loop when attempting to do a callout where the
5973 host was specified on an smtp transport, and looking it up yielded more
5974 than one IP address.
5976 31. Re-factored the code for checking spool and log partition space into a
5977 function that finds that data and another that does the check. The former
5978 is then used to implement four new variables: $spool_space, $log_space,
5979 $spool_inodes, and $log_inodes.
5981 32. The RFC2047 encoding function was originally intended for short strings
5982 such as real names; it was not keeping to the 75-character limit for
5983 encoded words that the RFC imposes. It now respects the limit, and
5984 generates multiple encoded words if necessary. To be on the safe side, I
5985 have increased the buffer size for the ${rfc2047: expansion operator from
5988 33. It is now permitted to omit both strings after an "if" condition; if the
5989 condition is true, the result is "true". As before, when the second string
5990 is omitted, a false condition yields an empty string. This makes it less
5991 cumbersome to write custom ACL and router conditions.
5993 34. Failure to deliver a bounce message always caused it to be frozen, even if
5994 there was an errors_to setting on the router. The errors_to setting is now
5997 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5998 canonical form (fully expanded) before being placed in
5999 $sender_host_address.
6001 36. The table in the code that translates DNS record types into text (T_A to
6002 "A" for instance) was missing entries for NS and CNAME. It is just possible
6003 that this could have caused confusion if both these types were looked up
6004 for the same domain, because the text type is used as part of Exim's
6005 per-process caching. But the chance of anyone hitting this buglet seems
6008 37. The dnsdb lookup has been extended in a number of ways.
6010 (1) There is a new type, "zns", which walks up the domain tree until it
6011 finds some nameserver records. It should be used with care.
6013 (2) There is a new type, "mxh", which is like "mx" except that it returns
6014 just the host names, not the priorities.
6016 (3) It is now possible to give a list of domains (or IP addresses) to be
6017 looked up. The behaviour when one of the lookups defers can be
6018 controlled by a keyword.
6020 (4) It is now possible to specify the separator character for use when
6021 multiple records are returned.
6023 38. The dnslists ACL condition has been extended: it is now possible to supply
6024 a list of IP addresses and/or domains to be looked up in a particular DNS
6027 39. Added log_selector=+queue_time_overall.
6029 40. When running the queue in the test harness, wait just a tad after forking a
6030 delivery process, to get repeatability of debugging output.
6032 41. Include certificate and key file names in error message when GnuTLS fails
6033 to set them up, because the GnuTLS error message doesn't include the name
6034 of the failing file when there is a problem reading it.
6036 42. Allow both -bf and -bF in the same test run.
6038 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6040 44. The "Exiscan patch" is now merged into the mainline Exim source.
6042 45. Sometimes the final signoff response after QUIT could fail to get
6043 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6044 < 0 before doing a fflush(). This bug looks as though it goes back to the
6045 introduction of TLS in release 3.20, but "sometimes" must have been rare
6046 because the tests only now provoked it.
6048 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6049 (this can affect the format of dates).
6051 47. exim_tidydb, when checking for the continued existence of a message for
6052 which it has found a message-specific retry record, was not finding
6053 messages that were in split spool directories. Consequently, it was
6054 deleting retry records that should have stayed in existence.
6056 48. Steve fixed some bugs in eximstats.
6058 49. The SPA authentication driver was not abandoning authentication and moving
6059 on to the next authenticator when an expansion was forced to fail,
6060 contradicting the general specification for all authenticators. Instead it
6061 was generating a temporary error. It now behaves as specified.
6063 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6064 (the order specifies the preference for clients). The order is now AES256,
6065 AES128, 3DES, ARCFOUR128.
6067 51. Small patch to Sieve code - explicitly set From: when generating an
6070 52. Exim crashed if a remote delivery caused a very long error message to be
6071 recorded - for instance if somebody sent an entire SpamAssassin report back
6072 as a large number of 550 error lines. This bug was coincidentally fixed by
6073 increasing the size of one of Exim's internal buffers (big_buffer) that
6074 happened as part of the Exiscan merge. However, to be on the safe side, I
6075 have made the code more robust (and fixed the comments that describe what
6078 53. Now that there can be additional text after "Completed" in log lines (if
6079 the queue_time_overall log selector is set), a one-byte patch to exigrep
6080 was needed to allow it to recognize "Completed" as not the last thing in
6083 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6084 patch that reportedly fixes this has been added. I am not expert enough to
6085 create a test for it. This is what the patch creator wrote:
6087 "I found a little strange behaviour of ldap code when working with
6088 Windows 2003 AD Domain, where users was placed in more than one
6089 Organization Units. When I tried to give exim partial DN, the exit code
6090 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6091 But simultaneously result of request was absolutely normal ldap result,
6092 so I produce this patch..."
6094 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6095 REFERENCE, so I have modified the code to exclude the patch when that macro
6098 55. Some experimental protocols are using DNS PTR records for new purposes. The
6099 keys for these records are domain names, not reversed IP addresses. The
6100 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6101 leaves it alone. Component reversal etc. now happens only for IP addresses.
6104 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6106 57. Double the size of the debug message buffer (to 2048) so that more of very
6107 long debug lines gets shown.
6109 58. The exicyclog utility now does better if the number of log files to keep
6110 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6112 59. Two changes related to the smtp_active_hostname option:
6114 (1) $smtp_active_hostname is now available as a variable.
6115 (2) The default for smtp_banner uses $smtp_active_hostname instead
6116 of $primary_hostname.
6118 60. The host_aton() function is supposed to be passed a string that is known
6119 to be a valid IP address. However, in the case of IPv6 addresses, it was
6120 not checking this. This is a hostage to fortune. Exim now panics and dies
6121 if the condition is not met. A case was found where this could be provoked
6122 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6123 components; fortuitously, this particular loophole had already been fixed
6124 by change 4.50/55 above.
6126 If there are any other similar loopholes, the new check in host_aton()
6127 itself should stop them being exploited. The report I received stated that
6128 data on the command line could provoke the exploit when Exim was running as
6129 exim, but did not say which command line option was involved. All I could
6130 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6131 running as the user.
6134 61. There was a buffer overflow vulnerability in the SPA authentication code
6135 (which came originally from the Samba project). I have added a test to the
6136 spa_base64_to_bits() function which I hope fixes it.
6139 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6140 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6142 63. The daemon start-up calls getloadavg() while still root for those OS that
6143 need the first call to be done as root, but it missed one case: when
6144 deliver_queue_load_max is set with deliver_drop_privilege. This is
6145 necessary for the benefit of the queue runner, because there is no re-exec
6146 when deliver_drop_privilege is set.
6148 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6149 This has been fixed.
6151 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6152 was in use, was not putting the data itself into the right store pool;
6153 consequently, it could be overwritten for a subsequent message in the same
6154 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6157 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6159 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6160 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6161 the function were treating the return as a boolean value, which happened to
6162 work because 0=false and not-0=true, but is not correct code.
6164 68. The host_aton() function was not handling scoped IPv6 addresses (those
6165 with, for example, "%eth0" on the end) correctly.
6167 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6168 negated items (that is, ~something) in unsigned ints. Some compilers
6169 apparently mutter when there is no cast.
6171 70. If an address verification called from an ACL failed, and did not produce a
6172 user-specific message (i.e. there was only a "system" message), nothing was
6173 put in $acl_verify_message. In this situation, it now puts the system
6176 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6177 session; change 4.31/43 added the unwanted input to the log line - except
6178 that it did not do this in the start of session case. It now does.
6180 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6181 This is wrong and can cause the other end to generate a synchronization
6182 error if it is another Exim or anything else that does the synchronization
6183 check. A QUIT command is no longer sent after a timeout.
6185 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6186 during host lookups.
6188 74. The defer_ok option of callout verification was not working if it was used
6189 when verifying addresses in header lines, that is, for this case:
6191 verify = header_sender/callout=defer_ok
6193 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6194 those file descriptors could be used for SMTP connections. If anything
6195 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6196 could be sent to the SMTP client, causing chaos. The daemon now opens
6197 stdin, stdout, and stderr to /dev/null when it puts itself into the
6200 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6201 log by default. The user can override this with suitable Perl magic.
6203 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6204 the log message when discard triggers, was not working for the DATA ACL or
6205 for the non-SMTP ACL.
6207 78. Error message wording change in sieve.c.
6209 79. If smtp_accept_max_per_host was set, the number of connections could be
6210 restricted to fewer than expected, because the daemon was trying to set up
6211 a new connection before checking whether the processes handling previous
6212 connections had finished. The check for completed processes is now done
6213 earlier. On busy systems, this bug wouldn't be noticed because something
6214 else would have woken the daemon, and it would have reaped the completed
6217 80. If a message was submitted locally by a user whose login name contained one
6218 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6219 It caused a spool format error. I have fixed the spool reading code. A
6220 related problem was that the "from" clause in the Received: line became
6221 illegal because of the space(s). It is now covered by ${quote_local_part.
6223 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6226 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6228 83. Patch from Sieve maintainer to fix unterminated string problem in
6229 "vacation" handling.
6231 84. Some minor changes to the Linux configuration files to help with other
6232 OS variants using glibc.
6234 85. One more patch for Sieve to update vacation handling to latest spec.
6237 ----------------------------------------------------
6238 See the note above about the 4.44 and 4.50 releases.
6239 ----------------------------------------------------
6245 1. Change 4.43/35 introduced a bug that caused file counts to be
6246 incorrectly computed when quota_filecount was set in an appendfile
6249 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6250 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6253 3. Give more explanation in the error message when the command for a transport
6254 filter fails to execute.
6256 4. There are several places where Exim runs a non-Exim command in a
6257 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6258 was being done only for the command run by the queryprogram router. It is
6259 now done for all such subprocesses. The other cases are: ${run, transport
6260 filters, and the commands run by the lmtp and pipe transports.
6262 5. Some older OS have a limit of 256 on the maximum number of file
6263 descriptors. Exim was using setrlimit() to set 1000 as a large value
6264 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6265 systems. I've change it so that if it can't get 1000, it tries for 256.
6267 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6268 was an oversight, and furthermore, ever since the addition of extra
6269 controls (e.g. 4.43/32), the checks on when to allow different forms of
6270 "control" were broken. There should now be diagnostics for all cases when a
6271 control that does not make sense is encountered.
6273 7. $recipients is now available in the predata ACL (oversight).
6275 8. Tidy the search cache before the fork to do a delivery from a message
6276 received from the command line. Otherwise the child will trigger a lookup
6277 failure and thereby defer the delivery if it tries to use (for example) a
6278 cached ldap connection that the parent has called unbind on.
6280 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6281 of $address_data from the recipient verification was clobbered by the
6282 sender verification.
6284 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6285 was its contents. (It was OK if the option was not defined at all.)
6287 11. A "Completed" log line is now written for messages that are removed from
6288 the spool by the -Mrm option.
6290 12. $host_address is now set to the target address during the checking of
6291 ignore_target_hosts.
6293 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6294 being passed; this would have caused $sender_host_name to have been used if
6295 matching the list had actually called for a host name (not very likely,
6296 since this list is usually IP addresses). A host name is now passed as
6299 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6300 code that specifies a non-restarting handler (typically sigaction() in
6301 modern systems) in an attempt to fix a rare and obscure crash bug.
6303 15. Narrowed the window for a race in the daemon that could cause it to ignore
6304 SIGCHLD signals. This is not a major problem, because they are used only to
6305 wake it up if nothing else does.
6307 16. A malformed maildirsize file could cause Exim to calculate negative values
6308 for the mailbox size or file count. Odd effects could occur as a result.
6309 The maildirsize information is now recalculated if the size or filecount
6312 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6313 support for a long time. Removed HAVE_SYS_VFS_H.
6315 18. Updated exipick to current release from John Jetmore.
6317 19. Allow an empty sender to be matched against a lookup in an address list.
6318 Previously the only cases considered were a regular expression, or an
6321 20. Exim went into a mad DNS lookup loop when doing a callout where the
6322 host was specified on the transport, if the DNS lookup yielded more than
6325 21. The RFC2047 encoding function was originally intended for short strings
6326 such as real names; it was not keeping to the 75-character limit for
6327 encoded words that the RFC imposes. It now respects the limit, and
6328 generates multiple encoded words if necessary. To be on the safe side, I
6329 have increased the buffer size for the ${rfc2047: expansion operator from
6332 22. Failure to deliver a bounce message always caused it to be frozen, even if
6333 there was an errors_to setting on the router. The errors_to setting is now
6336 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6337 canonical form (fully expanded) before being placed in
6338 $sender_host_address.
6340 24. Updated eximstats to version 1.33
6342 25. Include certificate and key file names in error message when GnuTLS fails
6343 to set them up, because the GnuTLS error message doesn't include the name
6344 of the failing file when there is a problem reading it.
6346 26. Expand error message when OpenSSL has problems setting up cert/key files.
6349 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6350 (this can affect the format of dates).
6352 28. exim_tidydb, when checking for the continued existence of a message for
6353 which it has found a message-specific retry record, was not finding
6354 messages that were in split spool directories. Consequently, it was
6355 deleting retry records that should have stayed in existence.
6357 29. eximstats updated to version 1.35
6358 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6359 1.35 - bugfix such that pie charts by volume are generated correctly
6361 30. The SPA authentication driver was not abandoning authentication and moving
6362 on to the next authenticator when an expansion was forced to fail,
6363 contradicting the general specification for all authenticators. Instead it
6364 was generating a temporary error. It now behaves as specified.
6366 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6367 (the order specifies the preference for clients). The order is now AES256,
6368 AES128, 3DES, ARCFOUR128.
6370 31. Small patch to Sieve code - explicitly set From: when generating an
6373 32. Exim crashed if a remote delivery caused a very long error message to be
6374 recorded - for instance if somebody sent an entire SpamAssassin report back
6375 as a large number of 550 error lines. This bug was coincidentally fixed by
6376 increasing the size of one of Exim's internal buffers (big_buffer) that
6377 happened as part of the Exiscan merge. However, to be on the safe side, I
6378 have made the code more robust (and fixed the comments that describe what
6381 33. Some experimental protocols are using DNS PTR records for new purposes. The
6382 keys for these records are domain names, not reversed IP addresses. The
6383 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6384 leaves it alone. Component reversal etc. now happens only for IP addresses.
6387 34. The host_aton() function is supposed to be passed a string that is known
6388 to be a valid IP address. However, in the case of IPv6 addresses, it was
6389 not checking this. This is a hostage to fortune. Exim now panics and dies
6390 if the condition is not met. A case was found where this could be provoked
6391 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6392 components; fortuitously, this particular loophole had already been fixed
6393 by change 4.50/55 or 4.44/33 above.
6395 If there are any other similar loopholes, the new check in host_aton()
6396 itself should stop them being exploited. The report I received stated that
6397 data on the command line could provoke the exploit when Exim was running as
6398 exim, but did not say which command line option was involved. All I could
6399 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6400 running as the user.
6403 35. There was a buffer overflow vulnerability in the SPA authentication code
6404 (which came originally from the Samba project). I have added a test to the
6405 spa_base64_to_bits() function which I hope fixes it.
6408 36. The daemon start-up calls getloadavg() while still root for those OS that
6409 need the first call to be done as root, but it missed one case: when
6410 deliver_queue_load_max is set with deliver_drop_privilege. This is
6411 necessary for the benefit of the queue runner, because there is no re-exec
6412 when deliver_drop_privilege is set.
6414 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6415 was in use, was not putting the data itself into the right store pool;
6416 consequently, it could be overwritten for a subsequent message in the same
6417 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6420 38. Sometimes the final signoff response after QUIT could fail to get
6421 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6422 < 0 before doing a fflush(). This bug looks as though it goes back to the
6423 introduction of TLS in release 3.20, but "sometimes" must have been rare
6424 because the tests only now provoked it.
6430 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6431 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6432 Now it returns an integer. A number of calls were still expecting a T/F
6433 return. Fortuitously, in all cases, the tests worked in OK situations,
6434 which is the norm. However, things would have gone wrong on any write
6435 failures on the smtp file descriptor. This function is used when sending
6436 messages over SMTP and also when doing verify callouts.
6438 2. When Exim is called to do synchronous delivery of a locally submitted
6439 message (the -odf or -odi options), it no longer closes stderr before doing
6442 3. Implemented the mua_wrapper option.
6444 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6446 5. Implemented the functions header_remove(), header_testname(),
6447 header_add_at_position(), and receive_remove_recipient(), and exported them
6450 6. If an ACL "warn" statement specified the addition of headers, Exim already
6451 inserted X-ACL-Warn: at the start if there was no header name. However, it
6452 was not making this test for the second and subsequent header lines if
6453 there were newlines in the string. This meant that an invalid header could
6454 be inserted if Exim was badly configured.
6456 7. Allow an ACL "warn" statement to add header lines at the start or after all
6457 the Received: headers, as well as at the end.
6459 8. Added the rcpt_4xx retry error code.
6461 9. Added postmaster_mailfrom=xxx to callout verification option.
6463 10. Added mailfrom=xxxx to the callout verification option, for verify=
6466 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6467 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6469 12. Inserted some casts to stop certain compilers warning when using pointer
6470 differences as field lengths or precisions in printf-type calls (mostly
6471 affecting debugging statements).
6473 13. Added optional readline() support for -be (dynamically loaded).
6475 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6476 same clock tick as a message's arrival, so that its received time was the
6477 same as the "first fail" time on the retry record, and that message
6478 remained on the queue past the ultimate address timeout, every queue runner
6479 would try a delivery (because it was past the ultimate address timeout) but
6480 after another failure, the ultimate address timeout, which should have then
6481 bounced the address, did not kick in. This was a "< instead of <=" error;
6482 in most cases the first failure would have been in the next clock tick
6483 after the received time, and all would be well.
6485 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6486 being recognized when the domain list was tested by the match_domain
6487 condition in an expansion string.
6489 16. Added the ${str2b64: operator.
6491 17. Exim was always calling setrlimit() to set a large limit for the number of
6492 processes, without checking whether the existing limit was already
6493 adequate. (It did check for the limit on file descriptors.) Furthermore,
6494 errors from getrlimit() and setrlimit() were being ignored. Now they are
6495 logged to the main and panic logs, but Exim does carry on, to try to do its
6496 job under whatever limits there are.
6498 18. Imported PCRE 5.0.
6500 19. Trivial typo in log message " temporarily refused connection" (the leading
6503 20. If the log selector return_path_on_delivery was set and an address was
6504 redirected to /dev/null, the delivery process crashed because it assumed
6505 that a return path would always be set for a "successful" delivery. In this
6506 case, the whole delivery is bypassed as an optimization, and therefore no
6509 21. Internal re-arrangement: the function for sending a challenge and reading
6510 a response while authentication was assuming a zero-terminated challenge
6511 string. It's now changed to take a pointer and a length, to allow for
6512 binary data in such strings.
6514 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6516 23. Exim was not respecting finduser_retries when seeking the login of the
6517 uid under which it was called; it was always trying 10 times. (The default
6518 setting of finduser_retries is zero.) Also, it was sleeping after the final
6519 failure, which is pointless.
6521 24. Implemented tls_on_connect_ports.
6523 25. Implemented acl_smtp_predata.
6525 26. If the domain in control=submission is set empty, Exim assumes that the
6526 authenticated id is a complete email address when it generates From: or
6527 Sender: header lines.
6529 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6530 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6531 chown and chgrp in /bin and hostname in /usr/bin.
6533 28. Exim was keeping the "process log" file open after each use, just as it
6534 does for the main log. This opens the possibility of it remaining open for
6535 long periods when the USR1 signal hits a daemon. Occasional processlog
6536 errors were reported, that could have been caused by this. Anyway, it seems
6537 much more sensible not to leave this file open at all, so that is what now
6540 29. The long-running daemon process does not normally write to the log once it
6541 has entered its main loop, and it closes the log before doing so. This is
6542 so that log files can straightforwardly be renamed and moved. However,
6543 there are a couple of unusual error situations where the daemon does write
6544 log entries, and I had neglected to close the log afterwards.
6546 30. The text of an SMTP error response that was received during a remote
6547 delivery was being truncated at 512 bytes. This is too short for some of
6548 the long messages that one sometimes sees. I've increased the limit to
6551 31. It is now possible to make retry rules that apply only when a message has a
6552 specific sender, in particular, an empty sender.
6554 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6555 it possible to be selective about when SMTP synchronization is enforced.
6557 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6559 32. Implemented hosts_connection_nolog.
6561 33. Added an ACL for QUIT.
6563 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6566 35. Added mailbox_size and mailbox_filecount to appendfile.
6568 36. Added control = no_multiline_responses to ACLs.
6570 37. There was a bug in the logic of the code that waits for the clock to tick
6571 in the case where the clock went backwards by a substantial amount such
6572 that the microsecond fraction of "now" was more than the microsecond
6573 fraction of "then" (but the whole seconds number was less).
6575 38. Added support for the libradius Radius client library this is found on
6576 FreeBSD (previously only the radiusclient library was supported).
6582 1. When certain lookups returned multiple values in the form name=value, the
6583 quoting of the values was not always being done properly. Specifically:
6584 (a) If the value started with a double quote, but contained no whitespace,
6586 (b) If the value contained whitespace other than a space character (i.e.
6587 tabs or newlines or carriage returns) it was not quoted.
6588 This fix has been applied to the mysql and pgsql lookups by writing a
6589 separate quoting function and calling it from the lookup code. The fix
6590 should probably also be applied to nisplus, ibase and oracle lookups, but
6591 since I cannot test any of those, I have not disturbed their existing code.
6593 2. A hit in the callout cache for a specific address caused a log line with no
6594 reason for rejecting RCPT. Now it says "Previous (cached) callout
6595 verification failure".
6597 3. There was an off-by-one bug in the queryprogram router. An over-long
6598 return line was truncated at 256 instead of 255 characters, thereby
6599 overflowing its buffer with the terminating zero. As well as fixing this, I
6600 have increased the buffer size to 1024 (and made a note to document this).
6602 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6603 when Exim is waiting for an SMTP response from a remote server, Exim
6604 restarts its select() call on the socket, thereby resetting its timeout.
6605 This is not a problem when such interrupts are rare. Somebody set up a cron
6606 job to run exiwhat every 2 minutes, which is less than the normal select()
6607 timeout (5 or 10 minutes). This meant that the select() timeout never
6608 kicked in because it was always reset. I have fixed this by comparing the
6609 time when an interrupt arrives with the time at the start of the first call
6610 to select(). If more time than the timeout has elapsed, the interrupt is
6611 treated as a timeout.
6613 5. Some internal re-factoring in preparation for the addition of Sieve
6614 extensions (by MH). In particular, the "personal" test is moved to a
6615 separate function, and given an option for scanning Cc: and Bcc: (which is
6616 not set for Exim filters).
6618 6. When Exim created an email address using the login of the caller as the
6619 local part (e.g. when creating a From: or Sender: header line), it was not
6620 quoting the local part when it contained special characters such as @.
6622 7. Installed new OpenBSD configuration files.
6624 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6625 try to make them clearer.
6627 9. Callout options, other than the timeout value, were being ignored when
6628 verifying sender addresses in header lines. For example, when using
6630 verify = header_sender/callout=no_cache
6632 the cache was (incorrectly) being used.
6634 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6635 only the Cygwin environment.
6637 11. When return_path_on_delivery was set as a log selector, if different remote
6638 addresses in the same message used different return paths and parallel
6639 remote delivery occurred, the wrong values would sometimes be logged.
6640 (Whenever a remote delivery process finished, the return path value from
6641 the most recently started remote delivery process was logged.)
6643 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6644 lines when AUTH and/or TLS are in use. This is the "received protocol"
6645 field. Exim used to use "asmtp" for authenticated SMTP, without any
6646 indication (in the protocol name) for TLS use. Now it follows the RFC and
6647 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6648 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6649 names appear in log lines as well as in Received: header lines.
6651 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6652 extensions, and comparison tests, and to fix some bugs.
6654 14. Changes to the "personal" filter test:
6656 (1) The test was buggy in that it was just doing the equivalent of
6657 "contains" tests on header lines. For example, if a user's address was
6658 anne@some.where, the "personal" test would incorrectly be true for
6660 To: susanne@some.where
6662 This test is now done by extracting each address from the header in turn,
6663 and checking the entire address. Other tests that are part of "personal"
6664 are now done using regular expressions (for example, to check local parts
6665 of addresses in From: header lines).
6667 (2) The list of non-personal local parts in From: addresses has been
6668 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6669 taken from the Sieve specification recommendations.
6671 (3) If the message contains any header line starting with "List-" it is
6672 treated as non-personal.
6674 (4) The test for "circular" in the Subject: header line has been removed
6675 because it now seems ill-conceived.
6677 15. Minor typos in src/EDITME comments corrected.
6679 16. Installed latest exipick from John Jetmore.
6681 17. If headers_add on a router specified a text string that was too long for
6682 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6683 of string_sprintf() is now avoided.
6685 18. $message_body_size was not set (it was always zero) when running the DATA
6686 ACL and the local_scan() function.
6688 19. For the "mail" command in an Exim filter, no default was being set for
6689 the once_repeat time, causing a random time value to be used if "once" was
6690 specified. (If the value happened to be <= 0, no repeat happened.) The
6691 default is now 0s, meaning "never repeat". The "vacation" command was OK
6692 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6693 (I found it when inspecting the code).
6695 20. There is now an overall timeout for performing a callout verification. It
6696 defaults to 4 times the callout timeout, which applies to individual SMTP
6697 commands during the callout. The overall timeout applies when there is more
6698 than one host that can be tried. The timeout is checked before trying the
6699 next host. This prevents very long delays if there are a large number of
6700 hosts and all are timing out (e.g. when the network connections are timing
6701 out). The value of the overall timeout can be changed by specifying an
6702 additional sub-option for "callout", called "maxwait". For example:
6704 verify = sender/callout=5s,maxwait=20s
6706 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6707 to the end before writing, but this should make it even safer).
6709 22. Exim was forgetting that it had advertised PIPELINING for the second and
6710 subsequent messages on an SMTP connection. It was also not resetting its
6711 memory on STARTTLS and an internal HELO.
6713 23. When Exim logs an SMTP synchronization error within a session, it now
6714 records whether PIPELINING has been advertised or not.
6716 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6717 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6718 rather than long int.
6720 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6722 26. Added the never_mail option to autoreply.
6728 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6729 crash if the getsockname() call failed; this can happen if a connection is
6730 closed very soon after it is established. The problem was simply in the
6731 order in which certain operations were done, causing Exim to try to write
6732 to the SMTP stream before it had set up the file descriptor. The bug has
6733 been fixed by making things happen in the correct order.
6739 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6740 before the connection was closed, thus losing the rejection response.
6742 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6743 some early Solaris releases, but causes trouble in current releases where
6744 socklen_t is defined.
6746 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6749 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6752 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6753 that releasing the top of it at the end releases what was used for sub-
6754 expansions (unless the block got too big). However, discard this block if
6755 the first thing is a variable or header, so that we can use its block when
6756 it is dynamic (useful for very large $message_headers, for example).
6758 6. Lookups now cache *every* query, not just the most recent. A new, separate
6759 store pool is used for this. It can be recovered when all lookup caches are
6760 flushed. Lookups now release memory at the end of their result strings.
6761 This has involved some general refactoring of the lookup sources.
6763 7. Some code has been added to the store_xxx() functions to reduce the amount
6764 of flapping under certain conditions.
6766 8. log_incoming_interface used to affect only the <= reception log lines. Now
6767 it causes the local interface and port to be added to several more SMTP log
6768 lines, for example "SMTP connection from", and rejection lines.
6770 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6772 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6774 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6775 could be overwritten at the end of the current message (or the start of a
6776 new message if it was set in a HELO ACL). The value is now preserved for
6777 the duration of the SMTP connection.
6779 12. If a transport had a headers_rewrite setting, and a matching header line
6780 contained an unqualified address, that address was qualified, even if it
6781 did not match any rewriting rules. The underlying bug was that the values
6782 of the flags that permit the existence of unqualified sender and recipient
6783 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6784 non-local messages, and by -bnq for local messages) were not being
6785 preserved with the message after it was received.
6787 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6788 "next input=" as part of the text comprising the host identity instead of
6789 the correct text. The code was using the same buffer for two different
6790 strings. However, depending on which order the printing function evaluated
6791 its arguments, the bug did not always show up. Under Linux, for example, my
6792 test suite worked just fine.
6794 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6795 doesn't work with some older versions of Perl. It has been changed to "my",
6796 which in any case is probably the better facility to use.
6798 15. A really picky compiler found some instances of statements for creating
6799 error messages that either had too many or two few arguments for the format
6802 16. The size of the buffer for calls to the DNS resolver has been increased
6803 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6804 for addresses that have a lot of PTR records. This alleviates a problem; it
6805 does not fully solve it.
6807 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6808 buffer now truncates the list and logs the incident, which is the same
6809 action as happens when Exim is looking up a host name and its aliases.
6810 Previously in this situation something unpredictable would happen;
6811 sometimes it was "internal error: store_reset failed".
6813 18. If a server dropped the connection unexpectedly when an Exim client was
6814 using GnuTLS and trying to read a response, the client delivery process
6815 crashed while trying to generate an error log message.
6817 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6818 string, for example:
6820 warn message = H1: something\nH2: something
6822 the text was added as a single header line from Exim's point of view
6823 though it ended up OK in the delivered message. However, searching for the
6824 second and subsequent header lines using $h_h2: did not work. This has been
6825 fixed. Similarly, if a system filter added multiple headers in this way,
6826 the routers could not see them.
6828 20. Expanded the error message when iplsearch is called with an invalid key to
6829 suggest using net-iplsearch in a host list.
6831 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6832 ACLs are no longer actually imposed (and a message to that effect is
6835 22. If a "gecos" field in a passwd entry contained escaped characters, in
6836 particular, if it contained a \" sequence, Exim got it wrong when building
6837 a From: or a Sender: header from that name. A second bug also caused
6838 incorrect handling when an unquoted " was present following a character
6839 that needed quoting.
6841 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6842 was not being matched caselessly.
6844 24. Arranged for all hyphens in the exim.8 source to be escaped with
6847 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6848 properly. Recipient callout cache records were still being keyed to include
6849 the sender, even when use_sender was set false. This led to far more
6850 callouts that were necessary. The sender is no longer included in the key
6851 when use_sender is false.
6853 26. Added "control = submission" modifier to ACLs.
6855 27. Added the ${base62d: operator to decode base 62 numbers.
6857 28. dnsdb lookups can now access SRV records.
6859 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6860 the configuration file.
6862 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6863 (-v) mode. This makes the output for a verbose queue run more intelligible.
6865 31. Added a use_postmaster feature to recipient callouts.
6867 32. Added the $body_zerocount variable, containing the number of binary zero
6868 bytes in the message body.
6870 33. The time of last modification of the "new" subdirectory is now used as the
6871 "mailbox time last read" when there is a quota error for a maildir
6874 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6876 35. Added +ignore_unknown as a special item in host lists.
6878 36. Code for decoding IPv6 addresses in host lists is now included, even if
6879 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6880 address was recognized as an IP address, but was then not correctly decoded
6881 into binary, causing unexpected and incorrect effects when compared with
6888 1. Very minor rewording of debugging text in manualroute to say "list of
6889 hosts" instead of "hostlist".
6891 2. If verify=header_syntax was set, and a header line with an unqualified
6892 address (no domain) and a large number of spaces between the end of the
6893 name and the colon was received, the reception process suffered a buffer
6894 overflow, and (when I tested it) crashed. This was caused by some obsolete
6895 code that should have been removed. The fix is to remove it!
6897 3. When running in the test harness, delay a bit after writing a bounce
6898 message to get a bit more predictability in the log output.
6900 4. Added a call to search_tidyup() just before forking a reception process. In
6901 theory, someone could use a lookup in the expansion of smtp_accept_max_
6902 per_host which, without the tidyup, could leave open a database connection.
6904 5. Added the variables $recipient_data and $sender_data which get set from a
6905 lookup success in an ACL "recipients" or "senders" condition, or a router
6906 "senders" option, similar to $domain_data and $local_part_data.
6908 6. Moved the writing of debug_print from before to after the "senders" test
6911 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6912 problems for message scanning, either using a data ACL, or using
6913 local_scan() because the Received: header was not generated till after they
6914 were called (in order to set the time as the time of reception completion).
6915 I have revised the way this works. The header is now generated after the
6916 body is received, but before the ACL or local_scan() are called. After they
6917 are run, the timestamp in the header is updated.
6923 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6924 before starting a queue runner without re-exec. This happened only when
6925 deliver_drop_privilege was set or when the Exim user was set to root. The
6926 effect of the bug was that timeouts during subsequent deliveries caused
6927 crashes instead of being properly handled. The handler is now left at its
6928 default (and expected) setting.
6930 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6931 message, again when deliver_drop_privilege is set or Exim is run as root.
6932 The bug described in (1) was not present in this case, but the tidying up
6933 of the other signals was missing. I have made the two cases consistent.
6935 3. The ignore_target_hosts setting on a manualroute router was being ignored
6936 for hosts that were looked up using the /MX notation.
6938 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6941 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6942 operated on the sender address. After changing the $sender_address to <>
6943 for the sender address verify, Exim was re-instated it as the original
6944 (before rewriting) address, but remembering that it had rewritten it, so it
6945 wasn't rewriting it again. This bug also had the effect of breaking the
6946 sender address verification caching when the sender address was rewritten.
6948 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6949 This has been changed so that if the ip literal address matches
6950 ignore_target_hosts, the router declines.
6952 7. Added expansion conditions match_domain, match_address, and match_local_
6953 part (NOT match_host).
6955 8. The placeholder for the Received: header didn't have a length field set.
6957 9. Added code to Exim itself and to exim_lock to test for a specific race
6958 condition that could lead to file corruption when using MBX delivery. The
6959 issue is with the lockfile that is created in /tmp. If this file is removed
6960 after a process has opened it but before that process has acquired a lock,
6961 there is the potential for a second process to recreate the file and also
6962 acquire a lock. This could lead to two Exim processes writing to the file
6963 at the same time. The added code performs the same test as UW imapd; it
6964 checks after acquiring the lock that its file descriptor still refers to
6965 the same named file.
6967 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6968 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6971 11. Added the smtp_active_hostname option. If used, this will typically be made
6972 to depend on the incoming interface address. Because $interface_address is
6973 not set up until the daemon has forked a reception process, error responses
6974 that can happen earlier (such as "too many connections") no longer contain
6977 12. If an expansion in a condition on a "warn" statement fails because a lookup
6978 defers, the "warn" statement is abandoned, and the next ACL statement is
6979 processed. Previously this caused the whole ACL to be aborted.
6981 13. Added the iplsearch lookup type.
6983 14. Added ident_timeout as a log selector.
6985 15. Added tls_certificate_verified as a log selector.
6987 16. Added a global option tls_require_ciphers (compare the smtp transport
6988 option of the same name). This controls incoming TLS connections.
6990 17. I finally figured out how to make tls_require_ciphers do a similar thing
6991 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6992 before starting the TLS session.
6994 18. Tabs are now shown as \t in -bP output.
6996 19. If the log selector return_path_on_delivery was set, Exim crashed when
6997 bouncing a message because it had too many Received: header lines.
6999 20. If two routers both had headers_remove settings, and the first one included
7000 a superfluous trailing colon, the final name in the first list and the
7001 first name in the second list were incorrectly joined into one item (with a
7002 colon in the middle).
7008 1. Added -C and -D options to the exinext utility, mainly to make it easier
7009 to include in the automated testing, but these could be helpful when
7010 multiple configurations are in use.
7012 2. The exinext utility was not formatting the output nicely when there was
7013 an alternate port involved in the retry record key, nor when there was a
7014 message id as well (for retries that were specific to a specific message
7015 and a specific host). It was also confused by IPv6 addresses, because of
7016 the additional colons they contain. I have fixed the IPv4 problem, and
7017 patched it up to do a reasonable job for IPv6.
7019 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7020 delivery, the log line now contains "pipelined" if PIPELINING was used.
7022 4. An SMTP transport process used to panic and die if the bind() call to set
7023 an explicit outgoing interface failed. This has been changed; it is now
7024 treated in the same way as a connect() failure.
7026 5. A reference to $sender_host_name in the part of a conditional expansion
7027 that was being skipped was still causing a DNS lookup. This no longer
7030 6. The def: expansion condition was not recognizing references to header lines
7031 that used bh_ and bheader_.
7033 7. Added the _cache feature to named lists.
7035 8. The code for checking quota_filecount in the appendfile transport was
7036 allowing one more file than it should have been.
7038 9. For compatibility with Sendmail, the command line option
7046 and sets the incoming protocol and host name (for trusted callers). The
7047 host name and its colon can be omitted when only the protocol is to be set.
7048 Note the Exim already has two private options, -pd and -ps, that refer to
7049 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7050 "s", but I don't think that's a major issue.
7052 10. A number of refactoring changes to the code, none of which should affect
7055 (a) The number of logging options was getting close to filling up the
7056 32-bit word that was used as a bit map. I have split them into two classes:
7057 those that are passed in the argument to log_write(), and those that are
7058 only ever tested independently outside of that function. These are now in
7059 separate 32-bit words, so there is plenty of room for expansion again.
7060 There is no change in the user interface or the logging behaviour.
7062 (b) When building, for example, log lines, the code previously used a
7063 macro that called string_cat() twice, in order to add two strings. This is
7064 not really sufficiently general. Furthermore, there was one instance where
7065 it was actually wrong because one of the argument was used twice, and in
7066 one call a function was used. (As it happened, calling the function twice
7067 did not affect the overall behaviour.) The macro has been replaced by a
7068 function that can join an arbitrary number of extra strings onto a growing
7071 (c) The code for expansion conditions now uses a table and a binary chop
7072 instead of a serial search (which was left over from when there were very
7073 few conditions). Also, it now recognizes conditions like "pam" even when
7074 the relevant support is not compiled in: a suitably worded error message is
7075 given if an attempt is made to use such a condition.
7077 11. Added ${time_interval:xxxxx}.
7079 12. A bug was causing one of the ddress fields not to be passed back correctly
7080 from remote delivery subprocesses. The field in question was not being
7081 subsequently used, so this caused to problems in practice.
7083 13. Added new log selectors queue_time and deliver_time.
7085 14. Might have fixed a bug in maildirsizefile handling that threw up
7086 "unexpected character" debug warnings, and recalculated the data
7087 unnecessarily. In any case, I expanded the warning message to give more
7090 15. Added the message "Restricted characters in address" to the statements in
7091 the default ACL that block characters like @ and % in local parts.
7093 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7094 Three changes have been made:
7096 (a) There was a serious bug; a negative response to MAIL caused the whole
7097 recipient domain to be cached as invalid, thereby blocking all messages
7098 to all local parts at the same domain, from all senders. This bug has
7099 been fixed. The domain is no longer cached after a negative response to
7100 MAIL if the sender used is not empty.
7102 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7105 (c) A new callout option, "use_sender" has been added for people who want
7106 the modified behaviour.
7112 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7115 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7116 indeed breaks things for older releases.
7118 3. Added additional logging to the case where there is a problem reading data
7119 from a filter that is running in a subprocess using a pipe, in order to
7120 try to track down a specific problem.
7122 4. Testing facility fudge: when running in the test harness and attempting
7123 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7124 getting "No route to host". Convert this to a timeout.
7126 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7129 6. Some OS don't have socklen_t but use size_t instead. This affects the
7130 fifth argument of getsockopt() amongst other things. This is now
7131 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7132 can be set for individual OS. I have set it for SunOS5, OSF1, and
7133 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7134 some earlier ones do not.
7136 7. Change 4.30/15 was not doing the test caselessly.
7138 8. The standard form for an IPv6 address literal was being rejected by address
7139 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7140 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7141 this, as well as the form without the "IPv6" on the front (but only when
7142 address literals are enabled, of course).
7144 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7146 10. Exim crashed if a message with an empty sender address specified by -f
7147 encountered a router with an errors_to setting. This could be provoked only
7148 by a command such as
7152 where an empty string was supplied; "<>" did not hit this bug.
7154 11. Installed PCRE release 4.5.
7156 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7157 remained set. It is now erased.
7159 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7160 times from message ids (which are base 36 rather than the normal 62).
7162 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7163 were being counted as actual protocol errors, and logged if the log
7164 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7165 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7166 and DATA following a set of rejected RCPTs do not count as protocol errors.
7167 In other words, Exim assumes they were pipelined, though this may not
7168 actually be the case. Of course, in all cases the client gets an
7169 appropriate error code.
7171 15. If a lookup fails in an ACL condition, a message about the failure may
7172 be available; it is used if testing the ACL cannot continue, because most
7173 such messages specify what the cause of the deferral is. However, some
7174 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7175 that caused an old message to be retained and used if a later statement
7176 caused a defer, replacing the real cause of the deferral.
7178 16. If an IP address had so many PTR records that the DNS lookup buffer
7179 was not large enough to hold them, Exim could crash while trying to process
7180 the truncated data. It now detects and logs this case.
7182 17. Further to 4.21/58, another change has been made: if (and only if) the
7183 first line of a message (the first header line) ends with CRLF, a bare LF
7184 in a subsequent header line has a space inserted after it, so as not to
7185 terminate the header.
7187 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7188 unnecessarily, used atoi() instead of strtol(), and didn't check the
7189 termination when getting file sizes from file names by regex.
7191 19. Completely re-implemented the support for maildirsize files, in the light
7192 of a number of problems with the previous contributed implementation
7193 (4.30/29). In particular:
7195 . If the quota is zero, the maildirsize file is maintained, but no quota is
7198 . If the maildir directory does not exist, it is created before any attempt
7199 to write a maildirsize file.
7201 . The quota value in the file is just a cache; if the quota is changed in
7202 the transport, the new value overrides.
7204 . A regular expression is available for excluding directories from the
7207 20. The autoreply transport checks the characters in options that define the
7208 message's headers; it allows continued headers, but it was checking with
7209 isspace() after an embedded newline instead of explicitly looking for a
7212 21. If all the "regular" hosts to which an address was routed had passed their
7213 expiry times, and had not reached their retry times, the address was
7214 bounced, even if fallback hosts were defined. Now Exim should go on to try
7217 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7218 equivalent code in the SMTP transport. Some hosts send humungous responses
7219 to HELO/EHLO, more than 1024 it seems.
7221 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7222 gives compiler warnings in some environments. I've now done it "properly",
7225 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7226 (because of problems with the built-in one) was declared to return uschar *
7227 instead of char *, causing compiler failure.
7229 25. Fixed a file descriptor leak when processing alias/forward files.
7231 26. Fixed a minor format string issue in dbfn.c.
7233 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7235 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7236 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7239 29. When a sender address is verified, it is cached, to save repeating the test
7240 when there is more than one recipient in a message. However, when the
7241 verification involves a callout, it is possible for different callout
7242 options to be set for different recipients. It is too complicated to keep
7243 track of this in the cache, so now Exim always runs a verification when a
7244 callout is required, relying on the callout cache for the optimization.
7245 The overhead is duplication of the address routing, but this should not be
7248 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7249 to be verified with callout=postmaster, and the main callout worked but the
7250 postmaster check failed, the verification correctly failed. However, if a
7251 subsequent RCPT command asked for sender verification *without* the
7252 postmaster check, incorrect caching caused this verification also to fail,
7255 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7256 it was not caching the DNS options (qualify_single, search_parents) that
7257 were used when the lookup failed. A subsequent lookup with different
7258 options therefore always gave the same answer, though there were cases
7259 where it should not have. (Example: a "domains = !$mx_any" option on a
7260 dnslookup router: the "domains" option is always processed without any
7261 widening, but the router might have qualify_single set.) Now Exim uses the
7262 cached value only when the same options are set.
7264 32. Added John Jetmore's "exipick" utility to the distribution.
7266 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7267 than a timeout (e.g. a certificate is required, and is not provided), an
7268 Exim server now closes the connection immediately. Previously it waited for
7269 the client to close - but if the client is SSL, it seems that they each
7270 wait for each other, leading to a delay before one of them times out.
7272 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7273 maintained 0.8.x compatibility because I don't think many are using it, and
7274 it is clearly obsolete.
7276 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7279 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7280 client certificate was expired. A simple patch fixes this, though I don't
7281 understand the full logic of why the verify callback is called multiple
7284 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7285 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7286 which causes problems with some clients (such as the Certicom SSL Plus
7287 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7288 disables the coutermeasure allowing Eudora to connect."
7290 38. Exim was not checking that a write() to a log file succeeded. This could
7291 lead to Bad Things if a log got too big, in particular if it hit a file
7292 size limit. Exim now panics and dies if it cannot write to a log file, just
7293 as it does if it cannot open a log file.
7295 39. Modified OS/Makefile-Linux so that it now contains
7297 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7299 The two -D definitions ensure that Exim is compiled with large file
7300 support, which makes it possible to handle log files that are bigger than
7303 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7304 instance) a domain was checked against a named list that involved a lookup,
7305 causing $domain_data to be set, then another domain was checked against the
7306 same list, then the first domain was re-checked, the value of $domain_data
7307 after the final check could be wrong. In particular, if the second check
7308 failed, it could be set empty. This bug probably also applied to
7311 41. The strip_trailing_dot option was not being applied to the address given
7312 with the -f command-line option.
7314 42. The code for reading a message's header from the spool was incrementing
7315 $received_count, but never initializing it. This meant that the value was
7316 incorrect (doubled) while delivering a message in the same process in which
7317 it was received. In the most common configuration of Exim, this never
7318 happens - a fresh exec is done - but it can happen when
7319 deliver_drop_privilege is set.
7321 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7322 it now includes up to 150 characters of the unexpected data in the log
7325 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7326 and building data strings. The size of both of these buffers was 10 000
7327 bytes - far larger than anybody would *ever* want, thought I. Needless to
7328 say, somebody hit the limit. I have increased the maximum line length to
7329 20 000 and the maximum data length of concatenated lines to 100 000. I have
7330 also fixed two bugs, because there was no checking on these buffers. Tsk,
7331 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7332 buffer is too small.
7334 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7335 lsearch lookups. Now it does.
7337 46. When parsing a route_list item in a manualroute router, a fixed-length
7338 buffer was used for the list of hosts. I made this 1024 bytes long,
7339 thinking that nobody would ever have a list of hosts that long. Wrong.
7340 Somebody had a whole pile of complicated expansion conditions, and the
7341 string was silently truncated, leading to an expansion error. It turns out
7342 that it is easier to change to an unlimited length (owing to other changes
7343 that have happened since this code was originally written) than to build
7344 structure for giving a limitation error. The length of the item that
7345 expands into the list of hosts is now unlimited.
7347 47. The lsearch lookup could not handle data where the length of text line was
7348 more than 4095 characters. Such lines were truncated, leading to shortened
7349 data being returned. It should now handle lines of any length.
7351 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7352 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7355 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7356 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7357 no longer happens in comment lines. A list of replacements is now placed
7358 at the head of all of the source files, except those whose only change is
7359 to replace PERL_COMMAND in the very first #! line.
7361 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7362 messages on the queue, with a bottom-up merge sort, using code contributed
7363 by Michael Haardt. This should make operations like -bp somewhat faster on
7364 large queues. It won't affect queue runners, except when queue_run_in_order
7367 51. Installed eximstats 1.31 in the distribution.
7369 52. Added support for SRV lookups to the dnslookup router.
7371 53. If an ACL referred to $message_body or $message_body_end, the value was not
7372 reset for any messages that followed in the same SMTP session.
7374 54. The store-handling optimization for building very long strings was not
7375 differentiating between the different store pools. I don't think this
7376 actually made any difference in practice, but I've tidied it.
7378 55. While running the routers to verify a sender address, $sender_address
7379 was still set to the sender address. This is wrong, because when routing to
7380 send a bounce to the sender, it would be empty. Therefore, I have changed
7381 it so that, while verifying a sender address, $sender_address is set to <>.
7382 (There is no change to what happens when verifying a recipient address.)
7384 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7385 target A or AAAA records (if not already returned) without resetting the
7386 qualify_single or search_parents options of the DNS resolver. These are
7387 inappropriate in this case because the targets of MX and SRV records must
7388 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7389 target that, when qualified, matched something in the local domain. These
7390 two options are now turned off when doing these lookups.
7392 57. It seems that at least some releases of Reiserfs (which does not have the
7393 concept of a fixed number of inodes) returns zero and not -1 for the
7394 number of available inodes. This interacted badly with check_spool_inodes,
7395 which assumed that -1 was the "no such thing" setting. What I have done is
7396 to check that the total number of inodes is greater than zero before doing
7397 the test of how many are available.
7399 58. When a "warn" ACL statement has a log_message modifier, the message is
7400 remembered, and not repeated. This is to avoid a lot of repetition when a
7401 message has many recipients that cause the same warning to be written.
7402 However, Exim was preserving the list of already written lines for an
7403 entire SMTP session, which doesn't seem right. The memory is now reset if a
7404 new message is started.
7406 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7407 debugging output unless log_rewrite was also set.
7409 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7410 of (int)(handle) when we know that handle contains (void *)(-1).
7412 61. The Exim daemon panic-logs an error return when it closes the incoming
7413 connection. However "connection reset by peer" seems to be common, and
7414 isn't really an error worthy of noting specially, so that particular error
7417 62. When Exim is trying to find all the local interfaces, it used to panic and
7418 die if the ioctl to get the interface flags failed. However, it seems that
7419 on at least one OS (Solaris 9) it is possible to have an interface that is
7420 included in the list of interfaces, but for which you get a failure error
7421 for this call. This happens when the interface is not "plumbed" into a
7422 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7423 failure of the "get flags" call assumes that the interface is down.
7425 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7426 makes life easier for people who are doing arithmetic on fields extracted
7427 from dates, where you often get leading zeros that should not be
7428 interpreted as octal.
7430 64. Added qualify_domain to the redirect router, to override the global
7433 65. If a pathologically long header line contained very many addresses (the
7434 report of this problem mentioned 10 000) and each of them was rewritten,
7435 Exim could use up a very large amount of memory. (It kept on making new
7436 copies of the header line as it rewrote, and never released the old ones.)
7437 At the expense of a bit more processing, the header rewriting function has
7438 been changed so that it no longer eats memory in this way.
7440 66. The generation of the Received: header has been moved from the time that a
7441 message starts to be received, to the time that it finishes. The timestamp
7442 in the Received: header should now be very close to that of the <= log
7443 line. There are two side-effects of this change:
7445 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7446 logged header lines no longer include the local Received: line, because
7447 it has not yet been created. The same applies to a copy of the message
7448 that is returned to a non-SMTP sender when a message is rejected.
7450 (b) When a filter file is tested using -bf, no additional Received: header
7451 is added to the test message. After some thought, I decided that this
7454 This change does not affect the value of $received_for. It is still set
7455 after address rewriting, but before local_scan() is called.
7457 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7459 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7460 gave an unhelpful panic error message, and a defer error. I have managed to
7461 change this behaviour so that it now rejects any supplied certificate,
7462 which seems right, as the list of acceptable certificates is empty.
7464 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7465 gave an unhelpful defer error. I have not managed to make this reject any
7466 supplied certificates, but the error message it gives is "no certificate
7467 supplied", which is not helpful.
7469 70. exigrep's output now also includes lines that are not associated with any
7470 message, but which match the given pattern. Implemented by a patch from
7471 Martin Sluka, which also tidied up the Perl a bit.
7473 71. Recipient callout verification, like sender verification, was using <> in
7474 the MAIL FROM command. This isn't really the right thing, since the actual
7475 sender may affect whether the remote host accepts the recipient or not. I
7476 have changed it to use the actual sender in the callout; this means that
7477 the cache record is now keyed on a recipient/sender pair, not just the
7478 recipient address. There doesn't seem to be a real danger of callout loops,
7479 since a callout by the remote host to check the sender would use <>.
7480 [SEE ABOVE: changed after hitting problems.]
7482 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7483 temporary errors. However, in the case of such a code being given after
7484 the end of a data transmission (i.e. after ".") Exim was failing to write
7485 a retry record for the message. (Yes, there was some broken host that was
7486 actually sending 8xx at this point.)
7488 73. An unknown lookup type in a host list could cause Exim to panic-die when
7489 the list was checked. (An example that provoked this was putting <; in the
7490 middle of a list instead of at the start.) If this happened during a DATA
7491 ACL check, a -D file could be left lying around. This kind of configuration
7492 error no longer causes Exim to die; instead it causes a defer error. The
7493 incident is still logged to the main and panic logs.
7495 74. Buglet left over from Exim 3 conversion. The message "too many messages
7496 in one connection" was written to the rejectlog but not the mainlog, except
7497 when address rewriting (yes!) was being logged.
7499 75. Added write_rejectlog option.
7501 76. When a system filter was run not as root (that is, when system_filter_user
7502 was set), the values of the $n variables were not being returned to the
7503 main process; thus, they were not subsequently available in the $sn
7506 77. Added +return_path_on_delivery log selector.
7508 78. A connection timeout was being treated differently from recipients deferred
7509 when testing hosts_max_try with a message that was older than the host's
7510 retry timeout. (The host should not be counted, thus allowing all hosts to
7511 be tried at least once before bouncing.) This may have been the cause of an
7512 occasionally reported bug whereby a message would remain on the queue
7513 longer than the retry timeout, but would be bounced if a delivery was
7514 forced. I say "may" because I never totally pinned down the problem;
7515 setting up timeout/retry tests is difficult. See also the next item.
7517 79. The ultimate address timeout was not being applied to errors that involved
7518 a combination of host plus message (for example, a timeout on a MAIL
7519 command). When an address resolved to a number of possible hosts, and they
7520 were not all tried for each delivery (e.g. because of hosts_max_try), a
7521 message could remain on the queue longer than the retry timeout.
7523 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7526 81. Fixed an obscure SMTP outgoing bug which required at least the following
7527 conditions: (a) there was another message waiting for the same server;
7528 (b) the server returned 5xx to all RCPT commands in the first message so
7529 that the message was not completed; (c) the server dropped the connection
7530 or gave a negative response to the RSET that Exim sends to abort the
7531 transaction. The observed case was a dropped connection after DATA that had
7532 been sent in pipelining mode. That is, the server had advertised PIPELINING
7533 but was not implementing it correctly. The effect of the bug was incorrect
7534 behaviour, such as trying another host, and this could lead to a crash.
7540 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7541 and daemon.c were passed as pointers to ints; they should have been
7542 pointers to socklen_t variables (which are typically unsigned ints).
7544 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7547 3. Fixed a really odd bug that affected only the testing scheme; patching a
7548 certain fixed string in the binary changed the value of another string that
7549 happened to be identical to the end of the original first string.
7551 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7552 name", it returns that address as the IP address. On some operating
7553 systems (e.g. Solaris), it also passes back the IP address string as the
7554 "host name". However, on others (e.g. Linux), it passes back an empty
7555 string. Exim wasn't checking for this, and was changing the host name to an
7556 empty string, assuming it had been canonicalized.
7558 5. Although rare, it is permitted to have more than one PTR record for a given
7559 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7560 all the names associated with an address, because they do in Solaris.
7561 However, it seems that they do not in Linux for data that comes from the
7562 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7563 I found this out when I moved to a new Linux workstation and tried to run
7564 the Exim test suite.
7566 To get round this problem I have changed the code so that it now does its
7567 own call to the DNS to look up PTR records when searching for a host name.
7568 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7569 addresses that are only in /etc/hosts are still found.
7571 This behaviour is, however, controlled by an option called host_lookup_
7572 order, which defaults to "bydns:byaddr". If people want to use the other
7573 order, or indeed, just use one or the other means of lookup, they can
7574 specify it in this variable.
7576 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7577 some operating systems, this comes back from gethostbyaddr() as an empty
7578 string, and this is what Exim used to test for. However, it seems that in
7579 other systems, "." is yielded. Exim now tests for this case too.
7581 7. The values of check_spool_space and check_log_space are now held internally
7582 as a number of kilobytes instead of an absolute number of bytes. If a
7583 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7584 kilobyte. This means that much larger values can be stored.
7586 8. Exim monitor: an attempt to get the action menu when not actually pointing
7587 at a message produces an empty menu entitled "No message selected". This
7588 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7589 no entries in it ("Shell widget menu has zero width and/or height"). So I
7590 have added a single, blank menu entry in this case.
7592 9. Added ${quote_local_part.
7594 10. MIME decoding is now applied to the contents of Subject: header lines when
7597 11. Now that a reference to $sender_host_address automatically causes a reverse
7598 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7599 host lookup before query-style lookups in lists that might use this
7600 variable. This has therefore been abolished, and the "net-" prefix is no
7601 longer necessary for query-style lookups.
7603 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7604 have been a typo for LFLAGS, so it has been changed.
7606 13. The install script calls Exim with "-C /dev/null" in order to find the
7607 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7608 to be output. However, since Exim outputs its version number before the
7609 error, it didn't break the script. It just looked ugly. I fixed this by
7610 always allowing "-C /dev/null" if the caller is root.
7612 14. Ignore overlarge ACL variable number when reading spool file - insurance
7613 against a later release with more variables having written the file.
7615 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7616 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7617 this, as well as the form without the "IPv6" on the front.
7619 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7620 OS/Makefile-Darwin file.
7622 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7623 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7625 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7626 variable. However, some OS use ints for inodes. I've added cast to long int
7627 to get rid of the compiler warning.
7629 19. I had forgotten to lock out "/../" in configuration file names when
7630 ALT_CONFIG_PREFIX was set.
7632 20. Routers used for verification do not need to specify transports. However,
7633 if such a router generated a host list, and callout was configured, Exim
7634 crashed, because it could not find a port number from the (non-existent)
7635 transport. It now assumes port 25 in this circumstance.
7637 21. Added the -t option to exigrep.
7639 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7640 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7641 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7642 them is set, LOOKUP_LSEARCH is forced.
7644 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7646 24. Added sender and host information to the "rejected by local_scan()" log
7647 line; previously there was no indication of these.
7649 25. Added .include_if_exists.
7651 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7652 Linux. It turns out that not all file systems support this. Apparently some
7653 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7654 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7655 not supported on the file descriptor, is now ignored when Exim is trying to
7656 sync a directory. This applies only to Linux.
7658 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7660 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7661 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7664 29. Added support for maildirsize files from supplied patch (modified a bit).
7666 30. The use of :fail: followed by an empty string could lead Exim to respond to
7667 sender verification failures with (e.g.):
7669 550 Verification failed for <xxx>
7670 550 Sender verify failed
7672 where the first response line was missing the '-' that indicates it is not
7673 the final line of the response.
7675 31. The loop for finding the name of the user that called Exim had a hardwired
7676 limit of 10; it now uses the value of finduser_retries, which is used for
7677 all other user lookups.
7679 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7682 33. Exim was neglecting to zero errno before one call of strtol() when
7683 expanding a string and expecting an integer value. On some systems this
7684 resulted in spurious "integer overflow" errors. Also, it was casting the
7685 result into an int without checking.
7687 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7688 did not work. The code looks as if it has *never* worked, though it appears
7689 to have been documented since at least release 1.62. I have made it work.
7691 35. The "timeout_DNS" error in retry rules, also documented since at least
7692 1.62, also never worked. As it isn't clear exactly what this means, and
7693 clearly it isn't a major issue, I have abolished the feature by treating it
7694 as "timeout", and writing a warning to the main and panic logs.
7696 36. The display of retry rules for -brt wasn't always showing the error code
7699 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7700 timeout_connect_A, timeout_connect_MX.
7702 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7703 to the empty sender.
7705 39. The daemon was not analysing the content of -oX till after it had closed
7706 stderr and disconnected from the controlling terminal. This meant that any
7707 syntax errors were only noted on the panic log, and the return code from
7708 the command was 0. By re-arranging the code a little, I've made the
7709 decoding happen first, so such errors now appear on stderr, and the return
7710 code is 1. However, the actual setting up of the sockets still happens in
7711 the disconnected process, so errors there are still only recorded on the
7714 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7715 connections (as happens on some IP stacks) was logged at start up time as
7716 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7717 it from "IPv6 and IPv4", which means that two separate sockets are being
7720 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7721 says whether AF_INET or AF_INET6 was passed as an argument.
7723 42. Exiwhat output was messed up when time zones were included in log
7726 43. Exiwhat now gives more information about the daemon's listening ports,
7727 and whether -tls-on-connect was used.
7729 44. The "port" option of the smtp transport is now expanded.
7731 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7732 silently ignored. Now an error message is written to the main and panic
7735 46. There's a new ACL modifier called "logwrite" which writes to a log file
7736 as soon as it is encountered.
7738 47. Added $local_user_uid and $local_user_gid at routing time.
7740 48. Exim crashed when trying to verify a sender address that was being
7743 49. Exim was recognizing only a space character after ".include". It now also
7744 recognizes a tab character.
7746 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7747 extracting the relevant information from the specification. The man page no
7748 longer contains scrambled data for the -d option, and I've added a section
7749 at the front about calling Exim under different names.
7751 51. Added "extra_headers" argument to the "mail" command in filter files.
7753 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7756 53. Installed eximstats 1.29.
7758 54. Added transport_filter_timeout as a generic transport option.
7760 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7761 Cc: header lines. This was required by RFC 822, but it not required by RFC
7764 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7765 incoming messages that did not have them. Now it does so only if the
7766 message originates locally, that is, if there is no associated remote host
7767 address. When Resent- header lines are present, this applies to the Resent-
7768 lines rather than the non-Resent- lines.
7770 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7771 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7773 58. Messages for configuration errors now include the name of the main
7774 configuration file - useful now that there may be more than one file in a
7775 list (.included file names were always shown).
7777 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7778 for those rare installations that do not start the daemon as root or run it
7779 setuid root. I've cut out the call to initgroups() if the daemon is not
7782 60. The Exim user and group can now be bound into the binary as text strings
7783 that are looked up at the start of Exim's processing.
7785 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7787 62. Added $mailstore_basename variable.
7789 63. Installed patch to sieve.c from Michael Haardt.
7791 64. When Exim failed to open the panic log after failing to open the main log,
7792 the original message it was trying to log was written to stderr and debug
7793 output, but if they were not available (the usual case in production), it
7794 was lost. Now it is written to syslog before the two lines that record the
7795 failures to open the logs.
7797 65. Users' Exim filters run in subprocesses under the user's uid. It is
7798 possible for a "deliver" command or an alias in a "personal" command to
7799 provoke an address rewrite. If logging of address rewriting is configured,
7800 this fails because the process is not running as root or exim. There may be
7801 a better way of dealing with this, but for the moment (because 4.30 needs
7802 to be released), I have disabled address rewrite logging when running a
7803 filter in a non-root, non-exim process.
7809 1. The buildconfig auxiliary program wasn't quoting the value set for
7810 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7811 not defined. This bug was masked in 4.22 by the effect that was fixed in
7814 2. Some messages that were rejected after a message id was allocated were
7815 shown as "incomplete" by exigrep. It no longer does this for messages that
7816 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7818 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7819 have allow_domain_literals set, the ID did not get logged in the <= line.
7820 Domain literals are now always recognized in Message-ID: header lines.
7822 4. The first argument for a ${extract expansion item is the key name or field
7823 number. Leading and trailing spaces in this item were not being ignored,
7824 causing some misleading effects.
7826 5. When deliver_drop_privilege was set, single queue runner processes started
7827 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7828 same command in the process it spins off) were not dropping privilege.
7830 6. When the daemon running as "exim" started a queue runner, it always
7831 re-executed Exim in the spun-off process. This is a waste of effort when
7832 deliver_drop_privilege is set. The new process now just calls the
7833 queue-runner function directly.
7839 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7842 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7843 ignored. Though the use of -oP was forcing the writing of a pid file, it
7844 was always written to the default place.
7846 3. If the message "no IP address found for host xxxx" is generated during
7847 incoming verification, it is now followed by identification of the incoming
7848 connection (so you can more easily find what provoked it).
7850 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7852 5. Added some features to "harden" Exim a bit more against certain attacks:
7854 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7855 be put in Local/Makefile. This is like the never_users runtime option,
7856 but it cannot be overridden. The default setting is "root".
7858 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7859 prefix string with which any file named in a -C command line option
7862 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7863 is retained for -C and -D only if the caller of Exim is root. Without
7864 it, the exim user may also use -C and -D and retain privilege.
7866 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7867 command line option is disabled.
7869 6. Macro names set by the -D option must start with an upper case letter, just
7870 like macro names defined in the configuration file.
7872 7. Added "dereference=" facility to LDAP.
7874 8. Two instances of the typo "uknown" in the source files are fixed.
7876 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7877 the Configure-Makefile script screwed up while processing it.
7879 10. Incorporated PCRE 4.4.
7881 11. The SMTP synchronization check was not operating right at the start of an
7882 SMTP session. For example, it could not catch a HELO sent before the client
7883 waited for the greeting. There is now a check for outstanding input at the
7884 point when the greeting is written. Because of the duplex, asynchronous
7885 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7886 way, but not yet received, when the check is performed.
7888 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7889 on TCP/IP sockets, because this apparently causes some broken clients to
7892 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7893 unchanged) from the Cygwin maintainer.
7895 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7896 was supported instead of testing for mbx. Effectively a typo.
7898 15. The spa authenticator server code was not checking that the input it
7899 received was valid base64.
7901 16. The debug output line for the "set" modifier in ACLs was not showing the
7902 name of the variable that was being set.
7904 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7906 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7907 lookup on its own. Something else was needed to trigger the lookup. For
7908 example, a match in host_lookup or the need for a host name in a host list.
7909 Now, if $sender_host_name is referenced and the host name has not yet been
7910 looked up, a lookup is performed. If the lookup fails, the variable remains
7911 empty, and $host_lookup_failed is set to "1".
7913 19. Added "eqi" as a case-independent comparison operator.
7915 20. The saslauthd authentication condition could segfault if neither service
7916 nor realm was specified.
7918 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7919 error message that was logged was misleading, and incoming SMTP
7920 connections were dropped. The message is now more accurate, and temporary
7921 errors are given to SMTP connections.
7923 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7924 (except RSET) with a 421 error, until QUIT is received. However, it was
7925 failing to send a response to QUIT.
7927 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7928 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7929 ACL is now run after the helo_try_verify_hosts code.
7931 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7932 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7933 case-independent, so other case variants are also recognized). Apparently
7934 some systems use these upper case variants.
7936 25. If more than two messages were waiting for the same host, and a transport
7937 filter was specified for the transport, Exim sent two messages over the
7938 same TCP/IP connection, and then failed with "socket operation on non-
7939 socket" when it tried to send the third.
7941 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7943 27. The extern definition of crypt16() in expand.c was not being excluded when
7944 the OS had its own crypt16() function.
7946 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7947 as a preferred synonym for return_size_limit, both as an option and as an
7950 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7952 30. Changed the default configuration ACL to relax the local part checking rule
7953 for addresses that are not in any local domains. For these addresses,
7954 slashes and pipe symbols are allowed within local parts, but the sequence
7955 /../ is explicitly forbidden.
7957 31. SPA server authentication was not clearing the challenge buffer before
7960 32. log_message in a "warn" ACL statement was writing to the reject log as
7961 well as to the main log, which contradicts the documentation and doesn't
7962 seem right (because no rejection is happening). So I have stopped it.
7964 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7965 However, I am unable to do any testing of this.
7967 34. Fixed an infelicity in the appendfile transport. When checking directories
7968 for a mailbox, to see if any needed to be created, it was accidentally
7969 using path names with one or more superfluous leading slashes; tracing
7970 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7972 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7973 contents are added to the log line that is written for every discarded
7974 recipient. (Previously a log_message setting was ignored.)
7976 36. The ${quote: operator now quotes the string if it is empty.
7978 37. The install script runs exim in order to find its version number. If for
7979 some reason other than non-existence or emptiness, which it checks, it
7980 could not run './exim', it was installing it with an empty version number,
7981 i.e. as "exim-". This error state is now caught, and the installation is
7984 38. An argument was missing from the function that creates an error message
7985 when Exim fails to connect to the socket for saslauthd authentication.
7986 This could cause Exim to crash, or give a corrupted message.
7988 39. Added isip, isip4, and isip6 to ${if conditions.
7990 40. The ACL variables $acl_xx are now saved with the message, and can be
7991 accessed later in routers, transports, and filters.
7993 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7994 strings in the file are not string-expanded.
7996 42. If a MAIL command specified a SIZE value that was too large to fit into an
7997 int variable, the check against message_size_limit failed. Such values are
7998 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7999 day this will have to be increased, but I don't think I want to be around
8000 when emails are that large.
8007 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8008 iconv() is not standard in FreeBSD.
8010 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8011 IPv6 enabled. The observed symptom was a segmentation fault on return from
8012 the function os_common_find_running_interfaces() in src/os.c.
8014 3. In the check_special_case() function in daemon.c I had used "errno" as an
8015 argument name, which causes warnings on some systems. This was basically a
8016 typo, since it was named "eno" in the comments!
8018 4. The code that waits for the clock to tick (at a resolution of some fraction
8019 of a second) so as to ensure message-id uniqueness was always waiting for
8020 at least one whole tick, when it could have waited for less. [This is
8021 almost certainly not relevant at current processor speeds, where it is
8022 unlikely to ever wait at all. But we try to future-proof.]
8024 5. The function that sleeps for a time interval that includes fractions of a
8025 second contained a race. It did not block SIGALRM between setting the
8026 timer, and suspending (a couple of lines later). If the interval was short
8027 and the sigsuspend() was delayed until after it had expired, the suspension
8028 never ended. On busy systems this could lead to processes getting stuck for
8031 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8032 process, before it forks any delivery processes. The open lookup caching
8033 mechanism meant that the open file or database connection was passed into
8034 the delivery process. The problem was that delivery processes always tidy
8035 up cached lookup data. This could cause a problem for the next delivery
8036 process started by the queue runner, because the external queue runner
8037 process does not know about the closure. So the next delivery process
8038 still has data in the lookup cache. In the case of a file lookup, there was
8039 no problem because closing a file descriptor in a subprocess doesn't affect
8040 the parent. However, if the lookup was caching a connection to a database,
8041 the connection was closed, and the second delivery process was likely to
8042 see errors such as "PGSQL: query failed: server closed the connection
8043 unexpectedly". The problem has been fixed by closing all cached lookups
8044 in a queue runner before running a delivery process.
8046 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8047 seem to have the "const" qualifier which it has on other OS. I've
8050 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8051 *of the same type* (client or server) with the same public name that an
8052 error should be diagnosed.
8054 9. When Exim looked up a host name for an IP address, but failed to find the
8055 original IP address when looking up the host name (a safety check), it
8056 output the message "<ip address> does not match any IP for NULL", which was
8057 confusing, to say the least. The bug was that the host name should have
8058 appeared instead of "NULL".
8060 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8061 user that is built into the binary, and the -C or -D options is used, root
8062 privilege is dropped before the configuration file is read. In addition,
8063 logging is switched to stderr instead of the normal log files. If the
8064 configuration then re-defines the Exim user, the unprivileged environment
8065 is probably not what is expected, so Exim logs a panic warning message (but
8068 However, if deliver_drop_privilege is set, the unprivileged state may well
8069 be exactly what is intended, so the warning has been cut out in that case,
8070 and Exim is allowed to try to write to its normal log files.
8076 1. smtp_return_error_details was not giving details for temporary sender
8077 or receiver verification errors.
8079 2. Diagnose a configuration error if two authenticators have the same public
8082 3. Exim used not to create the message log file for a message until the first
8083 delivery attempt. This could be confusing when incoming messages were held
8084 for policy or load reasons. The message log file is now created at the time
8085 the message is received, and an initial "Received" line is written to it.
8087 4. The automatically generated man page for command line options had a minor
8088 bug that caused no ill effects; however, a more serious problem was that
8089 the procedure for building the man page automatically didn't always
8090 operate. Consequently, release 4.20 contains an out-of-date version. This
8091 shouldn't happen again.
8093 5. When building Exim with embedded Perl support, the script that builds the
8094 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8095 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8097 6. The freeze_tell option was not being used for messages that were frozen on
8098 arrival, either by an ACL or by local_scan().
8100 7. Added the smtp_incomplete_transaction log selector.
8102 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8103 was accepting AUTH without a new EHLO.
8105 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8106 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8109 10. Logging of TCP/IP connections (when configured) now happens in the main
8110 daemon process instead of the child process, so that the TCP/IP connection
8111 count is more accurate (but it can never be perfect).
8113 11. The use of "drop" in a nested ACL was not being handled correctly in the
8114 outer ACL. Now, if condition failure induced by the nested "drop" causes
8115 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8116 or "require"), the connection is dropped.
8118 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8119 that yield "discard" can now be used with an "accept" or a "discard" verb,
8120 but an error is generated for any others (because I can't see a useful way
8121 to define what should happen).
8123 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8124 are now processed in the same way as lines in the Exim configuration file.
8125 In particular, continuation lines are supported.
8127 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8129 15. Added -ti meaning -t -i.
8131 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8132 domains, and warn by logging if others are found.
8134 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8135 returned from GIFCONF when Exim is trying to find the list of interfaces on
8136 a host. The code in os.c has been modified to copy each ifreq to an aligned
8137 structure in all cases.
8139 Also, in some cases, the returned ifreq's were being copied to a 'struct
8140 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8141 means the last couple of bytes of an IPv6 address could be chopped if the
8142 ifreq contained only a normal sockaddr (14 bytes storage).
8144 18. Named domain lists were not supported in the hosts_treat_as_local option.
8145 An entry such as +xxxx was not recognized, and was treated as a literal
8148 19. Ensure that header lines added by a DATA ACL are included in the reject log
8149 if the ACL subsequently rejects the message.
8151 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8152 MD5 (which is deprecated).
8154 21. When testing a filter file using -bf, Exim was writing a message when it
8155 took the sender from a "From " line in the message, but it was not doing so
8156 when it took $return_path from a Return-Path: header line. It now does.
8158 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8159 with a valid header line field name (a series of printing characters
8160 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8162 23. Changed "disc" in the source to "disk" to conform to the documentation and
8163 the book and for uniformity.
8165 24. Ignore Sendmail's -Ooption=value command line item.
8167 25. When execve() failed while trying to run a command in a pipe transport,
8168 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8169 could be confused with a return value of 69 from the command itself. This
8170 has been changed to 127, the value the shell returns if it is asked to run
8171 a non-existent command. The wording for the related log line suggests a
8172 non-existent command as the problem.
8174 26. If received_header_text expands to an empty string, do not add a Received:
8175 header line to the message. (Well, it adds a token one on the spool, but
8176 marks it "old" so that it doesn't get used or transmitted.)
8178 27. Installed eximstats 1.28 (addition of -nt option).
8180 28. There was no check for failure on the call to getsockname() in the daemon
8181 code. This can fail if there is a shortage of resources on the system, with
8182 ENOMEM, for example. A temporary error is now given on failure.
8184 29. Contrary to the C standard, it seems that in some environments, the
8185 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8186 program. Exim now does this explicitly; it affects the formatting of
8187 timestamps using strftime().
8189 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8190 complaints. I've now initialized all the variables, to avoid this.
8192 32. Header lines added by a system filter were not being "seen" during
8193 transport-time rewrites.
8195 33. The info_callback() function passed to OpenSSL is set up with type void
8196 (*)(SSL *, int, int), as described somewhere. However, when calling the
8197 function (actually a macro) that sets it up, the type void(*)() is
8198 expected. I've put in a cast to prevent warnings from picky compilers.
8200 34. If a DNS black list lookup found a CNAME record, but there were no A
8201 records associated with the domain it pointed at, Exim crashed.
8203 35. If a DNS black list lookup returned more than one A record, Exim ignored
8204 all but the first. It now scans all returned addresses if a particular IP
8205 value is being sought. In this situation, the contents of the
8206 $dnslist_value variable are a list of all the addresses, separated by a
8209 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8210 to accept a host name and all its aliases if the forward lookup for any of
8211 them yielded the IP address of the incoming connection. Now it accepts only
8212 those names whose forward lookup yields the correct IP address. Any other
8213 names are discarded. This closes a loophole whereby a rogue DNS
8214 administrator could create reverse DNS records to break through a
8215 wildcarded host restriction in an ACL.
8217 37. If a user filter or a system filter that ran in a subprocess used any of
8218 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8219 the wrong values were passed to the pipe command ($thisaddress had the
8220 value of $0, $0 had the value of $1, etc). This bug was introduced by
8221 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8223 38. Improved the line breaking for long SMTP error messages from ACLs.
8224 Previously, if there was no break point between 40 and 75 characters, Exim
8225 left the rest of the message alone. Two changes have been made: (a) I've
8226 reduced the minimum length to 35 characters; (b) if it can't find a break
8227 point between 35 and 75 characters, it looks ahead and uses the first one
8228 that it finds. This may give the occasional overlong line, but at least the
8229 remaining text gets split now.
8231 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8232 file descriptors might be low, and that setting 1000 would always raise it.
8233 It turns out that in some environments, the limit is already over 1000 and
8234 that lowering it causes trouble. So now Exim takes care not to decrease it.
8236 40. When delivering a message, the value of $return_path is set to $sender_
8237 address at the start of routing (routers may change the value). By an
8238 oversight, this default was not being set up when an address was tested by
8239 -bt or -bv, which affected the outcome if any router or filter referred to
8242 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8243 log only when "message" or "log_message" are set. However, if one of the
8244 conditions was an address verification, or a call to a nested ACL, the
8245 messages generated by the underlying test were being passed through. This
8246 no longer happens. The underlying message is available in $acl_verify_
8247 message for both "message" and "log_message" expansions, so it can be
8248 passed through if needed.
8250 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8251 new expansion $bh_ to give the encoded byte string without charset
8252 translation. Translation happens only if iconv() is available; HAVE_ICONV
8253 indicates this at build time. HEADERS_CHARSET gives the charset to
8254 translate to; headers_charset can change it in the configuration, and
8255 "headers charset" can change it in an individual filter file.
8257 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8258 that creates RFC 2047 encoded "words" labels them as that charset instead
8259 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8260 expansion operator; (ii) when Exim creates a From: line for a local
8261 message; (iii) when a header line is rewritten to include a "phrase" part.
8263 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8264 buggy, causing it to skip the first lines of messages whose message ID
8265 ended in 'D'. This would not have bitten before Exim release 4.14, because
8266 message IDs were unlikely to end in 'D' before then. The effect was to have
8267 incorrect size information for certain domains.
8269 45. #include "config.h" was missing at the start of the crypt16.c module. This
8270 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8273 46. If there was a timeout during a "random" callout check, Exim treated it as
8274 a failure of the random address, and carried on sending RSET and the real
8275 address. If the delay was just some slowness somewhere, the response to the
8276 original RCPT would be taken as a response to RSET and so on, causing
8277 mayhem of various kinds.
8279 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8280 when I implemented it. It didn't allow for the fact that some option values
8281 may legitimately be negative (e.g. size_addition), and it didn't even do
8282 the right test for positive values.
8284 48. Domain names in DNS records are case-independent. Exim always looks them up
8285 in lower case. Some resolvers return domain names in exactly the case they
8286 appear in the zone file, that is, they may contain uppercase letters. Not
8287 all resolvers do this - some return always lower case. Exim was treating a
8288 change of case by a resolver as a change of domain, similar to a widening
8289 of a domain abbreviation. This triggered its re-routing code and so it was
8290 trying to route what was effectively the same domain again. This normally
8291 caused routing to fail (because the router wouldn't handle the domain
8292 twice). Now Exim checks for this case specially, and just changes the
8293 casing of the domain that it ultimately uses when it transmits the message
8296 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8299 50. If a filter generated a file delivery with a non-absolute name (possible if
8300 no home directory exists for the router), the forbid_file option was not
8303 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8304 the existing equality matching.
8306 52. Exim was using ints instead of ino_t variables in some places where it was
8307 dealing with inode numbers.
8309 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8310 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8311 called TMPDIR, and if it finds it is different, it changes its value.
8313 54. The smtp_printf() function is now made available to local_scan() so
8314 additional output lines can be written before returning. There is also an
8315 smtp_fflush() function to enable the detection of a dropped connection.
8316 The variables smtp_input and smtp_batched_input are exported to
8319 55. Changed the default runtime configuration: the message "Unknown user"
8320 has been removed from the ACL, and instead placed on the localuser router,
8321 using the cannot_route_message feature. This means that any verification
8322 failures that generate their own messages won't get overridden. Similarly,
8323 the "Unrouteable address" message that was in the ACL for unverifiable
8324 relay addresses has also been removed.
8326 56. Added hosts_avoid_esmtp to the smtp transport.
8328 57. The exicyclog script was not checking for the esoteric option
8329 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8330 will work only if exicyclog is run under the appropriate euid.
8332 58. Following a discussion on the list, the rules by which Exim recognises line
8333 endings on incoming messages have been changed. The -dropcr and drop_cr
8334 options are now no-ops, retained only for backwards compatibility. The
8335 following line terminators are recognized: LF CRLF CR. However, special
8336 processing applies to CR:
8338 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8339 nor a local message in the state where . is a terminator.
8341 (ii) If a bare CR is encountered in a header line, an extra space is added
8342 after the line terminator so as not to end the header. The reasoning
8343 behind this is that bare CRs in header lines are most likely either
8344 to be mistakes, or people trying to play silly games.
8346 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8347 was being incorrectly given as 18 bytes larger than it should have been.
8348 This is a VOB (very old bug).
8350 60. This may never have affected anything current, but just in case it has:
8351 When the local host is found other than at the start of a list of hosts,
8352 the local host, those with the same MX, and any that follow, are discarded.
8353 When the list in question was part of a longer list of hosts, the following
8354 hosts (not currently being processed) were also being discarded. This no
8355 longer happens. I'm not sure if this situation could ever has previously
8358 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8361 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8362 header. This is something that is recommended in a new Internet Draft, and
8363 is something that is documented as being done by Sendmail. There are two
8364 possible values. For messages generated by the autoreply transport, Exim
8367 Auto-Submitted: auto-replied
8369 whereas for all other generated messages (e.g. bounces) it adds
8371 Auto-Submitted: auto-generated
8373 63. The "personal" condition in filters now includes a test for the
8374 Auto-Submitted: header. If it contains the string "auto-" the message it
8375 not considered personal.
8377 64. Added rcpt_include_affixes as a generic transport option.
8379 65. Added queue_only_override (default true).
8381 66. Added the syslog_duplication option.
8383 67. If what should have been the first header line of a message consisted of
8384 a space followed by a colon, Exim was mis-interpreting it as a header line.
8385 It isn't of course - it is syntactically invalid and should therefore be
8386 treated as the start of the message body. The misbehaviour could have
8387 caused a number of strange effects, including loss of data in subsequent
8388 header lines, and spool format errors.
8390 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8391 client host had authenticated. This control can now be exercised by an ACL
8392 for more flexibility.
8394 69. By default, callouts do not happen when testing with -bh. There is now a
8395 variant, -bhc, which does actually run the callout code, including
8396 consulting and updating the callout cache.
8398 70. Added support for saslauthd authentication, courtesy of Alexander
8401 71. If statvfs() failed on the spool or log directories while checking their
8402 size for availability, Exim confusingly gave the error "space shortage".
8403 Furthermore, in debugging mode it crashed with a floating point exception.
8404 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8405 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8406 serious problem, Exim now writes to the main and panic logs when this
8407 happens, with details of the failure. It then refuses to accept the
8408 incoming message, giving the message "spool directory problem" or "log
8409 directory problem" with a 421 code for SMTP messages.
8411 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8412 0, 1, and 2 exist, because some OS complain for execs without them (see
8413 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8414 descriptors. However, the code omitted to check that the open succeeded,
8415 causing mysterious errors if for some reason the permissions on /dev/null
8416 got screwed. Now Exim writes a message to the main and panic logs, and
8417 bombs out if it can't open /dev/null.
8419 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8420 interact so that it is all more flexible. It is supposed to remain
8421 backwards compatible. Also added extra_local_interfaces.
8423 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8424 to bomb out with an assertion failure - to the client this appears as a
8425 connection drop. This problem occurs in the part of the code that was taken
8426 from the Samba project. Fortunately, the assertion is in a very simple
8427 function, so I have fixed this by reproducing the function inline in the
8428 one place where it is called, and arranging for authentication to fail
8429 instead of killing the process with assert().
8431 75. The SPA client code was not working when the server requested OEM rather
8432 than Unicode encoding.
8434 76. Added code to make require_files with a specific uid setting more usable in
8435 the case where statting the file as root fails - usually a non-root-mounted
8436 NFS file system. When this happens and the failure is EACCES, Exim now
8437 forks a subprocess and does the per-uid checking as the relevant uid.
8439 77. Added process_log_path.
8441 78. If log_file_path was not explicitly set, a setting of check_log_space or
8442 check_log_inodes was ignored.
8444 79. If a space check for the spool or log partitions fails, the incident is now
8445 logged. Of course, in the latter case the data may get lost...
8447 80. Added the %p formatting code to string_format() so that it can be used to
8448 print addresses in debug_print(). Adjusted all the address printing in the
8449 debugging in store.c to use %p rather than %d.
8451 81. There was a concern that a line of code in smtp_in.c could overflow a
8452 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8453 initially expressed, the concern was not well-founded, because trailing
8454 spaces are removed early. However, if the trailing spaces were followed by
8455 a NULL, they did not get removed, so the overflow was possible. Two fixes
8458 (a) I re-wrote the offending code in a cleaner fashion.
8459 (b) If an incoming SMTP command contains a NULL character, it is rejected
8462 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8463 runs initgroups(), so that if the Exim user is in any additional groups,
8464 they will be used during message reception.
8470 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob