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.
192 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
193 after reception to before a subsequent reception. This should
194 mean slightly faster delivery, and also confirmation of reception
197 JH/02 Move from using the pcre library to pcre2. The former is no longer
198 being developed or supported (by the original developer).
200 JH/03 Constification work in the filters module required a major version
201 bump for the local-scan API. Specifically, the "headers_charset"
202 global which is visible via the API is now const and may therefore
203 not be modified by local-scan code.
205 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
206 sendfile() didi not account for the way the ClamAV driver code called it.
208 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
209 time check was being done for every character; replace that with one
212 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
213 sent was prefixed with a length byte.
215 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
216 RFC 5321. Previously Dovecot (at least) would log errors during
219 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
220 in 4.61 without a reason logged. Binaries will be bigger, which might
221 matter on diskspace-constrained systems, but debug is easier.
223 JH/09 Fix macro-definition during "-be" expansion testing. The move to
224 write-protected store for macros had not accounted for these runtime
225 additions; fix by removing this protection for "-be" mode.
227 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
228 handing out large-numbered file descriptors, violating the usual Unix
229 assumption (and required by Posix) that the lowest possible number will be
230 allocated by the kernel when a new one is needed. In the daemon, and any
231 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
232 are not useable for FD_SET() [and hence select()] and overwrite the stack.
233 Assorted crashes happen.
235 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
236 main-section options or in a connect ACL, the value from the first ever
237 connection was never replaced for subsequent connections. Found by
240 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
241 though only once PCRE2 was introduced: the memory accounting used under
242 debug offset allocations by an int, giving a hard trap in early startup.
243 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
245 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
246 with underbars is given. The write-protection of configuration introduced
247 in 4.95 trapped when normalisation was applied to an option not needing
250 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
252 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
253 failing to close the certificates directory, every hour or any time it
256 JH/16 Debugging initiated by an ACL control now continues through into routing
257 and transport processes. Previously debugging stopped any time Exim
258 re-execs, or for processing a queued message.
260 JH/17 The "expand" debug selector now gives more detail, specifically on the
261 result of expansion operators and items.
263 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
264 bad comparison between the option value and the name of the file to
265 be included was done, and a mismatch was wrongly identified.
266 4.88 to 4.95 are affected.
268 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
270 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
271 supplied as the db file-pair basename. Previously, if a directory
272 path was given, for example via the autoreply "once" option, the DB
273 file.pag and file.dir files would be created in that directory's
276 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
277 "taint" log_selector. These were previously deprecated.
279 JH/22 Fix static address-list lookups to properly return the matched item.
280 Previously only the domain part was returned.
282 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
283 the call into OpenSSL to send a TLS Close was being repeated; this
284 resulted in the library waiting for the peer's Close. If that was never
285 sent we waited forever. Fix by tracking send calls.
287 JH/24 The ${run} expansion item now expands its command string elements after
288 splitting. Previously it was before; the new ordering makes handling
289 zero-length arguments simpler. The old ordering can be obtained by
290 appending a new option "preexpand", after a comma, to the "run".
292 JH/25 Taint-check exec arguments for transport-initiated external processes.
293 Previously, tainted values could be used. This affects "pipe", "lmtp" and
294 "queryprogram" transport, transport-filter, and ETRN commands.
295 The ${run} expansion is also affected: in "preexpand" mode no part of
296 the command line may be tainted, in default mode the executable name
299 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
300 the the facility was not passed across execs, and only the first message
301 passed over a connection could use BDAT; any further ones using DATA.
303 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
304 uses $sending_ip_address and an interface is specified.
305 Previously any use of the local address in the EHLO name disabled
306 PIPECONNECT, the common case being to use the rDNS of it.
308 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
309 resumption. Previously verify failed because no certificate status is
310 passed on the wire for the restarted session. Fix by using the recorded
311 ocsp status of the stored session for the new connection.
313 JH/29 TLS resumption: the key for session lookup in the client now includes
314 more info that a server could potentially use in configuring a TLS
315 session, avoiding oferring mismatching sessions to such a server.
316 Previously only the server IP was used.
318 JH/30 Fix string_copyn() for limit greater than actual string length.
319 Previously the copied amount was the limit, which could result in a
320 overlapping memcpy for newly allocated destination soon after a
321 source string shorter than the limit. Found/investigated by KM.
323 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
324 close; it may be needed for a subsequent connection. This caused a
325 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
327 JH/32 Fix CHUNKING for a second message on a connection when the first was
328 rejected. Previously we did not reset the chunking-offered state, and
329 erroneously rejected the BDAT command. Investigation help from
332 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
333 an empty address. Previously the expansion returned an error.
335 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
336 proxy. Previously these were misparsed, leading to paniclog entries.
342 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
343 related applications. Previously an "H" was used where available info
344 says that "M" should be, so change to match.
346 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
347 as arguments, so an implementation trying to copy these into a local
348 buffer was taking a taint-enforcement trap. Fix by using dynamically
349 created buffers. Similar fix for radius expansion condition.
351 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
352 reasonable, eg. to count headers. Fix by using dynamically created
353 buffers rather than a local. Do similar fixes for ACL actions "dcc",
354 "log_reject_target", "malware" and "spam"; the arguments are expanded
355 so could be handling tainted values.
357 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
358 broken the (no-op) support for this sendmail command. Restore it
359 to doing nothing, silently, and returning good status.
361 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
362 record path was given (or the default used) without a leading directory
363 path, an error occurred on trying to open it. Use the transport's working
366 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
367 transport. Previously it was the name on the DNS A-record; use instead
368 the head of the CNAME chain leading there (if there is one). This seems
369 to align better with RFC 6125.
371 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
372 smtp_accept_max_per_host allocated resources which were not released
373 when the limit was exceeded. This eventually crashed the daemon. Fix
374 by adding a release action in that path.
376 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
377 expanded; previously using tainted values was rejected. Fix by using
378 dynamically-created buffers.
380 JH/09 Relax restrictions on ACL verify condition needing access to message
381 headers. Previously they were only permitted in data and non-smtp ACLs;
382 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
383 not_blind, header_sender and header_names_ascii verification.
385 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
386 Previously a macro used one argument twice; when called with the
387 argument as an expression having side-effects, incorrect operation
388 resulted. Use an inlineable function.
390 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
391 held open for a verify callout. Previously this wan not accounted for
392 and a corrupt onward SMTP conversation resulted.
394 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
395 Previously they were embedded naively in the constructed address; when
396 needed, strip the quoting and quote the entire local_part.
397 Also make the inbound_srs expansion condition handle quoting.
399 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
400 excluded, not matching the documentation.
402 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
403 was given for the sqlite_dbfile a trap resulted.
405 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
406 "name" argument resulted in a trap. There is no reason to disallow such;
407 this was a coding error.
409 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
410 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
411 spent suspended, ignoring the POSIX definition. Previously we assumed
412 it did and a constant offset from real time could be used as a correction.
413 Change to using the same clock source for the start-of-message and the
414 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
415 exists, just to get a clock slightly more aligned to reality.
417 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
418 RFC says it is optional some validators care. The missing char was not
419 intended but triggered by a line-wrap alignment. Discovery and fix by
420 Guillaume Outters, hacked on by JH.
422 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
423 name being quoted was tainted a trap would be taken. Fix by using
424 dynamically created buffers. The routine could have been called by a
425 rewrite with the "h" flag, by using the "-F" command-line option, or
426 by using a "name=" option on a control=submission ACL modifier.
428 JH/19 SPF: change the Authentication-Results expansion component to give
429 smtp.helo when the sender domain is empty. Previously it gave
432 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
433 not in 127.0.0.0/8 to help in spotting list domains taken over by a
434 domain-parking registrar.
436 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
437 Previously when a whitespace character was specified it was not inserted
438 after removing the newline.
440 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
441 the domain part of the recipient address. This overrides any tls_sni
442 option set, which was previously used.
444 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
447 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
448 is_tainted() had an off-by-one error in the overenthusiastic direction.
449 Find and fix by Gavan. Although NetBSD is not a supported platform for
450 4.94 this bug could affect other platforms.
452 PP/01 Fix default prime selection to be consistent.
453 One path used ike23 still, instead of exim.dev.20160529.3; now both
454 execution flows will use the same DH primes (currently
455 exim.dev.20160529.3).
457 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
458 option in smtp transport, to match the documentation. Previously
459 verification was not being done in some cases where it should have been.
461 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
462 than one server was defined and depending on the platform memory layout
463 details, an internal consistency trap could be hit while walking the list
466 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
467 files. The value is used by the authresults expansion item. Previously
468 if this was used in a router or transport, a crash could result.
470 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
471 have a sufficient incidence of EINTR returns from select that an
472 interaction with other operations done by the main daemon loop exposed
473 a bug in the error-handling. This was benign apart from the log
476 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
477 for consistency with delivered "=>" and failed "**" lines. While we're
478 there, handle PRX and TFO.
480 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
481 applied. This resulted, if any header-line rewrite rules were configured,
482 in a panic-log triggerable by sending a message with a long address in
483 a header. Fix by increasing the arbitrary limit to larger than a single
484 (dewrapped) 5322 header line maximum size.
486 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
487 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
488 RFC 6648 which deprecates X- options in protocols as a general practice.
489 Changeover between the implementations is handled by the mechanisms
492 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
493 a remote which uses callout/hold. Previously the local was queued.
495 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
498 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
499 log line. The asterisk is now consistently appended to the remote IP
500 (and port, if given), and will also be provided on defer and fail log
501 lines. Previously it could be placed on the local IP if that was being
502 logged, and was only provided on delivery lines.
504 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
506 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
507 authenticator client_send option. Previously the next char, after a pair
508 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
509 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
510 documentation. There is still no way to get a leading ^ immediately
511 after a NUL (ie. for the password of a PLAIN method authenticator.
513 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
514 files. For bad sizes read, delete the record and whine to paniclog.
516 JH/38 When logging an AUTH failure, as server, do not include sensitive
517 information. Previously, the credentials would be included if given
518 as part of the AUTH command line and an ACL denied authentication.
520 JH/39 Bug 2691: fix $local_part_data. When the matching list element
521 referred to a file, bad data was returned. This likely also affected
524 JH/40 The gsasl authenticator now supports caching of the salted password
525 generated by the client-side implementation. This required the addition
526 of a new variable: $auth4.
528 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
529 left undeleted; the attempt to re-create it then failed - resulting in
530 the usual "SIGHUP tp have daemon reload configuration" to not work.
531 This affected any platform not supporting "abstract" Unix-domain
532 sockets (i.e. not Linux).
534 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
535 recipients" response to RCPT in a later response, with a 250. The
536 previous coding assumed this would not happen, and under PIPELINING
537 would result in both lost and duplicate recipients for a message.
539 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
540 Previously the weighting was incorrectly applied. Similar fix for socks
541 proxies. Found and fixed by Heiko Schlichting.
543 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
544 not handle sub-lists included using the +namedlist syntax. While
545 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
546 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
549 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
550 Previously the permanent pool was used, so the sore could not be freed.
551 This meant a connection with many messages would use continually-growing
554 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
555 per-pool so as not to waste too much space. Previously a constant size
556 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
557 DOS attacks harder. The cost is wasted memory use in the larger blocks.
559 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
560 same space repeatedly, and a smaller process footprint.
562 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
563 for GnuTLS. Previously the name started "%s" which, while not a bug,
564 looked as if if might be one.
566 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
567 "continued connection") make the $proxy_* variables available. Previously
568 the information was not passed across the exec() call for subsequent
569 transport executions. This also mean that the log lines for the
570 messages can show the proxy information.
572 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
573 receive time. With modern systems the difference is significant.
574 The historical behaviour can be restored by disabling (a new) log_selector
575 "queue_time_exclusive".
577 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
578 content) were specifically tested for. Now, also cover expansions
579 resulting in ACL names and inline ACL content.
581 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
582 making it unusable in complex expressions.
584 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
585 fail with a taint-check on the spoolfile name, and leave the message
588 HS/01 Enforce absolute PID file path name.
590 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
592 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
594 PP/02 Bug 2643: Correct TLS DH constants.
595 A missing NUL termination in our code-generation tool had led to some
596 incorrect Diffie-Hellman constants in the Exim source.
597 Reported by kylon94, code-gen tool fix by Simon Arlott.
599 PP/03 Impose security length checks on various command-line options.
600 Fixes CVE-2020-SPRSS reported by Qualys.
602 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
603 better. Reported by Qualys.
605 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
606 providing a particularly obnoxious sender full name.
609 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
611 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
614 PP/08 Change default for recipients_max from unlimited to 50,000.
616 PP/09 Fix security issue with too many recipients on a message (to remove a
617 known security problem if someone does set recipients_max to unlimited,
618 or if local additions add to the recipient list).
619 Fixes CVE-2020-RCPTL reported by Qualys.
621 PP/10 Fix security issue in SMTP verb option parsing
622 Fixes CVE-2020-EXOPT reported by Qualys.
624 PP/11 Fix security issue in BDAT state confusion.
625 Ensure we reset known-good where we know we need to not be reading BDAT
626 data, as a general case fix, and move the places where we switch to BDAT
627 mode until after various protocol state checks.
628 Fixes CVE-2020-BDATA reported by Qualys.
630 HS/03 Die on "/../" in msglog file names
632 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
633 the Exim runtime user are allowed to create files.
635 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
638 QS/03 When reading the output from interpreted forward files we do not
639 pass the pipe between the parent and the interpreting process to
640 executed child processes (if any).
642 QS/04 Always die if requested from internal logging, even is logging is
645 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
646 the API; compilation noo longer completes with DMARC support included.
647 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
648 been reported on other platforms.
650 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
652 JH/56 Make the majority of info read from config files readonly, for defence-in-
653 depth against exploits. Suggestion by Qualys.
654 Not supported on Solaris 10.
656 JH/57 Fix control=fakreject for a custom message containing tainted data.
657 Previously this resulted in a log complaint, due to a re-expansion present
658 since fakereject was originally introduced.
660 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
661 modifier was given, a loop resulted.
663 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
664 pipelined SMTP command followed the BDAT LAST then it would be
665 incorrectly treated as part of the message body, causing a verification
668 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
669 require looser validation rules than those for 821-level addresses,
670 which only permit IP addresses.
676 JH/01 Avoid costly startup code when not strictly needed. This reduces time
677 for some exim process initialisations. It does mean that the logging
678 of TLS configuration problems is only done for the daemon startup.
680 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
682 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
683 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
686 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
687 DKIM signing is being done. Previously a transport_filter would always
688 disable CHUNKING, falling back to traditional DATA.
690 JH/05 Regard command-line recipients as tainted.
692 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
694 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
695 PAM library frees one of the arguments given to it, despite the
696 documentation. Therefore a plain malloc must be used.
698 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
699 on-stack buffers were used, resulting in a taint trap when DSN information
700 copied from a received message was written into the buffer.
702 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
703 the ordering of its ARC headers. This caused a crash.
705 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
706 a new record was being constructed with information from the peer, a trap
709 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
710 installation would get error messages from DMARC verify, when it hit the
711 nonexistent file indicated by the default. Distros wanting DMARC enabled
712 should both provide the file and set the option.
713 Also enforce no DMARC verification for command-line sourced messages.
715 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
716 could, depending on the platform, hang at the STARTTLS response.
718 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
719 message on a connection. Previously if one message had ARC headers and
720 the following one did not, a crash could result when adding an
721 Authentication-Results: header.
723 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
724 Exim main code and Exim-related utities. The introduction of taint
725 tracking also did many adjustments to string handling. Since then, eximon
726 frequently terminated with an assert failure.
728 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
729 check for 452 responses. This slightly helps the inefficieny of doing
730 a large alias-expansion into a recipient-limited target. The max_rcpt
731 transport option still applies (and at the current default, will override
732 the new feature). The check is done for either cause of synch, and forces
733 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
734 connection. The new facility is not tunable at this time.
736 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
737 library live data was being used, so the results became garbage. Make
738 copies while it is still usable.
740 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
741 on delivery deferred (==) and failed (**) lines (if a delivery was
742 attemtped). Previously it was only on completion (=>) lines.
744 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
745 for the expansion of the server_scram_iter and server_scram_salt options.
747 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
748 are now specifically given a NO_DATA response without hitting the system
749 resolver. The library goes on to do the now-standard TXT lookup.
750 Use of dnsdb lookups is not affected.
752 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
753 only retrieve the errormessage once. Previously two calls to dlerror()
754 were used, and the second one (for mainlog/paniclog) retrieved null
757 JH/20 Taint checking: disallow use of tainted data for
758 - the appendfile transport file and directory options
759 - the pipe transport command
760 - the autoreply transport file, log and once options
761 - file names used by the redirect router (including filter files)
763 - paths used by single-key lookups
764 Previously this was permitted.
766 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
767 adjusted the size of a major service buffer; this failed because the
768 buffer was in use at the time. Change to a compile-time increase in the
769 buffer size, when this authenticator is compiled into exim.
771 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
772 previous fast-mode was untenable in the face of glibs using mmap to
773 support larger malloc requests.
775 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
776 New values supported, if defined on system where compiled:
777 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
778 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
780 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
781 running a limited number of proceses in parallel, a benefit is gained. The
782 amount varies with the platform hardware and load. The use of the option
783 queue_run_in_order means we cannot do this, as ordering becomes
786 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
787 had introduced a string-copy (for ensuring NUL-termination) which was not
788 appropriate for that case, which can include embedded NUL bytes in the
789 block of data. Investigation showed the copy to actually be needless, the
790 data being length-specified.
792 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
793 done during a receiving connection, and both used TLS, global info was
794 used rather than per-connection info for tracking the state of data
795 queued for transmission. This could result in a connection hang.
797 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
798 Previously, when delivering serveral messages down a single connection
799 only the first would provide a SIZE. This was due to the size information
800 not being properly tracked.
802 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
803 TAI (at 37 seconds currently), pretend to be in UTC for time-related
804 expansion and logging. Previously, spurious values such as a future
805 minute could be seen.
807 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
808 it could crash from a null-deref. This could also affect the
809 ${addresses: } operator and ${readsock } item.
811 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
812 message following a mime one, the variable was not reset.
814 JH/30 When an pipelined-connect fails at the first response, assume incorrect
815 cached capability (perhaps the peer reneged?) and immediately retry in
818 JH/31 Fix spurious detection of timeout while writing to transport filter.
820 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
821 an attempt to copy the string was made before checking it.
823 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
824 taint of the lookup key was maintained; we now regard the presence in the
825 filesystem as sufficient validation.
827 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
828 argument is supplied.
830 JH/35 The dsearch lookup now requires that the directory is an absolute path.
831 Previously this was not checked, and nonempty relative paths made an
832 access under Exim's current working directory.
834 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
835 Previously no event was raised.
837 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
838 parameter supplied by the sender MAIL FROM command. Previously it was
839 ignored, and only the check_spool_space option value for the required
842 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
843 the size of the signing public-key. Previously it was instead giving
844 the size of the signature hash.
846 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
847 the default. See the (new) dkim_verify_min_keysizes option.
849 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
850 and an ACL use a lookup for checking either the local_part or domain,
851 stale data could be accessed. Ensure that variable references are
852 dropped between messages.
854 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
855 by the client was not checked as pointing within response data before
856 being used. A malicious client could thus cause an out-of-bounds read and
857 possibly gain authentication. Fix by adding the check.
859 JH/42 Internationalisation: change the default for downconversion in the smtp
860 transport to be "if needed". Previously it was "as previously set" for
861 the message, which usually meant "if needed" for message-submission but
862 "no" for everything else. However, MTAs have been seen using SMTPUTF8
863 even when the envelope addresses did not need it, resulting in forwarding
864 failures to non-supporting MTAs. A downconvert in such cases will be
865 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
866 transport. The change does mean that addresses needing conversion will
867 be converted when previously a delivery failure would occur.
869 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
870 response was received it would be used unchecked in a fail-DSN, violating
871 standards on line-length limits. Truncate if needed.
873 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
874 doesn't work. (Additionally add a new main config option to configure the
881 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
882 side, to decode a TLS 1.3 packet capture.
884 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
885 Previously the default library behaviour applied, sending two, each in
888 JH/03 Debug output for ACL now gives the config file name and line number for
891 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
893 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
895 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
896 buffer overrun for (non-chunking) other transports.
898 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
899 TLS1.3, means that a server rejecting a client certificate is not visible
900 to the client until the first read of encrypted data (typically the
901 response to EHLO). Add detection for that case and treat it as a failed
902 TLS connection attempt, so that the normal retry-in-clear can work (if
903 suitably configured).
905 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
906 and/or domain. Found and fixed by Jason Betts.
908 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
909 configuration). If a CNAME target was not a wellformed name pattern, a
912 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
913 the OS reports them interleaved with other addresses.
915 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
916 used both for input and for a verify callout, both encrypted, SMTP
917 responses being sent by the server could be lost. This resulted in
918 dropped connections and sometimes bounces generated by a peer sending
921 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
922 string. Previously it was possible to cause undefined behaviour in a
923 library routine (usually a crash). Found by "zerons".
925 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
928 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
929 API was removed, so update to use the newer ones.
931 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
932 any timeout set, is taking a long time. Previously we would hang on to a
933 rotated logfile "forever" if the input was arriving with long gaps
934 (a previous attempt to fix addressed lack, for a long time, of initial
937 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
938 shared (NFS) environment. The length of the tempfile name is now
939 4 + 16 ("hdr.$message_exim_id") which might break on file
940 systems which restrict the file name length to lower values.
943 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
944 shared (NFS) environment.
946 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
947 did for all versions <4.90). Notably -M, -m, --invert, -I may be
950 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
951 on some platforms for bit 31.
953 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
954 to changes apparently associated with TLS1.3 handling some of the APIs
955 previously used were either nonfunctional or inappropriate. Strings
956 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
957 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
958 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
959 This affects log line X= elements, the $tls_{in,out}_cipher variables,
960 and the use of specific cipher names in the encrypted= ACL condition.
962 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
964 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
965 verification result was not updated unless hosts_require_ocsp applied.
967 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
968 queue_list_requires_admin set to false, non-admin users were denied the
971 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
972 directory-of-certs mode. Previously they were advertised despite the
975 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
976 A single TCP connection by a client will now hold a TLS connection open
977 for multiple message deliveries, by default. Previously the default was to
980 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
981 default. If built with the facility, DANE will be used. The facility
982 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
984 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
985 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
986 must be defined and you must still, unless you define DISABLE_TLS, manage
987 the the include-dir and library-file requirements that go with that
988 choice. Non-TLS builds are still supported.
990 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
993 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
994 default. If the platform supports and has the facility enabled, it will
995 be requested on all coneections.
997 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
998 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1000 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1002 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1003 success-DSN messages. Previously the From: header was always the default
1004 one for these; the option was ignored.
1006 JH/28 Fix the timeout on smtp response to apply to the whole response.
1007 Previously it was reset for every read, so a teergrubing peer sending
1008 single bytes within the time limit could extend the connection for a
1009 long time. Credit to Qualsys Security Advisory Team for the discovery.
1011 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1012 delivery address, which leaked information of the results of local
1013 forwarding. Change to the original envelope recipient address, per
1016 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1017 requested. Previously not bounce was generated and a log entry of
1018 error ignored was made.
1020 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1022 JH/32 Introduce a general tainting mechanism for values read from the input
1023 channel, and values derived from them. Refuse to expand any tainted
1024 values, to catch one form of exploit.
1026 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1027 was unused and the unexpanded text used for the test. Found and
1028 fixed by Ruben Jenster.
1030 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1031 an attempt to use a TLS library read routine dereffed a nul pointer,
1034 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1035 them in our smtp response.
1037 JH/36 Have the general router option retry_use_local_part default to true when
1038 any of the restrictive preconditions are set (to anything). Previously it
1039 was only for check_local user. The change removes one item of manual
1040 configuration which is required for proper retries when a remote router
1041 handles a subset of addresses for a domain.
1043 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1044 link count into consideration.
1046 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1047 caused the extension of big_buffer, the following lines were ignored.
1049 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1050 accordance with RFC 2308. Previously there was no expiry, so a longlived
1051 receive process (eg. due to ACL delays) versus a short SOA value could
1054 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1056 JH/39 Promote DMARC support to mainline.
1058 JH/40 Bug 2452: Add a References: header to DSNs.
1060 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1061 parameters. The relevant library call is documented as "Deprecated: This
1062 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1063 3.6.0, DH parameters are negotiated following RFC7919."
1065 HS/06 Change the default of dnssec_request_domains to "*"
1067 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1068 carried on and emitted a BDAT command, even when PIPELINING was not
1071 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1072 buffer was used for the filename, resulting in a trap when tainted
1073 arguments (eg. $domain) were used.
1075 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1076 recommended to avoid a possible server-load attack. The feature can be
1077 re-enabled via the openssl_options main cofiguration option.
1079 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1080 for version 4.90 - adding a "more data" boolean to the arguments.
1081 Bumped the ABI version number also, this having been missed previously;
1082 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1083 respect of usage of smtp_printf() by either local_scan code or libraries
1084 accessed via the ${dlfunc } expansion item. Both will need coding
1085 adjustment for any calls to smtp_printf() to match the new function
1086 signature; a FALSE value for the new argument is always safe.
1088 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1089 the file-offset (which the Linux syscall does, and exim expects); this
1090 resulted in an indefinite loop.
1092 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1093 to do ARC verification. The Authentication-Results: header line added
1094 by the configuration then had no ARC item.
1096 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1097 an incorrect mode was used for reading the result, resulting in it being
1104 JH/01 Remove code calling the customisable local_scan function, unless a new
1105 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1107 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1108 non-signal-safe functions being used.
1110 JH/03 Bug 2269: When presented with a received message having a stupidly large
1111 number of DKIM-Signature headers, disable DKIM verification to avoid
1112 a resource-consumption attack. The limit is set at twenty.
1114 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1115 report of oldest_pass in ${authres } in consequence, and separate out
1116 some descriptions of reasons for verification fail.
1118 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1119 files in the spool were present and unlocked. A queue-runner could spot
1120 them, resulting in a duplicate delivery. Fix that by doing the unlock
1121 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1122 add more error-checking on spoolfile handling while that code is being
1125 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1126 No known attacks, no CVE, this is defensive hardening.
1128 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1129 a queue-runner could start a delivery while other operations were ongoing.
1130 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1131 Found and investigated by Tim Stewart. Fix by using the open message data
1132 file handle rather than opening another, and not locally closing it (which
1133 releases a lock) for that case, while creating the temporary .eml format
1134 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1136 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1137 $sender_verify_failure/$recipient_verify_failure to "random".
1139 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1142 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1143 Previously this would segfault.
1145 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1148 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1149 like zero, since the resolver should be doing this for us, But we need one
1150 as a CNAME but no MX presence gets the CNAME returned; we need to check
1151 that doesn't point to an MX to declare it "no MX returned" rather than
1152 "error, loop". A new main option is added so the older capability of
1153 following some limited number of chain links is maintained.
1155 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1157 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1158 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1159 now-deprecated earlier definitions used only specified the range up to TLS
1160 1.2 (in the older-version library docs).
1162 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1164 JH/15 Rework TLS client-side context management. Stop using a global, and
1165 explicitly pass a context around. This enables future use of TLS for
1166 connections to service-daemons (eg. malware scanning) while a client smtp
1167 connection is using TLS; with cutthrough connections this is quite likely.
1169 JH/16 Fix ARC verification to do AS checks in reverse order.
1171 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1173 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1174 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1175 Previously the "utf8" would be re-prepended for every additional message.
1177 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1178 Previously thery were accepted, resulting in issues when attempting to
1179 forward messages to a non-supporting MTA.
1181 PP/02 Let -n work with printing macros too, not just options.
1183 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1184 one parent address was copied, and bogus data was used at delivery-logging
1185 time. Either a crash (after delivery) or bogus log data could result.
1186 Discovery and analysis by Tim Stewart.
1188 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1189 Previously if the string ended mid-character, we did not insert the
1190 promised '?' replacement.
1192 PP/04 Documentation: current string operators work on bytes, not codepoints.
1194 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1195 should pack well giving a smaller memory footprint so better caching and
1196 therefore performance. Group the declarations where this can't be done so
1197 that the byte-sized flag variables are not interspersed among pointer
1198 variables, giving a better chance of good packing by the compiler.
1200 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1201 non-null, to avoid issues with sites running BATV. Previously reports were
1202 sent with an empty envelope sender so looked like bounces.
1204 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1205 The ignore_error flag wasn't being returned from the filter subprocess so
1206 was not set for later routers. Investigation and fix by Matthias Kurz.
1208 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1209 and a msg:complete for the whole, when a message is manually removed using
1210 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1212 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1213 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1214 hope that is portable enough.
1216 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1217 requiring a known-CA anchor certificate; make it now rely entirely on the
1218 TLSA as an anchor. Checking the name on the leaf cert against the name
1219 on the A-record for the host is still done for TA (but not for EE mode).
1221 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1222 included in delivery lines for non-proxied connections, when compiled with
1223 SUPPORT_SOCKS and running with proxy logging enabled.
1225 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1226 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1227 move the existing event to fire before the normal logging of message
1228 failure so that custom logging is bracketed by normal logging.
1230 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1231 msg:fail:internal event. Developement by Matthias Kurz.
1233 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1234 far too small for todays use of crypto signatures stored there. Go all
1235 the way to the max DNS message size of 64kB, even though this might be
1236 overmuch for IOT constrained device use.
1238 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1239 copy a string over itself. The library routine is documented as not
1240 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1242 JH/32 For main options check_spool_space and check_inode_space, where the
1243 platform supports 64b integers, support more than the previous 2^31 kB
1244 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1245 the previous G, M, k.
1247 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1248 $authenticated_fail_id variable on authentication failure. Previously
1251 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1252 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1253 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1254 GNUTLS_SEC_PARAM_MEDIUM.
1256 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1257 side. Previously we would continue as if no SNI had been received.
1259 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1260 "<" character, which should be a list-separator specification, we walked
1261 off past the nul-terimation.
1263 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1264 causes) even when the retry time is not yet met. Previously they were
1265 not, meaning that when (say) an account was over-quota and temp-rejecting,
1266 and multiple senders' messages were queued, only one sender would get
1267 notified on each configured delay_warning cycle.
1269 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1271 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1272 cipher-suites, an error can be left on the stack even for a succeeding
1273 accept; this results in impossible error messages when a later operation
1276 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1277 return error codes indicating retry. Under TLS1.3 this becomes required.
1279 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1280 it only wrote the new authenticators, resulting in a lack of tracking of
1281 peer changes of ESMTP extensions until the next cache flush.
1283 JH/41 Fix the loop reading a message header line to check for integer overflow,
1284 and more-often against header_maxsize. Previously a crafted message could
1285 induce a crash of the recive process; now the message is cleanly rejected.
1287 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1288 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1294 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1295 When redis_servers is set to a list of > 1 element, and the Redis servers
1296 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1297 case of MOVED into a DEFER case instead, thus moving the query onto the
1298 next server in the list. For a cluster of N elements, all N servers must
1299 be defined in redis_servers.
1301 GF/02 Catch and remove uninitialized value warning in exiqsumm
1302 Check for existence of @ARGV before looking at $ARGV[0]
1304 JH/01 Replace the store_release() internal interface with store_newblock(),
1305 which internalises the check required to safely use the old one, plus
1306 the allocate and data copy operations duplicated in both (!) of the
1307 extant use locations.
1309 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1310 modifier. This matches the restriction on the commandline.
1312 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1313 Previously only the last row was returned.
1315 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1316 we assumed that tags in the header were well-formed, and parsed the
1317 element content after inspecting only the first char of the tag.
1318 Assumptions at that stage could crash the receive process on malformed
1321 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1322 While running the DKIM ACL we operate on the Permanent memory pool so that
1323 variables created with "set" persist to the DATA ACL. Also (at any time)
1324 DNS lookups that fail create cache records using the Permanent pool. But
1325 expansions release any allocations made on the current pool - so a dnsdb
1326 lookup expansion done in the DKIM ACL releases the memory used for the
1327 DNS negative-cache, and bad things result. Solution is to switch to the
1328 Main pool for expansions.
1329 While we're in that code, add checks on the DNS cache during store_reset,
1330 active in the testsuite.
1331 Problem spotted, and debugging aided, by Wolfgang Breyha.
1333 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1334 When none of the hosts presented to a transport match an already-open
1335 connection, close it and proceed with the list. Previously we would
1336 queue the message. Spotted by Lena with Yahoo, probably involving
1339 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1340 Previously a spurious "250 OK id=" response was appended to the proper
1343 JH/08 The "support for" informational output now, which built with Content
1344 Scanning support, has a line for the malware scanner interfaces compiled
1345 in. Interface can be individually included or not at build time.
1347 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1348 by the template makefile "src/EDITME". The "STREAM" support for an older
1349 ClamAV interface method is removed.
1351 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1352 rows affected is given instead).
1354 JH/11 The runtime Berkeley DB library version is now additionally output by
1355 "exim -d -bV". Previously only the compile-time version was shown.
1357 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1358 SMTP connection. Previously, when one had more recipients than the
1359 first, an abortive onward connection was made. Move to full support for
1360 multiple onward connections in sequence, handling cutthrough connection
1361 for all multi-message initiating connections.
1363 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1364 routers. Previously, a multi-recipient message would fail to match the
1365 onward-connection opened for the first recipient, and cause its closure.
1367 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1368 a timeout on read on a GnuTLS initiating connection, resulting in the
1369 initiating connection being dropped. This mattered most when the callout
1370 was marked defer_ok. Fix to keep the two timeout-detection methods
1373 JH/15 Relax results from ACL control request to enable cutthrough, in
1374 unsupported situations, from error to silently (except under debug)
1375 ignoring. This covers use with PRDR, frozen messages, queue-only and
1378 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1380 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1381 metadata, resulting in a crash in free().
1383 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1384 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1385 Broken also in d185889f4, with init system revamp.
1387 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1388 Previously we abruptly closed the connection after reading a malware-
1389 found indication; now we go on to read the "scan ok" response line,
1392 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1393 ACL. Previously, a crash would result.
1395 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1396 macro text after a replacement (previously it was only once per line) and
1397 by skipping builtin macros when searching for an uppercase lead character.
1399 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1400 for the build is renamed.
1402 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1403 was allocated for every new TLS startup, meaning one per message. Fix
1404 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1406 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1407 reported the original. Fix to report (as far as possible) the ACL
1408 result replacing the original.
1410 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1411 OpenSSL. Certificate information is loaded for every new TLS startup,
1412 and the resources needed to be freed.
1414 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1416 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1419 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1420 DATA response info to the (existing) per-recipient response info for
1421 the "C=" log element. It can have useful tracking info from the
1422 destination system. Patch from Simon Arlott.
1424 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1425 length value. Previously this would segfault.
1427 HS/02 Support Avast multiline protoocol, this allows passing flags to
1428 newer versions of the scanner.
1430 JH/28 Ensure that variables possibly set during message acceptance are marked
1431 dead before release of memory in the daemon loop. This stops complaints
1432 about them when the debug_store option is enabled. Discovered specifically
1433 for sender_rate_period, but applies to a whole set of variables.
1434 Do the same for the queue-runner and queue-list loops, for variables set
1435 from spool message files. Do the same for the SMTP per-message loop, for
1436 certain variables indirectly set in ACL operations.
1438 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1439 as a multi-recipient message from a mailinglist manager). The coding had
1440 an arbitrary cutoff number of characters while checking for more input;
1441 enforced by writing a NUL into the buffer. This corrupted long / fast
1442 input. The problem was exposed more widely when more pipelineing of SMTP
1443 responses was introduced, and one Exim system was feeding another.
1444 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1445 receiving system, and refused recipients seen by the sending system
1446 (propating to people being dropped from mailing lists).
1447 Discovered and pinpointed by David Carter.
1449 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1450 replaced by the ${authresults } expansion.
1452 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1454 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1455 allows proper process termination in container environments.
1457 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1458 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1460 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1461 and "err_perm", deprecated since 4.83 when the RFC-defined words
1462 "temperror" and "permerror" were introduced.
1464 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1465 transport-filters or DKIM-signing. The restriction was lost in the
1466 consolidation of verify-callout and delivery SMTP handling.
1467 Extend the restriction to also cover ARC-signing.
1469 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1470 in defer=pass mode supply a 450 to the initiator. Previously the message
1473 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1474 tls_require_ciphers is used as before.
1476 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1477 "pass_unscanned". Only tmpfails from the scanner are written to
1478 the paniclog, as they may require admin intervention (permission
1479 denied, license issues). Other scanner errors (like decompression
1480 bombs) do not cause a paniclog entry.
1482 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1483 Previously it was possible to log spurious information in receive log
1486 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1487 triggered odd behaviour from Outlook Express clients.
1489 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1492 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1493 since the IETF WG has not yet settled on that versus the original
1494 "bare" representation.
1496 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1497 Previously the millisecond value corrupted the output.
1498 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1499 corrupted the output.
1505 JH/01 Rework error string handling in TLS interface so that the caller in
1506 more cases is responsible for logging. This permits library-sourced
1507 string to be attached to addresses during delivery, and collapses
1508 pairs of long lines into single ones.
1510 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1511 during configuration. Wildcards are allowed and expanded.
1513 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1514 This permits better logging.
1516 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1517 do not close it down and have the child transport start it up again on
1518 the passed-on TCP connection. Instead, proxy the child (and any
1519 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1520 affected: the continued delivery log lines do not have any DNSSEC, TLS
1521 Certificate or OCSP information. TLS cipher information is still logged.
1523 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1524 identical IP addresses on different listening ports. Will also affect
1527 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1528 add noisy ifdef guards to special-case this sillyness.
1529 Patch from Bernd Kuhls.
1531 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1532 than 255 are no longer allowed.
1534 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1535 Disable the session-cache too, which might reduce our load. Since we
1536 currrectly use a new context for every connection, both as server and
1537 client, there is no benefit for these.
1538 GnuTLS appears to not support tickets server-side by default (we don't
1539 call gnutls_session_ticket_enable_server()) but client side is enabled
1540 by default on recent versions (3.1.3 +) unless the PFS priority string
1543 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1544 <https://reproducible-builds.org/specs/source-date-epoch/>.
1546 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1547 the check for any unsuccessful recipients did not notice the limit, and
1548 erroneously found still-pending ones.
1550 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1551 MSG_MORE. Only in-clear (not on TLS connections).
1553 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1554 transport-filter is involved we can buffer the headers in memory for
1555 creating the signature, and read the spool data file once for the
1556 signature and again for transmission.
1558 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1559 4.77 as the kernel support then wasn't solid, having issues in 64bit
1560 mode. Now, it's been long enough. Add support for FreeBSD also.
1562 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1563 case where the routing stage had gathered several addresses to send to
1564 a host before calling the transport for the first, we previously failed
1565 to close down TLS in the old transport process before passing the TCP
1566 connection to the new process. The new one sent a STARTTLS command
1567 which naturally failed, giving a failed delivery and bloating the retry
1568 database. Investigation and fix prototype from Wolfgang Breyha.
1570 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1571 false-negatives in the check that the sender had not preempted a response
1572 or prompt from Exim (running as a server), due to that code's lack of
1573 awareness of the SMTP input buffering.
1575 PP/04 Add commandline_checks_require_admin option.
1576 Exim drops privileges sanely, various checks such as -be aren't a
1577 security problem, as long as you trust local users with access to their
1578 own account. When invoked by services which pass untrusted data to
1579 Exim, this might be an issue. Set this option in main configuration
1580 AND make fixes to the calling application, such as using `--` to stop
1583 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1584 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1585 responses to those, into a single TLS record each way (this usually means
1586 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1589 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1590 affects you only if you're dancing at the edge of the param size limits.
1591 If you are, and this message makes sense to you, then: raise the
1592 configured limit or use OpenSSL 1.1. Nothing we can do for older
1595 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1596 cmdline element to get the documented default one. Previously it was
1599 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1600 are unsuable for non-message-specific reasons - eg. connection timeout,
1601 banner-time rejection.
1603 JH/16 Fix logging of delivery remote port, when specified by router, under
1606 PP/06 Repair manualroute's ability to take options in any order, even if one
1607 is the name of a transport.
1610 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1612 JH/17 Change the list-building routines interface to use the expanding-string
1613 triplet model, for better allocation and copying behaviour.
1615 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1616 Previously it was constructed the first time a possibly-matching string
1617 was met in the configuration file input during startup; now it is done
1620 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1621 compatible one, to avoid the (poorly documented) possibility of a config
1622 file in the working directory redirecting the DB files, possibly correpting
1623 some existing file. CVE-2017-10140 assigned for BDB.
1625 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1626 cache-hot. Previously, although the result was properly cached, the
1627 initial verify call returned a defer.
1629 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1630 the main verify for receipient in uncached-mode.
1632 JH/22 Retire historical build files to an "unsupported" subdir. These are
1633 defined as "ones for which we have no current evidence of testing".
1635 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1636 if present. Previously it was ignored.
1638 JH/24 Start using specified-initialisers in C structure init coding. This is
1639 a C99 feature (it's 2017, so now considered safe).
1641 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1642 if was a fixed-sized field and bitmask ops via macros; it is now more
1645 PP/07 GitHub PR 56: Apply MariaDB build fix.
1646 Patch provided by Jaroslav Škarvada.
1648 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1649 during Coverity cleanups [4.87 JH/47]
1650 Diagnosis and fix provided by Michael Fischer v. Mollard.
1652 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1653 the right size to place the terminating semicolon on its own folded
1654 line, the header hash was calculated to an incorrect value thanks to
1655 the (relaxed) space the fold became.
1657 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1658 and confused the parent.
1660 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1661 which could crash as a result. This could lead to undeliverable messages.
1663 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1666 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1667 matters on fast-turnover and PID-randomising systems, which were getting
1668 out-of-order delivery.
1670 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1671 a possibly-overlapping copy. The symptom was that "Remote host closed
1672 connection in response to HELO" was logged instead of the actual 4xx
1675 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1676 Previously only that bufferd was discarded, resulting in SYMTP command
1679 JH/32 DKIM: when a message has multiple signatures matching an identity given
1680 in dkim_verify_signers, run the dkim acl once for each. Previously only
1681 one run was done. Bug 2189.
1683 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1684 "panic the current process" to "deliberately defer". The panic log is
1685 still written with the problem list name; the mail and reject logs now
1686 get a temp-reject line for the message that was being handled, saying
1687 something like "domains check lookup or other defer". The SMTP 451
1688 message is still "Temporary local problem".
1690 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1691 A crafted sequence of BDAT commands could result in in-use memory beeing
1692 freed. CVE-2017-16943.
1694 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1695 from SMTP input. Previously it was always done; now only done for DATA
1696 and not BDAT commands. CVE-2017-16944.
1698 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1699 to the message (such as an overlong header line). Previously this was
1700 not done and we did not exit BDAT mode. Followon from the previous item
1701 though a different problem.
1707 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1708 than -2003 did; needs libidn2 in addition to libidn.
1710 JH/02 The path option on a pipe transport is now expanded before use.
1712 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1713 Patch provided by "Björn", documentation fix added too.
1715 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1716 missing a wire-to-host endian conversion.
1718 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1719 close after a BDAT command line could be taken as a following command,
1720 giving a synch failure. Fix by only checking for synch immediately
1721 before acknowledging the chunk.
1723 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1724 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1725 macro. Patches provided by Josh Soref.
1727 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1728 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1729 by IANA as a service extension. However, John Klensin suggests that we
1732 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1733 the dkim code may be unix-mode line endings rather than smtp wire-format
1734 CRLF, so prepend a CR to any bare LF.
1736 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1737 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1739 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1740 the first were themselves being wrongly included in the feed into dkim
1741 processing; with most chunk sizes in use this resulted in an incorrect
1742 body hash calculated value.
1744 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1745 DKIM signature block, for verification. Although advised against by
1746 standards it is specifically not ruled illegal.
1748 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1750 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1751 missing a body hash (the bh= tag).
1753 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1754 It seems that HAProxy sends the Proxy Protocol information in clear and
1755 only then does a TLS startup, so do the same.
1757 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1758 TCP connections (such as for Spamd) unless the daemon successfully set
1759 Fast Open mode on its listening sockets. This fixes breakage seen on
1760 too-old kernels or those not configured for Fast Open, at the cost of
1761 requiring both directions being enabled for TFO, and TFO never being used
1762 by non-daemon-related Exim processes.
1764 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1765 endings, at least on the first header line. Try to canonify any that get
1766 past that check, despite the cost.
1768 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1769 now limited to an arbitrary five deep, while parsing addresses with the
1770 strip_excess_angle_brackets option enabled.
1772 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1773 instead leave the unprompted TLS handshake in socket buffer for the
1774 TLS library to consume.
1776 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1778 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1780 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1781 region in one operation, for speed, and this leaves assigned pointers
1782 dangling. Add checks run only under the testsuite which checks all
1783 variables at a store-reset and panics on a dangling pointer; add code
1784 explicitly nulling out all the variables discovered. Fixes one known
1785 bug: a transport crash, where a dangling pointer for $sending_ip_address
1786 originally assigned in a verify callout, is re-used.
1788 PP/06 Drop '.' from @INC in various Perl scripts.
1790 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1792 PP/08 Reduce a number of compilation warnings under clang; building with
1793 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1794 should be warning-free.
1796 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1798 HS/01 Fix portability problems introduced by PP/08 for platforms where
1799 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1801 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1802 chunk. This allows us to accept broken chunked messages. We need a more
1803 general solution here.
1805 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1806 already-broken messages in the queue.
1808 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1810 JH/19 Fix reference counting bug in routing-generated-address tracking.
1816 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1817 supports it and a size is available (ie. the sending peer gave us one).
1819 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1820 years of being deprecated). The replacements are the ACLs
1821 acl_smtp_mime and acl_not_smtp_mime.
1823 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1824 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1825 or even an in-clear connection were permitted. Now, if the host lookup
1826 was dnssec and dane was requested then the host is only used if the
1827 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1828 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1829 if one fails this test.
1830 This means that a poorly-configured remote DNS will make it incommunicado;
1831 but it protects against a DNS-interception attack on it.
1833 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1834 non-noisy when a race steals the message being considered.
1836 JH/05 If main configuration option tls_certificate is unset, generate a
1837 self-signed certificate for inbound TLS connections.
1839 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1840 in rewrites and routers.
1842 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1843 and logged a warning sing 4.83; now they are a configuration file error.
1845 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1846 (lacking @domain). Apply the same qualification processing as RCPT.
1848 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1850 JH/10 Support ${sha256:} applied to a string (as well as the previous
1853 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1854 a cutthrough deliver is pending, as we always want to make a connection.
1855 This also avoids re-routing the message when later placing the cutthrough
1856 connection after a verify cache hit.
1857 Do not update it with the verify result either.
1859 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1860 when routing results in more than one destination address.
1862 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1863 signing (which inhibits the cutthrough capability). Previously only
1864 the presence of an option was tested; now an expansion evaluating as
1865 empty is permissible (obviously it should depend only on data available
1866 when the cutthrough connection is made).
1868 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1869 the relevant preceding SMTP command did not note the pipelining mode.
1871 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1872 Previously they were not counted.
1874 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1875 as one having no matching records. Previously we deferred the message
1876 that needed the lookup.
1878 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1879 distinguished as "(=".
1881 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1882 for missing MX records. Previously it only worked for missing A records.
1884 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1886 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1887 after the data-go-ahead and data-ack. Patch from Jason Betts.
1889 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1890 even for a "none" policy. Patch from Tony Meyer.
1892 JH/22 Fix continued use of a connection for further deliveries. If a port was
1893 specified by a router, it must also match for the delivery to be
1896 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1897 When one of the recipients of a message was unsuitable for the connection
1898 (has no matching addresses), we lost track of needing to mark it
1899 deferred. As a result mail would be lost.
1901 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1903 JH/25 Decoding ACL controls is now done using a binary search; the source code
1904 takes up less space and should be simpler to maintain. Merge the ACL
1905 condition decode tables also, with similar effect.
1907 JH/26 Fix problem with one_time used on a redirect router which returned the
1908 parent address unchanged. A retry would see the parent address marked as
1909 delivered, so not attempt the (identical) child. As a result mail would
1912 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1913 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1914 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1917 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1918 Main config options check_{log,spool}_{inodes,space} are now
1919 100 inodes, 10MB unless set otherwise in the configuration.
1921 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1922 Previously it only applied to the main-section connection policy
1925 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1927 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1928 by me. Added RFC7919 DH primes as an alternative.
1930 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1931 are not in the system include path.
1933 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1934 GnuTLS, when a session startup failed (eg because the client disconnected)
1935 Exim did stdio operations after fclose. This was exposed by a recent
1936 change which nulled out the file handle after the fclose.
1938 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1939 signed directly by the cert-signing cert, rather than an intermediate
1940 OCSP-signing cert. This is the model used by LetsEncrypt.
1942 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1944 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1945 an incoming connection.
1947 HS/02 Bug 1802: Do not half-close the connection after sending a request
1950 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1951 fallback to "prime256v1".
1953 JH/34 SECURITY: Use proper copy of DATA command in error message.
1954 Could leak key material. Remotely exploitable. CVE-2016-9963.
1960 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1961 and 3.4.4 - once the server is enabled to respond to an OCSP request
1962 it does even when not requested, resulting in a stapling non-aware
1963 client dropping the TLS connection.
1965 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1966 support variable-length bit vectors. No functional change.
1968 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1969 The I= interface field on outgoing lines is now after the H= remote
1970 host field, same as incoming lines. There is a separate
1971 outgoing_interface log selector which allows you to disable the
1974 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1975 If not running log_selector +smtp_connection the mainlog would be held
1976 open indefinitely after a "too many connections" event, including to a
1977 deleted file after a log rotate. Leave the per net connection logging
1978 leaving it open for efficiency as that will be quickly detected by the
1979 check on the next write.
1981 HS/01 Bug 1671: Fix post transport crash.
1982 Processing the wait-<transport> messages could crash the delivery
1983 process if the message IDs didn't exist for some reason. When
1984 using 'split_spool_directory=yes' the construction of the spool
1985 file name failed already, exposing the same netto behaviour.
1987 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1988 mime_regex ACL conditions.
1990 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1991 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1992 to HELO, local diagnostic string.
1994 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1995 log line to debug. Even when configured with a tls authenticator many
1996 client connections are expected to not authenticate in this way, so
1997 an authenticate fail is not an error.
1999 HS/02 Add the Exim version string to the process info. This way exiwhat
2000 gives some more detail about the running daemon.
2002 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2003 matter for fast-change records such as DNSBLs.
2005 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2006 retry record, even if constant. There may be multiple transports with
2007 different interface settings and the retry behaviour needs to be kept
2010 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2012 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2014 JH/10 Bug 840: fix log_defer_output option of pipe transport
2016 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2017 hosts_randomize. This matters a lot when combined with mua_wrapper.
2019 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2020 ${quote_pgsql:<string>} operator.
2022 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2024 JH/14 Change header file name for freeradius-client. Relevant if compiling
2025 with Radius support; from the Gentoo tree and checked under Fedora.
2027 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2029 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2030 option transport_filter, meaning no filtering.
2032 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2034 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2035 defaults to "*" (all hosts). The variable is now available when not built
2036 with TLS, default unset, mainly to enable keeping the testsuite sane.
2037 If a server certificate is not supplied (via tls_certificate) an error is
2038 logged, and clients will find TLS connections fail on startup. Presumably
2039 they will retry in-clear.
2040 Packagers of Exim are strongly encouraged to create a server certificate
2041 at installation time.
2043 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2044 with the $config_file variable.
2046 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2047 in transport context, after the attempt, and per-recipient. The latter type
2048 is per host attempted. The event data is the error message, and the errno
2049 information encodes the lookup type (A vs. MX) used for the (first) host,
2050 and the trailing two digits of the smtp 4xx response.
2052 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2053 to write to mainlog (or rejectlog, paniclog) in the window between file
2054 creation and permissions/ownership being changed. Particularly affects
2055 installations where exicyclog is run as root, rather than exim user;
2056 result is that the running daemon panics and dies.
2058 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2060 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2061 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2062 "pri" and "weight". Note that the previous implicit priority given by the
2063 list order is no longer honoured.
2065 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2066 for DKIM processing.
2068 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2069 by defining SUPPORT_SOCKS.
2071 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2072 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2073 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2074 variables are renamed to proxy_{local,external}_{address,port}.
2076 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2077 for a build by defining SUPPORT_I18N
2079 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2080 of the query string, and make ${quote_redis:} do that quoting.
2082 JH/29 Move Events support from Experimental to mainline, enabled by default
2083 and removable for a build by defining DISABLE_EVENT.
2085 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2087 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2088 cached by the daemon.
2090 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2091 by defining LOOKUP_REDIS. The libhiredis library is required.
2093 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2094 keys are given for lookup.
2096 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2097 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2098 only supported when built with TLS support. The PolarSSL SHA routines
2099 are still used when the TLS library is too old for convenient support.
2101 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2102 openssl_options), for security. OpenSSL forces this from version 1.1.0
2103 server-side so match that on older versions.
2105 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2106 allocation for $value could be released as the expansion processing
2107 concluded, but leaving the global pointer active for it.
2109 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2110 and to use the domains and local_parts ACL conditions.
2112 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2113 incorrectly not doubled on cutthrough transmission, hence seen as a
2114 body-termination at the receiving system - resulting in truncated mails.
2115 Commonly the sender saw a TCP-level error, and retransmitted the message
2116 via the normal store-and-forward channel. This could result in duplicates
2117 received - but deduplicating mailstores were liable to retain only the
2118 initial truncated version.
2120 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2122 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2124 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2125 we're in there, support oversigning also; bug 1309.
2127 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2129 HS/04 Add support for keep_environment and add_environment options.
2131 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2132 either intentional arithmetic overflow during PRNG, or testing config-
2135 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2136 delivery resulted in actual delivery. Cancel cutthrough before DATA
2139 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2140 rejects a recipient: pass the reject to the originator.
2142 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2143 Many were false-positives and ignorable, but it's worth fixing the
2146 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2147 for the new environment-manipulation done at startup. Move the routines
2148 from being local to tls.c to being global via the os.c file.
2150 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2151 an extract embedded as result-arg for a map, the first arg for extract
2152 is unavailable so we cannot tell if this is a numbered or keyed
2153 extraction. Accept either.
2159 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2162 JH/02 The smtp transport option "multi_domain" is now expanded.
2164 JH/03 The smtp transport now requests PRDR by default, if the server offers
2167 JH/04 Certificate name checking on server certificates, when exim is a client,
2168 is now done by default. The transport option tls_verify_cert_hostnames
2169 can be used to disable this per-host. The build option
2170 EXPERIMENTAL_CERTNAMES is withdrawn.
2172 JH/05 The value of the tls_verify_certificates smtp transport and main options
2173 default to the word "system" to access the system default CA bundle.
2174 For GnuTLS, only version 3.0.20 or later.
2176 JH/06 Verification of the server certificate for a TLS connection is now tried
2177 (but not required) by default. The verification status is now logged by
2178 default, for both outbound TLS and client-certificate supplying inbound
2181 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2184 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2185 Status Notification (bounce) messages are now MIME format per RFC 3464.
2186 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2187 under the control of the dsn_advertise_hosts option, and routers may
2188 have a dsn_lasthop option.
2190 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2191 default, modifiable by a malware= option. The list separator for
2192 the options can now be changed in the usual way. Bug 68.
2194 JH/10 The smtp_receive_timeout main option is now expanded before use.
2196 JH/11 The incoming_interface log option now also enables logging of the
2197 local interface on delivery outgoing connections.
2199 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2200 if the interface and destination host and port all match.
2202 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2205 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2206 Patch from Andrew Lewis.
2208 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2209 now supports optional time-restrictions, weighting, and priority
2210 modifiers per server. Patch originally by <rommer@active.by>.
2212 JH/16 The spamd_address main option now supports a mixed list of local
2213 and remote servers. Remote servers can be IPv6 addresses, and
2214 specify a port-range.
2216 JH/17 Bug 68: The spamd_address main option now supports an optional
2217 timeout value per server.
2219 JH/18 Bug 1581: Router and transport options headers_add/remove can
2220 now have the list separator specified.
2222 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2225 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2228 JH/21 Support for the A6 type of dns record is withdrawn.
2230 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2231 rather than the verbs used.
2233 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2234 from 255 to 1024 chars.
2236 JH/24 Verification callouts now attempt to use TLS by default.
2238 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2239 are generic router options now. The defaults didn't change.
2241 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2242 Original patch from Alexander Shikoff, worked over by JH.
2244 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2245 files if ZCAT_COMMAND is not executable.
2247 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2249 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2251 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2252 Normally benign, it bites when the pair was led to by a CNAME;
2253 modern usage is to not canonicalize the domain to a CNAME target
2254 (and we were inconsistent anyway for A-only vs AAAA+A).
2256 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2258 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2259 when evaluating $sender_host_dnssec.
2261 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2262 $sender_helo_dnssec variable.
2264 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2266 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2268 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2270 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2271 documented as working, but never had. Support all but $spam_report.
2273 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2274 added for tls authenticator.
2276 HS/03 Add perl_taintmode main config option
2282 TL/01 When running the test suite, the README says that variables such as
2283 no_msglog_check are global and can be placed anywhere in a specific
2284 test's script, however it was observed that placement needed to be near
2285 the beginning for it to behave that way. Changed the runtest perl
2286 script to read through the entire script once to detect and set these
2287 variables, reset to the beginning of the script, and then run through
2288 the script parsing/test process like normal.
2290 TL/02 The BSD's have an arc4random API. One of the functions to induce
2291 adding randomness was arc4random_stir(), but it has been removed in
2292 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2293 function when detected.
2295 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2296 cause callback expansion.
2298 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2299 syntax errors in an expansion can be treated as a string instead of
2300 logging or causing an error, due to the internal use of bool_lax
2301 instead of bool when processing it.
2303 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2304 server certificates when making smtp deliveries.
2306 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2308 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2310 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2312 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2313 Merged patch from Sebastian Wiedenroth.
2315 JH/05 Fix results-pipe from transport process. Several recipients, combined
2316 with certificate use, exposed issues where response data items split
2317 over buffer boundaries were not parsed properly. This eventually
2318 resulted in duplicates being sent. This issue only became common enough
2319 to notice due to the introduction of connection certificate information,
2320 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2322 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2323 size buffer was used, resulting in syntax errors when an expansion
2326 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2327 version 3.3.6 or later.
2329 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2330 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2331 both become "event_action", the variables become $event_name, $event_data
2332 and $event_defer_errno. There is a new variable $verify_mode, usable in
2333 routers, transports and related events. The tls:cert event is now also
2334 raised for inbound connections, if the main configuration event_action
2337 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2338 early OCSP support, but no stapling (appears to be less than 1.0.0).
2340 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2341 server certificate names available under the smtp transport option
2342 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2345 JH/10 Time-related extraction expansions from certificates now use the main
2346 option "timezone" setting for output formatting, and are consistent
2347 between OpenSSL and GnuTLS compilations. Bug 1541.
2349 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2350 encoded parameter in the incoming message. Bug 1558.
2352 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2353 include certificate info, eximon was claiming there were spoolfile
2356 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2358 JH/14 Log delivery-related information more consistently, using the sequence
2359 "H=<name> [<ip>]" wherever possible.
2361 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2362 are problematic for Debian distribution, omit them from the release
2365 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2367 JH/16 Fix string representation of time values on 64bit time_t architectures.
2370 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2371 output list separator was used.
2376 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2377 checkers that were complaining about end of non-void function with no
2380 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2381 This was a regression introduced in 4.83 by another bugfix.
2383 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2385 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2386 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2392 TF/01 Correctly close the server side of TLS when forking for delivery.
2394 When a message was received over SMTP with TLS, Exim failed to clear up
2395 the incoming connection properly after forking off the child process to
2396 deliver the message. In some situations the subsequent outgoing
2397 delivery connection happened to have the same fd number as the incoming
2398 connection previously had. Exim would try to use TLS and fail, logging
2399 a "Bad file descriptor" error.
2401 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2402 utilities have not been installed.
2404 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2405 temporary space as the ACL may create new global variables.
2407 TL/01 LDAP support uses per connection or global context settings, depending
2408 upon the detected version of the libraries at build time.
2410 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2411 to extract and use the src ip:port in logging and expansions as if it
2412 were a direct connection from the outside internet. PPv2 support was
2413 updated based on HAProxy spec change in May 2014.
2415 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2417 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2418 Properly escape header and check for NULL return.
2420 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2423 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2425 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2426 characters in header names, implemented as a verify condition.
2427 Contributed by Michael Fischer v. Mollard.
2429 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2430 results permerror and temperror. Previous values are deprecated but
2431 still accepted. In a future release, err_perm and err_temp will be
2432 completely removed, which will be a backward incompatibility if the
2433 ACL tests for either of these two old results. Patch contributed by
2434 user bes-internal on the mailing list.
2436 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2438 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2439 selectors, in both main and reject logs.
2441 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2444 JH/07 Add malware type "sock" for talking to simple daemon.
2446 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2448 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2449 routers/transports under cutthrough routing.
2451 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2452 numbers. Touch up "bool" conditional to keep the same definition.
2454 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2456 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2458 JH/12 Expand items in router/transport headers_add or headers_remove lists
2459 individually rather than the list as a whole. Bug 1452.
2461 Required for reasonable handling of multiple headers_ options when
2462 they may be empty; requires that headers_remove items with embedded
2463 colons must have them doubled (or the list-separator changed).
2465 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2466 view the policy declared in the DMARC record. Currently, $dmarc_status
2467 is a combined value of both the record presence and the result of the
2470 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2472 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2473 dnslookup router and the smtp transport (applying to the forward
2476 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2477 of ldap servers used for a specific lookup. Patch provided by Heiko
2480 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2481 New variable $lookup_dnssec_authenticated for observability.
2483 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2484 Patch submitted by Lars Timman.
2486 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2488 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2489 Requires trusted mode and valid format message id, aborts otherwise.
2490 Patch contributed by Heiko Schlichting.
2492 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2493 certextract with support for various fields. Bug 1358.
2495 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2496 is requested by default, modifiable by smtp transport option
2499 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2500 operate on certificate variables to give certificate fingerprints
2501 Also new ${sha256:cert_variable}.
2503 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2505 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2506 Christian Aistleitner.
2508 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2510 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2511 file. Patch from Wolfgang Breyha.
2513 JH/25 Expand the coverage of the delivery $host and $host_address to
2514 client authenticators run in verify callout. Bug 1476.
2516 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2517 align with daemon_smtp_ports. Bug 72.
2519 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2520 support and error reporting did not work properly.
2522 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2523 and is readable. Patch from Andrew Colin Kissa.
2525 TL/14 Enhance documentation of ${run expansion and how it parses the
2526 commandline after expansion, particularly in the case when an
2527 unquoted variable expansion results in an empty value.
2529 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2531 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2532 by renaming away from it. Addresses GH issue 15, reported by
2535 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2537 TL/15 SECURITY: prevent double expansion in math comparison functions
2538 (can expand unsanitized data). Not remotely exploitable.
2545 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2547 PP/02 Make -n do something, by making it not do something.
2548 When combined with -bP, the name of an option is not output.
2550 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2553 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2554 $sender_host_name and config options to manage this, and basic check
2557 PP/05 DSCP support for outbound connections and control modifier for inbound.
2559 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2560 (Only plugin which currently uses this is kerberos4, which nobody should
2561 be using, but we should make it available and other future plugins might
2562 conceivably use it, even though it would break NAT; stuff *should* be
2563 using channel bindings instead).
2565 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2566 name; added for Sendmail compatibility; requires admin caller.
2567 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2568 just ignore it); requires trusted caller.
2569 Also parse but ignore: -Ac -Am -X<logfile>
2572 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2574 TL/02 Add +smtp_confirmation as a default logging option.
2576 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2577 Patch by Magnus Holmgren from 2007-02-20.
2579 TL/04 Bugzilla 1281 - Spec typo.
2580 Bugzilla 1283 - Spec typo.
2581 Bugzilla 1290 - Spec grammar fixes.
2583 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2585 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2587 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2588 mailing list by by Dmitry Isaikin.
2590 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2592 JH/02 Support "G" suffix to numbers in ${if comparisons.
2594 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2596 NM/01 Bugzilla 1197 - Spec typo
2597 Bugzilla 1196 - Spec examples corrections
2599 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2601 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2602 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2605 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2606 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2608 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2609 "acl {{name}{arg}...}", and optional args on acl condition
2612 JH/05 Permit multiple router/transport headers_add/remove lines.
2614 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2616 JH/07 Avoid using a waiting database for a single-message-only transport.
2617 Performance patch from Paul Fisher. Bugzilla 1262.
2619 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2622 JH/09 Add $headers_added variable, with content from use of ACL modifier
2623 add_header (but not yet added to the message). Bugzilla 199.
2625 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2626 Pulled from Bugzilla 817 by Wolfgang Breyha.
2628 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2630 (nb: this is the same fix as in Exim 4.80.1)
2632 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2635 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2638 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2640 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2641 advertise SMTP AUTH mechanism to us, instead of a generic
2642 protocol violation error. Also, make Exim more robust to bad
2643 data from the Dovecot auth socket.
2645 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2647 When a queue runner is handling a message, Exim first routes the
2648 recipient addresses, during which it prunes them based on the retry
2649 hints database. After that it attempts to deliver the message to
2650 any remaining recipients. It then updates the hints database using
2653 So if a recipient address works intermittently, it can get repeatedly
2654 deferred at routing time. The retry hints record remains fresh so the
2655 address never reaches the final cutoff time.
2657 This is a fairly common occurrence when a user is bumping up against
2658 their storage quota. Exim had some logic in its local delivery code
2659 to deal with this. However it did not apply to per-recipient defers
2660 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2662 This change adds a proper retry rule check during routing so that the
2663 final cutoff time is checked against the message's age. We only do
2664 this check if there is an address retry record and there is not a
2665 domain retry record; this implies that previous attempts to handle
2666 the address had the retry_use_local_parts option turned on. We use
2667 this as an approximation for the destination being like a local
2668 delivery, as in LMTP.
2670 I suspect this new check makes the old local delivery cutoff check
2671 redundant, but I have not verified this so I left the code in place.
2673 TF/02 Correct gecos expansion when From: is a prefix of the username.
2675 Test 0254 submits a message to Exim with the header
2679 When I ran the test suite under the user fanf2, Exim expanded
2680 the header to contain my full name, whereas it should have added
2681 a Resent-Sender: header. It erroneously treats any prefix of the
2682 username as equal to the username.
2684 This change corrects that bug.
2686 GF/01 DCC debug and logging tidyup
2687 Error conditions log to paniclog rather than rejectlog.
2688 Debug lines prefixed by "DCC: " to remove any ambiguity.
2690 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2692 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2693 Bug spotted by Jeremy Harris; was flawed since initial commit.
2694 Would have resulted in OCSP responses post-SNI triggering an Exim
2695 NULL dereference and crash.
2697 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2699 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2700 Bug detection, analysis and fix by Samuel Thibault.
2701 Bugzilla 1331, Debian bug #698092.
2703 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2705 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2706 Server implementation by Todd Lyons, client by JH.
2707 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2708 config variable "prdr_enable" controls whether the server
2709 advertises the facility. If the client requests PRDR a new
2710 acl_data_smtp_prdr ACL is called once for each recipient, after
2711 the body content is received and before the acl_smtp_data ACL.
2712 The client is controlled by both of: a hosts_try_prdr option
2713 on the smtp transport, and the server advertisement.
2714 Default client logging of deliveries and rejections involving
2715 PRDR are flagged with the string "PRDR".
2717 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2718 fclose(). Diagnosis by Todd Lyons.
2720 PP/17 Update configure.default to handle IPv6 localhost better.
2721 Patch by Alain Williams (plus minor tweaks).
2724 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2725 This is now consistent with GnuTLS, and is now documented: the
2726 previous undocumented portable approach to treating the option as
2727 unset was to force an expansion failure. That still works, and
2728 an empty string is now equivalent.
2730 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2731 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2732 not performing validation itself.
2734 PP/20 Added force_command boolean option to pipe transport.
2735 Patch from Nick Koston, of cPanel Inc.
2737 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2740 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2742 PP/21 Fix eximon continuous updating with timestamped log-files.
2743 Broken in a format-string cleanup in 4.80, missed when I repaired the
2744 other false fix of the same issue.
2745 Report and fix from Heiko Schlichting.
2748 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2749 Report from Prashanth Katuri.
2751 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2752 It's SecureTransport, so affects any MacOS clients which use the
2753 system-integrated TLS libraries, including email clients.
2755 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2756 using a MIME ACL for non-SMTP local injection.
2757 Report and assistance in diagnosis by Warren Baker.
2759 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2761 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2763 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2764 last id that failed so it may be referenced in subsequent ACL's.
2766 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2769 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2770 ldap library initialization, allowing self-signed CA's to be
2771 used. Also properly sets require_cert option later in code by
2772 using NULL (global ldap config) instead of ldap handle (per
2773 session). Bug diagnosis and testing by alxgomz.
2775 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2776 the src/util/ subdirectory.
2778 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2779 renamed to Transport Post Delivery Action by Jeremy Harris, as
2782 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2783 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2784 redis_servers = needs to be configured which will be used by the redis
2785 lookup. Patch from Warren Baker, of The Packet Hub.
2787 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2789 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2790 hostname or reverse DNS when processing a host list. Used suggestions
2791 from multiple comments on this bug.
2793 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2795 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2796 Missed a few lines, added it to make the runtest require no keyboard
2799 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2800 contains upper case chars. Make router use caseful_local_part.
2802 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2803 support when GnuTLS has been built with p11-kit.
2809 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2811 This, or similar/improved, will also be change PP/11 of 4.82.
2817 PP/01 Handle short writes when writing local log-files.
2818 In practice, only affects FreeBSD (8 onwards).
2819 Bugzilla 1053, with thanks to Dmitry Isaikin.
2821 NM/01 Bugzilla 949 - Documentation tweak
2823 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2826 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2828 PP/02 Implemented gsasl authenticator.
2830 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2832 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2833 `pkg-config foo` for cflags/libs.
2835 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2836 with rest of GSASL and with heimdal_gssapi.
2838 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2839 `pkg-config foo` for cflags/libs for the TLS implementation.
2841 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2842 properties get this fed in as external SSF. A number of robustness
2843 and debugging improvements to the cyrus_sasl authenticator.
2845 PP/08 cyrus_sasl server now expands the server_realm option.
2847 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2848 Patch by Jeremy Harris.
2850 PP/10 Added dbmjz lookup type.
2852 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2854 PP/12 MAIL args handles TAB as well as SP, for better interop with
2855 non-compliant senders.
2856 Analysis and variant patch by Todd Lyons.
2858 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2859 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2860 Patch from Dirk Mueller <dmueller@suse.com>
2862 PP/13 tls_peerdn now print-escaped for spool files.
2863 Observed some $tls_peerdn in wild which contained \n, which resulted
2864 in spool file corruption.
2866 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2867 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2868 or write after TLS renegotiation, which otherwise led to messages
2871 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2872 as a tracking header (ie: a signed header comes before the signature).
2873 Patch from Wolfgang Breyha.
2875 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2876 comma-sep list; embedded commas doubled.
2878 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2880 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2882 Report and patch from Dmitry Banschikov.
2884 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2885 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2886 failures. We appear to now support TLS1.1+ with Exim.
2888 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2889 lets Exim select keys and certificates based upon TLS SNI from client.
2890 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2891 before an outbound SMTP session. New log_selector, +tls_sni.
2893 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2894 NULL dereference. Report and patch from Alun Jones.
2896 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2897 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2898 Not seeing resolver debug output on NetBSD, but suspect this is a
2899 resolver implementation change.
2901 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2902 Left warnings. Added "eximon gdb" invocation mode.
2904 PP/21 Defaulting "accept_8bitmime" to true, not false.
2906 PP/22 Added -bw for inetd wait mode support.
2908 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2909 locate the relevant includes and libraries. Made this the default.
2911 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2912 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2914 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2915 This may cause build issues on older platforms.
2917 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2918 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2919 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2920 Added SNI support via GnuTLS too.
2921 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2923 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2925 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2926 Applied second patch from Janne, implementing suggestion to default
2927 multiple-strings-in-record handling to match SPF spec.
2929 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2931 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2932 read-only, out of scope).
2933 Patch from Wolfgang Breyha, report from Stuart Northfield.
2935 PP/29 Fix three issues highlighted by clang analyser static analysis.
2936 Only crash-plausible issue would require the Cambridge-specific
2937 iplookup router and a misconfiguration.
2938 Report from Marcin Mirosław.
2940 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2942 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2943 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2944 As part of this, removing so much warning spew let me fix some minor
2945 real issues in debug logging.
2947 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2948 assignment on my part. Fixed.
2950 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2951 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2952 Janne Snabb (who went above and beyond: thank you).
2954 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2955 string otherwise requires a connection and a bunch more work and it's
2956 relatively easy to get wrong. Should also expose TLS library linkage
2959 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2960 64-bit ${eval} (JH/03).
2962 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2963 GNU libc to support some of the 64-bit stuff, should not lead to
2964 conflicts. Defined before os.h is pulled in, so if a given platform
2965 needs to override this, it can.
2967 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2968 protection layer was required, which is not implemented.
2969 Bugzilla 1254, patch from Wolfgang Breyha.
2971 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2972 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2973 tls_dhparam take prime identifiers. Also unbreak combination of
2974 OpenSSL+DH_params+TLSSNI.
2976 PP/39 Disable SSLv2 by default in OpenSSL support.
2982 PP/01 Solaris build fix for Oracle's LDAP libraries.
2983 Bugzilla 1109, patch from Stephen Usher.
2985 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2987 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2990 TF/02 Fix a couple more cases where we did not log the error message
2991 when unlink() failed. See also change 4.74-TF/03.
2993 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2994 lock up or crash if it happened to be inside a call to libc when it
2995 got a SIGUSR1 from exiwhat.
2997 The SIGUSR1 handler appends the current process status to the process
2998 log which is later printed by exiwhat. It used to use the general
2999 purpose logging code to do this, but several functions it calls are
3000 not safe for signals.
3002 The new output code in the SIGUSR1 handler is specific to the process
3003 log, and simple enough that it's easy to inspect for signal safety.
3004 Removing some special cases also simplifies the general logging code.
3005 Removing the spurious timestamps from the process log simplifies
3008 TF/04 Improved ratelimit ACL condition.
3010 The /noupdate option has been deprecated in favour of /readonly which
3011 has clearer semantics. The /leaky, /strict, and /readonly update modes
3012 are mutually exclusive. The update mode is no longer included in the
3013 database key; it just determines when the database is updated. (This
3014 means that when you upgrade Exim will forget old rate measurements.)
3016 Exim now checks that the per_* options are used with an update mode that
3017 makes sense for the current ACL. For example, when Exim is processing a
3018 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3019 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3020 must specify per_mail/readonly. If you omit the update mode it defaults to
3021 /leaky where that makes sense (as before) or /readonly where required.
3023 The /noupdate option is now undocumented but still supported for
3024 backwards compatibility. It is equivalent to /readonly except that in
3025 ACLs where /readonly is required you may specify /leaky/noupdate or
3026 /strict/noupdate which are treated the same as /readonly.
3028 A useful new feature is the /count= option. This is a generalization
3029 of the per_byte option, so that you can measure the throughput of other
3030 aggregate values. For example, the per_byte option is now equivalent
3031 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3033 The per_rcpt option has been generalized using the /count= mechanism
3034 (though it's more complicated than the per_byte equivalence). When it is
3035 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3036 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3037 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3038 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3039 Note that using per_rcpt with a non-readonly update mode in more than
3040 one ACL will cause the recipients to be double-counted. (The per_mail
3041 and per_byte options don't have this problem.)
3043 The handling of very low rates has changed slightly. If the computed rate
3044 is less than the event's count (usually one) then this event is the first
3045 after a long gap. In this case the rate is set to the same as this event's
3046 count, so that the first message of a spam run is counted properly.
3048 The major new feature is a mechanism for counting the rate of unique
3049 events. The new per_addr option counts the number of different
3050 recipients that someone has sent messages to in the last time period. It
3051 behaves like per_rcpt if all the recipient addresses are different, but
3052 duplicate recipient addresses do not increase the measured rate. Like
3053 the /count= option this is a general mechanism, so the per_addr option
3054 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3055 example, measure the rate that a client uses different sender addresses
3056 with the options per_mail/unique=$sender_address. There are further
3057 details in the main documentation.
3059 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3061 TF/06 Removed a few PCRE remnants.
3063 TF/07 Automatically extract Exim's version number from tags in the git
3064 repository when doing development or release builds.
3066 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3067 Bugzilla 879. Patch from Paul Fisher.
3069 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3070 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3073 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3075 PP/05 Variable $av_failed, true if the AV scanner deferred.
3076 Bugzilla 1078. Patch from John Horne.
3078 PP/06 Stop make process more reliably on build failure.
3079 Bugzilla 1087. Patch from Heiko Schlittermann.
3081 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3082 Bugzilla 1089. Patch from Heiko Schlittermann.
3084 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3085 Bugzilla 1131. Patch from Holger Weiß.
3087 PP/09 Handle IPv6 addresses with SPF.
3088 Bugzilla 860. Patch from Wolfgang Breyha.
3090 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3092 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3095 PP/11 match_* no longer expand right-hand-side by default.
3096 New compile-time build option, EXPAND_LISTMATCH_RHS.
3097 New expansion conditions, "inlist", "inlisti".
3099 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3101 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3103 PP/14 fix log_write() format string regression from TF/03.
3104 Bugzilla 1152. Patch from Dmitry Isaikin.
3110 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3112 PP/02 Harmonised TLS library version reporting; only show if debugging.
3113 Layout now matches that introduced for other libraries in 4.74 PP/03.
3115 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3117 PP/04 New "dns_use_edns0" global option.
3119 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3122 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3123 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3125 TK/01 Updated PolarSSL code to 0.14.2.
3126 Bugzilla 1097. Patch from Andreas Metzler.
3128 PP/07 Catch divide-by-zero in ${eval:...}.
3129 Fixes bugzilla 1102.
3131 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3134 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3135 format-string attack -- SECURITY: remote arbitrary code execution.
3137 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3138 time unintentionally subject to list matching rules, letting the header
3139 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3140 arbitrary string expansion). This allowed for information disclosure.
3142 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3143 INT_MIN/-1 -- value coerced to INT_MAX.
3149 NM/01 Workaround for PCRE version dependency in version reporting
3152 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3153 This fixes portability to compilers other than gcc, notably
3154 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3156 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3157 makefiles for portability to HP-UX and POSIX correctness.
3159 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3160 Also via indented variable definition in the Makefile.
3161 (Debugging by Oliver Heesakkers).
3163 PP/02 Restore caching of spamd results with expanded spamd_address.
3164 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3166 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3167 Improves build reliability. Fix from: Frank Elsner
3169 NM/02 Fix wide character breakage in the rfc2047 coding
3170 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3172 NM/03 Allow underscore in dnslist lookups
3173 Fixes bug 1026. Patch from Graeme Fowler
3175 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3176 Code patches from Adam Ciarcinski of NetBSD.
3178 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3181 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3182 is logged, to avoid truncation. Patch from John Horne.
3184 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3185 Patch from Jakob Hirsch.
3187 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3188 SQL string expansion failure details.
3189 Patch from Andrey Oktyabrski.
3191 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3192 Patch from Simon Arlott.
3194 PP/09 New lookups functionality failed to compile on old gcc which rejects
3195 extern declarations in function scope.
3196 Patch from Oliver Fleischmann
3198 PP/10 Use sig_atomic_t for flags set from signal handlers.
3199 Check getgroups() return and improve debugging.
3200 Fixed developed for diagnosis in bug 927 (which turned out to be
3203 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3204 Patch from Mark Zealey.
3206 PP/12 Bugzilla 1056: Improved spamd server selection.
3207 Patch from Mark Zealey.
3209 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3210 Based on patch from Heiko Schlittermann.
3212 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3213 Patch from Uwe Doering, sign-off by Michael Haardt.
3215 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3216 variable declaration deep within a block. Bug and patch from
3219 PP/15 lookups-Makefile IRIX compatibility coercion.
3221 PP/16 Make DISABLE_DKIM build knob functional.
3223 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3224 Patch by Simon Arlott
3226 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3227 variable argument macros. Our copy now differs from upstream.
3233 TF/01 Failure to get a lock on a hints database can have serious
3234 consequences so log it to the panic log.
3236 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3237 controlled using the smtp_confirmation log selector.
3239 TF/03 Include the error message when we fail to unlink a spool file.
3241 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3242 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3243 for maintaining out-of-tree patches for some time.
3245 PP/01 Bugzilla 139: Documentation and portability issues.
3246 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3247 Handle per-OS dynamic-module compilation flags.
3249 PP/02 Let /dev/null have normal permissions.
3250 The 4.73 fixes were a little too stringent and complained about the
3251 permissions on /dev/null. Exempt it from some checks.
3252 Reported by Andreas M. Kirchwitz.
3254 PP/03 Report version information for many libraries, including
3255 Exim version information for dynamically loaded libraries. Created
3256 version.h, now support a version extension string for distributors
3257 who patch heavily. Dynamic module ABI change.
3259 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3260 privilege escalation vulnerability whereby the Exim run-time user
3261 can cause root to append content of the attacker's choosing to
3264 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3267 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3268 If dropping privileges for untrusted macros, we disabled normal logging
3269 on the basis that it would fail; for the Exim run-time user, this is not
3270 the case, and it resulted in successful deliveries going unlogged.
3271 Fixed. Reported by Andreas Metzler.
3277 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3278 only prepend for the Resent-* case. Fixes regression introduced in
3279 Exim 4.70 by NM/22 for Bugzilla 607.
3281 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3282 increasing numbers of administrators be bitten by this.
3284 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3286 PP/03 Bugzilla 994: added openssl_options main configuration option.
3288 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3290 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3292 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3294 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3295 without a peer certificate, leading to a segfault because of an
3296 assumption that peers always have certificates. Be a little more
3297 paranoid. Problem reported by Martin Tscholak.
3299 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3300 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3301 NB: ClamAV planning to remove STREAM in "middle of 2010".
3302 CL also introduces -bmalware, various -d+acl logging additions and
3303 more caution in buffer sizes.
3305 PP/09 Implemented reverse_ip expansion operator.
3307 PP/10 Bugzilla 937: provide a "debug" ACL control.
3309 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3311 PP/12 Bugzilla 973: Implement --version.
3313 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3315 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3317 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3319 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3320 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3321 ignore trailing whitespace.
3323 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3325 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3328 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3329 Notification from Dr Andrew Aitchison.
3331 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3332 ExtendedDetectionInfo response format.
3333 Notification from John Horne.
3335 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3338 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3339 XSL and documented dependency on system catalogs, with examples of how
3342 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3345 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3346 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3347 configuration file which is writeable by the Exim user or group.
3349 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3350 of configuration files to cover files specified with the -C option if
3351 they are going to be used with root privileges, not just the default
3354 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3355 option (effectively making it always true).
3357 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3358 files to be used while preserving root privileges.
3360 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3361 that rogue child processes cannot use them.
3363 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3364 run-time user, instead of root.
3366 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3367 Exim run-time user without dropping privileges.
3369 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3370 result string, instead of calling string_vformat() twice with the same
3373 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3374 for other users. Others should always drop root privileges if they use
3375 -C on the command line, even for a whitelisted configure file.
3377 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3379 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3385 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3386 $data_path, and $header_path variables; fixed documentation bugs and
3389 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3390 exipick to access non-standard spools, including the "frozen" queue
3393 NM/01 Bugzilla 965: Support mysql stored procedures.
3394 Patch from Alain Williams
3396 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3398 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3399 Patch from Andreas Metzler
3401 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3402 Patch from Kirill Miazine
3404 NM/05 Bugzilla 671: Added umask to procmail example.
3406 JJ/03 installed exipick 20100323.0, fixing doc bug
3408 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3409 directory. Notification and patch from Dan Rosenberg.
3411 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3413 TK/02 Improve log output when DKIM signing operation fails.
3415 MH/01 Treat the transport option dkim_domain as a colon separated
3416 list, not as a single string, and sign the message with each element,
3417 omitting multiple occurences of the same signer.
3419 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3420 Bugzilla 985, 986. Patch by Simon Arlott
3422 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3423 Patch by Simon Arlott
3425 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3426 MBX locking. Notification from Dan Rosenberg.
3432 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3434 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3436 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3438 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3440 NM/04 Bugzilla 915: Flag broken perl installation during build.
3446 TK/01 Added patch by Johannes Berg that expands the main option
3447 "spamd_address" if it starts with a dollar sign.
3449 TK/02 Write list of recipients to X-Envelope-Sender header when building
3450 the mbox-format spool file for content scanning (suggested by Jakob
3453 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3454 (http://www.dcc-servers.net/) support via dccifd. Activated by
3455 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3457 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3458 by Mark Daniel Reidel <mr@df.eu>.
3460 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3461 When building exim an external PCRE library is now needed -
3462 PCRE is a system library on the majority of modern systems.
3463 See entry on PCRE_LIBS in EDITME file.
3465 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3466 conversation. Added nologin parameter to request.
3467 Patch contributed by Kirill Miazine.
3469 TF/01 Do not log submission mode rewrites if they do not change the address.
3471 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3473 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3474 log files in place. Contributed by Roberto Lima.
3476 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3478 TF/03 Bugzilla 615: When checking the local_parts router precondition
3479 after a local_part_suffix or local_part_prefix option, Exim now
3480 does not use the address's named list lookup cache, since this
3481 contains cached lookups for the whole local part.
3483 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3484 Robert Millan. Documentation is in experimental-spec.txt.
3486 TF/04 Bugzilla 668: Fix parallel build (make -j).
3488 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3490 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3491 Patch provided by Jan Srzednicki.
3493 TF/05 Leading white space used to be stripped from $spam_report which
3494 wrecked the formatting. Now it is preserved.
3496 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3497 that they are available at delivery time.
3499 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3501 TF/08 TLS error reporting now respects the incoming_interface and
3502 incoming_port log selectors.
3504 TF/09 Produce a more useful error message if an SMTP transport's hosts
3505 setting expands to an empty string.
3507 NM/06 Bugzilla 744: EXPN did not work under TLS.
3508 Patch provided by Phil Pennock.
3510 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3511 Patch provided by Richard Godbee.
3513 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3514 acl_smtp_notquit, added index entry.
3516 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3517 Patch provided by Eugene Bujak.
3519 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3520 accept(). Patch provided by Maxim Dounin.
3522 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3523 Patch provided by Phil Pennock.
3525 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3527 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3528 Patch provided by Brad "anomie" Jorsch.
3530 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3531 Patch provided by Dean Brooks.
3533 TK/05 Add native DKIM support (does not depend on external libraries).
3535 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3536 Patch provided by Graeme Fowler.
3538 NM/16 Bugzilla 851: Documentation example syntax fix.
3540 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3542 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3545 NM/19 Bugzilla 745: TLS version reporting.
3546 Patch provided by Phil Pennock.
3548 NM/20 Bugzilla 167: bool: condition support.
3549 Patch provided by Phil Pennock.
3551 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3552 clients. Patch provided by Phil Pennock.
3554 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3555 Patch provided by Brad "anomie" Jorsch.
3557 NM/23 Bugzilla 687: Fix misparses in eximstats.
3558 Patch provided by Heiko Schlittermann.
3560 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3561 Patch provided by Heiko Schlittermann.
3563 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3564 plus update to original patch.
3566 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3568 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3569 Patch provided by David Brownlee.
3571 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3573 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3575 NM/30 Bugzilla 888: TLS documentation bugfixes.
3577 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3579 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3580 Unlike the original bugzilla I have changed all shell scripts in src tree.
3582 NM/33 Bugzilla 898: Transport filter timeout fix.
3583 Patch by Todd Rinaldo.
3585 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3586 Patch by Serge Demonchaux.
3588 NM/35 Bugzilla 39: Base64 decode bug fixes.
3589 Patch by Jakob Hirsch.
3591 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3593 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3595 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3597 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3603 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3604 ALT-N's libdkim that I have put here:
3605 http://duncanthrax.net/exim-experimental/
3607 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3608 were called 'true' and it seems that C99 defines that as a reserved
3609 keyword to be used with 'bool' variable types. That means you could
3610 not include C99-style headers which use bools without triggering
3611 build errors in sieve.c.
3613 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3614 as mailq or other aliases. Changed the --help handling significantly
3615 to do whats expected. exim_usage() emits usage/help information.
3617 SC/01 Added the -bylocaldomain option to eximstats.
3619 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3621 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3623 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3629 PH/01 Another patch from the Sieve maintainer.
3631 PH/02 When an IPv6 address is converted to a string for single-key lookup
3632 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3633 dots are used instead of colons so that keys in lsearch files need not
3634 contain colons. This was done some time before quoting was made available
3635 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3636 (notated using the quote facility) so as to distinguish them from IPv4
3637 keys. This meant that lookups for IP addresses in host lists did not work
3638 for iplsearch lookups.
3640 This has been fixed by arranging for IPv6 addresses to be expressed with
3641 colons if the lookup type is iplsearch. This is not incompatible, because
3642 previously such lookups could never work.
3644 The situation is now rather anomalous, since one *can* have colons in
3645 ordinary lsearch keys. However, making the change in all cases is
3646 incompatible and would probably break a number of configurations.
3648 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3651 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3652 conversion specification without a maximum field width, thereby enabling
3653 a rogue spamd server to cause a buffer overflow. While nobody in their
3654 right mind would setup Exim to query an untrusted spamd server, an
3655 attacker that gains access to a server running spamd could potentially
3656 exploit this vulnerability to run arbitrary code as the Exim user.
3658 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3659 $primary_hostname instead of what libspf2 thinks the hosts name is.
3661 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3662 a directory entry by the name of the lookup key. Previously, if a
3663 symlink pointed to a non-existing file or a file in a directory that
3664 Exim lacked permissions to read, a lookup for a key matching that
3665 symlink would fail. Now it is enough that a matching directory entry
3666 exists, symlink or not. (Bugzilla 503.)
3668 PH/03 The body_linecount and body_zerocount variables are now exported in the
3671 PH/04 Added the $dnslist_matched variable.
3673 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3674 This means they are set thereafter only if the connection becomes
3677 PH/06 Added the client_condition to authenticators so that some can be skipped
3678 by clients under certain conditions.
3680 PH/07 The error message for a badly-placed control=no_multiline_responses left
3681 "_responses" off the end of the name.
3683 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3685 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3686 (without spaces) instead of just copying the configuration text.
3688 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3690 PH/11 Added $max_received_linelength.
3692 PH/12 Added +ignore_defer and +include_defer to host lists.
3694 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3695 way in which PCRE > 7.0 is built.
3697 PH/14 Implemented queue_only_load_latch.
3699 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3700 MAIL command. The effect was to mangle the value on 64-bit systems.
3702 PH/16 Another patch from the Sieve maintainer.
3704 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3706 PH/18 If a system quota error occurred while trying to create the file for
3707 a maildir delivery, the message "Mailbox is full" was not appended to the
3708 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3709 only to a quota excession during the actual writing of the file.
3711 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3712 characters?) which causes problems in log lines. The DN values are now
3713 passed through string_printing() before being added to log lines.
3715 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3716 and InterBase are left for another time.)
3718 PH/21 Added message_body_newlines option.
3720 PH/22 Guard against possible overflow in moan_check_errorcopy().
3722 PH/23 POSIX allows open() to be a macro; guard against that.
3724 PH/24 If the recipient of an error message contained an @ in the local part
3725 (suitably quoted, of course), incorrect values were put in $domain and
3726 $local_part during the evaluation of errors_copy.
3732 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3733 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3736 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3737 issue a MAIL command.
3739 PH/02 In an ACL statement such as
3741 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3743 if a client was not listed at all, or was listed with a value other than
3744 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3745 the condition was not true (as it should be), so access was not denied.
3746 The bug was that the ! inversion was incorrectly passed on to the second
3747 item. This has been fixed.
3749 PH/03 Added additional dnslists conditions == and =& which are different from
3750 = and & when the dns lookup returns more than one IP address.
3752 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3753 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3755 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3756 FSYNC, which compiles an option called disable_fsync that allows for
3757 bypassing fsync(). The documentation is heavily laced with warnings.
3759 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3761 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3762 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3763 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3764 including adding "make clean"; (3) Added -fPIC when compiling the test
3765 dynamically loaded module, to get rid of a warning.
3767 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3768 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3769 = 0s. The bug is otherwise harmless.
3771 PH/07 There was a bug in the dovecot authenticator such that the value of
3772 $auth1 could be overwritten, and so not correctly preserved, after a
3773 successful authentication. This usually meant that the value preserved by
3774 the server_setid option was incorrect.
3776 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3778 PH/09 Installed PCRE release 7.0.
3780 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3781 run for batched SMTP input. It is now run at the start of every message
3782 in the batch. While fixing this I discovered that the process information
3783 (output by running exiwhat) was not always getting set for -bs and -bS
3784 input. This is fixed, and it now also says "batched" for BSMTP.
3786 PH/11 Added control=no_pipelining.
3788 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3789 patch, slightly modified), and move the expansion of helo_data till after
3790 the connection is made in the smtp transport (so it can use these
3793 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3795 PH/14 Added log_selector = +pid.
3797 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3799 PH/16 Add ${if forany and ${if forall.
3801 PH/17 Added dsn_from option to vary the From: line in DSNs.
3803 PH/18 Flush SMTP output before performing a callout, unless control =
3804 no_callout_flush is set.
3806 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3807 was true (the default) a successful delivery failed to delete the retry
3808 item, thus causing premature timeout of the address. The bug is now
3811 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3813 PH/21 Long custom messages for fakedefer and fakereject are now split up
3814 into multiline responses in the same way that messages for "deny" and
3815 other ACL rejections are.
3817 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3818 with slight modification.
3820 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3821 draft, changing the syntax and factoring some duplicate code".
3823 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3824 for deliveries of the second and subsequent messages over the same SMTP
3827 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3828 ${reduce, with only minor "tidies".
3830 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3832 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3833 expansion side effects.
3835 PH/27 When a message times out after an over-quota error from an Exim-imposed
3836 quota, the bounce message says "mailbox is full". This message was not
3837 being given when it was a system quota that was exceeded. It now should
3840 MH/03 Made $recipients available in local_scan(). local_scan() already has
3841 better access to the recipient list through recipients_list[], but
3842 $recipients can be useful in postmaster-provided expansion strings.
3844 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3845 in the case of a MAIL command with additional options following the
3846 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3847 were accidentally chopped off.
3849 PH/29 SMTP synchronization checks are implemented when a command is read -
3850 there is a check that no more input is waiting when there shouldn't be
3851 any. However, for some commands, a delay in an ACL can mean that it is
3852 some time before the response is written. In this time, more input might
3853 arrive, invalidly. So now there are extra checks after an ACL has run for
3854 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3855 pipelining has not been advertised.
3857 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3859 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3860 returned in the SMTP response when smtp_return_error_details was false.
3861 This has been fixed.
3863 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3864 instead of the C I/O that was originally supplied, because problems were
3865 reported on Solaris.
3867 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3868 Exim which did not show up earlier: it was assuming that a call to
3869 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3870 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3871 was a macro that became an assignment, so it seemed to work. This has
3872 changed to a proper function call with a void return, hence the compile
3873 error. Exim's code has been fixed.
3875 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3878 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3880 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3882 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3883 because it turns out to be too big or there is a timeout) there is no
3884 "Completed" line in the log. When some messages of this type were
3885 selected by exigrep, they were listed as "not completed". Others were
3886 picked up by some special patterns. I have improved the selection
3887 criteria to be more general.
3889 PH/38 The host_find_failed option in the manualroute router can now be set
3890 to "ignore", to completely ignore a host whose IP address cannot be
3891 found. If all hosts are ignored, the behaviour is controlled by the new
3892 host_all_ignored option.
3894 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3895 homing or because of multiple MX records with /mx) generated more than
3896 one IP address, and the following item turned out to be the local host,
3897 all the secondary addresses of the first item were incorrectly removed
3898 from the list, along with the local host and any following hosts (which
3899 is what is supposed to happen).
3901 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3902 whoever called Exim into the -H file. In the case of the daemon it was
3903 behaving confusingly. When first started, it used values for whoever
3904 started the daemon, but after a SIGHUP it used the Exim user (because it
3905 calls itself on a restart). I have changed the code so that it now always
3908 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3909 message are rejected with the same error (e.g. no authentication or bad
3910 sender address), and a DATA command is nevertheless sent (as can happen
3911 with PIPELINING or a stupid MUA), the error message that was given to the
3912 RCPT commands is included in the rejection of the DATA command. This is
3913 intended to be helpful for MUAs that show only the final error to their
3916 PH/42 Another patch from the Sieve maintainer.
3918 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3919 Eximstats - Fixed some broken HTML links and added missing column headers
3921 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3922 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3924 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3926 PH/43 Yet another patch from the Sieve maintainer.
3928 PH/44 I found a way to check for a TCP/IP connection going away before sending
3929 the response to the final '.' that terminates a message, but only in the
3930 case where the client has not sent further data following the '.'
3931 (unfortunately, this is allowed). However, in many cases there won't be
3932 any further data because there won't be any more messages to send. A call
3933 to select() can be used: if it shows that the input is "ready", there is
3934 either input waiting, or the socket has been closed. An attempt to read
3935 the next input character can distinguish the two cases. Previously, Exim
3936 would have sent an OK response which the client would never have see.
3937 This could lead to message repetition. This fix should cure that, at
3938 least in a lot of common cases.
3940 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3941 advertised in response to EHLO.
3947 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3948 fixed by 4.65/MH/01 (is this a record?) are fixed:
3950 (i) An empty string was always treated as zero by the numeric comparison
3951 operators. This behaviour has been restored.
3953 (ii) It is documented that the numeric comparison operators always treat
3954 their arguments as decimal numbers. This was broken in that numbers
3955 starting with 0 were being interpreted as octal.
3957 While fixing these problems I realized that there was another issue that
3958 hadn't been noticed. Values of message_size_limit (both the global option
3959 and the transport option) were treated as octal if they started with 0.
3960 The documentation was vague. These values are now always treated as
3961 decimal, and I will make that clear in the documentation.
3967 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3968 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3971 MH/01 Don't check that the operands of numeric comparison operators are
3972 integers when their expansion is in "skipping" mode (fixes bug
3973 introduced by 4.64-PH/07).
3975 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3976 child addresses, Exim now panics and dies. Previously, because the count
3977 is held in a short int, deliveries were likely to be lost. As such a
3978 large number of recipients for a single message is ridiculous
3979 (performance will be very, very poor), I have chosen to impose a limit
3980 rather than extend the field.
3986 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3987 leftover -K file (the existence of which was triggered by #402).
3988 While we were at it, introduced process PID as part of the -K
3989 filename. This should rule out race conditions when creating
3992 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3993 processing considerably. Previous code took too long for large mails,
3994 triggering a timeout which in turn triggers #401.
3996 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3997 in the DK code in transports.c. sendfile() is not really portable,
3998 hence the _LINUX specificness.
4000 TF/01 In the add_headers option to the mail command in an Exim filter,
4001 there was a bug that Exim would claim a syntax error in any
4002 header after the first one which had an odd number of characters
4005 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4006 callout verification, Exim cached a "reject" for the entire domain. This
4007 is correct for most verifications, but it is not correct for a recipient
4008 verification with use_sender or use_postmaster set, because in that case
4009 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4010 case of MAIL FROM:<> rejection from other early rejections (e.g.
4011 rejection of HELO). When verifying a recipient using a non-null MAIL
4012 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4013 Whatever the result of the callout, the value of the domain cache is
4014 left unchanged (for any other kind of callout, getting as far as trying
4015 RCPT means that the domain itself is ok).
4017 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4020 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4021 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4022 session, the daemon ignores ECONNECT errors and logs others; it now
4023 ignores EPIPE as well.
4025 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4026 (quoted-printable decoding).
4028 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4029 later the small subsequent patch to fix an introduced bug.
4031 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4033 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4035 PH/08 An error is now given if message_size_limit is specified negative.
4037 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4038 to be given (somewhat) arbitrary names.
4040 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4043 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4044 miscellaneous code fixes
4046 PH/10 Added the log_reject_target ACL modifier to specify where to log
4049 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4050 hostname. This is wrong, because it relates to the incoming message (and
4051 probably the interface on which it is arriving) and not to the outgoing
4052 callout (which could be using a different interface). This has been
4053 changed to use the value of the helo_data option from the smtp transport
4054 instead - this is what is used when a message is actually being sent. If
4055 there is no remote transport (possible with a router that sets up host
4056 addresses), $smtp_active_hostname is used.
4058 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4059 tweaks were necessary in order to get it to work (see also 21 below):
4060 (a) The code assumed that strncpy() returns a negative number on buffer
4061 overflow, which isn't the case. Replaced with Exim's string_format()
4063 (b) There were several signed/unsigned issues. I just did the minimum
4064 hacking in of casts. There is scope for a larger refactoring.
4065 (c) The code used strcasecmp() which is not a standard C function.
4066 Replaced with Exim's strcmpic() function.
4067 (d) The code set only $1; it now sets $auth1 as well.
4068 (e) A simple test gave the error "authentication client didn't specify
4069 service in request". It would seem that Dovecot has changed its
4070 interface. Fortunately there's a specification; I followed it and
4071 changed what the client sends and it appears to be working now.
4073 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4076 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4077 address A is aliased to B and C, where B exists and C does not. Without
4078 -v the output is "A verified" because verification stops after a
4079 successful redirection if more than one address is generated. However,
4080 with -v the child addresses are also verified. Exim was outputting "A
4081 failed to verify" and then showing the successful verification for C,
4082 with its parentage. It now outputs "B failed to verify", showing B's
4083 parentage before showing the successful verification of C.
4085 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4086 look up a TXT record in a specific list after matching in a combined
4089 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4090 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4091 they consult the DNS. I had assumed they would set it the way they
4092 wanted; and indeed my experiments on Linux seem to show that in some
4093 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4094 To be on the safe side, however, I have now made the interface to
4095 host_find_byname() similar to host_find_bydns(), with an argument
4096 containing the DNS resolver options. The host_find_byname() function now
4097 sets these options at its start, just as host_find_bydns() does. The smtp
4098 transport options dns_qualify_single and dns_search_parents are passed to
4099 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4100 of host_find_byname() use the default settings of RES_DEFNAMES
4101 (qualify_single) but not RES_DNSRCH (search_parents).
4103 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4104 spool_read_header() do less string testing, by means of a preliminary
4105 switch on the second character of optional "-foo" lines. (This is
4106 overdue, caused by the large number of possibilities that now exist.
4107 Originally there were few.) While I was there, I also converted the
4108 str(n)cmp tests so they don't re-test the leading "-" and the first
4109 character, in the hope this might squeeze out yet more improvement.
4111 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4112 flag allowing group syntax was set by the header_syntax check but not
4113 turned off, possible causing trouble later; (2) The flag was not being
4114 set at all for the header_verify test, causing "group"-style headers to
4115 be rejected. I have now set it in this case, and also caused header_
4116 verify to ignore an empty address taken from a group. While doing this, I
4117 came across some other cases where the code for allowing group syntax
4118 while scanning a header line wasn't quite right (mostly, not resetting
4119 the flag correctly in the right place). These bugs could have caused
4120 trouble for malformed header lines. I hope it is now all correct.
4122 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4123 with the "reply" argument non-NULL. The code, however (which originally
4124 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4125 but it didn't always do it. This confused somebody who was copying the
4126 code for some other use. I have removed all the tests.
4128 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4129 feature that was used to support insecure browsers during the U.S. crypto
4130 embargo. It requires special client support, and Exim is probably the
4131 only MTA that supported it -- and would never use it because real RSA is
4132 always available. This code has been removed, because it had the bad
4133 effect of slowing Exim down by computing (never used) parameters for the
4134 RSA_EXPORT functionality.
4136 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4137 authenticator to fail if there's a tab character in the incoming data
4138 (there should never be unless someone is messing about, as it's supposed
4139 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4140 if the connection is using TLS or if the remote IP is the same as the
4141 local IP, and the "valid-client-cert option" if a client certificate has
4144 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4145 authenticators. This can be used for authorization after authentication
4146 succeeds. (In the case of plaintext, it servers for both authentication
4149 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4150 if any retry times were supplied.
4152 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4153 connection, where there is no client IP address to check. In this
4154 situation, the verify now always succeeds.
4156 PH/25 Applied John Jetmore's -Mset patch.
4158 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4160 PH/27 In a string expansion for a processed (not raw) header when multiple
4161 headers of the same name were present, leading whitespace was being
4162 removed from all of them, but trailing whitespace was being removed only
4163 from the last one. Now trailing whitespace is removed from each header
4164 before concatenation. Completely empty headers in a concatenation (as
4165 before) are ignored.
4167 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4168 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4170 PH/29 [Removed. This was a change that I later backed out, and forgot to
4171 correct the ChangeLog entry (that I had efficiently created) before
4172 committing the later change.]
4174 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4175 address errors (4xx response to RCPT) over the same connection as other
4176 messages routed to the same hosts. Such deliveries are always "forced",
4177 so retry times are not inspected. This resulted in far too many retries
4178 for the affected addresses. The effect occurred only when there were more
4179 hosts than the hosts_max_try setting in the smtp transport when it had
4180 the 4xx errors. Those hosts that it had tried were not added to the list
4181 of hosts for which the message was waiting, so if all were tried, there
4182 was no problem. Two fixes have been applied:
4184 (i) If there are any address or message errors in an SMTP delivery, none
4185 of the hosts (tried or untried) are now added to the list of hosts
4186 for which the message is waiting, so the message should not be a
4187 candidate for sending over the same connection that was used for a
4188 successful delivery of some other message. This seems entirely
4189 reasonable: after all the message is NOT "waiting for some host".
4190 This is so "obvious" that I'm not sure why it wasn't done
4191 previously. Hope I haven't missed anything, but it can't do any
4192 harm, as the worst effect is to miss an optimization.
4194 (ii) If, despite (i), such a delivery is accidentally attempted, the
4195 routing retry time is respected, so at least it doesn't keep
4196 hammering the server.
4198 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4199 in ${readsocket because some servers need this prod.
4201 PH/32 Added some extra debug output when updating a wait-xxx database.
4203 PH/33 The hint "could be header name not terminated by colon", which has been
4204 given for certain expansion errors for a long time, was not being given
4205 for the ${if def:h_colon_omitted{... case.
4207 PH/34 The spec says: "With one important exception, whenever a domain list is
4208 being scanned, $domain contains the subject domain." There was at least
4209 one case where this was not true.
4211 PH/35 The error "getsockname() failed: connection reset by peer" was being
4212 written to the panic log as well as the main log, but it isn't really
4213 panic-worthy as it just means the connection died rather early on. I have
4214 removed the panic log writing for the ECONNRESET error when getsockname()
4217 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4218 runs only) independently of the message's sender address. This meant
4219 that, if the 4xx error was in fact related to the sender, a different
4220 message to the same recipient with a different sender could confuse
4221 things. In particular, this can happen when sending to a greylisting
4222 server, but other circumstances could also provoke similar problems.
4223 I have changed the default so that the retry time for these errors is now
4224 based a combination of the sender and recipient addresses. This change
4225 can be overridden by setting address_retry_include_sender=false in the
4228 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4229 remote server are returned as part of bounce messages. This was not
4230 happening for LMTP over a pipe (the lmtp transport), but now it is the
4231 same for both kinds of LMTP.
4233 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4234 in header lines that were in fact CNAMEs. This is no longer the case.
4236 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4237 and queue runs started by the daemon processed all messages. This has
4238 been fixed so that -R and -S can now usefully be given with -q<time>.
4240 PH/40 Import PCRE release 6.7 (fixes some bugs).
4242 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4244 PH/42 Give an error if -q is specified more than once.
4246 PH/43 Renamed the variables $interface_address and $interface_port as
4247 $received_ip_address and $received_port, to make it clear that these
4248 values apply to message reception, and not to the outgoing interface when
4249 a message is delivered. (The old names remain recognized, of course.)
4251 PH/44 There was no timeout on the connect() call when using a Unix domain
4252 socket in the ${readsocket expansion. There now is.
4254 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4255 be meaningful with "accept".
4257 SC/01 Eximstats V1.43
4258 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4260 SC/02 Eximstats V1.44
4261 Use a glob alias rather than an array ref in the generated
4262 parser. This improves both readability and performance.
4264 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4265 Collect SpamAssassin and rejection statistics.
4266 Don't display local sender or destination tables unless
4267 there is data to show.
4268 Added average volumes into the top table text output.
4270 SC/04 Eximstats V1.46
4271 Collect data on the number of addresses (recipients)
4272 as well as the number of messages.
4274 SC/05 Eximstats V1.47
4275 Added 'Message too big' to the list of mail rejection
4276 reasons (thanks to Marco Gaiarin).
4278 SC/06 Eximstats V1.48
4279 Mainlog lines which have GMT offsets and are too short to
4280 have a flag are now skipped.
4282 SC/07 Eximstats V1.49 (Alain Williams)
4283 Added the -emptyok flag.
4285 SC/08 Eximstats V1.50
4286 Fixes for obtaining the IP address from reject messages.
4288 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4289 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4290 whitespace changes from 4.64-PH/27
4292 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4295 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4296 are found, allow negative numbers in numeric criteria)
4298 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4300 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4301 to match changes made in 4.64-PH/43
4303 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4305 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4306 "domain" is actually an IP address, and force a failure. This locks out
4307 those revolvers/nameservers that support "A-for-A" lookups, in
4308 contravention of the specifications.
4310 PH/48 When a host name was looked up from an IP address, and the subsequent
4311 forward lookup of the name timed out, the host name was left in
4312 $sender_host_name, contrary to the specification.
4314 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4315 restricted to single-key lookups, Exim was not diagnosing an error if
4316 * or *@ was used with a query-style lookup.
4318 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4320 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4321 long ago, but noone interested enough thought of it. Let's just say that
4322 the "1.1" means that there are some new functions that weren't there at
4323 some point in the past.
4325 PH/51 Error processing for expansion failure of helo_data from an smtp
4326 transport during callout processing was broken.
4328 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4329 tested/used via the -bh/-bhc/-bs options.
4331 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4332 bug, fixed in subsequent PCRE releases).
4334 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4335 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4337 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4343 SC/01 Use a glob alias rather than an array ref in eximstats generated
4344 parser. This improves both readability and performance.
4346 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4347 Don't display local sender or destination tables in eximstats unless
4348 there is data to show.
4349 Added average volumes into the eximstats top table text output.
4351 SC/03 Collect data on the number of addresses (recipients) as well
4352 as the number of messages in eximstats.
4354 TF/01 Correct an error in the documentation for the redirect router. Exim
4355 does (usually) call initgroups() when daemonizing.
4357 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4358 with consistent privilege compared to when running as a daemon.
4360 TF/03 Note in the spec that $authenticated_id is not set for local
4361 submissions from trusted users.
4363 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4364 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4366 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4367 by adding some example configuration directives to the default
4368 configuration file. A little bit of work is required to uncomment the
4369 directives and define how usernames and passwords are checked, but
4370 there is now a framework to start from.
4372 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4373 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4374 without this. I don't know how relevant this is to other LDAP libraries.
4376 PH/02 Add the verb name to the "unknown ACL verb" error.
4378 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4380 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4382 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4383 directory not expanded when it should be if an expanded home directory
4384 was set for the address (which is overridden by the transport).
4386 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4389 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4390 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4391 because it is too late at that time, and has no effect.
4393 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4394 security issue with \' (bugzilla #107). I could not use the
4395 PQescapeStringConn() function, because it needs a PGconn value as one of
4398 PH/08 When testing addresses using -bt, indicate those final addresses that
4399 are duplicates that would not cause an additional delivery. At least one
4400 person was confused, thinking that -bt output corresponded to deliveries.
4401 (Suppressing duplicates isn't a good idea as you lose the information
4402 about possibly different redirections that led to the duplicates.)
4404 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4405 systems where poll() doesn't work, in particular OS X.
4407 PH/10 Added more information to debugging output for retry time not reached.
4409 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4410 operations in malware.c.
4412 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4415 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4416 syslog_duplication set false, log lines that would normally be written
4417 both the the main log and to the reject log were not written to syslog at
4420 PH/14 In the default configuration, change the use of "message" in ACL warn
4421 statements to "add_header".
4423 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4424 not followed by a command (e.g. "seen endif").
4426 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4427 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4430 PH/17 Added extra conditions to the default value of delay_warning_condition
4434 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4435 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4436 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4439 The Auto-Submitted: and various List- headers are standardised, whereas I
4440 don't think Precedence: ever was.
4442 PH/18 Refactored debugging code in route_finduser() to show more information,
4443 in particular, the error code if getpwnam() issues one.
4445 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4446 This is apparently needed in addition to the PH/07 change above to avoid
4447 any possible encoding problems.
4449 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4450 but not after initializing Perl.
4452 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4453 output them only if debugging. By default they are written stderr,
4454 apparently, which is not desirable.
4456 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4459 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4462 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4464 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4465 authenticated or an ident call has been made. Suppress the default
4466 values for $authenticated_id and $authenticated_sender (but permit -oMai
4467 and -oMas) when testing with -bh.
4469 PH/24 Re-jigged the order of the tests in the default configuration so that the
4470 tests for valid domains and recipients precede the DNS black list and CSA
4471 tests, on the grounds that those ones are more expensive.
4473 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4474 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4475 This bug exists in every version of Exim that I still have, right back to
4478 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4479 However, an attempt to turn on case-sensitivity in a regex key by
4480 including (?-i) didn't work because the subject string was already
4481 lowercased, and the effects were non-intuitive. It turns out that a
4482 one-line patch can be used to allow (?-i) to work as expected.
4488 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4489 other effects) broke the use of negated acl sub-conditions.
4491 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4494 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4495 "Deny" causes Exim to reject the incoming connection with a 554 error.
4496 Unfortunately, if there is a major crisis, such as a disk failure,
4497 tcp-wrappers gives "deny", whereas what one would like would be some
4498 kind of temporary error. A kludge has been added to help with this.
4499 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4500 554 error is used if errno is still zero or contains ENOENT (which occurs
4501 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4504 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4506 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4507 errors. Otherwise a message that provokes a temporary error (when other
4508 messages do not) can cause a whole host to time out.
4510 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4511 addresses were routed directly to files or pipes from a redirect router.
4512 File deliveries just didn't batch; pipe deliveries might have suffered
4515 PH/06 A failure to get a lock for a hints database would erroneously always say
4516 "Failed to get write lock", even when it was really a read lock.
4518 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4519 of 0600. This has been changed to use the value of the lockfile_mode
4520 option (which defaults to 0600).
4522 PH/08 Applied small patch from the Sieve maintainer.
4524 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4525 folder from quota calculations, a direct delivery into this folder messed
4526 up the contents of the maildirsize file. This was because the regex was
4527 used only to exclude .Trash (or whatever) when the size of the mailbox
4528 was calculated. There was no check that a delivery was happening into an
4529 excluded directory. This bug has been fixed by ignoring all quota
4530 processing for deliveries into excluded directories.
4532 PH/10 Added the maildirfolder_create_regex option to appendfile.
4538 PH/01 The code for finding all the local interface addresses on a FreeBSD
4539 system running IPv6 was broken. This may well have applied to all BSD
4540 systems, as well as to others that have similar system calls. The broken
4541 code found IPv4 interfaces correctly, but gave incorrect values for the
4542 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4543 that it would not match correctly against @[] and not recognize the IPv6
4546 PH/02 The ipliteral router was not recognizing addresses of the form user@
4547 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4549 PH/03 Added disable_ipv6.
4551 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4552 decoded form, because it is most often used to construct To: headers
4553 lines in autoreplies, and the decoded form may well be syntactically
4554 invalid. However, $reply_address has leading white space removed, and all
4555 newlines turned into spaces so that the autoreply transport does not
4558 PH/05 If group was specified without a user on a router, and no group or user
4559 was specified on a transport, the group from the router was ignored.
4561 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4562 for visible compile-time settings that can be used to change these
4563 numbers, for those that want even more. Backwards compatibility with old
4564 spool files has been maintained. However, going back to a previous Exim
4565 release will lost any variables that are in spool files.
4567 PH/07 Two small changes when running in the test harness: increase delay when
4568 passing a TCP/IP connection to a new process, in case the original
4569 process has to generate a bounce, and remove special handling of
4570 127.0.0.2 (sic), which is no longer necessary.
4572 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4573 be the same on different OS.
4575 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4578 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4579 whether --show-vars was specified or not
4581 JJ/02 exipick: Added support for new ACL variable spool format introduced
4584 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4585 syntactically invalid From: or Reply-to: line, and a filter used this to
4586 generate an autoreply, and therefore failed to obtain an address for the
4587 autoreply, Exim could try to deliver to a non-existent relative file
4588 name, causing unrelated and misleading errors. What now happens is that
4589 it logs this as a hard delivery error, but does not attempt to create a
4592 PH/11 The exinext utility has a -C option for testing purposes, but although
4593 the given file was scanned by exinext itself; it wasn't being passed on
4594 when Exim was called.
4596 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4597 an end-of-file indication when reading a command response.
4599 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4600 compiled. In many other places in Exim, IPv6 addresses are always
4601 recognized, so I have changed this. It also means that IPv4 domain
4602 literals of the form [IPV4:n.n.n.n] are now always recognized.
4604 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4605 used if the router is not running as root, for example, when verifying at
4606 ACL time, or when using -bh. The debugging output from this situation was
4607 non-existent - all you got was a failure to exec. I have made two
4610 (a) Failures to set uid/gid, the current directory, or a process leader
4611 in a subprocess such as that created by queryprogram now generate
4612 suitable debugging output when -d is set.
4614 (b) The queryprogram router detects when it is not running as root,
4615 outputs suitable debugging information if -d is set, and then runs
4616 the subprocess without attempting to change uid/gid.
4618 PH/15 Minor change to Makefile for building test_host (undocumented testing
4621 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4622 additional section of a DNS packet that returns MX or SRV records.
4623 Instead, it always explicitly searches for A/AAAA records. This avoids
4624 major problems that occur when a DNS server includes only records of one
4625 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4626 fixed another bug: if SRV records were looked up and the corresponding
4627 address records were *not* found in the additional section, the port
4628 values from the SRV records were lost.
4630 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4631 using the correct key (the original address) when searching the retry
4632 rules in order to find which one to use for generating the retry hint.
4634 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4635 adding the default one. Similarly, if it contains a Reply-To: header, the
4636 errors_reply_to option, if set, is not used.
4638 PH/19 When calculating a retry time, Exim used to measure the "time since
4639 failure" by looking at the "first failed" field in the retry record. Now
4640 it does not use this if it is later than than the arrival time of the
4641 message. Instead it uses the arrival time. This makes for better
4642 behaviour in cases where some deliveries succeed, thus re-setting the
4643 "first failed" field. An example is a quota failure for a huge message
4644 when small messages continue to be delivered. Without this change, the
4645 "time since failure" will always be short, possible causing more frequent
4646 delivery attempts for the huge message than are intended.
4647 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4649 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4650 $1, $2, $3) because the numerical variables can be reset during some
4651 expansion items (e.g. "match"), thereby losing the authentication data.
4653 PH/21 Make -bV show the size of off_t variables so that the test suite can
4654 decide whether to run tests for quotas > 2G.
4656 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4657 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4658 filecount value is greater than 2G or if a quota value is greater than 2G
4659 on a system where the size of off_t is not greater than 4, a panic error
4662 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4663 never match. The debug and -bh output now contains an explicit error
4664 message indicating a malformed IPv4 address or mask.
4666 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4667 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4668 PH/23 above applies.
4670 PH/25 Do not write to syslog when running in the test harness. The only
4671 occasion when this arises is a failure to open the main or panic logs
4672 (for which there is an explicit test).
4674 PH/26 Added the /no_tell option to "control=freeze".
4676 PH/27 If a host name lookup failed very early in a connection, for example, if
4677 the IP address matched host_lookup and the reverse lookup yielded a name
4678 that did not have a forward lookup, an error message of the form "no IP
4679 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4680 could be logged. Now it outputs the IP address instead of "NULL".
4682 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4683 allows the sender and the authenticated sender to be set when
4684 submitting a message from within Exim. Since child_open_exim() is
4685 documented for local_scan(), the new function should be too.
4687 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4688 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4689 results in an empty string is now treated as unset.
4691 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4693 PH/31 Added sender_verify_fail logging option.
4695 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4696 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4697 tidied the source and removed it altogether.
4699 PH/33 When a queue run was abandoned because the load average was too high, a
4700 log line was always written; now it is written only if the queue_run log
4701 selector is set. In addition, the log line for abandonment now contains
4702 information about the queue run such as the pid. This is always present
4703 in "start" and "stop" lines but was omitted from the "abandon" line.
4705 PH/34 Omit spaces between a header name and the colon in the error message that
4706 is given when verify = headers_syntax fails (if there are lots of them,
4707 the message gets confusing).
4709 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4710 names, as there are now some PTR records that contain slashes. This check
4711 is only to protect against broken name servers that fall over on strange
4712 characters, so the fact that it applies to all lookups doesn't matter.
4714 PH/36 Now that the new test suite is complete, we can remove some of the
4715 special code in Exim that was needed for the old test suite. For example,
4716 sorting DNS records because real resolvers return them in an arbitrary
4717 order. The new test suite's fake resolver always returns records in the
4720 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4721 bounces) except when queue_only is set, to avoid logging races between
4722 the different processes.
4724 PH/38 Panic-die if .include specifies a non-absolute path.
4726 PH/39 A tweak to the "H" retry rule from its user.
4728 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4729 a label. They prevented compilation on older perls.
4731 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4732 a warning to be raised on newish perls.
4734 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4735 on queue. Changes to match documented behaviour of showing count of
4736 messages matching specified criteria.
4738 PH/40 Changed the default ident timeout from 30s to 5s.
4740 PH/41 Added support for the use of login_cap features, on those BSD systems
4741 that have them, for controlling the resources used by pipe deliveries.
4743 PH/42 The content-scanning code uses fopen() to create files in which to put
4744 message data. Previously it was not paying any attention to the mode of
4745 the files. Exim runs with umask(0) because the rest of the code creates
4746 files with open(), and sets the required mode explicitly. Thus, these
4747 files were ending up world-writeable. This was not a big issue, because,
4748 being within the spool directory, they were not world-accessible. I have
4749 created a function called modefopen, which takes an additional mode
4750 argument. It sets umask(777), creates the file, chmods it to the required
4751 mode, then resets the umask. All the relevant calls to fopen() in the
4752 content scanning code have been changed to use this function.
4754 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4755 to 24 hours. This avoids potential overflow problems when processing G
4756 and H retry rules. I suspect nobody ever tinkers with this value.
4758 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4760 PH/45 When the plaintext authenticator is running as a client, the server's
4761 challenges are checked to ensure they are valid base64 strings. By
4762 default, the authentication attempt is cancelled if an invalid string is
4763 received. Setting client_ignore_invalid_base64 true ignores these errors.
4764 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4765 they are received. Thus, the responses can be made to depend on the
4766 challenges. If an invalid string is ignored, an empty string is placed in
4769 PH/46 Messages that are created by the autoreply transport now contains a
4770 References: header, in accordance with RFCs 2822 and 3834.
4772 PH/47 Added authenticated_sender_force to the smtp transport.
4774 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4776 PH/49 Installed latest patch from the Sieve maintainer.
4778 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4779 was also set, the appendfile transport was unnecessarily scanning a
4780 directory of message files (e.g. for maildir delivery) to find the count
4781 of files (along with the size), even though it did not need this
4782 information. It now does the scan only if it needs to find either the
4783 size of the count of files.
4785 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4787 PH/52 Two bugs concerned with error handling when the smtp transport is
4790 (i) Exim was not creating retry information for temporary errors given
4791 for individual recipients after the DATA command when the smtp transport
4792 was used in LMTP mode. This meant that they could be retried too
4793 frequently, and not timed out correctly.
4795 (ii) Exim was setting the flag that allows error details to be returned
4796 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4797 recipients that were returned after the DATA command.
4799 PH/53 This is related to PH/52, but is more general: for any failing address,
4800 when detailed error information was permitted to be returned to the
4801 sender, but the error was temporary, then after the final timeout, only
4802 "retry timeout exceeded" was returned. Now it returns the full error as
4803 well as "retry timeout exceeded".
4805 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4806 do this, and (what is worse) MTAs that accept it.
4808 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4809 will now be deprecated.
4811 PH/56 New os.c-cygwin from the Cygwin maintainer.
4813 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4814 formats (previously only available in exim formats via -bpr, -bpru,
4815 and -bpra. Now also available in native and exiqgrep formats)
4817 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4818 with very large, slow to parse queues
4820 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4822 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4824 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4825 responses to authentication challenges, though it was showing the
4826 challenges; (ii) I've removed the CR characters from the debug output for
4829 PH/58 Allow for the insertion of a newline as well as a space when a string
4830 is turned into more than one encoded-word during RFC 2047 encoding. The
4831 Sieve code now uses this.
4833 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4834 data_4xx, lost_connection, tls_required.
4836 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4837 message was being sent as an SMTP response.
4839 PH/61 Add -l and -k options to exicyclog.
4841 PH/62 When verifying, if an address was redirected to one new address, so that
4842 verification continued, and the new address failed or deferred after
4843 having set something in $address_data, the value of $address_data was not
4844 passed back to the ACL. This was different to the case when no
4845 redirection occurred. The value is now passed back in both cases.
4847 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4848 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4849 use login_cap.h, so on its own it isn't the distinguishing feature. The
4850 new name refers directly to the setclassresources() function.
4852 PH/65 Added configuration files for NetBSD3.
4854 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4856 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4857 is preferred over IPv4.
4859 PH/68 The bounce_return_message and bounce_return_body options were not being
4860 honoured for bounces generated during the reception of non-SMTP messages.
4861 In particular, this applied to messages rejected by the ACL. This bug has
4862 been fixed. However, if bounce_return_message is true and bounce_return_
4863 body is false, the headers that are returned for a non-SMTP message
4864 include only those that have been read before the error was detected.
4865 (In the case of an ACL rejection, they have all been read.)
4867 PH/69 The HTML version of the specification is now built in a directory called
4868 spec_html instead of spec.html, because the latter looks like a path with
4869 a MIME-type, and this confuses some software.
4871 PH/70 Catch two compiler warnings in sieve.c.
4873 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4874 function verify_get_ident() calls ip_connect() to connect a socket, but
4875 if the "connect()" function timed out, ip_connect() used to close the
4876 socket. However, verify_get_ident() also closes the socket later, and in
4877 between Exim writes to the log, which may get opened at this point. When
4878 the socket was closed in ip_connect(), the log could get the same file
4879 descriptor number as the socket. This naturally causes chaos. The fix is
4880 not to close the socket in ip_connect(); the socket should be closed by
4881 the function that creates it. There was only one place in the code where
4882 this was missing, in the iplookup router, which I don't think anybody now
4883 uses, but I've fixed it anyway.
4885 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4886 well as to direct DNS lookups. Otherwise the handling of names in host
4887 lists is inconsistent and therefore confusing.
4893 PH/01 Two changes to the default runtime configuration:
4895 (1) Move the checks for relay_from_hosts and authenticated clients from
4896 after to before the (commented out) DNS black list checks.
4898 (2) Add control=submission to the relay_from_hosts and authenticated
4899 clients checks, on the grounds that messages accepted by these
4900 statements are most likely to be submissions.
4902 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4904 (1) Generate an error if the third argument for the ${prvs expansion is
4907 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4910 (3) Reset the variables that are obtained from the first argument of
4911 ${prvscheck and used in the second argument before leaving the code,
4912 because their memory is reclaimed, so using them afterwards may do
4915 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4916 one (it's much easier than Tom thought :-).
4918 (5) Because of (4), we can now allow for the use of $prvscheck_result
4919 inside the third argument.
4921 PH/03 For some reason, the default setting of PATH when running a command from
4922 a pipe transport was just "/usr/bin". I have changed it to
4925 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4926 anything to be listed in the output from -bV.
4928 PH/05 When a filter generated an autoreply, the entire To: header line was
4929 quoted in the delivery log line, like this:
4931 => >A.N.Other <ano@some.domain> <original@ddress> ...
4933 This has been changed so that it extracts the operative address. There
4934 may be more than one such address. If so, they are comma-separated, like
4937 => >ano@some.domain,ona@other.domain <original@ddress> ...
4939 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4940 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4941 not being looked up in the rDNS to get a host name, Exim was showing the
4942 IP address twice in Received: lines, even though the IP addresses were
4943 identical. For example:
4945 Received: from [1.2.3.4] (helo=[1.2.3.4])
4947 However, if the real host name was known, it was omitting the HELO data
4948 if it matched the actual IP address. This has been tidied up so that it
4949 doesn't show the same IP address twice.
4951 PH/07 When both +timestamp and +memory debugging was on, the value given by
4952 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4953 called by the debug printing, thereby overwriting the timestamp buffer.
4954 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4956 PH/08 When the original message was included in an autoreply transport, it
4957 always said "this is a copy of the message, including all the headers",
4958 even if body_only or headers_only was set. It now gives an appropriate
4961 PH/09 Applied a patch from the Sieve maintainer which:
4963 o fixes some comments
4964 o adds the (disabled) notify extension core
4965 o adds some debug output for the result of if/elsif tests
4966 o points to the current vacation draft in the documentation
4967 and documents the missing references header update
4971 o fixes a bug in processing the envelope test (when testing
4972 multiple envelope elements, the last element determined the
4975 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4976 Electronic Mail") by including:
4978 Auto-submitted: auto-generated
4980 in the messages that it generates (bounce messages and others, such as
4981 warnings). In the case of bounce messages for non-SMTP messages, there was
4982 also a typo: it was using "Auto_submitted" (underscore instead of
4983 hyphen). Since every message generated by Exim is necessarily in response
4984 to another message, thes have all been changed to:
4986 Auto-Submitted: auto-replied
4988 in accordance with these statements in the RFC:
4990 The auto-replied keyword:
4992 - SHOULD be used on messages sent in direct response to another
4993 message by an automatic process,
4995 - MUST NOT be used on manually-generated messages,
4997 - MAY be used on Delivery Status Notifications (DSNs) and Message
4998 Disposition Notifications (MDNs),
5000 - MUST NOT be used on messages generated by automatic or periodic
5001 processes, except for messages which are automatic responses to
5004 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5005 to the default Received: header definition.
5007 PH/12 Added log selector acl_warn_skipped (default on).
5009 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5010 variables because (a) they are in the wrong storage pool and (b) even if
5011 they were copied, it wouldn't work properly because of the caching.
5013 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5014 checking when decoding. Apparently there are clients that generate
5015 overlong encoded strings. Why am I not surprised?
5017 PH/15 If the first argument of "${if match_address" was not empty, but did not
5018 contain an "@" character, Exim crashed. Now it writes a panic log message
5019 and treats the condition as false.
5021 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5023 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5024 extension "envelope-auth". The code is finished and in agreement with
5025 other implementations, but there is no documentation so far and in fact,
5026 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5027 not changing the active code.
5029 Print executed "if" and "elsif" statements when debugging is used. This
5030 helps a great deal to understand what a filter does.
5032 Document more things not specified clearly in RFC3028. I had all this
5033 sorted out, when out of a sudden new issues came to my mind. Oops."
5035 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5038 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5039 canonical form (as documented). However, after a host name lookup from
5040 the IP address, check_host() was doing a simple string comparison with
5041 addresses acquired from the DNS when checking that the found name did
5042 have the original IP as one of its addresses. Since any found IPv6
5043 addresses are likely to be in abbreviated form, the comparison could
5044 fail. Luckily, there already exists a function for doing the comparison
5045 by converting both addresses to binary, so now that is used instead of
5046 the text comparison.
5048 PH/20 There was another similar case to PH/19, when a complete host name was
5049 given in a host list; looking up its IP address could give an abbreviated
5050 form, whereas the current host's name might or might not be abbreviated.
5051 The same fix has been applied.
5057 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5058 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5061 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5062 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5064 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5066 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5067 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5068 Also, for "Auto-Submitted", treat anything other than "no" as
5069 non-personal, in accordance with RFC 3834. (Previously it treated
5070 anything starting "auto-" as non-personal.)
5072 TF/01 The control=submission/name=... option had a problem with syntax
5073 errors if the name included a slash character. The /name= option
5074 now slurps the rest of the string, so it can include any characters
5075 but it must come last in the list of options (after /sender_retain
5078 PH/05 Some modifications to the interface to the fake nameserver for the new
5086 TK/01 Added the "success_on_redirect" address verification option. See
5087 NewStuff for rationale and an example.
5089 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5091 PH/02 Patch to exigrep to allow it to work on syslog lines.
5093 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5094 fread() to skip over the body file's header line, because in Cygwin the
5095 header line is locked and is inaccessible.
5097 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5098 co-exist for some time) to make it clear that it is the Exim ID that is
5099 referenced, not the Message-ID: header line.
5101 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5102 string_format() function, because snprintf() does not exist on all
5105 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5106 ${stat: expansion item.
5108 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5109 protocol synchronization error", to keep the pedants happy.
5111 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5112 well as for IRIX systems, when gcc is being used. See the host.c source
5115 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5117 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5120 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5121 transport and to the smtp transport in LMTP mode.
5123 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5125 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5126 run a filter in a subprocess. This could lead to confusion in subsequent
5127 lookups in the parent process. There should also be a search_tidyup() at
5128 the end of the subprocess.
5130 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5131 only if the host matched helo_try_verify_hosts, which caused the
5132 verification to occur when the EHLO/HELO command was issued. The ACL just
5133 tested the remembered result. Now, if a previous verification attempt has
5134 not happened, "verify = helo" does it there and then.
5136 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5138 TK/03 Fix log output including CR from clamd.
5140 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5141 exist provoked a memory error which could cause a segfault.
5143 PH/15 Installed PCRE 6.2
5145 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5147 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5148 of the problem. Specifically, suggested +O2 rather than +O1 for the
5151 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5153 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5154 record, the port was not correctly logged when the outgoing_port log
5155 selector was set (it logged the transort's default port).
5157 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5158 fallback_hosts, and "hosts" in the smtp transport.
5160 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5161 host errors such as "Connection refused".
5163 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5164 authentication with radiusclient 0.4.9:
5166 - Error returned from rc_read_config was caught wrongly
5167 - Username/password not passed on to radius server due to wrong length.
5169 The presumption is that some radiusclient API changes for 4.51/PH/17
5170 were not taken care of correctly. The code is still untested by me (my
5171 Linux distribution still has 0.3.2 of radiusclient), but it was
5172 contributed by a Radius user.
5174 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5175 expanding the "port" option of the smtp transport.
5177 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5178 while reading a MIME header. Thanks to Tom Hughes for a patch.
5180 PH/24 Include config.h inside local_scan.h so that configuration settings are
5183 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5184 that in an ACL for RCPT (for example), you can examine exactly what was
5187 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5188 commands, but it was not correctly comparing the address with the actual
5189 client host address. Thus, it would show the EHLO address in Received:
5190 header lines when this was not necessary.
5192 PH/27 Added the % operator to ${eval:}.
5194 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5195 it should be ignoring failures (because with -C, for example, it has lost
5196 privilege). It wasn't ignoring creation failures other than "already
5199 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5202 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5203 because an input line was too long, either on its own, or by virtue of
5204 too many continuations, the temporary file was not being removed, and the
5205 return code was incorrect.
5207 PH/31 Missing "BOOL" in function definition in filtertest.c.
5209 PH/32 Applied Sieve patches from the maintainer.
5211 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5213 PH/33 Added "verify = not_blind".
5215 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5216 Local/Makefile (with some defaults set). These are used in built scripts
5217 such as exicyclog, but they have never been used in the exim_install
5218 script (though there are many overriding facilities there). I have
5219 arranged that the exim_install script now takes note of these two
5222 PH/35 Installed configuration files for Dragonfly.
5224 PH/36 When a locally submitted message by a trusted user did not contain a
5225 From: header, and the sender address was obtained from -f or from an SMTP
5226 MAIL command, and the trusted user did not use -F to supply a sender
5227 name, $originator_name was incorrectly used when constructing a From:
5228 header. Furthermore, $originator_name was used for submission mode
5229 messages from external hosts without From: headers in a similar way,
5230 which is clearly wrong.
5232 PH/37 Added control=suppress_local_fixups.
5234 PH/38 When log_selector = +received_sender was set, and the addition of the
5235 sender made the log line's construction buffer exactly full, or one byte
5236 less than full, an overflow happened when the terminating "\n" was
5239 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5240 when the result of a list match is failure because a DNS lookup failed.
5242 PH/40 RM_COMMAND is now used in the building process.
5244 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5245 the "build-* directories that it finds.
5247 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5248 address in a domain literal was a prefix of an interface address.
5250 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5251 when verifying a sender address, unless rewrite_headers is false.
5253 PH/44 Wrote a long comment about why errors_to addresses are verified as
5254 recipients, not senders.
5256 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5257 the ratelimit ACL was added.
5259 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5261 PH/46 Added extra information about PostgreSQL errors to the error string.
5263 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5264 avoiding the need to install special zones in a real server. This is
5265 backwards compatible; if it can't find the fake resolver, it drops back.
5266 Thus, both old and new test suites can be run.
5268 TF/02 Added util/ratelimit.pl
5270 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5271 clock is set back in time.
5273 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5274 Candler <B.Candler@pobox.com>.
5276 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5277 OK for addresses that are the result of redirections.
5279 PH/48 A number of further additions for the benefit of the new test suite,
5280 including a fake gethostbyname() that interfaces to the fake DNS resolver
5283 TF/06 The fix for widen_domains has also been applied to qualify_single and
5284 search_parents which are the other dnslookup options that can cause
5287 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5290 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5292 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5293 service" if it encountered an MX record with an empty target hostname.
5294 The message is now "an MX or SRV record indicated no SMTP service".
5296 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5297 if the DNS of the sending site is misconfigured. This is quite a
5298 common situation. This change restores the behaviour of treating a
5299 helo verification defer as a failure.
5301 PH/51 If self=fail was set on a router, the bounce message did not include the
5302 actual error message.
5308 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5310 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5311 command itself ended in error, the underlying message about the transport
5312 filter timeout was being overwritten with the pipe command error. Now the
5313 underlying error message should be appended to the second error message.
5315 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5317 PH/02 Reduce the amount of output that "make" produces by default. Full output
5318 can still be requested.
5320 PH/03 The warning log line about a condition test deferring for a "warn" verb
5321 was being output only once per connection, rather than after each
5322 occurrence (because it was using the same function as for successful
5323 "warn" verbs). This seems wrong, so I have changed it.
5325 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5326 it should not have, which might have caused a crash in the right
5327 circumstances, but probably never did.
5329 PH/04 Installed a modified version of Tony Finch's patch to make submission
5330 mode fix the return path as well as the Sender: header line, and to
5331 add a /name= option so that you can make the user's friendly name appear
5334 TF/03 Added the control = fakedefer ACL modifier.
5336 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5337 Mark Lowes for thorough testing.
5339 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5341 TK/03 Merged latest SRS patch from Miles Wilton.
5343 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5344 with the definition in sysexits.h (which is #included earlier).
5345 Fortunately, Exim does not actually use EX_OK. The code used to try to
5346 preserve the sysexits.h value, by assuming that macro definitions were
5347 scanned for macro replacements. I have been disabused of this notion,
5348 so now the code just undefines EX_OK before #including unistd.h.
5350 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5351 in the smtp transport. When a block could not be written in a single
5352 write() function, the timeout was being re-applied to each part-write.
5353 This seems wrong - if the receiver was accepting one byte at a time it
5354 would take for ever. The timeout is now adjusted when this happens. It
5355 doesn't have to be particularly precise.
5357 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5358 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5360 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5361 without a domain if the check to <postmaster@domain> fails.
5363 SC/01 Eximstats: added -xls and the ability to specify output files
5364 (patch written by Frank Heydlauf).
5366 SC/02 Eximstats: use FileHandles for outputting results.
5368 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5370 SC/04 Eximstats: fixed display of large numbers with -nvr option
5372 SC/05 Eximstats: fixed merging of reports with empty tables.
5374 SC/06 Eximstats: added the -include_original_destination flag
5376 SC/07 Eximstats: removed tabs and trailing whitespace.
5378 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5380 TK/06 MBOX spool code: Add real "From " MBOX separator line
5381 so the .eml file is really in mbox format (even though
5382 most programs do not really care). Patch from Alex Miller.
5384 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5385 The latter is generated from $received_to and is only set if the
5386 message has one envelope recipient. SA can use these headers,
5387 obviously out-of-the-box. Patch from Alex Miller.
5389 PH/08 The ${def test on a variable was returning false if the variable's
5390 value was "0", contrary to what the specification has always said!
5391 The result should be true unless the variable is empty.
5393 PH/09 The syntax error of a character other than { following "${if
5394 def:variable_name" (after optional whitespace) was not being diagnosed.
5395 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5396 accidental colon was present, for example, could give incorrect results.
5398 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5399 result is used (not including appendfile, where other changes are about
5402 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5403 This involved changing a lot of size variables from int to off_t. It
5404 should work with maildirs and everything.
5406 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5407 spamd dying while we are connected to it.
5409 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5412 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5413 function for BDB 4.3.
5415 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5417 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5418 the sources for PCRE have been re-arranged and more files are now
5421 PH/15 The code I had for printing potentially long long variables in PH/11
5422 above was not the best (it lost precision). The length of off_t variables
5423 is now inspected at build time, and an appropriate printing format (%ld
5424 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5425 to be "long long int" or "long int". This is needed for the internal
5426 formatting function string_vformat().
5428 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5429 the configuration file to be ":syslog", then the script "guesses" where
5430 the logs files are, rather than using the compiled in default. In our
5431 case the guess is not the same as the compiled default, so the script
5432 suddenly stopped working when I started to use syslog. The patch checks
5433 to see if log_file_path is "". If so, it attempts to read it from exim
5434 with no configuration file to get the compiled in version, before it
5435 falls back to the previous guessing code."
5437 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5438 implementing BATV in an Exim configuration. See NewStuff for the gory
5441 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5442 Makefile that are specific to HP-UX.
5444 PH/18 If the "use_postmaster" option was set for a recipient callout together
5445 with the "random" option, the postmaster address was used as the MAIL
5446 FROM address for the random test, but not for the subsequent recipient
5447 test. It is now used for both.
5449 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5450 patch removes a few documentation additions to RFC 3028, because the
5451 latest draft now contains them. It adds the new en;ascii-case comparator
5452 and a new error check for 8bit text in MIME parts. Comparator and
5453 require names are now matched exactly. I enabled the subaddress
5454 extension, but it is not well tested yet (read: it works for me)."
5456 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5457 rework some of the code of TK/09 above to avoid the hardwired use of
5458 "%lld" and "long long". Replaced the call to snprintf() with a call to
5461 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5462 records point to non-existent hosts", "retry timeout exceeded", and
5463 "retry time not reached for any host after a long failure period".
5465 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5466 experimental DomainKeys support:
5468 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5469 (2) On an error such as an illegally used "control", the wrong name for
5470 the control was given.
5472 These problems did NOT occur unless DomainKeys support was compiled.
5474 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5476 PH/24 Added ${if match_ip condition.
5478 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5479 databases so that it will be absolutely obvious if a crash occurs in the
5480 DB library. This is a regular occurrence (often caused by mis-matched
5483 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5484 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5485 picked up on a user's system that detects such things. There doesn't seem
5486 to be a gcc warning option for this - only an attribute that has to be
5487 put on the function's prototype. It seems that in Fedora Core 4 they have
5488 set this on a number of new functions. No doubt there will be more in due
5491 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5492 specify a transport. However, if an address that was verified by such a
5493 router was the subject of a callout, Exim crashed because it tried to
5494 read the rcpt_include_affixes from the non-existent transport. Now it
5495 just assumes that the setting of that option is false. This bug was
5496 introduced by 4.51/PH/31.
5498 PH/28 Changed -d+all to exclude +memory, because that information is very
5499 rarely of interest, but it makes the output a lot bigger. People tend to
5500 do -d+all out of habit.
5502 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5503 code in os-type was giving problems when libc.so lives in lib64, like on
5506 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5507 aren't the modern standard, and it seems that some systems' include files
5508 don't always have them. Exim was already checking for some of the newer
5509 ones like T_AAAA, and defining it itself. I've added checks for all the
5510 record types that Exim uses.
5512 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5513 not automatically generating a new one, as it is supposed to. This
5514 prevented TLS from working. If the file did exist, but contained invalid
5515 data, a new version was generated, as expected. It was only the case of a
5516 non-existent file that was broken.
5518 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5519 with a change in libdomainkeys > 0.64.
5521 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5522 from DNS. If the selector record carries the flag, it now has
5523 precedence over the domain-wide flag.
5525 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5527 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5528 the use of a port name that isn't defined in /etc/services) occurred, the
5529 message was deferred as in a normal delivery, and thus remained on the
5530 spool, instead of being failed because of the mua_wrapper setting. This
5531 is now fixed, and I tidied up some of the mua_wrapper messages at the
5534 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5535 the messages in a hash of arrays rather than using individual hashes.
5536 This is a bit cleaner and results in dramatic memory savings, albeit
5537 at a slight CPU cost.
5539 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5540 as requested by Marc Sherman.
5542 SC/10 Eximstats: added histograms for user specified patterns as requested
5545 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5547 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5548 fopen() in the content-scanning modules that did not already have it.
5554 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5555 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5557 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5559 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5561 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5562 to test Sieve filters that use "vacation".
5564 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5565 that changes the way the GnuTLS parameters are stored in the cache file.
5566 The new format can be generated externally. For backward compatibility,
5567 if the data in the cache doesn't make sense, Exim assumes it has read an
5568 old-format file, and it generates new data and writes a new file. This
5569 means that you can't go back to an older release without removing the
5572 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5573 work if there are any delivery delays because "one_time" forces the
5574 parent to be marked "delivered", so its unseen clone is never tried
5575 again. For this reason, Exim now forbids the simultaneous setting of
5578 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5579 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5580 entry if you want to know the details. The fix, however, neglected to
5581 consider the case when local delivery batching is involved. The test for
5582 "previously delivered" was not happening when checking to see if an
5583 address could be batched with a previous (undelivered) one; under
5584 certain circumstances this could lead to multiple deliveries to the same
5587 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5588 in its include files, and this causes problems building Exim.
5590 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5591 header_syntax) but Exim was just ignoring anything given after a slash.
5592 In particular, this caused confusion with an attempt to use "verify =
5593 reverse_host_lookup/defer_ok". An error is now given when options are
5594 supplied for verify items that do not have them. (Maybe reverse_host_
5595 lookup should have a defer_ok option, but that's a different point.)
5597 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5598 defined by RFC 821) to 2048, because there were problems with some AUTH
5599 commands, and RFC 1869 says the size should be increased for extended
5600 SMTP commands that take arguments.
5602 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5605 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5606 "unknown" error; now it says that the functionality isn't in the binary.
5608 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5609 an address' error message when a string expansion fails (syntax or
5610 whatever). Otherwise the password may appear in the log. Following change
5611 PH/42 below, there is no longer a chance of it appearing in a bounce
5614 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5616 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5617 of its addresses was ever tried. (Bugzilla bug #2.)
5619 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5620 the result incorrectly in the debug output. (It correctly added a newline
5621 to what was transported.)
5623 TF/01 Added $received_time.
5625 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5626 commented out examples of how to interface to a virus scanner and to
5627 SpamAssassin. Also added commented examples of av_scanner and
5628 spamd_address settings.
5630 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5631 and controls are allowed in which ACLs. There were a couple of minor
5632 errors. Some of the entries in the conditions table (which is a table of
5633 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5634 negation of where the condition IS allowed.
5636 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5638 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5639 header file does not have a version number, so I've had to invent a new
5640 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5641 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5642 radiusclient), but it was contributed by a Radius user.
5644 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5645 files or directories, for OpenSSL.
5647 PH/19 When an Exim process that is running as root has to create an Exim log
5648 file, it does so in a subprocess that runs as exim:exim so as to get the
5649 ownership right at creation (otherwise, other Exim processes might see
5650 the file with the wrong ownership). There was no test for failure of this
5651 fork() call, which would lead to the process getting stuck as it waited
5652 for a non-existent subprocess. Forks do occasionally fail when resources
5653 run out. I reviewed all the other calls to fork(); they all seem to check
5656 PH/20 When checking for unexpected SMTP input at connect time (before writing
5657 the banner), Exim was not dealing correctly with a non-positive return
5658 from the read() function. If the client had disconnected by this time,
5659 the result was a log entry for a synchronization error with an empty
5660 string after "input=" when read() returned zero. If read() returned -1
5661 (an event I could not check), uninitialized data bytes were printed.
5662 There were reports of junk text (parts of files, etc) appearing after
5665 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5667 PH/22 Added support for macro redefinition, and (re)definition in between
5668 driver and ACL definitions.
5670 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5671 forgetting to use the resulting value; it was using the unexpanded value.
5673 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5674 hadn't been configured. The fix is from Juergen Kreileder, who
5675 understands it better than I do:
5677 "Here's what I see happening with three configured cyrus_sasl
5678 authenticators configured (plain, login, cram-md5):
5680 On startup auth_cyrus_sasl_init() gets called for each of these.
5681 This means three calls to sasl_listmech() without a specified mech_list.
5682 => SASL tests which mechs of all available mechs actually work
5683 => three warnings about OTP not working
5684 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5686 With the patch, sasl_listmech() also gets called three times. But now
5687 SASL's mech_list option is set to the server_mech specified in the the
5688 authenticator. Or in other words, the answer from sasl_listmech()
5689 gets limited to just the mech you're testing for (which is different
5691 => the return list contains just 'plain' or 'login', 'cram-md5' or
5692 nothing depending on the value of ob->server_mech.
5694 I've just tested the patch: Authentication still works fine,
5695 unavailable mechs specified in the exim configuration are still
5696 caught, and the auth.log warnings about OTP are gone."
5698 PH/25 When debugging is enabled, the contents of the command line are added
5699 to the debugging output, even when log_selector=+arguments is not
5702 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5703 answer is "GNU", and only if the return is "GNU/something" is the answer
5706 PH/27 $acl_verify_message is now set immediately after the failure of a
5707 verification in an ACL, and so is available in subsequent modifiers. In
5708 particular, the message can be preserved by coding like this:
5710 warn !verify = sender
5711 set acl_m0 = $acl_verify_message
5713 Previously, $acl_verify_message was set only while expanding "message"
5714 and "log_message" when a very denied access.
5716 PH/28 Modified OS/os.c-Linux with
5718 -#ifndef OS_LOAD_AVERAGE
5719 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5721 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5722 nomenclature these days.)
5724 PH/29 Installed patch from the Sieve maintainer that adds the options
5725 sieve_useraddress and sieve_subaddress to the redirect router.
5727 PH/30 In these circumstances:
5728 . Two addresses routed to the same list of hosts;
5729 . First host does not offer TLS;
5730 . First host accepts first address;
5731 . First host gives temporary error to second address;
5732 . Second host offers TLS and a TLS session is established;
5733 . Second host accepts second address.
5734 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5735 and peerdn, if requested) that were in fact used only for the second
5738 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5739 attention to any local part prefix or suffix that was matched by the
5740 router that accepted the address. It now behaves in the same way as it
5741 does for delivery: the affixes are removed from the local part unless
5742 rcpt_include_affixes is set on the transport.
5744 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5745 timeout during the DATA phase of an incoming message.
5747 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5748 applied a patch sanctioned by the Sieve maintainer.
5750 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5751 the uid or gid is negative. A case of a negative gid caused this to be
5752 noticed. The fix allows for either to be negative.
5754 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5755 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5758 PH/36 The change PH/12 above was broken. Fixed it.
5760 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5761 the grounds that routing the same address twice would always produce the
5762 same answer. This might have been true once, but it is certainly no
5763 longer true now. Routing a child address may depend on the previous
5764 routing that produced that child. Some complicated redirection strategies
5765 went wrong when messages had multiple recipients, and made Exim's
5766 behaviour dependent on the order in which the addresses were given.
5768 I have moved the duplicate checking until after the routing is complete.
5769 Exim scans the addresses that are assigned to local and remote
5770 transports, and removes any duplicates. This means that more work will be
5771 done, as duplicates will always all be routed, but duplicates are
5772 presumably rare, so I don't expect this is of any significance.
5774 For deliveries to pipes, files, and autoreplies, the duplicate checking
5775 still happens during the routing process, since they are not going to be
5778 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5779 It corrects a timeout issue with spamd. This is Ian's comment: "The
5780 background is that sometimes spamd either never reads data from a
5781 connection it has accepted, or it never writes response data. The exiscan
5782 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5783 blindly assumes that writes won't block so it may never time out."
5785 PH/39 Allow G after quota size as well as K and M.
5787 PH/40 The value set for $authenticated_id in an authenticator may not contain
5788 binary zeroes or newlines because the value is written to log lines and
5789 to spool files. There was no check on this. Now the value is run through
5790 the string_printing() function so that such characters are converted to
5791 printable escape sequences.
5793 PH/41 $message_linecount is a new variable that contains the total number of
5794 lines in the message. Compare $body_linecount, which is the count for the
5797 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5798 bounce and delay warning messages, except in certain special cases, which
5801 (a) An SMTP error message from a remote host;
5802 (b) A message specified in a :fail: redirection;
5803 (c) A message specified in a "fail" command in a system filter;
5804 (d) A message specified in a FAIL return from the queryprogram router;
5805 (e) A message specified by the cannot_route_message router option.
5807 In these cases only, Exim does include the error details in bounce and
5808 warning messages. There are also a few cases where bland messages such
5809 as "unrouteable address" or "local delivery error" are given.
5811 PH/43 $value is now also set for the "else" part of a ${run expansion.
5813 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5814 being worked on, but at least Exim now implements the latest version to
5817 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5818 process to complete was treated as a delivery failure, a timeout while
5819 writing the message to the pipe was logged, but erroneously treated as a
5820 successful delivery. Such timeouts include transport filter timeouts. For
5821 consistency with the overall process timeout, these timeouts are now
5822 treated as errors, giving rise to delivery failures by default. However,
5823 there is now a new Boolean option for the pipe transport called
5824 timeout_defer, which, if set TRUE, converts the failures into defers for
5825 both kinds of timeout. A transport filter timeout is now identified in
5828 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5829 systems where "make" and "gmake" are different, calling "gmake" at top
5830 level broke things. I've arranged for the value of $(MAKE) to be passed
5831 from the Makefile to this script so that it can call the same version of
5835 A note about Exim versions 4.44 and 4.50
5836 ----------------------------------------
5838 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5839 changes of various kinds. As a consequence, a big documentation update was
5840 needed. This delayed the release for rather longer than seemed good, especially
5841 in the light of a couple of (minor) security issues. Therefore, the changes
5842 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5843 release. So 4.44 and 4.50 are in effect two different branches that both start
5846 I have left the 4.50 change log unchanged; it contains all the changes since
5847 4.43. The change log for 4.44 is below; many of its items are identical to
5848 those for 4.50. This seems to be the most sensible way to preserve the
5849 historical information.
5855 1. Minor wording change to the doc/README.SIEVE file.
5857 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5858 computation of the current number of files was incorrect.
5860 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5861 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5864 4. Give more explanation in the error message when the command for a transport
5865 filter fails to execute.
5867 5. There are several places where Exim runs a non-Exim command in a
5868 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5869 was being done only for the command run by the queryprogram router. It is
5870 now done for all such subprocesses. The other cases are: ${run, transport
5871 filters, and the commands run by the lmtp and pipe transports.
5873 6. Added CONFIGURE_GROUP build-time option.
5875 7. Some older OS have a limit of 256 on the maximum number of file
5876 descriptors. Exim was using setrlimit() to set 1000 as a large value
5877 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5878 systems. I've change it so that if it can't get 1000, it tries for 256.
5880 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5881 was an oversight, and furthermore, ever since the addition of extra
5882 controls (e.g. 4.43/32), the checks on when to allow different forms of
5883 "control" were broken. There should now be diagnostics for all cases when a
5884 control that does not make sense is encountered.
5886 9. Added the /retain_sender option to "control=submission".
5888 10. $recipients is now available in the predata ACL (oversight).
5890 11. Tidy the search cache before the fork to do a delivery from a message
5891 received from the command line. Otherwise the child will trigger a lookup
5892 failure and thereby defer the delivery if it tries to use (for example) a
5893 cached ldap connection that the parent has called unbind on.
5895 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5896 of $address_data from the recipient verification was clobbered by the
5897 sender verification.
5899 13. The value of address_data from a sender verification is now available in
5900 $sender_address_data in subsequent conditions in the ACL statement.
5902 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5904 15. Added a new option "connect=<time>" to callout options, to set a different
5907 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5908 was its contents. (It was OK if the option was not defined at all.)
5910 17. A "Completed" log line is now written for messages that are removed from
5911 the spool by the -Mrm option.
5913 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5914 information about exactly what failed.
5916 19. Added -dd to debug only the daemon process.
5918 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5919 handles timeouts, both on the server side and network timeouts. Renamed the
5920 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5922 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5923 It is now set to "smtps".
5925 22. $host_address is now set to the target address during the checking of
5926 ignore_target_hosts.
5928 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5929 being passed; this would have caused $sender_host_name to have been used if
5930 matching the list had actually called for a host name (not very likely,
5931 since this list is usually IP addresses). A host name is now passed as
5934 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5935 code that specifies a non-restarting handler (typically sigaction() in
5936 modern systems) in an attempt to fix a rare and obscure crash bug.
5938 25. Narrowed the window for a race in the daemon that could cause it to ignore
5939 SIGCHLD signals. This is not a major problem, because they are used only to
5940 wake it up if nothing else does.
5942 26. A malformed maildirsize file could cause Exim to calculate negative values
5943 for the mailbox size or file count. Odd effects could occur as a result.
5944 The maildirsize information is now recalculated if the size or filecount
5947 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5948 support for a long time. Removed HAVE_SYS_VFS_H.
5950 28. Installed the latest version of exipick from John Jetmore.
5952 29. In an address list, if the pattern was not a regular expression, an empty
5953 subject address (from a bounce message) matched only if the pattern was an
5954 empty string. Non-empty patterns were not even tested. This was the wrong
5955 because it is perfectly reasonable to use an empty address as part of a
5956 database query. An empty address is now tested by patterns that are
5957 lookups. However, all the other forms of pattern expect the subject to
5958 contain a local part and a domain, and therefore, for them, an empty
5959 address still always fails if the pattern is not itself empty.
5961 30. Exim went into a mad DNS loop when attempting to do a callout where the
5962 host was specified on an smtp transport, and looking it up yielded more
5963 than one IP address.
5965 31. Re-factored the code for checking spool and log partition space into a
5966 function that finds that data and another that does the check. The former
5967 is then used to implement four new variables: $spool_space, $log_space,
5968 $spool_inodes, and $log_inodes.
5970 32. The RFC2047 encoding function was originally intended for short strings
5971 such as real names; it was not keeping to the 75-character limit for
5972 encoded words that the RFC imposes. It now respects the limit, and
5973 generates multiple encoded words if necessary. To be on the safe side, I
5974 have increased the buffer size for the ${rfc2047: expansion operator from
5977 33. It is now permitted to omit both strings after an "if" condition; if the
5978 condition is true, the result is "true". As before, when the second string
5979 is omitted, a false condition yields an empty string. This makes it less
5980 cumbersome to write custom ACL and router conditions.
5982 34. Failure to deliver a bounce message always caused it to be frozen, even if
5983 there was an errors_to setting on the router. The errors_to setting is now
5986 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5987 canonical form (fully expanded) before being placed in
5988 $sender_host_address.
5990 36. The table in the code that translates DNS record types into text (T_A to
5991 "A" for instance) was missing entries for NS and CNAME. It is just possible
5992 that this could have caused confusion if both these types were looked up
5993 for the same domain, because the text type is used as part of Exim's
5994 per-process caching. But the chance of anyone hitting this buglet seems
5997 37. The dnsdb lookup has been extended in a number of ways.
5999 (1) There is a new type, "zns", which walks up the domain tree until it
6000 finds some nameserver records. It should be used with care.
6002 (2) There is a new type, "mxh", which is like "mx" except that it returns
6003 just the host names, not the priorities.
6005 (3) It is now possible to give a list of domains (or IP addresses) to be
6006 looked up. The behaviour when one of the lookups defers can be
6007 controlled by a keyword.
6009 (4) It is now possible to specify the separator character for use when
6010 multiple records are returned.
6012 38. The dnslists ACL condition has been extended: it is now possible to supply
6013 a list of IP addresses and/or domains to be looked up in a particular DNS
6016 39. Added log_selector=+queue_time_overall.
6018 40. When running the queue in the test harness, wait just a tad after forking a
6019 delivery process, to get repeatability of debugging output.
6021 41. Include certificate and key file names in error message when GnuTLS fails
6022 to set them up, because the GnuTLS error message doesn't include the name
6023 of the failing file when there is a problem reading it.
6025 42. Allow both -bf and -bF in the same test run.
6027 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6029 44. The "Exiscan patch" is now merged into the mainline Exim source.
6031 45. Sometimes the final signoff response after QUIT could fail to get
6032 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6033 < 0 before doing a fflush(). This bug looks as though it goes back to the
6034 introduction of TLS in release 3.20, but "sometimes" must have been rare
6035 because the tests only now provoked it.
6037 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6038 (this can affect the format of dates).
6040 47. exim_tidydb, when checking for the continued existence of a message for
6041 which it has found a message-specific retry record, was not finding
6042 messages that were in split spool directories. Consequently, it was
6043 deleting retry records that should have stayed in existence.
6045 48. Steve fixed some bugs in eximstats.
6047 49. The SPA authentication driver was not abandoning authentication and moving
6048 on to the next authenticator when an expansion was forced to fail,
6049 contradicting the general specification for all authenticators. Instead it
6050 was generating a temporary error. It now behaves as specified.
6052 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6053 (the order specifies the preference for clients). The order is now AES256,
6054 AES128, 3DES, ARCFOUR128.
6056 51. Small patch to Sieve code - explicitly set From: when generating an
6059 52. Exim crashed if a remote delivery caused a very long error message to be
6060 recorded - for instance if somebody sent an entire SpamAssassin report back
6061 as a large number of 550 error lines. This bug was coincidentally fixed by
6062 increasing the size of one of Exim's internal buffers (big_buffer) that
6063 happened as part of the Exiscan merge. However, to be on the safe side, I
6064 have made the code more robust (and fixed the comments that describe what
6067 53. Now that there can be additional text after "Completed" in log lines (if
6068 the queue_time_overall log selector is set), a one-byte patch to exigrep
6069 was needed to allow it to recognize "Completed" as not the last thing in
6072 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6073 patch that reportedly fixes this has been added. I am not expert enough to
6074 create a test for it. This is what the patch creator wrote:
6076 "I found a little strange behaviour of ldap code when working with
6077 Windows 2003 AD Domain, where users was placed in more than one
6078 Organization Units. When I tried to give exim partial DN, the exit code
6079 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6080 But simultaneously result of request was absolutely normal ldap result,
6081 so I produce this patch..."
6083 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6084 REFERENCE, so I have modified the code to exclude the patch when that macro
6087 55. Some experimental protocols are using DNS PTR records for new purposes. The
6088 keys for these records are domain names, not reversed IP addresses. The
6089 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6090 leaves it alone. Component reversal etc. now happens only for IP addresses.
6093 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6095 57. Double the size of the debug message buffer (to 2048) so that more of very
6096 long debug lines gets shown.
6098 58. The exicyclog utility now does better if the number of log files to keep
6099 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6101 59. Two changes related to the smtp_active_hostname option:
6103 (1) $smtp_active_hostname is now available as a variable.
6104 (2) The default for smtp_banner uses $smtp_active_hostname instead
6105 of $primary_hostname.
6107 60. The host_aton() function is supposed to be passed a string that is known
6108 to be a valid IP address. However, in the case of IPv6 addresses, it was
6109 not checking this. This is a hostage to fortune. Exim now panics and dies
6110 if the condition is not met. A case was found where this could be provoked
6111 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6112 components; fortuitously, this particular loophole had already been fixed
6113 by change 4.50/55 above.
6115 If there are any other similar loopholes, the new check in host_aton()
6116 itself should stop them being exploited. The report I received stated that
6117 data on the command line could provoke the exploit when Exim was running as
6118 exim, but did not say which command line option was involved. All I could
6119 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6120 running as the user.
6123 61. There was a buffer overflow vulnerability in the SPA authentication code
6124 (which came originally from the Samba project). I have added a test to the
6125 spa_base64_to_bits() function which I hope fixes it.
6128 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6129 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6131 63. The daemon start-up calls getloadavg() while still root for those OS that
6132 need the first call to be done as root, but it missed one case: when
6133 deliver_queue_load_max is set with deliver_drop_privilege. This is
6134 necessary for the benefit of the queue runner, because there is no re-exec
6135 when deliver_drop_privilege is set.
6137 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6138 This has been fixed.
6140 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6141 was in use, was not putting the data itself into the right store pool;
6142 consequently, it could be overwritten for a subsequent message in the same
6143 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6146 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6148 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6149 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6150 the function were treating the return as a boolean value, which happened to
6151 work because 0=false and not-0=true, but is not correct code.
6153 68. The host_aton() function was not handling scoped IPv6 addresses (those
6154 with, for example, "%eth0" on the end) correctly.
6156 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6157 negated items (that is, ~something) in unsigned ints. Some compilers
6158 apparently mutter when there is no cast.
6160 70. If an address verification called from an ACL failed, and did not produce a
6161 user-specific message (i.e. there was only a "system" message), nothing was
6162 put in $acl_verify_message. In this situation, it now puts the system
6165 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6166 session; change 4.31/43 added the unwanted input to the log line - except
6167 that it did not do this in the start of session case. It now does.
6169 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6170 This is wrong and can cause the other end to generate a synchronization
6171 error if it is another Exim or anything else that does the synchronization
6172 check. A QUIT command is no longer sent after a timeout.
6174 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6175 during host lookups.
6177 74. The defer_ok option of callout verification was not working if it was used
6178 when verifying addresses in header lines, that is, for this case:
6180 verify = header_sender/callout=defer_ok
6182 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6183 those file descriptors could be used for SMTP connections. If anything
6184 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6185 could be sent to the SMTP client, causing chaos. The daemon now opens
6186 stdin, stdout, and stderr to /dev/null when it puts itself into the
6189 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6190 log by default. The user can override this with suitable Perl magic.
6192 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6193 the log message when discard triggers, was not working for the DATA ACL or
6194 for the non-SMTP ACL.
6196 78. Error message wording change in sieve.c.
6198 79. If smtp_accept_max_per_host was set, the number of connections could be
6199 restricted to fewer than expected, because the daemon was trying to set up
6200 a new connection before checking whether the processes handling previous
6201 connections had finished. The check for completed processes is now done
6202 earlier. On busy systems, this bug wouldn't be noticed because something
6203 else would have woken the daemon, and it would have reaped the completed
6206 80. If a message was submitted locally by a user whose login name contained one
6207 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6208 It caused a spool format error. I have fixed the spool reading code. A
6209 related problem was that the "from" clause in the Received: line became
6210 illegal because of the space(s). It is now covered by ${quote_local_part.
6212 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6215 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6217 83. Patch from Sieve maintainer to fix unterminated string problem in
6218 "vacation" handling.
6220 84. Some minor changes to the Linux configuration files to help with other
6221 OS variants using glibc.
6223 85. One more patch for Sieve to update vacation handling to latest spec.
6226 ----------------------------------------------------
6227 See the note above about the 4.44 and 4.50 releases.
6228 ----------------------------------------------------
6234 1. Change 4.43/35 introduced a bug that caused file counts to be
6235 incorrectly computed when quota_filecount was set in an appendfile
6238 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6239 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6242 3. Give more explanation in the error message when the command for a transport
6243 filter fails to execute.
6245 4. There are several places where Exim runs a non-Exim command in a
6246 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6247 was being done only for the command run by the queryprogram router. It is
6248 now done for all such subprocesses. The other cases are: ${run, transport
6249 filters, and the commands run by the lmtp and pipe transports.
6251 5. Some older OS have a limit of 256 on the maximum number of file
6252 descriptors. Exim was using setrlimit() to set 1000 as a large value
6253 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6254 systems. I've change it so that if it can't get 1000, it tries for 256.
6256 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6257 was an oversight, and furthermore, ever since the addition of extra
6258 controls (e.g. 4.43/32), the checks on when to allow different forms of
6259 "control" were broken. There should now be diagnostics for all cases when a
6260 control that does not make sense is encountered.
6262 7. $recipients is now available in the predata ACL (oversight).
6264 8. Tidy the search cache before the fork to do a delivery from a message
6265 received from the command line. Otherwise the child will trigger a lookup
6266 failure and thereby defer the delivery if it tries to use (for example) a
6267 cached ldap connection that the parent has called unbind on.
6269 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6270 of $address_data from the recipient verification was clobbered by the
6271 sender verification.
6273 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6274 was its contents. (It was OK if the option was not defined at all.)
6276 11. A "Completed" log line is now written for messages that are removed from
6277 the spool by the -Mrm option.
6279 12. $host_address is now set to the target address during the checking of
6280 ignore_target_hosts.
6282 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6283 being passed; this would have caused $sender_host_name to have been used if
6284 matching the list had actually called for a host name (not very likely,
6285 since this list is usually IP addresses). A host name is now passed as
6288 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6289 code that specifies a non-restarting handler (typically sigaction() in
6290 modern systems) in an attempt to fix a rare and obscure crash bug.
6292 15. Narrowed the window for a race in the daemon that could cause it to ignore
6293 SIGCHLD signals. This is not a major problem, because they are used only to
6294 wake it up if nothing else does.
6296 16. A malformed maildirsize file could cause Exim to calculate negative values
6297 for the mailbox size or file count. Odd effects could occur as a result.
6298 The maildirsize information is now recalculated if the size or filecount
6301 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6302 support for a long time. Removed HAVE_SYS_VFS_H.
6304 18. Updated exipick to current release from John Jetmore.
6306 19. Allow an empty sender to be matched against a lookup in an address list.
6307 Previously the only cases considered were a regular expression, or an
6310 20. Exim went into a mad DNS lookup loop when doing a callout where the
6311 host was specified on the transport, if the DNS lookup yielded more than
6314 21. The RFC2047 encoding function was originally intended for short strings
6315 such as real names; it was not keeping to the 75-character limit for
6316 encoded words that the RFC imposes. It now respects the limit, and
6317 generates multiple encoded words if necessary. To be on the safe side, I
6318 have increased the buffer size for the ${rfc2047: expansion operator from
6321 22. Failure to deliver a bounce message always caused it to be frozen, even if
6322 there was an errors_to setting on the router. The errors_to setting is now
6325 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6326 canonical form (fully expanded) before being placed in
6327 $sender_host_address.
6329 24. Updated eximstats to version 1.33
6331 25. Include certificate and key file names in error message when GnuTLS fails
6332 to set them up, because the GnuTLS error message doesn't include the name
6333 of the failing file when there is a problem reading it.
6335 26. Expand error message when OpenSSL has problems setting up cert/key files.
6338 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6339 (this can affect the format of dates).
6341 28. exim_tidydb, when checking for the continued existence of a message for
6342 which it has found a message-specific retry record, was not finding
6343 messages that were in split spool directories. Consequently, it was
6344 deleting retry records that should have stayed in existence.
6346 29. eximstats updated to version 1.35
6347 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6348 1.35 - bugfix such that pie charts by volume are generated correctly
6350 30. The SPA authentication driver was not abandoning authentication and moving
6351 on to the next authenticator when an expansion was forced to fail,
6352 contradicting the general specification for all authenticators. Instead it
6353 was generating a temporary error. It now behaves as specified.
6355 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6356 (the order specifies the preference for clients). The order is now AES256,
6357 AES128, 3DES, ARCFOUR128.
6359 31. Small patch to Sieve code - explicitly set From: when generating an
6362 32. Exim crashed if a remote delivery caused a very long error message to be
6363 recorded - for instance if somebody sent an entire SpamAssassin report back
6364 as a large number of 550 error lines. This bug was coincidentally fixed by
6365 increasing the size of one of Exim's internal buffers (big_buffer) that
6366 happened as part of the Exiscan merge. However, to be on the safe side, I
6367 have made the code more robust (and fixed the comments that describe what
6370 33. Some experimental protocols are using DNS PTR records for new purposes. The
6371 keys for these records are domain names, not reversed IP addresses. The
6372 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6373 leaves it alone. Component reversal etc. now happens only for IP addresses.
6376 34. The host_aton() function is supposed to be passed a string that is known
6377 to be a valid IP address. However, in the case of IPv6 addresses, it was
6378 not checking this. This is a hostage to fortune. Exim now panics and dies
6379 if the condition is not met. A case was found where this could be provoked
6380 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6381 components; fortuitously, this particular loophole had already been fixed
6382 by change 4.50/55 or 4.44/33 above.
6384 If there are any other similar loopholes, the new check in host_aton()
6385 itself should stop them being exploited. The report I received stated that
6386 data on the command line could provoke the exploit when Exim was running as
6387 exim, but did not say which command line option was involved. All I could
6388 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6389 running as the user.
6392 35. There was a buffer overflow vulnerability in the SPA authentication code
6393 (which came originally from the Samba project). I have added a test to the
6394 spa_base64_to_bits() function which I hope fixes it.
6397 36. The daemon start-up calls getloadavg() while still root for those OS that
6398 need the first call to be done as root, but it missed one case: when
6399 deliver_queue_load_max is set with deliver_drop_privilege. This is
6400 necessary for the benefit of the queue runner, because there is no re-exec
6401 when deliver_drop_privilege is set.
6403 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6404 was in use, was not putting the data itself into the right store pool;
6405 consequently, it could be overwritten for a subsequent message in the same
6406 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6409 38. Sometimes the final signoff response after QUIT could fail to get
6410 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6411 < 0 before doing a fflush(). This bug looks as though it goes back to the
6412 introduction of TLS in release 3.20, but "sometimes" must have been rare
6413 because the tests only now provoked it.
6419 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6420 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6421 Now it returns an integer. A number of calls were still expecting a T/F
6422 return. Fortuitously, in all cases, the tests worked in OK situations,
6423 which is the norm. However, things would have gone wrong on any write
6424 failures on the smtp file descriptor. This function is used when sending
6425 messages over SMTP and also when doing verify callouts.
6427 2. When Exim is called to do synchronous delivery of a locally submitted
6428 message (the -odf or -odi options), it no longer closes stderr before doing
6431 3. Implemented the mua_wrapper option.
6433 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6435 5. Implemented the functions header_remove(), header_testname(),
6436 header_add_at_position(), and receive_remove_recipient(), and exported them
6439 6. If an ACL "warn" statement specified the addition of headers, Exim already
6440 inserted X-ACL-Warn: at the start if there was no header name. However, it
6441 was not making this test for the second and subsequent header lines if
6442 there were newlines in the string. This meant that an invalid header could
6443 be inserted if Exim was badly configured.
6445 7. Allow an ACL "warn" statement to add header lines at the start or after all
6446 the Received: headers, as well as at the end.
6448 8. Added the rcpt_4xx retry error code.
6450 9. Added postmaster_mailfrom=xxx to callout verification option.
6452 10. Added mailfrom=xxxx to the callout verification option, for verify=
6455 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6456 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6458 12. Inserted some casts to stop certain compilers warning when using pointer
6459 differences as field lengths or precisions in printf-type calls (mostly
6460 affecting debugging statements).
6462 13. Added optional readline() support for -be (dynamically loaded).
6464 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6465 same clock tick as a message's arrival, so that its received time was the
6466 same as the "first fail" time on the retry record, and that message
6467 remained on the queue past the ultimate address timeout, every queue runner
6468 would try a delivery (because it was past the ultimate address timeout) but
6469 after another failure, the ultimate address timeout, which should have then
6470 bounced the address, did not kick in. This was a "< instead of <=" error;
6471 in most cases the first failure would have been in the next clock tick
6472 after the received time, and all would be well.
6474 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6475 being recognized when the domain list was tested by the match_domain
6476 condition in an expansion string.
6478 16. Added the ${str2b64: operator.
6480 17. Exim was always calling setrlimit() to set a large limit for the number of
6481 processes, without checking whether the existing limit was already
6482 adequate. (It did check for the limit on file descriptors.) Furthermore,
6483 errors from getrlimit() and setrlimit() were being ignored. Now they are
6484 logged to the main and panic logs, but Exim does carry on, to try to do its
6485 job under whatever limits there are.
6487 18. Imported PCRE 5.0.
6489 19. Trivial typo in log message " temporarily refused connection" (the leading
6492 20. If the log selector return_path_on_delivery was set and an address was
6493 redirected to /dev/null, the delivery process crashed because it assumed
6494 that a return path would always be set for a "successful" delivery. In this
6495 case, the whole delivery is bypassed as an optimization, and therefore no
6498 21. Internal re-arrangement: the function for sending a challenge and reading
6499 a response while authentication was assuming a zero-terminated challenge
6500 string. It's now changed to take a pointer and a length, to allow for
6501 binary data in such strings.
6503 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6505 23. Exim was not respecting finduser_retries when seeking the login of the
6506 uid under which it was called; it was always trying 10 times. (The default
6507 setting of finduser_retries is zero.) Also, it was sleeping after the final
6508 failure, which is pointless.
6510 24. Implemented tls_on_connect_ports.
6512 25. Implemented acl_smtp_predata.
6514 26. If the domain in control=submission is set empty, Exim assumes that the
6515 authenticated id is a complete email address when it generates From: or
6516 Sender: header lines.
6518 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6519 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6520 chown and chgrp in /bin and hostname in /usr/bin.
6522 28. Exim was keeping the "process log" file open after each use, just as it
6523 does for the main log. This opens the possibility of it remaining open for
6524 long periods when the USR1 signal hits a daemon. Occasional processlog
6525 errors were reported, that could have been caused by this. Anyway, it seems
6526 much more sensible not to leave this file open at all, so that is what now
6529 29. The long-running daemon process does not normally write to the log once it
6530 has entered its main loop, and it closes the log before doing so. This is
6531 so that log files can straightforwardly be renamed and moved. However,
6532 there are a couple of unusual error situations where the daemon does write
6533 log entries, and I had neglected to close the log afterwards.
6535 30. The text of an SMTP error response that was received during a remote
6536 delivery was being truncated at 512 bytes. This is too short for some of
6537 the long messages that one sometimes sees. I've increased the limit to
6540 31. It is now possible to make retry rules that apply only when a message has a
6541 specific sender, in particular, an empty sender.
6543 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6544 it possible to be selective about when SMTP synchronization is enforced.
6546 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6548 32. Implemented hosts_connection_nolog.
6550 33. Added an ACL for QUIT.
6552 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6555 35. Added mailbox_size and mailbox_filecount to appendfile.
6557 36. Added control = no_multiline_responses to ACLs.
6559 37. There was a bug in the logic of the code that waits for the clock to tick
6560 in the case where the clock went backwards by a substantial amount such
6561 that the microsecond fraction of "now" was more than the microsecond
6562 fraction of "then" (but the whole seconds number was less).
6564 38. Added support for the libradius Radius client library this is found on
6565 FreeBSD (previously only the radiusclient library was supported).
6571 1. When certain lookups returned multiple values in the form name=value, the
6572 quoting of the values was not always being done properly. Specifically:
6573 (a) If the value started with a double quote, but contained no whitespace,
6575 (b) If the value contained whitespace other than a space character (i.e.
6576 tabs or newlines or carriage returns) it was not quoted.
6577 This fix has been applied to the mysql and pgsql lookups by writing a
6578 separate quoting function and calling it from the lookup code. The fix
6579 should probably also be applied to nisplus, ibase and oracle lookups, but
6580 since I cannot test any of those, I have not disturbed their existing code.
6582 2. A hit in the callout cache for a specific address caused a log line with no
6583 reason for rejecting RCPT. Now it says "Previous (cached) callout
6584 verification failure".
6586 3. There was an off-by-one bug in the queryprogram router. An over-long
6587 return line was truncated at 256 instead of 255 characters, thereby
6588 overflowing its buffer with the terminating zero. As well as fixing this, I
6589 have increased the buffer size to 1024 (and made a note to document this).
6591 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6592 when Exim is waiting for an SMTP response from a remote server, Exim
6593 restarts its select() call on the socket, thereby resetting its timeout.
6594 This is not a problem when such interrupts are rare. Somebody set up a cron
6595 job to run exiwhat every 2 minutes, which is less than the normal select()
6596 timeout (5 or 10 minutes). This meant that the select() timeout never
6597 kicked in because it was always reset. I have fixed this by comparing the
6598 time when an interrupt arrives with the time at the start of the first call
6599 to select(). If more time than the timeout has elapsed, the interrupt is
6600 treated as a timeout.
6602 5. Some internal re-factoring in preparation for the addition of Sieve
6603 extensions (by MH). In particular, the "personal" test is moved to a
6604 separate function, and given an option for scanning Cc: and Bcc: (which is
6605 not set for Exim filters).
6607 6. When Exim created an email address using the login of the caller as the
6608 local part (e.g. when creating a From: or Sender: header line), it was not
6609 quoting the local part when it contained special characters such as @.
6611 7. Installed new OpenBSD configuration files.
6613 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6614 try to make them clearer.
6616 9. Callout options, other than the timeout value, were being ignored when
6617 verifying sender addresses in header lines. For example, when using
6619 verify = header_sender/callout=no_cache
6621 the cache was (incorrectly) being used.
6623 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6624 only the Cygwin environment.
6626 11. When return_path_on_delivery was set as a log selector, if different remote
6627 addresses in the same message used different return paths and parallel
6628 remote delivery occurred, the wrong values would sometimes be logged.
6629 (Whenever a remote delivery process finished, the return path value from
6630 the most recently started remote delivery process was logged.)
6632 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6633 lines when AUTH and/or TLS are in use. This is the "received protocol"
6634 field. Exim used to use "asmtp" for authenticated SMTP, without any
6635 indication (in the protocol name) for TLS use. Now it follows the RFC and
6636 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6637 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6638 names appear in log lines as well as in Received: header lines.
6640 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6641 extensions, and comparison tests, and to fix some bugs.
6643 14. Changes to the "personal" filter test:
6645 (1) The test was buggy in that it was just doing the equivalent of
6646 "contains" tests on header lines. For example, if a user's address was
6647 anne@some.where, the "personal" test would incorrectly be true for
6649 To: susanne@some.where
6651 This test is now done by extracting each address from the header in turn,
6652 and checking the entire address. Other tests that are part of "personal"
6653 are now done using regular expressions (for example, to check local parts
6654 of addresses in From: header lines).
6656 (2) The list of non-personal local parts in From: addresses has been
6657 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6658 taken from the Sieve specification recommendations.
6660 (3) If the message contains any header line starting with "List-" it is
6661 treated as non-personal.
6663 (4) The test for "circular" in the Subject: header line has been removed
6664 because it now seems ill-conceived.
6666 15. Minor typos in src/EDITME comments corrected.
6668 16. Installed latest exipick from John Jetmore.
6670 17. If headers_add on a router specified a text string that was too long for
6671 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6672 of string_sprintf() is now avoided.
6674 18. $message_body_size was not set (it was always zero) when running the DATA
6675 ACL and the local_scan() function.
6677 19. For the "mail" command in an Exim filter, no default was being set for
6678 the once_repeat time, causing a random time value to be used if "once" was
6679 specified. (If the value happened to be <= 0, no repeat happened.) The
6680 default is now 0s, meaning "never repeat". The "vacation" command was OK
6681 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6682 (I found it when inspecting the code).
6684 20. There is now an overall timeout for performing a callout verification. It
6685 defaults to 4 times the callout timeout, which applies to individual SMTP
6686 commands during the callout. The overall timeout applies when there is more
6687 than one host that can be tried. The timeout is checked before trying the
6688 next host. This prevents very long delays if there are a large number of
6689 hosts and all are timing out (e.g. when the network connections are timing
6690 out). The value of the overall timeout can be changed by specifying an
6691 additional sub-option for "callout", called "maxwait". For example:
6693 verify = sender/callout=5s,maxwait=20s
6695 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6696 to the end before writing, but this should make it even safer).
6698 22. Exim was forgetting that it had advertised PIPELINING for the second and
6699 subsequent messages on an SMTP connection. It was also not resetting its
6700 memory on STARTTLS and an internal HELO.
6702 23. When Exim logs an SMTP synchronization error within a session, it now
6703 records whether PIPELINING has been advertised or not.
6705 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6706 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6707 rather than long int.
6709 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6711 26. Added the never_mail option to autoreply.
6717 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6718 crash if the getsockname() call failed; this can happen if a connection is
6719 closed very soon after it is established. The problem was simply in the
6720 order in which certain operations were done, causing Exim to try to write
6721 to the SMTP stream before it had set up the file descriptor. The bug has
6722 been fixed by making things happen in the correct order.
6728 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6729 before the connection was closed, thus losing the rejection response.
6731 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6732 some early Solaris releases, but causes trouble in current releases where
6733 socklen_t is defined.
6735 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6738 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6741 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6742 that releasing the top of it at the end releases what was used for sub-
6743 expansions (unless the block got too big). However, discard this block if
6744 the first thing is a variable or header, so that we can use its block when
6745 it is dynamic (useful for very large $message_headers, for example).
6747 6. Lookups now cache *every* query, not just the most recent. A new, separate
6748 store pool is used for this. It can be recovered when all lookup caches are
6749 flushed. Lookups now release memory at the end of their result strings.
6750 This has involved some general refactoring of the lookup sources.
6752 7. Some code has been added to the store_xxx() functions to reduce the amount
6753 of flapping under certain conditions.
6755 8. log_incoming_interface used to affect only the <= reception log lines. Now
6756 it causes the local interface and port to be added to several more SMTP log
6757 lines, for example "SMTP connection from", and rejection lines.
6759 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6761 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6763 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6764 could be overwritten at the end of the current message (or the start of a
6765 new message if it was set in a HELO ACL). The value is now preserved for
6766 the duration of the SMTP connection.
6768 12. If a transport had a headers_rewrite setting, and a matching header line
6769 contained an unqualified address, that address was qualified, even if it
6770 did not match any rewriting rules. The underlying bug was that the values
6771 of the flags that permit the existence of unqualified sender and recipient
6772 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6773 non-local messages, and by -bnq for local messages) were not being
6774 preserved with the message after it was received.
6776 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6777 "next input=" as part of the text comprising the host identity instead of
6778 the correct text. The code was using the same buffer for two different
6779 strings. However, depending on which order the printing function evaluated
6780 its arguments, the bug did not always show up. Under Linux, for example, my
6781 test suite worked just fine.
6783 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6784 doesn't work with some older versions of Perl. It has been changed to "my",
6785 which in any case is probably the better facility to use.
6787 15. A really picky compiler found some instances of statements for creating
6788 error messages that either had too many or two few arguments for the format
6791 16. The size of the buffer for calls to the DNS resolver has been increased
6792 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6793 for addresses that have a lot of PTR records. This alleviates a problem; it
6794 does not fully solve it.
6796 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6797 buffer now truncates the list and logs the incident, which is the same
6798 action as happens when Exim is looking up a host name and its aliases.
6799 Previously in this situation something unpredictable would happen;
6800 sometimes it was "internal error: store_reset failed".
6802 18. If a server dropped the connection unexpectedly when an Exim client was
6803 using GnuTLS and trying to read a response, the client delivery process
6804 crashed while trying to generate an error log message.
6806 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6807 string, for example:
6809 warn message = H1: something\nH2: something
6811 the text was added as a single header line from Exim's point of view
6812 though it ended up OK in the delivered message. However, searching for the
6813 second and subsequent header lines using $h_h2: did not work. This has been
6814 fixed. Similarly, if a system filter added multiple headers in this way,
6815 the routers could not see them.
6817 20. Expanded the error message when iplsearch is called with an invalid key to
6818 suggest using net-iplsearch in a host list.
6820 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6821 ACLs are no longer actually imposed (and a message to that effect is
6824 22. If a "gecos" field in a passwd entry contained escaped characters, in
6825 particular, if it contained a \" sequence, Exim got it wrong when building
6826 a From: or a Sender: header from that name. A second bug also caused
6827 incorrect handling when an unquoted " was present following a character
6828 that needed quoting.
6830 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6831 was not being matched caselessly.
6833 24. Arranged for all hyphens in the exim.8 source to be escaped with
6836 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6837 properly. Recipient callout cache records were still being keyed to include
6838 the sender, even when use_sender was set false. This led to far more
6839 callouts that were necessary. The sender is no longer included in the key
6840 when use_sender is false.
6842 26. Added "control = submission" modifier to ACLs.
6844 27. Added the ${base62d: operator to decode base 62 numbers.
6846 28. dnsdb lookups can now access SRV records.
6848 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6849 the configuration file.
6851 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6852 (-v) mode. This makes the output for a verbose queue run more intelligible.
6854 31. Added a use_postmaster feature to recipient callouts.
6856 32. Added the $body_zerocount variable, containing the number of binary zero
6857 bytes in the message body.
6859 33. The time of last modification of the "new" subdirectory is now used as the
6860 "mailbox time last read" when there is a quota error for a maildir
6863 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6865 35. Added +ignore_unknown as a special item in host lists.
6867 36. Code for decoding IPv6 addresses in host lists is now included, even if
6868 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6869 address was recognized as an IP address, but was then not correctly decoded
6870 into binary, causing unexpected and incorrect effects when compared with
6877 1. Very minor rewording of debugging text in manualroute to say "list of
6878 hosts" instead of "hostlist".
6880 2. If verify=header_syntax was set, and a header line with an unqualified
6881 address (no domain) and a large number of spaces between the end of the
6882 name and the colon was received, the reception process suffered a buffer
6883 overflow, and (when I tested it) crashed. This was caused by some obsolete
6884 code that should have been removed. The fix is to remove it!
6886 3. When running in the test harness, delay a bit after writing a bounce
6887 message to get a bit more predictability in the log output.
6889 4. Added a call to search_tidyup() just before forking a reception process. In
6890 theory, someone could use a lookup in the expansion of smtp_accept_max_
6891 per_host which, without the tidyup, could leave open a database connection.
6893 5. Added the variables $recipient_data and $sender_data which get set from a
6894 lookup success in an ACL "recipients" or "senders" condition, or a router
6895 "senders" option, similar to $domain_data and $local_part_data.
6897 6. Moved the writing of debug_print from before to after the "senders" test
6900 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6901 problems for message scanning, either using a data ACL, or using
6902 local_scan() because the Received: header was not generated till after they
6903 were called (in order to set the time as the time of reception completion).
6904 I have revised the way this works. The header is now generated after the
6905 body is received, but before the ACL or local_scan() are called. After they
6906 are run, the timestamp in the header is updated.
6912 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6913 before starting a queue runner without re-exec. This happened only when
6914 deliver_drop_privilege was set or when the Exim user was set to root. The
6915 effect of the bug was that timeouts during subsequent deliveries caused
6916 crashes instead of being properly handled. The handler is now left at its
6917 default (and expected) setting.
6919 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6920 message, again when deliver_drop_privilege is set or Exim is run as root.
6921 The bug described in (1) was not present in this case, but the tidying up
6922 of the other signals was missing. I have made the two cases consistent.
6924 3. The ignore_target_hosts setting on a manualroute router was being ignored
6925 for hosts that were looked up using the /MX notation.
6927 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6930 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6931 operated on the sender address. After changing the $sender_address to <>
6932 for the sender address verify, Exim was re-instated it as the original
6933 (before rewriting) address, but remembering that it had rewritten it, so it
6934 wasn't rewriting it again. This bug also had the effect of breaking the
6935 sender address verification caching when the sender address was rewritten.
6937 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6938 This has been changed so that if the ip literal address matches
6939 ignore_target_hosts, the router declines.
6941 7. Added expansion conditions match_domain, match_address, and match_local_
6942 part (NOT match_host).
6944 8. The placeholder for the Received: header didn't have a length field set.
6946 9. Added code to Exim itself and to exim_lock to test for a specific race
6947 condition that could lead to file corruption when using MBX delivery. The
6948 issue is with the lockfile that is created in /tmp. If this file is removed
6949 after a process has opened it but before that process has acquired a lock,
6950 there is the potential for a second process to recreate the file and also
6951 acquire a lock. This could lead to two Exim processes writing to the file
6952 at the same time. The added code performs the same test as UW imapd; it
6953 checks after acquiring the lock that its file descriptor still refers to
6954 the same named file.
6956 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6957 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6960 11. Added the smtp_active_hostname option. If used, this will typically be made
6961 to depend on the incoming interface address. Because $interface_address is
6962 not set up until the daemon has forked a reception process, error responses
6963 that can happen earlier (such as "too many connections") no longer contain
6966 12. If an expansion in a condition on a "warn" statement fails because a lookup
6967 defers, the "warn" statement is abandoned, and the next ACL statement is
6968 processed. Previously this caused the whole ACL to be aborted.
6970 13. Added the iplsearch lookup type.
6972 14. Added ident_timeout as a log selector.
6974 15. Added tls_certificate_verified as a log selector.
6976 16. Added a global option tls_require_ciphers (compare the smtp transport
6977 option of the same name). This controls incoming TLS connections.
6979 17. I finally figured out how to make tls_require_ciphers do a similar thing
6980 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6981 before starting the TLS session.
6983 18. Tabs are now shown as \t in -bP output.
6985 19. If the log selector return_path_on_delivery was set, Exim crashed when
6986 bouncing a message because it had too many Received: header lines.
6988 20. If two routers both had headers_remove settings, and the first one included
6989 a superfluous trailing colon, the final name in the first list and the
6990 first name in the second list were incorrectly joined into one item (with a
6991 colon in the middle).
6997 1. Added -C and -D options to the exinext utility, mainly to make it easier
6998 to include in the automated testing, but these could be helpful when
6999 multiple configurations are in use.
7001 2. The exinext utility was not formatting the output nicely when there was
7002 an alternate port involved in the retry record key, nor when there was a
7003 message id as well (for retries that were specific to a specific message
7004 and a specific host). It was also confused by IPv6 addresses, because of
7005 the additional colons they contain. I have fixed the IPv4 problem, and
7006 patched it up to do a reasonable job for IPv6.
7008 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7009 delivery, the log line now contains "pipelined" if PIPELINING was used.
7011 4. An SMTP transport process used to panic and die if the bind() call to set
7012 an explicit outgoing interface failed. This has been changed; it is now
7013 treated in the same way as a connect() failure.
7015 5. A reference to $sender_host_name in the part of a conditional expansion
7016 that was being skipped was still causing a DNS lookup. This no longer
7019 6. The def: expansion condition was not recognizing references to header lines
7020 that used bh_ and bheader_.
7022 7. Added the _cache feature to named lists.
7024 8. The code for checking quota_filecount in the appendfile transport was
7025 allowing one more file than it should have been.
7027 9. For compatibility with Sendmail, the command line option
7035 and sets the incoming protocol and host name (for trusted callers). The
7036 host name and its colon can be omitted when only the protocol is to be set.
7037 Note the Exim already has two private options, -pd and -ps, that refer to
7038 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7039 "s", but I don't think that's a major issue.
7041 10. A number of refactoring changes to the code, none of which should affect
7044 (a) The number of logging options was getting close to filling up the
7045 32-bit word that was used as a bit map. I have split them into two classes:
7046 those that are passed in the argument to log_write(), and those that are
7047 only ever tested independently outside of that function. These are now in
7048 separate 32-bit words, so there is plenty of room for expansion again.
7049 There is no change in the user interface or the logging behaviour.
7051 (b) When building, for example, log lines, the code previously used a
7052 macro that called string_cat() twice, in order to add two strings. This is
7053 not really sufficiently general. Furthermore, there was one instance where
7054 it was actually wrong because one of the argument was used twice, and in
7055 one call a function was used. (As it happened, calling the function twice
7056 did not affect the overall behaviour.) The macro has been replaced by a
7057 function that can join an arbitrary number of extra strings onto a growing
7060 (c) The code for expansion conditions now uses a table and a binary chop
7061 instead of a serial search (which was left over from when there were very
7062 few conditions). Also, it now recognizes conditions like "pam" even when
7063 the relevant support is not compiled in: a suitably worded error message is
7064 given if an attempt is made to use such a condition.
7066 11. Added ${time_interval:xxxxx}.
7068 12. A bug was causing one of the ddress fields not to be passed back correctly
7069 from remote delivery subprocesses. The field in question was not being
7070 subsequently used, so this caused to problems in practice.
7072 13. Added new log selectors queue_time and deliver_time.
7074 14. Might have fixed a bug in maildirsizefile handling that threw up
7075 "unexpected character" debug warnings, and recalculated the data
7076 unnecessarily. In any case, I expanded the warning message to give more
7079 15. Added the message "Restricted characters in address" to the statements in
7080 the default ACL that block characters like @ and % in local parts.
7082 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7083 Three changes have been made:
7085 (a) There was a serious bug; a negative response to MAIL caused the whole
7086 recipient domain to be cached as invalid, thereby blocking all messages
7087 to all local parts at the same domain, from all senders. This bug has
7088 been fixed. The domain is no longer cached after a negative response to
7089 MAIL if the sender used is not empty.
7091 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7094 (c) A new callout option, "use_sender" has been added for people who want
7095 the modified behaviour.
7101 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7104 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7105 indeed breaks things for older releases.
7107 3. Added additional logging to the case where there is a problem reading data
7108 from a filter that is running in a subprocess using a pipe, in order to
7109 try to track down a specific problem.
7111 4. Testing facility fudge: when running in the test harness and attempting
7112 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7113 getting "No route to host". Convert this to a timeout.
7115 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7118 6. Some OS don't have socklen_t but use size_t instead. This affects the
7119 fifth argument of getsockopt() amongst other things. This is now
7120 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7121 can be set for individual OS. I have set it for SunOS5, OSF1, and
7122 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7123 some earlier ones do not.
7125 7. Change 4.30/15 was not doing the test caselessly.
7127 8. The standard form for an IPv6 address literal was being rejected by address
7128 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7129 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7130 this, as well as the form without the "IPv6" on the front (but only when
7131 address literals are enabled, of course).
7133 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7135 10. Exim crashed if a message with an empty sender address specified by -f
7136 encountered a router with an errors_to setting. This could be provoked only
7137 by a command such as
7141 where an empty string was supplied; "<>" did not hit this bug.
7143 11. Installed PCRE release 4.5.
7145 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7146 remained set. It is now erased.
7148 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7149 times from message ids (which are base 36 rather than the normal 62).
7151 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7152 were being counted as actual protocol errors, and logged if the log
7153 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7154 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7155 and DATA following a set of rejected RCPTs do not count as protocol errors.
7156 In other words, Exim assumes they were pipelined, though this may not
7157 actually be the case. Of course, in all cases the client gets an
7158 appropriate error code.
7160 15. If a lookup fails in an ACL condition, a message about the failure may
7161 be available; it is used if testing the ACL cannot continue, because most
7162 such messages specify what the cause of the deferral is. However, some
7163 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7164 that caused an old message to be retained and used if a later statement
7165 caused a defer, replacing the real cause of the deferral.
7167 16. If an IP address had so many PTR records that the DNS lookup buffer
7168 was not large enough to hold them, Exim could crash while trying to process
7169 the truncated data. It now detects and logs this case.
7171 17. Further to 4.21/58, another change has been made: if (and only if) the
7172 first line of a message (the first header line) ends with CRLF, a bare LF
7173 in a subsequent header line has a space inserted after it, so as not to
7174 terminate the header.
7176 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7177 unnecessarily, used atoi() instead of strtol(), and didn't check the
7178 termination when getting file sizes from file names by regex.
7180 19. Completely re-implemented the support for maildirsize files, in the light
7181 of a number of problems with the previous contributed implementation
7182 (4.30/29). In particular:
7184 . If the quota is zero, the maildirsize file is maintained, but no quota is
7187 . If the maildir directory does not exist, it is created before any attempt
7188 to write a maildirsize file.
7190 . The quota value in the file is just a cache; if the quota is changed in
7191 the transport, the new value overrides.
7193 . A regular expression is available for excluding directories from the
7196 20. The autoreply transport checks the characters in options that define the
7197 message's headers; it allows continued headers, but it was checking with
7198 isspace() after an embedded newline instead of explicitly looking for a
7201 21. If all the "regular" hosts to which an address was routed had passed their
7202 expiry times, and had not reached their retry times, the address was
7203 bounced, even if fallback hosts were defined. Now Exim should go on to try
7206 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7207 equivalent code in the SMTP transport. Some hosts send humungous responses
7208 to HELO/EHLO, more than 1024 it seems.
7210 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7211 gives compiler warnings in some environments. I've now done it "properly",
7214 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7215 (because of problems with the built-in one) was declared to return uschar *
7216 instead of char *, causing compiler failure.
7218 25. Fixed a file descriptor leak when processing alias/forward files.
7220 26. Fixed a minor format string issue in dbfn.c.
7222 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7224 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7225 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7228 29. When a sender address is verified, it is cached, to save repeating the test
7229 when there is more than one recipient in a message. However, when the
7230 verification involves a callout, it is possible for different callout
7231 options to be set for different recipients. It is too complicated to keep
7232 track of this in the cache, so now Exim always runs a verification when a
7233 callout is required, relying on the callout cache for the optimization.
7234 The overhead is duplication of the address routing, but this should not be
7237 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7238 to be verified with callout=postmaster, and the main callout worked but the
7239 postmaster check failed, the verification correctly failed. However, if a
7240 subsequent RCPT command asked for sender verification *without* the
7241 postmaster check, incorrect caching caused this verification also to fail,
7244 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7245 it was not caching the DNS options (qualify_single, search_parents) that
7246 were used when the lookup failed. A subsequent lookup with different
7247 options therefore always gave the same answer, though there were cases
7248 where it should not have. (Example: a "domains = !$mx_any" option on a
7249 dnslookup router: the "domains" option is always processed without any
7250 widening, but the router might have qualify_single set.) Now Exim uses the
7251 cached value only when the same options are set.
7253 32. Added John Jetmore's "exipick" utility to the distribution.
7255 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7256 than a timeout (e.g. a certificate is required, and is not provided), an
7257 Exim server now closes the connection immediately. Previously it waited for
7258 the client to close - but if the client is SSL, it seems that they each
7259 wait for each other, leading to a delay before one of them times out.
7261 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7262 maintained 0.8.x compatibility because I don't think many are using it, and
7263 it is clearly obsolete.
7265 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7268 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7269 client certificate was expired. A simple patch fixes this, though I don't
7270 understand the full logic of why the verify callback is called multiple
7273 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7274 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7275 which causes problems with some clients (such as the Certicom SSL Plus
7276 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7277 disables the coutermeasure allowing Eudora to connect."
7279 38. Exim was not checking that a write() to a log file succeeded. This could
7280 lead to Bad Things if a log got too big, in particular if it hit a file
7281 size limit. Exim now panics and dies if it cannot write to a log file, just
7282 as it does if it cannot open a log file.
7284 39. Modified OS/Makefile-Linux so that it now contains
7286 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7288 The two -D definitions ensure that Exim is compiled with large file
7289 support, which makes it possible to handle log files that are bigger than
7292 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7293 instance) a domain was checked against a named list that involved a lookup,
7294 causing $domain_data to be set, then another domain was checked against the
7295 same list, then the first domain was re-checked, the value of $domain_data
7296 after the final check could be wrong. In particular, if the second check
7297 failed, it could be set empty. This bug probably also applied to
7300 41. The strip_trailing_dot option was not being applied to the address given
7301 with the -f command-line option.
7303 42. The code for reading a message's header from the spool was incrementing
7304 $received_count, but never initializing it. This meant that the value was
7305 incorrect (doubled) while delivering a message in the same process in which
7306 it was received. In the most common configuration of Exim, this never
7307 happens - a fresh exec is done - but it can happen when
7308 deliver_drop_privilege is set.
7310 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7311 it now includes up to 150 characters of the unexpected data in the log
7314 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7315 and building data strings. The size of both of these buffers was 10 000
7316 bytes - far larger than anybody would *ever* want, thought I. Needless to
7317 say, somebody hit the limit. I have increased the maximum line length to
7318 20 000 and the maximum data length of concatenated lines to 100 000. I have
7319 also fixed two bugs, because there was no checking on these buffers. Tsk,
7320 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7321 buffer is too small.
7323 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7324 lsearch lookups. Now it does.
7326 46. When parsing a route_list item in a manualroute router, a fixed-length
7327 buffer was used for the list of hosts. I made this 1024 bytes long,
7328 thinking that nobody would ever have a list of hosts that long. Wrong.
7329 Somebody had a whole pile of complicated expansion conditions, and the
7330 string was silently truncated, leading to an expansion error. It turns out
7331 that it is easier to change to an unlimited length (owing to other changes
7332 that have happened since this code was originally written) than to build
7333 structure for giving a limitation error. The length of the item that
7334 expands into the list of hosts is now unlimited.
7336 47. The lsearch lookup could not handle data where the length of text line was
7337 more than 4095 characters. Such lines were truncated, leading to shortened
7338 data being returned. It should now handle lines of any length.
7340 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7341 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7344 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7345 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7346 no longer happens in comment lines. A list of replacements is now placed
7347 at the head of all of the source files, except those whose only change is
7348 to replace PERL_COMMAND in the very first #! line.
7350 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7351 messages on the queue, with a bottom-up merge sort, using code contributed
7352 by Michael Haardt. This should make operations like -bp somewhat faster on
7353 large queues. It won't affect queue runners, except when queue_run_in_order
7356 51. Installed eximstats 1.31 in the distribution.
7358 52. Added support for SRV lookups to the dnslookup router.
7360 53. If an ACL referred to $message_body or $message_body_end, the value was not
7361 reset for any messages that followed in the same SMTP session.
7363 54. The store-handling optimization for building very long strings was not
7364 differentiating between the different store pools. I don't think this
7365 actually made any difference in practice, but I've tidied it.
7367 55. While running the routers to verify a sender address, $sender_address
7368 was still set to the sender address. This is wrong, because when routing to
7369 send a bounce to the sender, it would be empty. Therefore, I have changed
7370 it so that, while verifying a sender address, $sender_address is set to <>.
7371 (There is no change to what happens when verifying a recipient address.)
7373 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7374 target A or AAAA records (if not already returned) without resetting the
7375 qualify_single or search_parents options of the DNS resolver. These are
7376 inappropriate in this case because the targets of MX and SRV records must
7377 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7378 target that, when qualified, matched something in the local domain. These
7379 two options are now turned off when doing these lookups.
7381 57. It seems that at least some releases of Reiserfs (which does not have the
7382 concept of a fixed number of inodes) returns zero and not -1 for the
7383 number of available inodes. This interacted badly with check_spool_inodes,
7384 which assumed that -1 was the "no such thing" setting. What I have done is
7385 to check that the total number of inodes is greater than zero before doing
7386 the test of how many are available.
7388 58. When a "warn" ACL statement has a log_message modifier, the message is
7389 remembered, and not repeated. This is to avoid a lot of repetition when a
7390 message has many recipients that cause the same warning to be written.
7391 However, Exim was preserving the list of already written lines for an
7392 entire SMTP session, which doesn't seem right. The memory is now reset if a
7393 new message is started.
7395 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7396 debugging output unless log_rewrite was also set.
7398 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7399 of (int)(handle) when we know that handle contains (void *)(-1).
7401 61. The Exim daemon panic-logs an error return when it closes the incoming
7402 connection. However "connection reset by peer" seems to be common, and
7403 isn't really an error worthy of noting specially, so that particular error
7406 62. When Exim is trying to find all the local interfaces, it used to panic and
7407 die if the ioctl to get the interface flags failed. However, it seems that
7408 on at least one OS (Solaris 9) it is possible to have an interface that is
7409 included in the list of interfaces, but for which you get a failure error
7410 for this call. This happens when the interface is not "plumbed" into a
7411 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7412 failure of the "get flags" call assumes that the interface is down.
7414 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7415 makes life easier for people who are doing arithmetic on fields extracted
7416 from dates, where you often get leading zeros that should not be
7417 interpreted as octal.
7419 64. Added qualify_domain to the redirect router, to override the global
7422 65. If a pathologically long header line contained very many addresses (the
7423 report of this problem mentioned 10 000) and each of them was rewritten,
7424 Exim could use up a very large amount of memory. (It kept on making new
7425 copies of the header line as it rewrote, and never released the old ones.)
7426 At the expense of a bit more processing, the header rewriting function has
7427 been changed so that it no longer eats memory in this way.
7429 66. The generation of the Received: header has been moved from the time that a
7430 message starts to be received, to the time that it finishes. The timestamp
7431 in the Received: header should now be very close to that of the <= log
7432 line. There are two side-effects of this change:
7434 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7435 logged header lines no longer include the local Received: line, because
7436 it has not yet been created. The same applies to a copy of the message
7437 that is returned to a non-SMTP sender when a message is rejected.
7439 (b) When a filter file is tested using -bf, no additional Received: header
7440 is added to the test message. After some thought, I decided that this
7443 This change does not affect the value of $received_for. It is still set
7444 after address rewriting, but before local_scan() is called.
7446 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7448 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7449 gave an unhelpful panic error message, and a defer error. I have managed to
7450 change this behaviour so that it now rejects any supplied certificate,
7451 which seems right, as the list of acceptable certificates is empty.
7453 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7454 gave an unhelpful defer error. I have not managed to make this reject any
7455 supplied certificates, but the error message it gives is "no certificate
7456 supplied", which is not helpful.
7458 70. exigrep's output now also includes lines that are not associated with any
7459 message, but which match the given pattern. Implemented by a patch from
7460 Martin Sluka, which also tidied up the Perl a bit.
7462 71. Recipient callout verification, like sender verification, was using <> in
7463 the MAIL FROM command. This isn't really the right thing, since the actual
7464 sender may affect whether the remote host accepts the recipient or not. I
7465 have changed it to use the actual sender in the callout; this means that
7466 the cache record is now keyed on a recipient/sender pair, not just the
7467 recipient address. There doesn't seem to be a real danger of callout loops,
7468 since a callout by the remote host to check the sender would use <>.
7469 [SEE ABOVE: changed after hitting problems.]
7471 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7472 temporary errors. However, in the case of such a code being given after
7473 the end of a data transmission (i.e. after ".") Exim was failing to write
7474 a retry record for the message. (Yes, there was some broken host that was
7475 actually sending 8xx at this point.)
7477 73. An unknown lookup type in a host list could cause Exim to panic-die when
7478 the list was checked. (An example that provoked this was putting <; in the
7479 middle of a list instead of at the start.) If this happened during a DATA
7480 ACL check, a -D file could be left lying around. This kind of configuration
7481 error no longer causes Exim to die; instead it causes a defer error. The
7482 incident is still logged to the main and panic logs.
7484 74. Buglet left over from Exim 3 conversion. The message "too many messages
7485 in one connection" was written to the rejectlog but not the mainlog, except
7486 when address rewriting (yes!) was being logged.
7488 75. Added write_rejectlog option.
7490 76. When a system filter was run not as root (that is, when system_filter_user
7491 was set), the values of the $n variables were not being returned to the
7492 main process; thus, they were not subsequently available in the $sn
7495 77. Added +return_path_on_delivery log selector.
7497 78. A connection timeout was being treated differently from recipients deferred
7498 when testing hosts_max_try with a message that was older than the host's
7499 retry timeout. (The host should not be counted, thus allowing all hosts to
7500 be tried at least once before bouncing.) This may have been the cause of an
7501 occasionally reported bug whereby a message would remain on the queue
7502 longer than the retry timeout, but would be bounced if a delivery was
7503 forced. I say "may" because I never totally pinned down the problem;
7504 setting up timeout/retry tests is difficult. See also the next item.
7506 79. The ultimate address timeout was not being applied to errors that involved
7507 a combination of host plus message (for example, a timeout on a MAIL
7508 command). When an address resolved to a number of possible hosts, and they
7509 were not all tried for each delivery (e.g. because of hosts_max_try), a
7510 message could remain on the queue longer than the retry timeout.
7512 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7515 81. Fixed an obscure SMTP outgoing bug which required at least the following
7516 conditions: (a) there was another message waiting for the same server;
7517 (b) the server returned 5xx to all RCPT commands in the first message so
7518 that the message was not completed; (c) the server dropped the connection
7519 or gave a negative response to the RSET that Exim sends to abort the
7520 transaction. The observed case was a dropped connection after DATA that had
7521 been sent in pipelining mode. That is, the server had advertised PIPELINING
7522 but was not implementing it correctly. The effect of the bug was incorrect
7523 behaviour, such as trying another host, and this could lead to a crash.
7529 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7530 and daemon.c were passed as pointers to ints; they should have been
7531 pointers to socklen_t variables (which are typically unsigned ints).
7533 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7536 3. Fixed a really odd bug that affected only the testing scheme; patching a
7537 certain fixed string in the binary changed the value of another string that
7538 happened to be identical to the end of the original first string.
7540 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7541 name", it returns that address as the IP address. On some operating
7542 systems (e.g. Solaris), it also passes back the IP address string as the
7543 "host name". However, on others (e.g. Linux), it passes back an empty
7544 string. Exim wasn't checking for this, and was changing the host name to an
7545 empty string, assuming it had been canonicalized.
7547 5. Although rare, it is permitted to have more than one PTR record for a given
7548 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7549 all the names associated with an address, because they do in Solaris.
7550 However, it seems that they do not in Linux for data that comes from the
7551 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7552 I found this out when I moved to a new Linux workstation and tried to run
7553 the Exim test suite.
7555 To get round this problem I have changed the code so that it now does its
7556 own call to the DNS to look up PTR records when searching for a host name.
7557 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7558 addresses that are only in /etc/hosts are still found.
7560 This behaviour is, however, controlled by an option called host_lookup_
7561 order, which defaults to "bydns:byaddr". If people want to use the other
7562 order, or indeed, just use one or the other means of lookup, they can
7563 specify it in this variable.
7565 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7566 some operating systems, this comes back from gethostbyaddr() as an empty
7567 string, and this is what Exim used to test for. However, it seems that in
7568 other systems, "." is yielded. Exim now tests for this case too.
7570 7. The values of check_spool_space and check_log_space are now held internally
7571 as a number of kilobytes instead of an absolute number of bytes. If a
7572 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7573 kilobyte. This means that much larger values can be stored.
7575 8. Exim monitor: an attempt to get the action menu when not actually pointing
7576 at a message produces an empty menu entitled "No message selected". This
7577 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7578 no entries in it ("Shell widget menu has zero width and/or height"). So I
7579 have added a single, blank menu entry in this case.
7581 9. Added ${quote_local_part.
7583 10. MIME decoding is now applied to the contents of Subject: header lines when
7586 11. Now that a reference to $sender_host_address automatically causes a reverse
7587 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7588 host lookup before query-style lookups in lists that might use this
7589 variable. This has therefore been abolished, and the "net-" prefix is no
7590 longer necessary for query-style lookups.
7592 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7593 have been a typo for LFLAGS, so it has been changed.
7595 13. The install script calls Exim with "-C /dev/null" in order to find the
7596 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7597 to be output. However, since Exim outputs its version number before the
7598 error, it didn't break the script. It just looked ugly. I fixed this by
7599 always allowing "-C /dev/null" if the caller is root.
7601 14. Ignore overlarge ACL variable number when reading spool file - insurance
7602 against a later release with more variables having written the file.
7604 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7605 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7606 this, as well as the form without the "IPv6" on the front.
7608 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7609 OS/Makefile-Darwin file.
7611 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7612 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7614 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7615 variable. However, some OS use ints for inodes. I've added cast to long int
7616 to get rid of the compiler warning.
7618 19. I had forgotten to lock out "/../" in configuration file names when
7619 ALT_CONFIG_PREFIX was set.
7621 20. Routers used for verification do not need to specify transports. However,
7622 if such a router generated a host list, and callout was configured, Exim
7623 crashed, because it could not find a port number from the (non-existent)
7624 transport. It now assumes port 25 in this circumstance.
7626 21. Added the -t option to exigrep.
7628 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7629 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7630 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7631 them is set, LOOKUP_LSEARCH is forced.
7633 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7635 24. Added sender and host information to the "rejected by local_scan()" log
7636 line; previously there was no indication of these.
7638 25. Added .include_if_exists.
7640 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7641 Linux. It turns out that not all file systems support this. Apparently some
7642 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7643 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7644 not supported on the file descriptor, is now ignored when Exim is trying to
7645 sync a directory. This applies only to Linux.
7647 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7649 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7650 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7653 29. Added support for maildirsize files from supplied patch (modified a bit).
7655 30. The use of :fail: followed by an empty string could lead Exim to respond to
7656 sender verification failures with (e.g.):
7658 550 Verification failed for <xxx>
7659 550 Sender verify failed
7661 where the first response line was missing the '-' that indicates it is not
7662 the final line of the response.
7664 31. The loop for finding the name of the user that called Exim had a hardwired
7665 limit of 10; it now uses the value of finduser_retries, which is used for
7666 all other user lookups.
7668 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7671 33. Exim was neglecting to zero errno before one call of strtol() when
7672 expanding a string and expecting an integer value. On some systems this
7673 resulted in spurious "integer overflow" errors. Also, it was casting the
7674 result into an int without checking.
7676 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7677 did not work. The code looks as if it has *never* worked, though it appears
7678 to have been documented since at least release 1.62. I have made it work.
7680 35. The "timeout_DNS" error in retry rules, also documented since at least
7681 1.62, also never worked. As it isn't clear exactly what this means, and
7682 clearly it isn't a major issue, I have abolished the feature by treating it
7683 as "timeout", and writing a warning to the main and panic logs.
7685 36. The display of retry rules for -brt wasn't always showing the error code
7688 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7689 timeout_connect_A, timeout_connect_MX.
7691 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7692 to the empty sender.
7694 39. The daemon was not analysing the content of -oX till after it had closed
7695 stderr and disconnected from the controlling terminal. This meant that any
7696 syntax errors were only noted on the panic log, and the return code from
7697 the command was 0. By re-arranging the code a little, I've made the
7698 decoding happen first, so such errors now appear on stderr, and the return
7699 code is 1. However, the actual setting up of the sockets still happens in
7700 the disconnected process, so errors there are still only recorded on the
7703 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7704 connections (as happens on some IP stacks) was logged at start up time as
7705 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7706 it from "IPv6 and IPv4", which means that two separate sockets are being
7709 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7710 says whether AF_INET or AF_INET6 was passed as an argument.
7712 42. Exiwhat output was messed up when time zones were included in log
7715 43. Exiwhat now gives more information about the daemon's listening ports,
7716 and whether -tls-on-connect was used.
7718 44. The "port" option of the smtp transport is now expanded.
7720 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7721 silently ignored. Now an error message is written to the main and panic
7724 46. There's a new ACL modifier called "logwrite" which writes to a log file
7725 as soon as it is encountered.
7727 47. Added $local_user_uid and $local_user_gid at routing time.
7729 48. Exim crashed when trying to verify a sender address that was being
7732 49. Exim was recognizing only a space character after ".include". It now also
7733 recognizes a tab character.
7735 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7736 extracting the relevant information from the specification. The man page no
7737 longer contains scrambled data for the -d option, and I've added a section
7738 at the front about calling Exim under different names.
7740 51. Added "extra_headers" argument to the "mail" command in filter files.
7742 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7745 53. Installed eximstats 1.29.
7747 54. Added transport_filter_timeout as a generic transport option.
7749 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7750 Cc: header lines. This was required by RFC 822, but it not required by RFC
7753 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7754 incoming messages that did not have them. Now it does so only if the
7755 message originates locally, that is, if there is no associated remote host
7756 address. When Resent- header lines are present, this applies to the Resent-
7757 lines rather than the non-Resent- lines.
7759 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7760 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7762 58. Messages for configuration errors now include the name of the main
7763 configuration file - useful now that there may be more than one file in a
7764 list (.included file names were always shown).
7766 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7767 for those rare installations that do not start the daemon as root or run it
7768 setuid root. I've cut out the call to initgroups() if the daemon is not
7771 60. The Exim user and group can now be bound into the binary as text strings
7772 that are looked up at the start of Exim's processing.
7774 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7776 62. Added $mailstore_basename variable.
7778 63. Installed patch to sieve.c from Michael Haardt.
7780 64. When Exim failed to open the panic log after failing to open the main log,
7781 the original message it was trying to log was written to stderr and debug
7782 output, but if they were not available (the usual case in production), it
7783 was lost. Now it is written to syslog before the two lines that record the
7784 failures to open the logs.
7786 65. Users' Exim filters run in subprocesses under the user's uid. It is
7787 possible for a "deliver" command or an alias in a "personal" command to
7788 provoke an address rewrite. If logging of address rewriting is configured,
7789 this fails because the process is not running as root or exim. There may be
7790 a better way of dealing with this, but for the moment (because 4.30 needs
7791 to be released), I have disabled address rewrite logging when running a
7792 filter in a non-root, non-exim process.
7798 1. The buildconfig auxiliary program wasn't quoting the value set for
7799 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7800 not defined. This bug was masked in 4.22 by the effect that was fixed in
7803 2. Some messages that were rejected after a message id was allocated were
7804 shown as "incomplete" by exigrep. It no longer does this for messages that
7805 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7807 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7808 have allow_domain_literals set, the ID did not get logged in the <= line.
7809 Domain literals are now always recognized in Message-ID: header lines.
7811 4. The first argument for a ${extract expansion item is the key name or field
7812 number. Leading and trailing spaces in this item were not being ignored,
7813 causing some misleading effects.
7815 5. When deliver_drop_privilege was set, single queue runner processes started
7816 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7817 same command in the process it spins off) were not dropping privilege.
7819 6. When the daemon running as "exim" started a queue runner, it always
7820 re-executed Exim in the spun-off process. This is a waste of effort when
7821 deliver_drop_privilege is set. The new process now just calls the
7822 queue-runner function directly.
7828 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7831 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7832 ignored. Though the use of -oP was forcing the writing of a pid file, it
7833 was always written to the default place.
7835 3. If the message "no IP address found for host xxxx" is generated during
7836 incoming verification, it is now followed by identification of the incoming
7837 connection (so you can more easily find what provoked it).
7839 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7841 5. Added some features to "harden" Exim a bit more against certain attacks:
7843 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7844 be put in Local/Makefile. This is like the never_users runtime option,
7845 but it cannot be overridden. The default setting is "root".
7847 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7848 prefix string with which any file named in a -C command line option
7851 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7852 is retained for -C and -D only if the caller of Exim is root. Without
7853 it, the exim user may also use -C and -D and retain privilege.
7855 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7856 command line option is disabled.
7858 6. Macro names set by the -D option must start with an upper case letter, just
7859 like macro names defined in the configuration file.
7861 7. Added "dereference=" facility to LDAP.
7863 8. Two instances of the typo "uknown" in the source files are fixed.
7865 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7866 the Configure-Makefile script screwed up while processing it.
7868 10. Incorporated PCRE 4.4.
7870 11. The SMTP synchronization check was not operating right at the start of an
7871 SMTP session. For example, it could not catch a HELO sent before the client
7872 waited for the greeting. There is now a check for outstanding input at the
7873 point when the greeting is written. Because of the duplex, asynchronous
7874 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7875 way, but not yet received, when the check is performed.
7877 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7878 on TCP/IP sockets, because this apparently causes some broken clients to
7881 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7882 unchanged) from the Cygwin maintainer.
7884 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7885 was supported instead of testing for mbx. Effectively a typo.
7887 15. The spa authenticator server code was not checking that the input it
7888 received was valid base64.
7890 16. The debug output line for the "set" modifier in ACLs was not showing the
7891 name of the variable that was being set.
7893 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7895 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7896 lookup on its own. Something else was needed to trigger the lookup. For
7897 example, a match in host_lookup or the need for a host name in a host list.
7898 Now, if $sender_host_name is referenced and the host name has not yet been
7899 looked up, a lookup is performed. If the lookup fails, the variable remains
7900 empty, and $host_lookup_failed is set to "1".
7902 19. Added "eqi" as a case-independent comparison operator.
7904 20. The saslauthd authentication condition could segfault if neither service
7905 nor realm was specified.
7907 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7908 error message that was logged was misleading, and incoming SMTP
7909 connections were dropped. The message is now more accurate, and temporary
7910 errors are given to SMTP connections.
7912 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7913 (except RSET) with a 421 error, until QUIT is received. However, it was
7914 failing to send a response to QUIT.
7916 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7917 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7918 ACL is now run after the helo_try_verify_hosts code.
7920 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7921 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7922 case-independent, so other case variants are also recognized). Apparently
7923 some systems use these upper case variants.
7925 25. If more than two messages were waiting for the same host, and a transport
7926 filter was specified for the transport, Exim sent two messages over the
7927 same TCP/IP connection, and then failed with "socket operation on non-
7928 socket" when it tried to send the third.
7930 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7932 27. The extern definition of crypt16() in expand.c was not being excluded when
7933 the OS had its own crypt16() function.
7935 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7936 as a preferred synonym for return_size_limit, both as an option and as an
7939 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7941 30. Changed the default configuration ACL to relax the local part checking rule
7942 for addresses that are not in any local domains. For these addresses,
7943 slashes and pipe symbols are allowed within local parts, but the sequence
7944 /../ is explicitly forbidden.
7946 31. SPA server authentication was not clearing the challenge buffer before
7949 32. log_message in a "warn" ACL statement was writing to the reject log as
7950 well as to the main log, which contradicts the documentation and doesn't
7951 seem right (because no rejection is happening). So I have stopped it.
7953 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7954 However, I am unable to do any testing of this.
7956 34. Fixed an infelicity in the appendfile transport. When checking directories
7957 for a mailbox, to see if any needed to be created, it was accidentally
7958 using path names with one or more superfluous leading slashes; tracing
7959 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7961 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7962 contents are added to the log line that is written for every discarded
7963 recipient. (Previously a log_message setting was ignored.)
7965 36. The ${quote: operator now quotes the string if it is empty.
7967 37. The install script runs exim in order to find its version number. If for
7968 some reason other than non-existence or emptiness, which it checks, it
7969 could not run './exim', it was installing it with an empty version number,
7970 i.e. as "exim-". This error state is now caught, and the installation is
7973 38. An argument was missing from the function that creates an error message
7974 when Exim fails to connect to the socket for saslauthd authentication.
7975 This could cause Exim to crash, or give a corrupted message.
7977 39. Added isip, isip4, and isip6 to ${if conditions.
7979 40. The ACL variables $acl_xx are now saved with the message, and can be
7980 accessed later in routers, transports, and filters.
7982 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7983 strings in the file are not string-expanded.
7985 42. If a MAIL command specified a SIZE value that was too large to fit into an
7986 int variable, the check against message_size_limit failed. Such values are
7987 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7988 day this will have to be increased, but I don't think I want to be around
7989 when emails are that large.
7996 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7997 iconv() is not standard in FreeBSD.
7999 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8000 IPv6 enabled. The observed symptom was a segmentation fault on return from
8001 the function os_common_find_running_interfaces() in src/os.c.
8003 3. In the check_special_case() function in daemon.c I had used "errno" as an
8004 argument name, which causes warnings on some systems. This was basically a
8005 typo, since it was named "eno" in the comments!
8007 4. The code that waits for the clock to tick (at a resolution of some fraction
8008 of a second) so as to ensure message-id uniqueness was always waiting for
8009 at least one whole tick, when it could have waited for less. [This is
8010 almost certainly not relevant at current processor speeds, where it is
8011 unlikely to ever wait at all. But we try to future-proof.]
8013 5. The function that sleeps for a time interval that includes fractions of a
8014 second contained a race. It did not block SIGALRM between setting the
8015 timer, and suspending (a couple of lines later). If the interval was short
8016 and the sigsuspend() was delayed until after it had expired, the suspension
8017 never ended. On busy systems this could lead to processes getting stuck for
8020 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8021 process, before it forks any delivery processes. The open lookup caching
8022 mechanism meant that the open file or database connection was passed into
8023 the delivery process. The problem was that delivery processes always tidy
8024 up cached lookup data. This could cause a problem for the next delivery
8025 process started by the queue runner, because the external queue runner
8026 process does not know about the closure. So the next delivery process
8027 still has data in the lookup cache. In the case of a file lookup, there was
8028 no problem because closing a file descriptor in a subprocess doesn't affect
8029 the parent. However, if the lookup was caching a connection to a database,
8030 the connection was closed, and the second delivery process was likely to
8031 see errors such as "PGSQL: query failed: server closed the connection
8032 unexpectedly". The problem has been fixed by closing all cached lookups
8033 in a queue runner before running a delivery process.
8035 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8036 seem to have the "const" qualifier which it has on other OS. I've
8039 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8040 *of the same type* (client or server) with the same public name that an
8041 error should be diagnosed.
8043 9. When Exim looked up a host name for an IP address, but failed to find the
8044 original IP address when looking up the host name (a safety check), it
8045 output the message "<ip address> does not match any IP for NULL", which was
8046 confusing, to say the least. The bug was that the host name should have
8047 appeared instead of "NULL".
8049 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8050 user that is built into the binary, and the -C or -D options is used, root
8051 privilege is dropped before the configuration file is read. In addition,
8052 logging is switched to stderr instead of the normal log files. If the
8053 configuration then re-defines the Exim user, the unprivileged environment
8054 is probably not what is expected, so Exim logs a panic warning message (but
8057 However, if deliver_drop_privilege is set, the unprivileged state may well
8058 be exactly what is intended, so the warning has been cut out in that case,
8059 and Exim is allowed to try to write to its normal log files.
8065 1. smtp_return_error_details was not giving details for temporary sender
8066 or receiver verification errors.
8068 2. Diagnose a configuration error if two authenticators have the same public
8071 3. Exim used not to create the message log file for a message until the first
8072 delivery attempt. This could be confusing when incoming messages were held
8073 for policy or load reasons. The message log file is now created at the time
8074 the message is received, and an initial "Received" line is written to it.
8076 4. The automatically generated man page for command line options had a minor
8077 bug that caused no ill effects; however, a more serious problem was that
8078 the procedure for building the man page automatically didn't always
8079 operate. Consequently, release 4.20 contains an out-of-date version. This
8080 shouldn't happen again.
8082 5. When building Exim with embedded Perl support, the script that builds the
8083 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8084 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8086 6. The freeze_tell option was not being used for messages that were frozen on
8087 arrival, either by an ACL or by local_scan().
8089 7. Added the smtp_incomplete_transaction log selector.
8091 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8092 was accepting AUTH without a new EHLO.
8094 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8095 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8098 10. Logging of TCP/IP connections (when configured) now happens in the main
8099 daemon process instead of the child process, so that the TCP/IP connection
8100 count is more accurate (but it can never be perfect).
8102 11. The use of "drop" in a nested ACL was not being handled correctly in the
8103 outer ACL. Now, if condition failure induced by the nested "drop" causes
8104 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8105 or "require"), the connection is dropped.
8107 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8108 that yield "discard" can now be used with an "accept" or a "discard" verb,
8109 but an error is generated for any others (because I can't see a useful way
8110 to define what should happen).
8112 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8113 are now processed in the same way as lines in the Exim configuration file.
8114 In particular, continuation lines are supported.
8116 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8118 15. Added -ti meaning -t -i.
8120 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8121 domains, and warn by logging if others are found.
8123 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8124 returned from GIFCONF when Exim is trying to find the list of interfaces on
8125 a host. The code in os.c has been modified to copy each ifreq to an aligned
8126 structure in all cases.
8128 Also, in some cases, the returned ifreq's were being copied to a 'struct
8129 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8130 means the last couple of bytes of an IPv6 address could be chopped if the
8131 ifreq contained only a normal sockaddr (14 bytes storage).
8133 18. Named domain lists were not supported in the hosts_treat_as_local option.
8134 An entry such as +xxxx was not recognized, and was treated as a literal
8137 19. Ensure that header lines added by a DATA ACL are included in the reject log
8138 if the ACL subsequently rejects the message.
8140 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8141 MD5 (which is deprecated).
8143 21. When testing a filter file using -bf, Exim was writing a message when it
8144 took the sender from a "From " line in the message, but it was not doing so
8145 when it took $return_path from a Return-Path: header line. It now does.
8147 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8148 with a valid header line field name (a series of printing characters
8149 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8151 23. Changed "disc" in the source to "disk" to conform to the documentation and
8152 the book and for uniformity.
8154 24. Ignore Sendmail's -Ooption=value command line item.
8156 25. When execve() failed while trying to run a command in a pipe transport,
8157 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8158 could be confused with a return value of 69 from the command itself. This
8159 has been changed to 127, the value the shell returns if it is asked to run
8160 a non-existent command. The wording for the related log line suggests a
8161 non-existent command as the problem.
8163 26. If received_header_text expands to an empty string, do not add a Received:
8164 header line to the message. (Well, it adds a token one on the spool, but
8165 marks it "old" so that it doesn't get used or transmitted.)
8167 27. Installed eximstats 1.28 (addition of -nt option).
8169 28. There was no check for failure on the call to getsockname() in the daemon
8170 code. This can fail if there is a shortage of resources on the system, with
8171 ENOMEM, for example. A temporary error is now given on failure.
8173 29. Contrary to the C standard, it seems that in some environments, the
8174 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8175 program. Exim now does this explicitly; it affects the formatting of
8176 timestamps using strftime().
8178 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8179 complaints. I've now initialized all the variables, to avoid this.
8181 32. Header lines added by a system filter were not being "seen" during
8182 transport-time rewrites.
8184 33. The info_callback() function passed to OpenSSL is set up with type void
8185 (*)(SSL *, int, int), as described somewhere. However, when calling the
8186 function (actually a macro) that sets it up, the type void(*)() is
8187 expected. I've put in a cast to prevent warnings from picky compilers.
8189 34. If a DNS black list lookup found a CNAME record, but there were no A
8190 records associated with the domain it pointed at, Exim crashed.
8192 35. If a DNS black list lookup returned more than one A record, Exim ignored
8193 all but the first. It now scans all returned addresses if a particular IP
8194 value is being sought. In this situation, the contents of the
8195 $dnslist_value variable are a list of all the addresses, separated by a
8198 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8199 to accept a host name and all its aliases if the forward lookup for any of
8200 them yielded the IP address of the incoming connection. Now it accepts only
8201 those names whose forward lookup yields the correct IP address. Any other
8202 names are discarded. This closes a loophole whereby a rogue DNS
8203 administrator could create reverse DNS records to break through a
8204 wildcarded host restriction in an ACL.
8206 37. If a user filter or a system filter that ran in a subprocess used any of
8207 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8208 the wrong values were passed to the pipe command ($thisaddress had the
8209 value of $0, $0 had the value of $1, etc). This bug was introduced by
8210 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8212 38. Improved the line breaking for long SMTP error messages from ACLs.
8213 Previously, if there was no break point between 40 and 75 characters, Exim
8214 left the rest of the message alone. Two changes have been made: (a) I've
8215 reduced the minimum length to 35 characters; (b) if it can't find a break
8216 point between 35 and 75 characters, it looks ahead and uses the first one
8217 that it finds. This may give the occasional overlong line, but at least the
8218 remaining text gets split now.
8220 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8221 file descriptors might be low, and that setting 1000 would always raise it.
8222 It turns out that in some environments, the limit is already over 1000 and
8223 that lowering it causes trouble. So now Exim takes care not to decrease it.
8225 40. When delivering a message, the value of $return_path is set to $sender_
8226 address at the start of routing (routers may change the value). By an
8227 oversight, this default was not being set up when an address was tested by
8228 -bt or -bv, which affected the outcome if any router or filter referred to
8231 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8232 log only when "message" or "log_message" are set. However, if one of the
8233 conditions was an address verification, or a call to a nested ACL, the
8234 messages generated by the underlying test were being passed through. This
8235 no longer happens. The underlying message is available in $acl_verify_
8236 message for both "message" and "log_message" expansions, so it can be
8237 passed through if needed.
8239 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8240 new expansion $bh_ to give the encoded byte string without charset
8241 translation. Translation happens only if iconv() is available; HAVE_ICONV
8242 indicates this at build time. HEADERS_CHARSET gives the charset to
8243 translate to; headers_charset can change it in the configuration, and
8244 "headers charset" can change it in an individual filter file.
8246 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8247 that creates RFC 2047 encoded "words" labels them as that charset instead
8248 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8249 expansion operator; (ii) when Exim creates a From: line for a local
8250 message; (iii) when a header line is rewritten to include a "phrase" part.
8252 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8253 buggy, causing it to skip the first lines of messages whose message ID
8254 ended in 'D'. This would not have bitten before Exim release 4.14, because
8255 message IDs were unlikely to end in 'D' before then. The effect was to have
8256 incorrect size information for certain domains.
8258 45. #include "config.h" was missing at the start of the crypt16.c module. This
8259 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8262 46. If there was a timeout during a "random" callout check, Exim treated it as
8263 a failure of the random address, and carried on sending RSET and the real
8264 address. If the delay was just some slowness somewhere, the response to the
8265 original RCPT would be taken as a response to RSET and so on, causing
8266 mayhem of various kinds.
8268 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8269 when I implemented it. It didn't allow for the fact that some option values
8270 may legitimately be negative (e.g. size_addition), and it didn't even do
8271 the right test for positive values.
8273 48. Domain names in DNS records are case-independent. Exim always looks them up
8274 in lower case. Some resolvers return domain names in exactly the case they
8275 appear in the zone file, that is, they may contain uppercase letters. Not
8276 all resolvers do this - some return always lower case. Exim was treating a
8277 change of case by a resolver as a change of domain, similar to a widening
8278 of a domain abbreviation. This triggered its re-routing code and so it was
8279 trying to route what was effectively the same domain again. This normally
8280 caused routing to fail (because the router wouldn't handle the domain
8281 twice). Now Exim checks for this case specially, and just changes the
8282 casing of the domain that it ultimately uses when it transmits the message
8285 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8288 50. If a filter generated a file delivery with a non-absolute name (possible if
8289 no home directory exists for the router), the forbid_file option was not
8292 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8293 the existing equality matching.
8295 52. Exim was using ints instead of ino_t variables in some places where it was
8296 dealing with inode numbers.
8298 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8299 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8300 called TMPDIR, and if it finds it is different, it changes its value.
8302 54. The smtp_printf() function is now made available to local_scan() so
8303 additional output lines can be written before returning. There is also an
8304 smtp_fflush() function to enable the detection of a dropped connection.
8305 The variables smtp_input and smtp_batched_input are exported to
8308 55. Changed the default runtime configuration: the message "Unknown user"
8309 has been removed from the ACL, and instead placed on the localuser router,
8310 using the cannot_route_message feature. This means that any verification
8311 failures that generate their own messages won't get overridden. Similarly,
8312 the "Unrouteable address" message that was in the ACL for unverifiable
8313 relay addresses has also been removed.
8315 56. Added hosts_avoid_esmtp to the smtp transport.
8317 57. The exicyclog script was not checking for the esoteric option
8318 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8319 will work only if exicyclog is run under the appropriate euid.
8321 58. Following a discussion on the list, the rules by which Exim recognises line
8322 endings on incoming messages have been changed. The -dropcr and drop_cr
8323 options are now no-ops, retained only for backwards compatibility. The
8324 following line terminators are recognized: LF CRLF CR. However, special
8325 processing applies to CR:
8327 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8328 nor a local message in the state where . is a terminator.
8330 (ii) If a bare CR is encountered in a header line, an extra space is added
8331 after the line terminator so as not to end the header. The reasoning
8332 behind this is that bare CRs in header lines are most likely either
8333 to be mistakes, or people trying to play silly games.
8335 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8336 was being incorrectly given as 18 bytes larger than it should have been.
8337 This is a VOB (very old bug).
8339 60. This may never have affected anything current, but just in case it has:
8340 When the local host is found other than at the start of a list of hosts,
8341 the local host, those with the same MX, and any that follow, are discarded.
8342 When the list in question was part of a longer list of hosts, the following
8343 hosts (not currently being processed) were also being discarded. This no
8344 longer happens. I'm not sure if this situation could ever has previously
8347 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8350 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8351 header. This is something that is recommended in a new Internet Draft, and
8352 is something that is documented as being done by Sendmail. There are two
8353 possible values. For messages generated by the autoreply transport, Exim
8356 Auto-Submitted: auto-replied
8358 whereas for all other generated messages (e.g. bounces) it adds
8360 Auto-Submitted: auto-generated
8362 63. The "personal" condition in filters now includes a test for the
8363 Auto-Submitted: header. If it contains the string "auto-" the message it
8364 not considered personal.
8366 64. Added rcpt_include_affixes as a generic transport option.
8368 65. Added queue_only_override (default true).
8370 66. Added the syslog_duplication option.
8372 67. If what should have been the first header line of a message consisted of
8373 a space followed by a colon, Exim was mis-interpreting it as a header line.
8374 It isn't of course - it is syntactically invalid and should therefore be
8375 treated as the start of the message body. The misbehaviour could have
8376 caused a number of strange effects, including loss of data in subsequent
8377 header lines, and spool format errors.
8379 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8380 client host had authenticated. This control can now be exercised by an ACL
8381 for more flexibility.
8383 69. By default, callouts do not happen when testing with -bh. There is now a
8384 variant, -bhc, which does actually run the callout code, including
8385 consulting and updating the callout cache.
8387 70. Added support for saslauthd authentication, courtesy of Alexander
8390 71. If statvfs() failed on the spool or log directories while checking their
8391 size for availability, Exim confusingly gave the error "space shortage".
8392 Furthermore, in debugging mode it crashed with a floating point exception.
8393 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8394 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8395 serious problem, Exim now writes to the main and panic logs when this
8396 happens, with details of the failure. It then refuses to accept the
8397 incoming message, giving the message "spool directory problem" or "log
8398 directory problem" with a 421 code for SMTP messages.
8400 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8401 0, 1, and 2 exist, because some OS complain for execs without them (see
8402 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8403 descriptors. However, the code omitted to check that the open succeeded,
8404 causing mysterious errors if for some reason the permissions on /dev/null
8405 got screwed. Now Exim writes a message to the main and panic logs, and
8406 bombs out if it can't open /dev/null.
8408 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8409 interact so that it is all more flexible. It is supposed to remain
8410 backwards compatible. Also added extra_local_interfaces.
8412 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8413 to bomb out with an assertion failure - to the client this appears as a
8414 connection drop. This problem occurs in the part of the code that was taken
8415 from the Samba project. Fortunately, the assertion is in a very simple
8416 function, so I have fixed this by reproducing the function inline in the
8417 one place where it is called, and arranging for authentication to fail
8418 instead of killing the process with assert().
8420 75. The SPA client code was not working when the server requested OEM rather
8421 than Unicode encoding.
8423 76. Added code to make require_files with a specific uid setting more usable in
8424 the case where statting the file as root fails - usually a non-root-mounted
8425 NFS file system. When this happens and the failure is EACCES, Exim now
8426 forks a subprocess and does the per-uid checking as the relevant uid.
8428 77. Added process_log_path.
8430 78. If log_file_path was not explicitly set, a setting of check_log_space or
8431 check_log_inodes was ignored.
8433 79. If a space check for the spool or log partitions fails, the incident is now
8434 logged. Of course, in the latter case the data may get lost...
8436 80. Added the %p formatting code to string_format() so that it can be used to
8437 print addresses in debug_print(). Adjusted all the address printing in the
8438 debugging in store.c to use %p rather than %d.
8440 81. There was a concern that a line of code in smtp_in.c could overflow a
8441 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8442 initially expressed, the concern was not well-founded, because trailing
8443 spaces are removed early. However, if the trailing spaces were followed by
8444 a NULL, they did not get removed, so the overflow was possible. Two fixes
8447 (a) I re-wrote the offending code in a cleaner fashion.
8448 (b) If an incoming SMTP command contains a NULL character, it is rejected
8451 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8452 runs initgroups(), so that if the Exim user is in any additional groups,
8453 they will be used during message reception.
8459 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob