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 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
92 included (though probably not limited to) a process crash from stack
93 memory limit, or from excessive open files. Replace this with a paniclog
94 whine (as this is likely a configuration error), and returning
97 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
98 this always failed, probably leading to the usual downgrade to in-clear
101 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
102 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
103 to not checking that list for these lookups.
105 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
108 JH/23 Fix crash in string expansions. Previously, if an empty variable was
109 immediately followed by an expansion operator, a null-indirection read
110 was done, killing the process.
112 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
113 include an SMTP response string which is longer than that supported
114 by the delivering transport. Alleviate by wrapping such lines before
117 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
118 chars (RFC limit). Previously a limit of 12 items was made, which with
119 a not-impossible References: in the message being bounced could still
120 be over-large and get stopped in the transport.
122 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
123 close. Previously a bare socket close was done.
125 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
128 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
129 message being considered for delivery was already being handled by
130 another process, and having an SMTP connection already open, the function
131 to close it tried to use an uninitialized variable. This would afftect
132 high-volume sites more, especially when running mailing-list-style loads.
133 Pollution of logs was the major effect, as the other process delivered
134 the message. Found and partly investigated by Graeme Fowler.
136 JH/29 Change format of the internal ID used for message identification. The old
137 version only supported 31 bits for a PID element; the new 64 (on systems
138 which can use Base-62 encoding, which is all currently supported ones
139 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
140 and must use Base-36). The new ID is 23 characters rather than 16, and is
141 visible in various places - notably logs, message headers, and spool file
142 names. Various of the ancillary utilities also have to know the format.
143 As well as the expanded PID portion, the sub-second part of the time
144 recorded in the ID is expanded to support finer precision. Theoretically
145 this permits a receive rate from a single comms channel of better than the
147 The major timestamp part of the ID is not changed; at 6 characters it is
148 usable until about year 3700.
149 Updating from previously releases is fully supported: old-format spool
150 files are still usable, and the utilities support both formats. New
151 message will use the new format. The one hints-DB file type which uses
152 message-IDs (the transport wait- DB) will be discarded if an old-format ID
153 is seen; new ones will be built with only new-format IDs.
154 Optionally, a utility can be used to convert spool files from old to new,
155 but this is only an efficiency measure not a requirement for operation
156 Downgrading from new to old requires running a provided utility, having
157 first stopped all operations. This will convert any spool files from new
158 back to old (losing time-precision and PID information) and remove any
159 wait- hints databases.
161 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
162 we treated them as item separators when parsing for a list item, but they
163 need to be protected by the doublequotes. While there, add handling for
166 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
167 Found and fixed by Jasen Betts. No testcase for this as my usual text
168 editor insists on emitting only valid UTF-8.
170 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
171 a null-indirection SIGSEGV for the receive process.
173 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
174 -bh use. Internal checking would spot this and take a panic.
176 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
177 In 4.96 this would expand to empty.
179 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
180 certificate. Find and fix by Andreas Metzler.
182 JH/36 Add ARC info to DMARC hostory records.
184 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
185 or fakedefer. Previously the sender could discover that the message
186 had in fact been accepted.
188 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
189 sequences. Previously the input was not noted as being tainted; notably
190 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
191 bad coding of authenticators.
193 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
194 and ${tr...}. Found and diagnosed by Heiko Schlichting.
196 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
197 could be triggered by externally-supplied input. Found by Trend Micro.
200 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
201 be triggered by externally-controlled input. Found by Trend Micro.
204 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
205 be triggered by externally-controlled input. Found by Trend Micro.
212 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
213 after reception to before a subsequent reception. This should
214 mean slightly faster delivery, and also confirmation of reception
217 JH/02 Move from using the pcre library to pcre2. The former is no longer
218 being developed or supported (by the original developer).
220 JH/03 Constification work in the filters module required a major version
221 bump for the local-scan API. Specifically, the "headers_charset"
222 global which is visible via the API is now const and may therefore
223 not be modified by local-scan code.
225 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
226 sendfile() didi not account for the way the ClamAV driver code called it.
228 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
229 time check was being done for every character; replace that with one
232 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
233 sent was prefixed with a length byte.
235 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
236 RFC 5321. Previously Dovecot (at least) would log errors during
239 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
240 in 4.61 without a reason logged. Binaries will be bigger, which might
241 matter on diskspace-constrained systems, but debug is easier.
243 JH/09 Fix macro-definition during "-be" expansion testing. The move to
244 write-protected store for macros had not accounted for these runtime
245 additions; fix by removing this protection for "-be" mode.
247 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
248 handing out large-numbered file descriptors, violating the usual Unix
249 assumption (and required by Posix) that the lowest possible number will be
250 allocated by the kernel when a new one is needed. In the daemon, and any
251 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
252 are not useable for FD_SET() [and hence select()] and overwrite the stack.
253 Assorted crashes happen.
255 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
256 main-section options or in a connect ACL, the value from the first ever
257 connection was never replaced for subsequent connections. Found by
260 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
261 though only once PCRE2 was introduced: the memory accounting used under
262 debug offset allocations by an int, giving a hard trap in early startup.
263 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
265 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
266 with underbars is given. The write-protection of configuration introduced
267 in 4.95 trapped when normalisation was applied to an option not needing
270 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
272 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
273 failing to close the certificates directory, every hour or any time it
276 JH/16 Debugging initiated by an ACL control now continues through into routing
277 and transport processes. Previously debugging stopped any time Exim
278 re-execs, or for processing a queued message.
280 JH/17 The "expand" debug selector now gives more detail, specifically on the
281 result of expansion operators and items.
283 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
284 bad comparison between the option value and the name of the file to
285 be included was done, and a mismatch was wrongly identified.
286 4.88 to 4.95 are affected.
288 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
290 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
291 supplied as the db file-pair basename. Previously, if a directory
292 path was given, for example via the autoreply "once" option, the DB
293 file.pag and file.dir files would be created in that directory's
296 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
297 "taint" log_selector. These were previously deprecated.
299 JH/22 Fix static address-list lookups to properly return the matched item.
300 Previously only the domain part was returned.
302 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
303 the call into OpenSSL to send a TLS Close was being repeated; this
304 resulted in the library waiting for the peer's Close. If that was never
305 sent we waited forever. Fix by tracking send calls.
307 JH/24 The ${run} expansion item now expands its command string elements after
308 splitting. Previously it was before; the new ordering makes handling
309 zero-length arguments simpler. The old ordering can be obtained by
310 appending a new option "preexpand", after a comma, to the "run".
312 JH/25 Taint-check exec arguments for transport-initiated external processes.
313 Previously, tainted values could be used. This affects "pipe", "lmtp" and
314 "queryprogram" transport, transport-filter, and ETRN commands.
315 The ${run} expansion is also affected: in "preexpand" mode no part of
316 the command line may be tainted, in default mode the executable name
319 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
320 the the facility was not passed across execs, and only the first message
321 passed over a connection could use BDAT; any further ones using DATA.
323 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
324 uses $sending_ip_address and an interface is specified.
325 Previously any use of the local address in the EHLO name disabled
326 PIPECONNECT, the common case being to use the rDNS of it.
328 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
329 resumption. Previously verify failed because no certificate status is
330 passed on the wire for the restarted session. Fix by using the recorded
331 ocsp status of the stored session for the new connection.
333 JH/29 TLS resumption: the key for session lookup in the client now includes
334 more info that a server could potentially use in configuring a TLS
335 session, avoiding oferring mismatching sessions to such a server.
336 Previously only the server IP was used.
338 JH/30 Fix string_copyn() for limit greater than actual string length.
339 Previously the copied amount was the limit, which could result in a
340 overlapping memcpy for newly allocated destination soon after a
341 source string shorter than the limit. Found/investigated by KM.
343 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
344 close; it may be needed for a subsequent connection. This caused a
345 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
347 JH/32 Fix CHUNKING for a second message on a connection when the first was
348 rejected. Previously we did not reset the chunking-offered state, and
349 erroneously rejected the BDAT command. Investigation help from
352 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
353 an empty address. Previously the expansion returned an error.
355 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
356 proxy. Previously these were misparsed, leading to paniclog entries.
362 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
363 related applications. Previously an "H" was used where available info
364 says that "M" should be, so change to match.
366 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
367 as arguments, so an implementation trying to copy these into a local
368 buffer was taking a taint-enforcement trap. Fix by using dynamically
369 created buffers. Similar fix for radius expansion condition.
371 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
372 reasonable, eg. to count headers. Fix by using dynamically created
373 buffers rather than a local. Do similar fixes for ACL actions "dcc",
374 "log_reject_target", "malware" and "spam"; the arguments are expanded
375 so could be handling tainted values.
377 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
378 broken the (no-op) support for this sendmail command. Restore it
379 to doing nothing, silently, and returning good status.
381 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
382 record path was given (or the default used) without a leading directory
383 path, an error occurred on trying to open it. Use the transport's working
386 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
387 transport. Previously it was the name on the DNS A-record; use instead
388 the head of the CNAME chain leading there (if there is one). This seems
389 to align better with RFC 6125.
391 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
392 smtp_accept_max_per_host allocated resources which were not released
393 when the limit was exceeded. This eventually crashed the daemon. Fix
394 by adding a release action in that path.
396 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
397 expanded; previously using tainted values was rejected. Fix by using
398 dynamically-created buffers.
400 JH/09 Relax restrictions on ACL verify condition needing access to message
401 headers. Previously they were only permitted in data and non-smtp ACLs;
402 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
403 not_blind, header_sender and header_names_ascii verification.
405 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
406 Previously a macro used one argument twice; when called with the
407 argument as an expression having side-effects, incorrect operation
408 resulted. Use an inlineable function.
410 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
411 held open for a verify callout. Previously this wan not accounted for
412 and a corrupt onward SMTP conversation resulted.
414 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
415 Previously they were embedded naively in the constructed address; when
416 needed, strip the quoting and quote the entire local_part.
417 Also make the inbound_srs expansion condition handle quoting.
419 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
420 excluded, not matching the documentation.
422 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
423 was given for the sqlite_dbfile a trap resulted.
425 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
426 "name" argument resulted in a trap. There is no reason to disallow such;
427 this was a coding error.
429 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
430 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
431 spent suspended, ignoring the POSIX definition. Previously we assumed
432 it did and a constant offset from real time could be used as a correction.
433 Change to using the same clock source for the start-of-message and the
434 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
435 exists, just to get a clock slightly more aligned to reality.
437 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
438 RFC says it is optional some validators care. The missing char was not
439 intended but triggered by a line-wrap alignment. Discovery and fix by
440 Guillaume Outters, hacked on by JH.
442 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
443 name being quoted was tainted a trap would be taken. Fix by using
444 dynamically created buffers. The routine could have been called by a
445 rewrite with the "h" flag, by using the "-F" command-line option, or
446 by using a "name=" option on a control=submission ACL modifier.
448 JH/19 SPF: change the Authentication-Results expansion component to give
449 smtp.helo when the sender domain is empty. Previously it gave
452 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
453 not in 127.0.0.0/8 to help in spotting list domains taken over by a
454 domain-parking registrar.
456 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
457 Previously when a whitespace character was specified it was not inserted
458 after removing the newline.
460 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
461 the domain part of the recipient address. This overrides any tls_sni
462 option set, which was previously used.
464 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
467 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
468 is_tainted() had an off-by-one error in the overenthusiastic direction.
469 Find and fix by Gavan. Although NetBSD is not a supported platform for
470 4.94 this bug could affect other platforms.
472 PP/01 Fix default prime selection to be consistent.
473 One path used ike23 still, instead of exim.dev.20160529.3; now both
474 execution flows will use the same DH primes (currently
475 exim.dev.20160529.3).
477 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
478 option in smtp transport, to match the documentation. Previously
479 verification was not being done in some cases where it should have been.
481 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
482 than one server was defined and depending on the platform memory layout
483 details, an internal consistency trap could be hit while walking the list
486 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
487 files. The value is used by the authresults expansion item. Previously
488 if this was used in a router or transport, a crash could result.
490 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
491 have a sufficient incidence of EINTR returns from select that an
492 interaction with other operations done by the main daemon loop exposed
493 a bug in the error-handling. This was benign apart from the log
496 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
497 for consistency with delivered "=>" and failed "**" lines. While we're
498 there, handle PRX and TFO.
500 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
501 applied. This resulted, if any header-line rewrite rules were configured,
502 in a panic-log triggerable by sending a message with a long address in
503 a header. Fix by increasing the arbitrary limit to larger than a single
504 (dewrapped) 5322 header line maximum size.
506 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
507 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
508 RFC 6648 which deprecates X- options in protocols as a general practice.
509 Changeover between the implementations is handled by the mechanisms
512 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
513 a remote which uses callout/hold. Previously the local was queued.
515 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
518 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
519 log line. The asterisk is now consistently appended to the remote IP
520 (and port, if given), and will also be provided on defer and fail log
521 lines. Previously it could be placed on the local IP if that was being
522 logged, and was only provided on delivery lines.
524 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
526 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
527 authenticator client_send option. Previously the next char, after a pair
528 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
529 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
530 documentation. There is still no way to get a leading ^ immediately
531 after a NUL (ie. for the password of a PLAIN method authenticator.
533 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
534 files. For bad sizes read, delete the record and whine to paniclog.
536 JH/38 When logging an AUTH failure, as server, do not include sensitive
537 information. Previously, the credentials would be included if given
538 as part of the AUTH command line and an ACL denied authentication.
540 JH/39 Bug 2691: fix $local_part_data. When the matching list element
541 referred to a file, bad data was returned. This likely also affected
544 JH/40 The gsasl authenticator now supports caching of the salted password
545 generated by the client-side implementation. This required the addition
546 of a new variable: $auth4.
548 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
549 left undeleted; the attempt to re-create it then failed - resulting in
550 the usual "SIGHUP tp have daemon reload configuration" to not work.
551 This affected any platform not supporting "abstract" Unix-domain
552 sockets (i.e. not Linux).
554 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
555 recipients" response to RCPT in a later response, with a 250. The
556 previous coding assumed this would not happen, and under PIPELINING
557 would result in both lost and duplicate recipients for a message.
559 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
560 Previously the weighting was incorrectly applied. Similar fix for socks
561 proxies. Found and fixed by Heiko Schlichting.
563 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
564 not handle sub-lists included using the +namedlist syntax. While
565 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
566 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
569 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
570 Previously the permanent pool was used, so the sore could not be freed.
571 This meant a connection with many messages would use continually-growing
574 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
575 per-pool so as not to waste too much space. Previously a constant size
576 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
577 DOS attacks harder. The cost is wasted memory use in the larger blocks.
579 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
580 same space repeatedly, and a smaller process footprint.
582 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
583 for GnuTLS. Previously the name started "%s" which, while not a bug,
584 looked as if if might be one.
586 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
587 "continued connection") make the $proxy_* variables available. Previously
588 the information was not passed across the exec() call for subsequent
589 transport executions. This also mean that the log lines for the
590 messages can show the proxy information.
592 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
593 receive time. With modern systems the difference is significant.
594 The historical behaviour can be restored by disabling (a new) log_selector
595 "queue_time_exclusive".
597 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
598 content) were specifically tested for. Now, also cover expansions
599 resulting in ACL names and inline ACL content.
601 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
602 making it unusable in complex expressions.
604 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
605 fail with a taint-check on the spoolfile name, and leave the message
608 HS/01 Enforce absolute PID file path name.
610 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
612 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
614 PP/02 Bug 2643: Correct TLS DH constants.
615 A missing NUL termination in our code-generation tool had led to some
616 incorrect Diffie-Hellman constants in the Exim source.
617 Reported by kylon94, code-gen tool fix by Simon Arlott.
619 PP/03 Impose security length checks on various command-line options.
620 Fixes CVE-2020-SPRSS reported by Qualys.
622 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
623 better. Reported by Qualys.
625 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
626 providing a particularly obnoxious sender full name.
629 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
631 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
634 PP/08 Change default for recipients_max from unlimited to 50,000.
636 PP/09 Fix security issue with too many recipients on a message (to remove a
637 known security problem if someone does set recipients_max to unlimited,
638 or if local additions add to the recipient list).
639 Fixes CVE-2020-RCPTL reported by Qualys.
641 PP/10 Fix security issue in SMTP verb option parsing
642 Fixes CVE-2020-EXOPT reported by Qualys.
644 PP/11 Fix security issue in BDAT state confusion.
645 Ensure we reset known-good where we know we need to not be reading BDAT
646 data, as a general case fix, and move the places where we switch to BDAT
647 mode until after various protocol state checks.
648 Fixes CVE-2020-BDATA reported by Qualys.
650 HS/03 Die on "/../" in msglog file names
652 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
653 the Exim runtime user are allowed to create files.
655 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
658 QS/03 When reading the output from interpreted forward files we do not
659 pass the pipe between the parent and the interpreting process to
660 executed child processes (if any).
662 QS/04 Always die if requested from internal logging, even is logging is
665 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
666 the API; compilation noo longer completes with DMARC support included.
667 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
668 been reported on other platforms.
670 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
672 JH/56 Make the majority of info read from config files readonly, for defence-in-
673 depth against exploits. Suggestion by Qualys.
674 Not supported on Solaris 10.
676 JH/57 Fix control=fakreject for a custom message containing tainted data.
677 Previously this resulted in a log complaint, due to a re-expansion present
678 since fakereject was originally introduced.
680 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
681 modifier was given, a loop resulted.
683 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
684 pipelined SMTP command followed the BDAT LAST then it would be
685 incorrectly treated as part of the message body, causing a verification
688 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
689 require looser validation rules than those for 821-level addresses,
690 which only permit IP addresses.
696 JH/01 Avoid costly startup code when not strictly needed. This reduces time
697 for some exim process initialisations. It does mean that the logging
698 of TLS configuration problems is only done for the daemon startup.
700 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
702 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
703 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
706 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
707 DKIM signing is being done. Previously a transport_filter would always
708 disable CHUNKING, falling back to traditional DATA.
710 JH/05 Regard command-line recipients as tainted.
712 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
714 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
715 PAM library frees one of the arguments given to it, despite the
716 documentation. Therefore a plain malloc must be used.
718 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
719 on-stack buffers were used, resulting in a taint trap when DSN information
720 copied from a received message was written into the buffer.
722 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
723 the ordering of its ARC headers. This caused a crash.
725 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
726 a new record was being constructed with information from the peer, a trap
729 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
730 installation would get error messages from DMARC verify, when it hit the
731 nonexistent file indicated by the default. Distros wanting DMARC enabled
732 should both provide the file and set the option.
733 Also enforce no DMARC verification for command-line sourced messages.
735 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
736 could, depending on the platform, hang at the STARTTLS response.
738 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
739 message on a connection. Previously if one message had ARC headers and
740 the following one did not, a crash could result when adding an
741 Authentication-Results: header.
743 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
744 Exim main code and Exim-related utities. The introduction of taint
745 tracking also did many adjustments to string handling. Since then, eximon
746 frequently terminated with an assert failure.
748 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
749 check for 452 responses. This slightly helps the inefficieny of doing
750 a large alias-expansion into a recipient-limited target. The max_rcpt
751 transport option still applies (and at the current default, will override
752 the new feature). The check is done for either cause of synch, and forces
753 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
754 connection. The new facility is not tunable at this time.
756 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
757 library live data was being used, so the results became garbage. Make
758 copies while it is still usable.
760 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
761 on delivery deferred (==) and failed (**) lines (if a delivery was
762 attemtped). Previously it was only on completion (=>) lines.
764 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
765 for the expansion of the server_scram_iter and server_scram_salt options.
767 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
768 are now specifically given a NO_DATA response without hitting the system
769 resolver. The library goes on to do the now-standard TXT lookup.
770 Use of dnsdb lookups is not affected.
772 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
773 only retrieve the errormessage once. Previously two calls to dlerror()
774 were used, and the second one (for mainlog/paniclog) retrieved null
777 JH/20 Taint checking: disallow use of tainted data for
778 - the appendfile transport file and directory options
779 - the pipe transport command
780 - the autoreply transport file, log and once options
781 - file names used by the redirect router (including filter files)
783 - paths used by single-key lookups
784 Previously this was permitted.
786 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
787 adjusted the size of a major service buffer; this failed because the
788 buffer was in use at the time. Change to a compile-time increase in the
789 buffer size, when this authenticator is compiled into exim.
791 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
792 previous fast-mode was untenable in the face of glibs using mmap to
793 support larger malloc requests.
795 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
796 New values supported, if defined on system where compiled:
797 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
798 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
800 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
801 running a limited number of proceses in parallel, a benefit is gained. The
802 amount varies with the platform hardware and load. The use of the option
803 queue_run_in_order means we cannot do this, as ordering becomes
806 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
807 had introduced a string-copy (for ensuring NUL-termination) which was not
808 appropriate for that case, which can include embedded NUL bytes in the
809 block of data. Investigation showed the copy to actually be needless, the
810 data being length-specified.
812 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
813 done during a receiving connection, and both used TLS, global info was
814 used rather than per-connection info for tracking the state of data
815 queued for transmission. This could result in a connection hang.
817 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
818 Previously, when delivering serveral messages down a single connection
819 only the first would provide a SIZE. This was due to the size information
820 not being properly tracked.
822 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
823 TAI (at 37 seconds currently), pretend to be in UTC for time-related
824 expansion and logging. Previously, spurious values such as a future
825 minute could be seen.
827 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
828 it could crash from a null-deref. This could also affect the
829 ${addresses: } operator and ${readsock } item.
831 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
832 message following a mime one, the variable was not reset.
834 JH/30 When an pipelined-connect fails at the first response, assume incorrect
835 cached capability (perhaps the peer reneged?) and immediately retry in
838 JH/31 Fix spurious detection of timeout while writing to transport filter.
840 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
841 an attempt to copy the string was made before checking it.
843 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
844 taint of the lookup key was maintained; we now regard the presence in the
845 filesystem as sufficient validation.
847 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
848 argument is supplied.
850 JH/35 The dsearch lookup now requires that the directory is an absolute path.
851 Previously this was not checked, and nonempty relative paths made an
852 access under Exim's current working directory.
854 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
855 Previously no event was raised.
857 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
858 parameter supplied by the sender MAIL FROM command. Previously it was
859 ignored, and only the check_spool_space option value for the required
862 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
863 the size of the signing public-key. Previously it was instead giving
864 the size of the signature hash.
866 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
867 the default. See the (new) dkim_verify_min_keysizes option.
869 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
870 and an ACL use a lookup for checking either the local_part or domain,
871 stale data could be accessed. Ensure that variable references are
872 dropped between messages.
874 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
875 by the client was not checked as pointing within response data before
876 being used. A malicious client could thus cause an out-of-bounds read and
877 possibly gain authentication. Fix by adding the check.
879 JH/42 Internationalisation: change the default for downconversion in the smtp
880 transport to be "if needed". Previously it was "as previously set" for
881 the message, which usually meant "if needed" for message-submission but
882 "no" for everything else. However, MTAs have been seen using SMTPUTF8
883 even when the envelope addresses did not need it, resulting in forwarding
884 failures to non-supporting MTAs. A downconvert in such cases will be
885 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
886 transport. The change does mean that addresses needing conversion will
887 be converted when previously a delivery failure would occur.
889 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
890 response was received it would be used unchecked in a fail-DSN, violating
891 standards on line-length limits. Truncate if needed.
893 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
894 doesn't work. (Additionally add a new main config option to configure the
901 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
902 side, to decode a TLS 1.3 packet capture.
904 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
905 Previously the default library behaviour applied, sending two, each in
908 JH/03 Debug output for ACL now gives the config file name and line number for
911 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
913 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
915 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
916 buffer overrun for (non-chunking) other transports.
918 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
919 TLS1.3, means that a server rejecting a client certificate is not visible
920 to the client until the first read of encrypted data (typically the
921 response to EHLO). Add detection for that case and treat it as a failed
922 TLS connection attempt, so that the normal retry-in-clear can work (if
923 suitably configured).
925 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
926 and/or domain. Found and fixed by Jason Betts.
928 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
929 configuration). If a CNAME target was not a wellformed name pattern, a
932 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
933 the OS reports them interleaved with other addresses.
935 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
936 used both for input and for a verify callout, both encrypted, SMTP
937 responses being sent by the server could be lost. This resulted in
938 dropped connections and sometimes bounces generated by a peer sending
941 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
942 string. Previously it was possible to cause undefined behaviour in a
943 library routine (usually a crash). Found by "zerons".
945 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
948 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
949 API was removed, so update to use the newer ones.
951 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
952 any timeout set, is taking a long time. Previously we would hang on to a
953 rotated logfile "forever" if the input was arriving with long gaps
954 (a previous attempt to fix addressed lack, for a long time, of initial
957 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
958 shared (NFS) environment. The length of the tempfile name is now
959 4 + 16 ("hdr.$message_exim_id") which might break on file
960 systems which restrict the file name length to lower values.
963 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
964 shared (NFS) environment.
966 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
967 did for all versions <4.90). Notably -M, -m, --invert, -I may be
970 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
971 on some platforms for bit 31.
973 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
974 to changes apparently associated with TLS1.3 handling some of the APIs
975 previously used were either nonfunctional or inappropriate. Strings
976 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
977 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
978 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
979 This affects log line X= elements, the $tls_{in,out}_cipher variables,
980 and the use of specific cipher names in the encrypted= ACL condition.
982 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
984 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
985 verification result was not updated unless hosts_require_ocsp applied.
987 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
988 queue_list_requires_admin set to false, non-admin users were denied the
991 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
992 directory-of-certs mode. Previously they were advertised despite the
995 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
996 A single TCP connection by a client will now hold a TLS connection open
997 for multiple message deliveries, by default. Previously the default was to
1000 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1001 default. If built with the facility, DANE will be used. The facility
1002 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1004 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1005 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1006 must be defined and you must still, unless you define DISABLE_TLS, manage
1007 the the include-dir and library-file requirements that go with that
1008 choice. Non-TLS builds are still supported.
1010 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1013 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1014 default. If the platform supports and has the facility enabled, it will
1015 be requested on all coneections.
1017 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1018 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1020 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1022 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1023 success-DSN messages. Previously the From: header was always the default
1024 one for these; the option was ignored.
1026 JH/28 Fix the timeout on smtp response to apply to the whole response.
1027 Previously it was reset for every read, so a teergrubing peer sending
1028 single bytes within the time limit could extend the connection for a
1029 long time. Credit to Qualsys Security Advisory Team for the discovery.
1031 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1032 delivery address, which leaked information of the results of local
1033 forwarding. Change to the original envelope recipient address, per
1036 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1037 requested. Previously not bounce was generated and a log entry of
1038 error ignored was made.
1040 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1042 JH/32 Introduce a general tainting mechanism for values read from the input
1043 channel, and values derived from them. Refuse to expand any tainted
1044 values, to catch one form of exploit.
1046 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1047 was unused and the unexpanded text used for the test. Found and
1048 fixed by Ruben Jenster.
1050 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1051 an attempt to use a TLS library read routine dereffed a nul pointer,
1054 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1055 them in our smtp response.
1057 JH/36 Have the general router option retry_use_local_part default to true when
1058 any of the restrictive preconditions are set (to anything). Previously it
1059 was only for check_local user. The change removes one item of manual
1060 configuration which is required for proper retries when a remote router
1061 handles a subset of addresses for a domain.
1063 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1064 link count into consideration.
1066 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1067 caused the extension of big_buffer, the following lines were ignored.
1069 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1070 accordance with RFC 2308. Previously there was no expiry, so a longlived
1071 receive process (eg. due to ACL delays) versus a short SOA value could
1074 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1076 JH/39 Promote DMARC support to mainline.
1078 JH/40 Bug 2452: Add a References: header to DSNs.
1080 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1081 parameters. The relevant library call is documented as "Deprecated: This
1082 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1083 3.6.0, DH parameters are negotiated following RFC7919."
1085 HS/06 Change the default of dnssec_request_domains to "*"
1087 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1088 carried on and emitted a BDAT command, even when PIPELINING was not
1091 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1092 buffer was used for the filename, resulting in a trap when tainted
1093 arguments (eg. $domain) were used.
1095 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1096 recommended to avoid a possible server-load attack. The feature can be
1097 re-enabled via the openssl_options main cofiguration option.
1099 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1100 for version 4.90 - adding a "more data" boolean to the arguments.
1101 Bumped the ABI version number also, this having been missed previously;
1102 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1103 respect of usage of smtp_printf() by either local_scan code or libraries
1104 accessed via the ${dlfunc } expansion item. Both will need coding
1105 adjustment for any calls to smtp_printf() to match the new function
1106 signature; a FALSE value for the new argument is always safe.
1108 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1109 the file-offset (which the Linux syscall does, and exim expects); this
1110 resulted in an indefinite loop.
1112 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1113 to do ARC verification. The Authentication-Results: header line added
1114 by the configuration then had no ARC item.
1116 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1117 an incorrect mode was used for reading the result, resulting in it being
1124 JH/01 Remove code calling the customisable local_scan function, unless a new
1125 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1127 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1128 non-signal-safe functions being used.
1130 JH/03 Bug 2269: When presented with a received message having a stupidly large
1131 number of DKIM-Signature headers, disable DKIM verification to avoid
1132 a resource-consumption attack. The limit is set at twenty.
1134 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1135 report of oldest_pass in ${authres } in consequence, and separate out
1136 some descriptions of reasons for verification fail.
1138 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1139 files in the spool were present and unlocked. A queue-runner could spot
1140 them, resulting in a duplicate delivery. Fix that by doing the unlock
1141 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1142 add more error-checking on spoolfile handling while that code is being
1145 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1146 No known attacks, no CVE, this is defensive hardening.
1148 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1149 a queue-runner could start a delivery while other operations were ongoing.
1150 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1151 Found and investigated by Tim Stewart. Fix by using the open message data
1152 file handle rather than opening another, and not locally closing it (which
1153 releases a lock) for that case, while creating the temporary .eml format
1154 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1156 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1157 $sender_verify_failure/$recipient_verify_failure to "random".
1159 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1162 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1163 Previously this would segfault.
1165 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1168 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1169 like zero, since the resolver should be doing this for us, But we need one
1170 as a CNAME but no MX presence gets the CNAME returned; we need to check
1171 that doesn't point to an MX to declare it "no MX returned" rather than
1172 "error, loop". A new main option is added so the older capability of
1173 following some limited number of chain links is maintained.
1175 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1177 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1178 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1179 now-deprecated earlier definitions used only specified the range up to TLS
1180 1.2 (in the older-version library docs).
1182 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1184 JH/15 Rework TLS client-side context management. Stop using a global, and
1185 explicitly pass a context around. This enables future use of TLS for
1186 connections to service-daemons (eg. malware scanning) while a client smtp
1187 connection is using TLS; with cutthrough connections this is quite likely.
1189 JH/16 Fix ARC verification to do AS checks in reverse order.
1191 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1193 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1194 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1195 Previously the "utf8" would be re-prepended for every additional message.
1197 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1198 Previously thery were accepted, resulting in issues when attempting to
1199 forward messages to a non-supporting MTA.
1201 PP/02 Let -n work with printing macros too, not just options.
1203 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1204 one parent address was copied, and bogus data was used at delivery-logging
1205 time. Either a crash (after delivery) or bogus log data could result.
1206 Discovery and analysis by Tim Stewart.
1208 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1209 Previously if the string ended mid-character, we did not insert the
1210 promised '?' replacement.
1212 PP/04 Documentation: current string operators work on bytes, not codepoints.
1214 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1215 should pack well giving a smaller memory footprint so better caching and
1216 therefore performance. Group the declarations where this can't be done so
1217 that the byte-sized flag variables are not interspersed among pointer
1218 variables, giving a better chance of good packing by the compiler.
1220 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1221 non-null, to avoid issues with sites running BATV. Previously reports were
1222 sent with an empty envelope sender so looked like bounces.
1224 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1225 The ignore_error flag wasn't being returned from the filter subprocess so
1226 was not set for later routers. Investigation and fix by Matthias Kurz.
1228 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1229 and a msg:complete for the whole, when a message is manually removed using
1230 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1232 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1233 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1234 hope that is portable enough.
1236 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1237 requiring a known-CA anchor certificate; make it now rely entirely on the
1238 TLSA as an anchor. Checking the name on the leaf cert against the name
1239 on the A-record for the host is still done for TA (but not for EE mode).
1241 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1242 included in delivery lines for non-proxied connections, when compiled with
1243 SUPPORT_SOCKS and running with proxy logging enabled.
1245 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1246 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1247 move the existing event to fire before the normal logging of message
1248 failure so that custom logging is bracketed by normal logging.
1250 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1251 msg:fail:internal event. Developement by Matthias Kurz.
1253 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1254 far too small for todays use of crypto signatures stored there. Go all
1255 the way to the max DNS message size of 64kB, even though this might be
1256 overmuch for IOT constrained device use.
1258 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1259 copy a string over itself. The library routine is documented as not
1260 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1262 JH/32 For main options check_spool_space and check_inode_space, where the
1263 platform supports 64b integers, support more than the previous 2^31 kB
1264 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1265 the previous G, M, k.
1267 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1268 $authenticated_fail_id variable on authentication failure. Previously
1271 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1272 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1273 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1274 GNUTLS_SEC_PARAM_MEDIUM.
1276 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1277 side. Previously we would continue as if no SNI had been received.
1279 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1280 "<" character, which should be a list-separator specification, we walked
1281 off past the nul-terimation.
1283 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1284 causes) even when the retry time is not yet met. Previously they were
1285 not, meaning that when (say) an account was over-quota and temp-rejecting,
1286 and multiple senders' messages were queued, only one sender would get
1287 notified on each configured delay_warning cycle.
1289 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1291 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1292 cipher-suites, an error can be left on the stack even for a succeeding
1293 accept; this results in impossible error messages when a later operation
1296 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1297 return error codes indicating retry. Under TLS1.3 this becomes required.
1299 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1300 it only wrote the new authenticators, resulting in a lack of tracking of
1301 peer changes of ESMTP extensions until the next cache flush.
1303 JH/41 Fix the loop reading a message header line to check for integer overflow,
1304 and more-often against header_maxsize. Previously a crafted message could
1305 induce a crash of the recive process; now the message is cleanly rejected.
1307 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1308 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1314 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1315 When redis_servers is set to a list of > 1 element, and the Redis servers
1316 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1317 case of MOVED into a DEFER case instead, thus moving the query onto the
1318 next server in the list. For a cluster of N elements, all N servers must
1319 be defined in redis_servers.
1321 GF/02 Catch and remove uninitialized value warning in exiqsumm
1322 Check for existence of @ARGV before looking at $ARGV[0]
1324 JH/01 Replace the store_release() internal interface with store_newblock(),
1325 which internalises the check required to safely use the old one, plus
1326 the allocate and data copy operations duplicated in both (!) of the
1327 extant use locations.
1329 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1330 modifier. This matches the restriction on the commandline.
1332 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1333 Previously only the last row was returned.
1335 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1336 we assumed that tags in the header were well-formed, and parsed the
1337 element content after inspecting only the first char of the tag.
1338 Assumptions at that stage could crash the receive process on malformed
1341 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1342 While running the DKIM ACL we operate on the Permanent memory pool so that
1343 variables created with "set" persist to the DATA ACL. Also (at any time)
1344 DNS lookups that fail create cache records using the Permanent pool. But
1345 expansions release any allocations made on the current pool - so a dnsdb
1346 lookup expansion done in the DKIM ACL releases the memory used for the
1347 DNS negative-cache, and bad things result. Solution is to switch to the
1348 Main pool for expansions.
1349 While we're in that code, add checks on the DNS cache during store_reset,
1350 active in the testsuite.
1351 Problem spotted, and debugging aided, by Wolfgang Breyha.
1353 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1354 When none of the hosts presented to a transport match an already-open
1355 connection, close it and proceed with the list. Previously we would
1356 queue the message. Spotted by Lena with Yahoo, probably involving
1359 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1360 Previously a spurious "250 OK id=" response was appended to the proper
1363 JH/08 The "support for" informational output now, which built with Content
1364 Scanning support, has a line for the malware scanner interfaces compiled
1365 in. Interface can be individually included or not at build time.
1367 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1368 by the template makefile "src/EDITME". The "STREAM" support for an older
1369 ClamAV interface method is removed.
1371 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1372 rows affected is given instead).
1374 JH/11 The runtime Berkeley DB library version is now additionally output by
1375 "exim -d -bV". Previously only the compile-time version was shown.
1377 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1378 SMTP connection. Previously, when one had more recipients than the
1379 first, an abortive onward connection was made. Move to full support for
1380 multiple onward connections in sequence, handling cutthrough connection
1381 for all multi-message initiating connections.
1383 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1384 routers. Previously, a multi-recipient message would fail to match the
1385 onward-connection opened for the first recipient, and cause its closure.
1387 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1388 a timeout on read on a GnuTLS initiating connection, resulting in the
1389 initiating connection being dropped. This mattered most when the callout
1390 was marked defer_ok. Fix to keep the two timeout-detection methods
1393 JH/15 Relax results from ACL control request to enable cutthrough, in
1394 unsupported situations, from error to silently (except under debug)
1395 ignoring. This covers use with PRDR, frozen messages, queue-only and
1398 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1400 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1401 metadata, resulting in a crash in free().
1403 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1404 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1405 Broken also in d185889f4, with init system revamp.
1407 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1408 Previously we abruptly closed the connection after reading a malware-
1409 found indication; now we go on to read the "scan ok" response line,
1412 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1413 ACL. Previously, a crash would result.
1415 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1416 macro text after a replacement (previously it was only once per line) and
1417 by skipping builtin macros when searching for an uppercase lead character.
1419 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1420 for the build is renamed.
1422 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1423 was allocated for every new TLS startup, meaning one per message. Fix
1424 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1426 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1427 reported the original. Fix to report (as far as possible) the ACL
1428 result replacing the original.
1430 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1431 OpenSSL. Certificate information is loaded for every new TLS startup,
1432 and the resources needed to be freed.
1434 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1436 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1439 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1440 DATA response info to the (existing) per-recipient response info for
1441 the "C=" log element. It can have useful tracking info from the
1442 destination system. Patch from Simon Arlott.
1444 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1445 length value. Previously this would segfault.
1447 HS/02 Support Avast multiline protoocol, this allows passing flags to
1448 newer versions of the scanner.
1450 JH/28 Ensure that variables possibly set during message acceptance are marked
1451 dead before release of memory in the daemon loop. This stops complaints
1452 about them when the debug_store option is enabled. Discovered specifically
1453 for sender_rate_period, but applies to a whole set of variables.
1454 Do the same for the queue-runner and queue-list loops, for variables set
1455 from spool message files. Do the same for the SMTP per-message loop, for
1456 certain variables indirectly set in ACL operations.
1458 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1459 as a multi-recipient message from a mailinglist manager). The coding had
1460 an arbitrary cutoff number of characters while checking for more input;
1461 enforced by writing a NUL into the buffer. This corrupted long / fast
1462 input. The problem was exposed more widely when more pipelineing of SMTP
1463 responses was introduced, and one Exim system was feeding another.
1464 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1465 receiving system, and refused recipients seen by the sending system
1466 (propating to people being dropped from mailing lists).
1467 Discovered and pinpointed by David Carter.
1469 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1470 replaced by the ${authresults } expansion.
1472 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1474 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1475 allows proper process termination in container environments.
1477 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1478 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1480 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1481 and "err_perm", deprecated since 4.83 when the RFC-defined words
1482 "temperror" and "permerror" were introduced.
1484 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1485 transport-filters or DKIM-signing. The restriction was lost in the
1486 consolidation of verify-callout and delivery SMTP handling.
1487 Extend the restriction to also cover ARC-signing.
1489 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1490 in defer=pass mode supply a 450 to the initiator. Previously the message
1493 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1494 tls_require_ciphers is used as before.
1496 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1497 "pass_unscanned". Only tmpfails from the scanner are written to
1498 the paniclog, as they may require admin intervention (permission
1499 denied, license issues). Other scanner errors (like decompression
1500 bombs) do not cause a paniclog entry.
1502 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1503 Previously it was possible to log spurious information in receive log
1506 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1507 triggered odd behaviour from Outlook Express clients.
1509 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1512 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1513 since the IETF WG has not yet settled on that versus the original
1514 "bare" representation.
1516 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1517 Previously the millisecond value corrupted the output.
1518 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1519 corrupted the output.
1525 JH/01 Rework error string handling in TLS interface so that the caller in
1526 more cases is responsible for logging. This permits library-sourced
1527 string to be attached to addresses during delivery, and collapses
1528 pairs of long lines into single ones.
1530 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1531 during configuration. Wildcards are allowed and expanded.
1533 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1534 This permits better logging.
1536 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1537 do not close it down and have the child transport start it up again on
1538 the passed-on TCP connection. Instead, proxy the child (and any
1539 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1540 affected: the continued delivery log lines do not have any DNSSEC, TLS
1541 Certificate or OCSP information. TLS cipher information is still logged.
1543 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1544 identical IP addresses on different listening ports. Will also affect
1547 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1548 add noisy ifdef guards to special-case this sillyness.
1549 Patch from Bernd Kuhls.
1551 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1552 than 255 are no longer allowed.
1554 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1555 Disable the session-cache too, which might reduce our load. Since we
1556 currrectly use a new context for every connection, both as server and
1557 client, there is no benefit for these.
1558 GnuTLS appears to not support tickets server-side by default (we don't
1559 call gnutls_session_ticket_enable_server()) but client side is enabled
1560 by default on recent versions (3.1.3 +) unless the PFS priority string
1563 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1564 <https://reproducible-builds.org/specs/source-date-epoch/>.
1566 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1567 the check for any unsuccessful recipients did not notice the limit, and
1568 erroneously found still-pending ones.
1570 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1571 MSG_MORE. Only in-clear (not on TLS connections).
1573 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1574 transport-filter is involved we can buffer the headers in memory for
1575 creating the signature, and read the spool data file once for the
1576 signature and again for transmission.
1578 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1579 4.77 as the kernel support then wasn't solid, having issues in 64bit
1580 mode. Now, it's been long enough. Add support for FreeBSD also.
1582 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1583 case where the routing stage had gathered several addresses to send to
1584 a host before calling the transport for the first, we previously failed
1585 to close down TLS in the old transport process before passing the TCP
1586 connection to the new process. The new one sent a STARTTLS command
1587 which naturally failed, giving a failed delivery and bloating the retry
1588 database. Investigation and fix prototype from Wolfgang Breyha.
1590 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1591 false-negatives in the check that the sender had not preempted a response
1592 or prompt from Exim (running as a server), due to that code's lack of
1593 awareness of the SMTP input buffering.
1595 PP/04 Add commandline_checks_require_admin option.
1596 Exim drops privileges sanely, various checks such as -be aren't a
1597 security problem, as long as you trust local users with access to their
1598 own account. When invoked by services which pass untrusted data to
1599 Exim, this might be an issue. Set this option in main configuration
1600 AND make fixes to the calling application, such as using `--` to stop
1603 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1604 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1605 responses to those, into a single TLS record each way (this usually means
1606 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1609 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1610 affects you only if you're dancing at the edge of the param size limits.
1611 If you are, and this message makes sense to you, then: raise the
1612 configured limit or use OpenSSL 1.1. Nothing we can do for older
1615 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1616 cmdline element to get the documented default one. Previously it was
1619 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1620 are unsuable for non-message-specific reasons - eg. connection timeout,
1621 banner-time rejection.
1623 JH/16 Fix logging of delivery remote port, when specified by router, under
1626 PP/06 Repair manualroute's ability to take options in any order, even if one
1627 is the name of a transport.
1630 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1632 JH/17 Change the list-building routines interface to use the expanding-string
1633 triplet model, for better allocation and copying behaviour.
1635 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1636 Previously it was constructed the first time a possibly-matching string
1637 was met in the configuration file input during startup; now it is done
1640 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1641 compatible one, to avoid the (poorly documented) possibility of a config
1642 file in the working directory redirecting the DB files, possibly correpting
1643 some existing file. CVE-2017-10140 assigned for BDB.
1645 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1646 cache-hot. Previously, although the result was properly cached, the
1647 initial verify call returned a defer.
1649 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1650 the main verify for receipient in uncached-mode.
1652 JH/22 Retire historical build files to an "unsupported" subdir. These are
1653 defined as "ones for which we have no current evidence of testing".
1655 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1656 if present. Previously it was ignored.
1658 JH/24 Start using specified-initialisers in C structure init coding. This is
1659 a C99 feature (it's 2017, so now considered safe).
1661 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1662 if was a fixed-sized field and bitmask ops via macros; it is now more
1665 PP/07 GitHub PR 56: Apply MariaDB build fix.
1666 Patch provided by Jaroslav Škarvada.
1668 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1669 during Coverity cleanups [4.87 JH/47]
1670 Diagnosis and fix provided by Michael Fischer v. Mollard.
1672 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1673 the right size to place the terminating semicolon on its own folded
1674 line, the header hash was calculated to an incorrect value thanks to
1675 the (relaxed) space the fold became.
1677 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1678 and confused the parent.
1680 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1681 which could crash as a result. This could lead to undeliverable messages.
1683 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1686 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1687 matters on fast-turnover and PID-randomising systems, which were getting
1688 out-of-order delivery.
1690 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1691 a possibly-overlapping copy. The symptom was that "Remote host closed
1692 connection in response to HELO" was logged instead of the actual 4xx
1695 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1696 Previously only that bufferd was discarded, resulting in SYMTP command
1699 JH/32 DKIM: when a message has multiple signatures matching an identity given
1700 in dkim_verify_signers, run the dkim acl once for each. Previously only
1701 one run was done. Bug 2189.
1703 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1704 "panic the current process" to "deliberately defer". The panic log is
1705 still written with the problem list name; the mail and reject logs now
1706 get a temp-reject line for the message that was being handled, saying
1707 something like "domains check lookup or other defer". The SMTP 451
1708 message is still "Temporary local problem".
1710 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1711 A crafted sequence of BDAT commands could result in in-use memory beeing
1712 freed. CVE-2017-16943.
1714 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1715 from SMTP input. Previously it was always done; now only done for DATA
1716 and not BDAT commands. CVE-2017-16944.
1718 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1719 to the message (such as an overlong header line). Previously this was
1720 not done and we did not exit BDAT mode. Followon from the previous item
1721 though a different problem.
1727 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1728 than -2003 did; needs libidn2 in addition to libidn.
1730 JH/02 The path option on a pipe transport is now expanded before use.
1732 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1733 Patch provided by "Björn", documentation fix added too.
1735 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1736 missing a wire-to-host endian conversion.
1738 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1739 close after a BDAT command line could be taken as a following command,
1740 giving a synch failure. Fix by only checking for synch immediately
1741 before acknowledging the chunk.
1743 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1744 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1745 macro. Patches provided by Josh Soref.
1747 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1748 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1749 by IANA as a service extension. However, John Klensin suggests that we
1752 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1753 the dkim code may be unix-mode line endings rather than smtp wire-format
1754 CRLF, so prepend a CR to any bare LF.
1756 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1757 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1759 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1760 the first were themselves being wrongly included in the feed into dkim
1761 processing; with most chunk sizes in use this resulted in an incorrect
1762 body hash calculated value.
1764 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1765 DKIM signature block, for verification. Although advised against by
1766 standards it is specifically not ruled illegal.
1768 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1770 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1771 missing a body hash (the bh= tag).
1773 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1774 It seems that HAProxy sends the Proxy Protocol information in clear and
1775 only then does a TLS startup, so do the same.
1777 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1778 TCP connections (such as for Spamd) unless the daemon successfully set
1779 Fast Open mode on its listening sockets. This fixes breakage seen on
1780 too-old kernels or those not configured for Fast Open, at the cost of
1781 requiring both directions being enabled for TFO, and TFO never being used
1782 by non-daemon-related Exim processes.
1784 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1785 endings, at least on the first header line. Try to canonify any that get
1786 past that check, despite the cost.
1788 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1789 now limited to an arbitrary five deep, while parsing addresses with the
1790 strip_excess_angle_brackets option enabled.
1792 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1793 instead leave the unprompted TLS handshake in socket buffer for the
1794 TLS library to consume.
1796 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1798 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1800 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1801 region in one operation, for speed, and this leaves assigned pointers
1802 dangling. Add checks run only under the testsuite which checks all
1803 variables at a store-reset and panics on a dangling pointer; add code
1804 explicitly nulling out all the variables discovered. Fixes one known
1805 bug: a transport crash, where a dangling pointer for $sending_ip_address
1806 originally assigned in a verify callout, is re-used.
1808 PP/06 Drop '.' from @INC in various Perl scripts.
1810 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1812 PP/08 Reduce a number of compilation warnings under clang; building with
1813 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1814 should be warning-free.
1816 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1818 HS/01 Fix portability problems introduced by PP/08 for platforms where
1819 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1821 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1822 chunk. This allows us to accept broken chunked messages. We need a more
1823 general solution here.
1825 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1826 already-broken messages in the queue.
1828 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1830 JH/19 Fix reference counting bug in routing-generated-address tracking.
1836 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1837 supports it and a size is available (ie. the sending peer gave us one).
1839 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1840 years of being deprecated). The replacements are the ACLs
1841 acl_smtp_mime and acl_not_smtp_mime.
1843 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1844 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1845 or even an in-clear connection were permitted. Now, if the host lookup
1846 was dnssec and dane was requested then the host is only used if the
1847 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1848 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1849 if one fails this test.
1850 This means that a poorly-configured remote DNS will make it incommunicado;
1851 but it protects against a DNS-interception attack on it.
1853 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1854 non-noisy when a race steals the message being considered.
1856 JH/05 If main configuration option tls_certificate is unset, generate a
1857 self-signed certificate for inbound TLS connections.
1859 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1860 in rewrites and routers.
1862 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1863 and logged a warning sing 4.83; now they are a configuration file error.
1865 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1866 (lacking @domain). Apply the same qualification processing as RCPT.
1868 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1870 JH/10 Support ${sha256:} applied to a string (as well as the previous
1873 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1874 a cutthrough deliver is pending, as we always want to make a connection.
1875 This also avoids re-routing the message when later placing the cutthrough
1876 connection after a verify cache hit.
1877 Do not update it with the verify result either.
1879 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1880 when routing results in more than one destination address.
1882 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1883 signing (which inhibits the cutthrough capability). Previously only
1884 the presence of an option was tested; now an expansion evaluating as
1885 empty is permissible (obviously it should depend only on data available
1886 when the cutthrough connection is made).
1888 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1889 the relevant preceding SMTP command did not note the pipelining mode.
1891 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1892 Previously they were not counted.
1894 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1895 as one having no matching records. Previously we deferred the message
1896 that needed the lookup.
1898 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1899 distinguished as "(=".
1901 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1902 for missing MX records. Previously it only worked for missing A records.
1904 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1906 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1907 after the data-go-ahead and data-ack. Patch from Jason Betts.
1909 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1910 even for a "none" policy. Patch from Tony Meyer.
1912 JH/22 Fix continued use of a connection for further deliveries. If a port was
1913 specified by a router, it must also match for the delivery to be
1916 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1917 When one of the recipients of a message was unsuitable for the connection
1918 (has no matching addresses), we lost track of needing to mark it
1919 deferred. As a result mail would be lost.
1921 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1923 JH/25 Decoding ACL controls is now done using a binary search; the source code
1924 takes up less space and should be simpler to maintain. Merge the ACL
1925 condition decode tables also, with similar effect.
1927 JH/26 Fix problem with one_time used on a redirect router which returned the
1928 parent address unchanged. A retry would see the parent address marked as
1929 delivered, so not attempt the (identical) child. As a result mail would
1932 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1933 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1934 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1937 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1938 Main config options check_{log,spool}_{inodes,space} are now
1939 100 inodes, 10MB unless set otherwise in the configuration.
1941 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1942 Previously it only applied to the main-section connection policy
1945 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1947 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1948 by me. Added RFC7919 DH primes as an alternative.
1950 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1951 are not in the system include path.
1953 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1954 GnuTLS, when a session startup failed (eg because the client disconnected)
1955 Exim did stdio operations after fclose. This was exposed by a recent
1956 change which nulled out the file handle after the fclose.
1958 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1959 signed directly by the cert-signing cert, rather than an intermediate
1960 OCSP-signing cert. This is the model used by LetsEncrypt.
1962 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1964 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1965 an incoming connection.
1967 HS/02 Bug 1802: Do not half-close the connection after sending a request
1970 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1971 fallback to "prime256v1".
1973 JH/34 SECURITY: Use proper copy of DATA command in error message.
1974 Could leak key material. Remotely exploitable. CVE-2016-9963.
1980 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1981 and 3.4.4 - once the server is enabled to respond to an OCSP request
1982 it does even when not requested, resulting in a stapling non-aware
1983 client dropping the TLS connection.
1985 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1986 support variable-length bit vectors. No functional change.
1988 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1989 The I= interface field on outgoing lines is now after the H= remote
1990 host field, same as incoming lines. There is a separate
1991 outgoing_interface log selector which allows you to disable the
1994 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1995 If not running log_selector +smtp_connection the mainlog would be held
1996 open indefinitely after a "too many connections" event, including to a
1997 deleted file after a log rotate. Leave the per net connection logging
1998 leaving it open for efficiency as that will be quickly detected by the
1999 check on the next write.
2001 HS/01 Bug 1671: Fix post transport crash.
2002 Processing the wait-<transport> messages could crash the delivery
2003 process if the message IDs didn't exist for some reason. When
2004 using 'split_spool_directory=yes' the construction of the spool
2005 file name failed already, exposing the same netto behaviour.
2007 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2008 mime_regex ACL conditions.
2010 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2011 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2012 to HELO, local diagnostic string.
2014 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2015 log line to debug. Even when configured with a tls authenticator many
2016 client connections are expected to not authenticate in this way, so
2017 an authenticate fail is not an error.
2019 HS/02 Add the Exim version string to the process info. This way exiwhat
2020 gives some more detail about the running daemon.
2022 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2023 matter for fast-change records such as DNSBLs.
2025 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2026 retry record, even if constant. There may be multiple transports with
2027 different interface settings and the retry behaviour needs to be kept
2030 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2032 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2034 JH/10 Bug 840: fix log_defer_output option of pipe transport
2036 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2037 hosts_randomize. This matters a lot when combined with mua_wrapper.
2039 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2040 ${quote_pgsql:<string>} operator.
2042 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2044 JH/14 Change header file name for freeradius-client. Relevant if compiling
2045 with Radius support; from the Gentoo tree and checked under Fedora.
2047 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2049 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2050 option transport_filter, meaning no filtering.
2052 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2054 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2055 defaults to "*" (all hosts). The variable is now available when not built
2056 with TLS, default unset, mainly to enable keeping the testsuite sane.
2057 If a server certificate is not supplied (via tls_certificate) an error is
2058 logged, and clients will find TLS connections fail on startup. Presumably
2059 they will retry in-clear.
2060 Packagers of Exim are strongly encouraged to create a server certificate
2061 at installation time.
2063 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2064 with the $config_file variable.
2066 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2067 in transport context, after the attempt, and per-recipient. The latter type
2068 is per host attempted. The event data is the error message, and the errno
2069 information encodes the lookup type (A vs. MX) used for the (first) host,
2070 and the trailing two digits of the smtp 4xx response.
2072 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2073 to write to mainlog (or rejectlog, paniclog) in the window between file
2074 creation and permissions/ownership being changed. Particularly affects
2075 installations where exicyclog is run as root, rather than exim user;
2076 result is that the running daemon panics and dies.
2078 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2080 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2081 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2082 "pri" and "weight". Note that the previous implicit priority given by the
2083 list order is no longer honoured.
2085 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2086 for DKIM processing.
2088 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2089 by defining SUPPORT_SOCKS.
2091 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2092 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2093 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2094 variables are renamed to proxy_{local,external}_{address,port}.
2096 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2097 for a build by defining SUPPORT_I18N
2099 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2100 of the query string, and make ${quote_redis:} do that quoting.
2102 JH/29 Move Events support from Experimental to mainline, enabled by default
2103 and removable for a build by defining DISABLE_EVENT.
2105 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2107 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2108 cached by the daemon.
2110 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2111 by defining LOOKUP_REDIS. The libhiredis library is required.
2113 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2114 keys are given for lookup.
2116 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2117 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2118 only supported when built with TLS support. The PolarSSL SHA routines
2119 are still used when the TLS library is too old for convenient support.
2121 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2122 openssl_options), for security. OpenSSL forces this from version 1.1.0
2123 server-side so match that on older versions.
2125 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2126 allocation for $value could be released as the expansion processing
2127 concluded, but leaving the global pointer active for it.
2129 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2130 and to use the domains and local_parts ACL conditions.
2132 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2133 incorrectly not doubled on cutthrough transmission, hence seen as a
2134 body-termination at the receiving system - resulting in truncated mails.
2135 Commonly the sender saw a TCP-level error, and retransmitted the message
2136 via the normal store-and-forward channel. This could result in duplicates
2137 received - but deduplicating mailstores were liable to retain only the
2138 initial truncated version.
2140 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2142 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2144 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2145 we're in there, support oversigning also; bug 1309.
2147 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2149 HS/04 Add support for keep_environment and add_environment options.
2151 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2152 either intentional arithmetic overflow during PRNG, or testing config-
2155 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2156 delivery resulted in actual delivery. Cancel cutthrough before DATA
2159 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2160 rejects a recipient: pass the reject to the originator.
2162 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2163 Many were false-positives and ignorable, but it's worth fixing the
2166 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2167 for the new environment-manipulation done at startup. Move the routines
2168 from being local to tls.c to being global via the os.c file.
2170 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2171 an extract embedded as result-arg for a map, the first arg for extract
2172 is unavailable so we cannot tell if this is a numbered or keyed
2173 extraction. Accept either.
2179 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2182 JH/02 The smtp transport option "multi_domain" is now expanded.
2184 JH/03 The smtp transport now requests PRDR by default, if the server offers
2187 JH/04 Certificate name checking on server certificates, when exim is a client,
2188 is now done by default. The transport option tls_verify_cert_hostnames
2189 can be used to disable this per-host. The build option
2190 EXPERIMENTAL_CERTNAMES is withdrawn.
2192 JH/05 The value of the tls_verify_certificates smtp transport and main options
2193 default to the word "system" to access the system default CA bundle.
2194 For GnuTLS, only version 3.0.20 or later.
2196 JH/06 Verification of the server certificate for a TLS connection is now tried
2197 (but not required) by default. The verification status is now logged by
2198 default, for both outbound TLS and client-certificate supplying inbound
2201 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2204 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2205 Status Notification (bounce) messages are now MIME format per RFC 3464.
2206 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2207 under the control of the dsn_advertise_hosts option, and routers may
2208 have a dsn_lasthop option.
2210 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2211 default, modifiable by a malware= option. The list separator for
2212 the options can now be changed in the usual way. Bug 68.
2214 JH/10 The smtp_receive_timeout main option is now expanded before use.
2216 JH/11 The incoming_interface log option now also enables logging of the
2217 local interface on delivery outgoing connections.
2219 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2220 if the interface and destination host and port all match.
2222 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2225 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2226 Patch from Andrew Lewis.
2228 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2229 now supports optional time-restrictions, weighting, and priority
2230 modifiers per server. Patch originally by <rommer@active.by>.
2232 JH/16 The spamd_address main option now supports a mixed list of local
2233 and remote servers. Remote servers can be IPv6 addresses, and
2234 specify a port-range.
2236 JH/17 Bug 68: The spamd_address main option now supports an optional
2237 timeout value per server.
2239 JH/18 Bug 1581: Router and transport options headers_add/remove can
2240 now have the list separator specified.
2242 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2245 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2248 JH/21 Support for the A6 type of dns record is withdrawn.
2250 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2251 rather than the verbs used.
2253 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2254 from 255 to 1024 chars.
2256 JH/24 Verification callouts now attempt to use TLS by default.
2258 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2259 are generic router options now. The defaults didn't change.
2261 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2262 Original patch from Alexander Shikoff, worked over by JH.
2264 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2265 files if ZCAT_COMMAND is not executable.
2267 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2269 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2271 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2272 Normally benign, it bites when the pair was led to by a CNAME;
2273 modern usage is to not canonicalize the domain to a CNAME target
2274 (and we were inconsistent anyway for A-only vs AAAA+A).
2276 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2278 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2279 when evaluating $sender_host_dnssec.
2281 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2282 $sender_helo_dnssec variable.
2284 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2286 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2288 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2290 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2291 documented as working, but never had. Support all but $spam_report.
2293 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2294 added for tls authenticator.
2296 HS/03 Add perl_taintmode main config option
2302 TL/01 When running the test suite, the README says that variables such as
2303 no_msglog_check are global and can be placed anywhere in a specific
2304 test's script, however it was observed that placement needed to be near
2305 the beginning for it to behave that way. Changed the runtest perl
2306 script to read through the entire script once to detect and set these
2307 variables, reset to the beginning of the script, and then run through
2308 the script parsing/test process like normal.
2310 TL/02 The BSD's have an arc4random API. One of the functions to induce
2311 adding randomness was arc4random_stir(), but it has been removed in
2312 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2313 function when detected.
2315 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2316 cause callback expansion.
2318 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2319 syntax errors in an expansion can be treated as a string instead of
2320 logging or causing an error, due to the internal use of bool_lax
2321 instead of bool when processing it.
2323 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2324 server certificates when making smtp deliveries.
2326 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2328 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2330 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2332 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2333 Merged patch from Sebastian Wiedenroth.
2335 JH/05 Fix results-pipe from transport process. Several recipients, combined
2336 with certificate use, exposed issues where response data items split
2337 over buffer boundaries were not parsed properly. This eventually
2338 resulted in duplicates being sent. This issue only became common enough
2339 to notice due to the introduction of connection certificate information,
2340 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2342 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2343 size buffer was used, resulting in syntax errors when an expansion
2346 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2347 version 3.3.6 or later.
2349 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2350 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2351 both become "event_action", the variables become $event_name, $event_data
2352 and $event_defer_errno. There is a new variable $verify_mode, usable in
2353 routers, transports and related events. The tls:cert event is now also
2354 raised for inbound connections, if the main configuration event_action
2357 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2358 early OCSP support, but no stapling (appears to be less than 1.0.0).
2360 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2361 server certificate names available under the smtp transport option
2362 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2365 JH/10 Time-related extraction expansions from certificates now use the main
2366 option "timezone" setting for output formatting, and are consistent
2367 between OpenSSL and GnuTLS compilations. Bug 1541.
2369 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2370 encoded parameter in the incoming message. Bug 1558.
2372 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2373 include certificate info, eximon was claiming there were spoolfile
2376 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2378 JH/14 Log delivery-related information more consistently, using the sequence
2379 "H=<name> [<ip>]" wherever possible.
2381 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2382 are problematic for Debian distribution, omit them from the release
2385 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2387 JH/16 Fix string representation of time values on 64bit time_t architectures.
2390 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2391 output list separator was used.
2396 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2397 checkers that were complaining about end of non-void function with no
2400 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2401 This was a regression introduced in 4.83 by another bugfix.
2403 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2405 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2406 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2412 TF/01 Correctly close the server side of TLS when forking for delivery.
2414 When a message was received over SMTP with TLS, Exim failed to clear up
2415 the incoming connection properly after forking off the child process to
2416 deliver the message. In some situations the subsequent outgoing
2417 delivery connection happened to have the same fd number as the incoming
2418 connection previously had. Exim would try to use TLS and fail, logging
2419 a "Bad file descriptor" error.
2421 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2422 utilities have not been installed.
2424 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2425 temporary space as the ACL may create new global variables.
2427 TL/01 LDAP support uses per connection or global context settings, depending
2428 upon the detected version of the libraries at build time.
2430 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2431 to extract and use the src ip:port in logging and expansions as if it
2432 were a direct connection from the outside internet. PPv2 support was
2433 updated based on HAProxy spec change in May 2014.
2435 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2437 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2438 Properly escape header and check for NULL return.
2440 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2443 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2445 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2446 characters in header names, implemented as a verify condition.
2447 Contributed by Michael Fischer v. Mollard.
2449 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2450 results permerror and temperror. Previous values are deprecated but
2451 still accepted. In a future release, err_perm and err_temp will be
2452 completely removed, which will be a backward incompatibility if the
2453 ACL tests for either of these two old results. Patch contributed by
2454 user bes-internal on the mailing list.
2456 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2458 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2459 selectors, in both main and reject logs.
2461 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2464 JH/07 Add malware type "sock" for talking to simple daemon.
2466 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2468 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2469 routers/transports under cutthrough routing.
2471 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2472 numbers. Touch up "bool" conditional to keep the same definition.
2474 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2476 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2478 JH/12 Expand items in router/transport headers_add or headers_remove lists
2479 individually rather than the list as a whole. Bug 1452.
2481 Required for reasonable handling of multiple headers_ options when
2482 they may be empty; requires that headers_remove items with embedded
2483 colons must have them doubled (or the list-separator changed).
2485 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2486 view the policy declared in the DMARC record. Currently, $dmarc_status
2487 is a combined value of both the record presence and the result of the
2490 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2492 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2493 dnslookup router and the smtp transport (applying to the forward
2496 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2497 of ldap servers used for a specific lookup. Patch provided by Heiko
2500 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2501 New variable $lookup_dnssec_authenticated for observability.
2503 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2504 Patch submitted by Lars Timman.
2506 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2508 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2509 Requires trusted mode and valid format message id, aborts otherwise.
2510 Patch contributed by Heiko Schlichting.
2512 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2513 certextract with support for various fields. Bug 1358.
2515 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2516 is requested by default, modifiable by smtp transport option
2519 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2520 operate on certificate variables to give certificate fingerprints
2521 Also new ${sha256:cert_variable}.
2523 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2525 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2526 Christian Aistleitner.
2528 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2530 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2531 file. Patch from Wolfgang Breyha.
2533 JH/25 Expand the coverage of the delivery $host and $host_address to
2534 client authenticators run in verify callout. Bug 1476.
2536 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2537 align with daemon_smtp_ports. Bug 72.
2539 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2540 support and error reporting did not work properly.
2542 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2543 and is readable. Patch from Andrew Colin Kissa.
2545 TL/14 Enhance documentation of ${run expansion and how it parses the
2546 commandline after expansion, particularly in the case when an
2547 unquoted variable expansion results in an empty value.
2549 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2551 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2552 by renaming away from it. Addresses GH issue 15, reported by
2555 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2557 TL/15 SECURITY: prevent double expansion in math comparison functions
2558 (can expand unsanitized data). Not remotely exploitable.
2565 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2567 PP/02 Make -n do something, by making it not do something.
2568 When combined with -bP, the name of an option is not output.
2570 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2573 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2574 $sender_host_name and config options to manage this, and basic check
2577 PP/05 DSCP support for outbound connections and control modifier for inbound.
2579 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2580 (Only plugin which currently uses this is kerberos4, which nobody should
2581 be using, but we should make it available and other future plugins might
2582 conceivably use it, even though it would break NAT; stuff *should* be
2583 using channel bindings instead).
2585 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2586 name; added for Sendmail compatibility; requires admin caller.
2587 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2588 just ignore it); requires trusted caller.
2589 Also parse but ignore: -Ac -Am -X<logfile>
2592 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2594 TL/02 Add +smtp_confirmation as a default logging option.
2596 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2597 Patch by Magnus Holmgren from 2007-02-20.
2599 TL/04 Bugzilla 1281 - Spec typo.
2600 Bugzilla 1283 - Spec typo.
2601 Bugzilla 1290 - Spec grammar fixes.
2603 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2605 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2607 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2608 mailing list by by Dmitry Isaikin.
2610 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2612 JH/02 Support "G" suffix to numbers in ${if comparisons.
2614 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2616 NM/01 Bugzilla 1197 - Spec typo
2617 Bugzilla 1196 - Spec examples corrections
2619 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2621 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2622 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2625 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2626 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2628 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2629 "acl {{name}{arg}...}", and optional args on acl condition
2632 JH/05 Permit multiple router/transport headers_add/remove lines.
2634 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2636 JH/07 Avoid using a waiting database for a single-message-only transport.
2637 Performance patch from Paul Fisher. Bugzilla 1262.
2639 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2642 JH/09 Add $headers_added variable, with content from use of ACL modifier
2643 add_header (but not yet added to the message). Bugzilla 199.
2645 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2646 Pulled from Bugzilla 817 by Wolfgang Breyha.
2648 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2650 (nb: this is the same fix as in Exim 4.80.1)
2652 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2655 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2658 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2660 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2661 advertise SMTP AUTH mechanism to us, instead of a generic
2662 protocol violation error. Also, make Exim more robust to bad
2663 data from the Dovecot auth socket.
2665 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2667 When a queue runner is handling a message, Exim first routes the
2668 recipient addresses, during which it prunes them based on the retry
2669 hints database. After that it attempts to deliver the message to
2670 any remaining recipients. It then updates the hints database using
2673 So if a recipient address works intermittently, it can get repeatedly
2674 deferred at routing time. The retry hints record remains fresh so the
2675 address never reaches the final cutoff time.
2677 This is a fairly common occurrence when a user is bumping up against
2678 their storage quota. Exim had some logic in its local delivery code
2679 to deal with this. However it did not apply to per-recipient defers
2680 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2682 This change adds a proper retry rule check during routing so that the
2683 final cutoff time is checked against the message's age. We only do
2684 this check if there is an address retry record and there is not a
2685 domain retry record; this implies that previous attempts to handle
2686 the address had the retry_use_local_parts option turned on. We use
2687 this as an approximation for the destination being like a local
2688 delivery, as in LMTP.
2690 I suspect this new check makes the old local delivery cutoff check
2691 redundant, but I have not verified this so I left the code in place.
2693 TF/02 Correct gecos expansion when From: is a prefix of the username.
2695 Test 0254 submits a message to Exim with the header
2699 When I ran the test suite under the user fanf2, Exim expanded
2700 the header to contain my full name, whereas it should have added
2701 a Resent-Sender: header. It erroneously treats any prefix of the
2702 username as equal to the username.
2704 This change corrects that bug.
2706 GF/01 DCC debug and logging tidyup
2707 Error conditions log to paniclog rather than rejectlog.
2708 Debug lines prefixed by "DCC: " to remove any ambiguity.
2710 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2712 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2713 Bug spotted by Jeremy Harris; was flawed since initial commit.
2714 Would have resulted in OCSP responses post-SNI triggering an Exim
2715 NULL dereference and crash.
2717 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2719 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2720 Bug detection, analysis and fix by Samuel Thibault.
2721 Bugzilla 1331, Debian bug #698092.
2723 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2725 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2726 Server implementation by Todd Lyons, client by JH.
2727 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2728 config variable "prdr_enable" controls whether the server
2729 advertises the facility. If the client requests PRDR a new
2730 acl_data_smtp_prdr ACL is called once for each recipient, after
2731 the body content is received and before the acl_smtp_data ACL.
2732 The client is controlled by both of: a hosts_try_prdr option
2733 on the smtp transport, and the server advertisement.
2734 Default client logging of deliveries and rejections involving
2735 PRDR are flagged with the string "PRDR".
2737 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2738 fclose(). Diagnosis by Todd Lyons.
2740 PP/17 Update configure.default to handle IPv6 localhost better.
2741 Patch by Alain Williams (plus minor tweaks).
2744 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2745 This is now consistent with GnuTLS, and is now documented: the
2746 previous undocumented portable approach to treating the option as
2747 unset was to force an expansion failure. That still works, and
2748 an empty string is now equivalent.
2750 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2751 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2752 not performing validation itself.
2754 PP/20 Added force_command boolean option to pipe transport.
2755 Patch from Nick Koston, of cPanel Inc.
2757 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2760 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2762 PP/21 Fix eximon continuous updating with timestamped log-files.
2763 Broken in a format-string cleanup in 4.80, missed when I repaired the
2764 other false fix of the same issue.
2765 Report and fix from Heiko Schlichting.
2768 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2769 Report from Prashanth Katuri.
2771 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2772 It's SecureTransport, so affects any MacOS clients which use the
2773 system-integrated TLS libraries, including email clients.
2775 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2776 using a MIME ACL for non-SMTP local injection.
2777 Report and assistance in diagnosis by Warren Baker.
2779 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2781 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2783 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2784 last id that failed so it may be referenced in subsequent ACL's.
2786 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2789 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2790 ldap library initialization, allowing self-signed CA's to be
2791 used. Also properly sets require_cert option later in code by
2792 using NULL (global ldap config) instead of ldap handle (per
2793 session). Bug diagnosis and testing by alxgomz.
2795 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2796 the src/util/ subdirectory.
2798 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2799 renamed to Transport Post Delivery Action by Jeremy Harris, as
2802 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2803 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2804 redis_servers = needs to be configured which will be used by the redis
2805 lookup. Patch from Warren Baker, of The Packet Hub.
2807 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2809 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2810 hostname or reverse DNS when processing a host list. Used suggestions
2811 from multiple comments on this bug.
2813 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2815 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2816 Missed a few lines, added it to make the runtest require no keyboard
2819 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2820 contains upper case chars. Make router use caseful_local_part.
2822 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2823 support when GnuTLS has been built with p11-kit.
2829 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2831 This, or similar/improved, will also be change PP/11 of 4.82.
2837 PP/01 Handle short writes when writing local log-files.
2838 In practice, only affects FreeBSD (8 onwards).
2839 Bugzilla 1053, with thanks to Dmitry Isaikin.
2841 NM/01 Bugzilla 949 - Documentation tweak
2843 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2846 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2848 PP/02 Implemented gsasl authenticator.
2850 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2852 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2853 `pkg-config foo` for cflags/libs.
2855 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2856 with rest of GSASL and with heimdal_gssapi.
2858 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2859 `pkg-config foo` for cflags/libs for the TLS implementation.
2861 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2862 properties get this fed in as external SSF. A number of robustness
2863 and debugging improvements to the cyrus_sasl authenticator.
2865 PP/08 cyrus_sasl server now expands the server_realm option.
2867 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2868 Patch by Jeremy Harris.
2870 PP/10 Added dbmjz lookup type.
2872 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2874 PP/12 MAIL args handles TAB as well as SP, for better interop with
2875 non-compliant senders.
2876 Analysis and variant patch by Todd Lyons.
2878 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2879 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2880 Patch from Dirk Mueller <dmueller@suse.com>
2882 PP/13 tls_peerdn now print-escaped for spool files.
2883 Observed some $tls_peerdn in wild which contained \n, which resulted
2884 in spool file corruption.
2886 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2887 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2888 or write after TLS renegotiation, which otherwise led to messages
2891 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2892 as a tracking header (ie: a signed header comes before the signature).
2893 Patch from Wolfgang Breyha.
2895 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2896 comma-sep list; embedded commas doubled.
2898 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2900 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2902 Report and patch from Dmitry Banschikov.
2904 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2905 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2906 failures. We appear to now support TLS1.1+ with Exim.
2908 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2909 lets Exim select keys and certificates based upon TLS SNI from client.
2910 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2911 before an outbound SMTP session. New log_selector, +tls_sni.
2913 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2914 NULL dereference. Report and patch from Alun Jones.
2916 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2917 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2918 Not seeing resolver debug output on NetBSD, but suspect this is a
2919 resolver implementation change.
2921 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2922 Left warnings. Added "eximon gdb" invocation mode.
2924 PP/21 Defaulting "accept_8bitmime" to true, not false.
2926 PP/22 Added -bw for inetd wait mode support.
2928 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2929 locate the relevant includes and libraries. Made this the default.
2931 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2932 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2934 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2935 This may cause build issues on older platforms.
2937 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2938 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2939 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2940 Added SNI support via GnuTLS too.
2941 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2943 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2945 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2946 Applied second patch from Janne, implementing suggestion to default
2947 multiple-strings-in-record handling to match SPF spec.
2949 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2951 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2952 read-only, out of scope).
2953 Patch from Wolfgang Breyha, report from Stuart Northfield.
2955 PP/29 Fix three issues highlighted by clang analyser static analysis.
2956 Only crash-plausible issue would require the Cambridge-specific
2957 iplookup router and a misconfiguration.
2958 Report from Marcin Mirosław.
2960 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2962 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2963 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2964 As part of this, removing so much warning spew let me fix some minor
2965 real issues in debug logging.
2967 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2968 assignment on my part. Fixed.
2970 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2971 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2972 Janne Snabb (who went above and beyond: thank you).
2974 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2975 string otherwise requires a connection and a bunch more work and it's
2976 relatively easy to get wrong. Should also expose TLS library linkage
2979 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2980 64-bit ${eval} (JH/03).
2982 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2983 GNU libc to support some of the 64-bit stuff, should not lead to
2984 conflicts. Defined before os.h is pulled in, so if a given platform
2985 needs to override this, it can.
2987 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2988 protection layer was required, which is not implemented.
2989 Bugzilla 1254, patch from Wolfgang Breyha.
2991 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2992 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2993 tls_dhparam take prime identifiers. Also unbreak combination of
2994 OpenSSL+DH_params+TLSSNI.
2996 PP/39 Disable SSLv2 by default in OpenSSL support.
3002 PP/01 Solaris build fix for Oracle's LDAP libraries.
3003 Bugzilla 1109, patch from Stephen Usher.
3005 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3007 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3010 TF/02 Fix a couple more cases where we did not log the error message
3011 when unlink() failed. See also change 4.74-TF/03.
3013 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3014 lock up or crash if it happened to be inside a call to libc when it
3015 got a SIGUSR1 from exiwhat.
3017 The SIGUSR1 handler appends the current process status to the process
3018 log which is later printed by exiwhat. It used to use the general
3019 purpose logging code to do this, but several functions it calls are
3020 not safe for signals.
3022 The new output code in the SIGUSR1 handler is specific to the process
3023 log, and simple enough that it's easy to inspect for signal safety.
3024 Removing some special cases also simplifies the general logging code.
3025 Removing the spurious timestamps from the process log simplifies
3028 TF/04 Improved ratelimit ACL condition.
3030 The /noupdate option has been deprecated in favour of /readonly which
3031 has clearer semantics. The /leaky, /strict, and /readonly update modes
3032 are mutually exclusive. The update mode is no longer included in the
3033 database key; it just determines when the database is updated. (This
3034 means that when you upgrade Exim will forget old rate measurements.)
3036 Exim now checks that the per_* options are used with an update mode that
3037 makes sense for the current ACL. For example, when Exim is processing a
3038 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3039 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3040 must specify per_mail/readonly. If you omit the update mode it defaults to
3041 /leaky where that makes sense (as before) or /readonly where required.
3043 The /noupdate option is now undocumented but still supported for
3044 backwards compatibility. It is equivalent to /readonly except that in
3045 ACLs where /readonly is required you may specify /leaky/noupdate or
3046 /strict/noupdate which are treated the same as /readonly.
3048 A useful new feature is the /count= option. This is a generalization
3049 of the per_byte option, so that you can measure the throughput of other
3050 aggregate values. For example, the per_byte option is now equivalent
3051 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3053 The per_rcpt option has been generalized using the /count= mechanism
3054 (though it's more complicated than the per_byte equivalence). When it is
3055 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3056 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3057 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3058 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3059 Note that using per_rcpt with a non-readonly update mode in more than
3060 one ACL will cause the recipients to be double-counted. (The per_mail
3061 and per_byte options don't have this problem.)
3063 The handling of very low rates has changed slightly. If the computed rate
3064 is less than the event's count (usually one) then this event is the first
3065 after a long gap. In this case the rate is set to the same as this event's
3066 count, so that the first message of a spam run is counted properly.
3068 The major new feature is a mechanism for counting the rate of unique
3069 events. The new per_addr option counts the number of different
3070 recipients that someone has sent messages to in the last time period. It
3071 behaves like per_rcpt if all the recipient addresses are different, but
3072 duplicate recipient addresses do not increase the measured rate. Like
3073 the /count= option this is a general mechanism, so the per_addr option
3074 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3075 example, measure the rate that a client uses different sender addresses
3076 with the options per_mail/unique=$sender_address. There are further
3077 details in the main documentation.
3079 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3081 TF/06 Removed a few PCRE remnants.
3083 TF/07 Automatically extract Exim's version number from tags in the git
3084 repository when doing development or release builds.
3086 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3087 Bugzilla 879. Patch from Paul Fisher.
3089 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3090 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3093 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3095 PP/05 Variable $av_failed, true if the AV scanner deferred.
3096 Bugzilla 1078. Patch from John Horne.
3098 PP/06 Stop make process more reliably on build failure.
3099 Bugzilla 1087. Patch from Heiko Schlittermann.
3101 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3102 Bugzilla 1089. Patch from Heiko Schlittermann.
3104 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3105 Bugzilla 1131. Patch from Holger Weiß.
3107 PP/09 Handle IPv6 addresses with SPF.
3108 Bugzilla 860. Patch from Wolfgang Breyha.
3110 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3112 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3115 PP/11 match_* no longer expand right-hand-side by default.
3116 New compile-time build option, EXPAND_LISTMATCH_RHS.
3117 New expansion conditions, "inlist", "inlisti".
3119 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3121 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3123 PP/14 fix log_write() format string regression from TF/03.
3124 Bugzilla 1152. Patch from Dmitry Isaikin.
3130 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3132 PP/02 Harmonised TLS library version reporting; only show if debugging.
3133 Layout now matches that introduced for other libraries in 4.74 PP/03.
3135 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3137 PP/04 New "dns_use_edns0" global option.
3139 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3142 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3143 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3145 TK/01 Updated PolarSSL code to 0.14.2.
3146 Bugzilla 1097. Patch from Andreas Metzler.
3148 PP/07 Catch divide-by-zero in ${eval:...}.
3149 Fixes bugzilla 1102.
3151 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3154 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3155 format-string attack -- SECURITY: remote arbitrary code execution.
3157 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3158 time unintentionally subject to list matching rules, letting the header
3159 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3160 arbitrary string expansion). This allowed for information disclosure.
3162 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3163 INT_MIN/-1 -- value coerced to INT_MAX.
3169 NM/01 Workaround for PCRE version dependency in version reporting
3172 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3173 This fixes portability to compilers other than gcc, notably
3174 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3176 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3177 makefiles for portability to HP-UX and POSIX correctness.
3179 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3180 Also via indented variable definition in the Makefile.
3181 (Debugging by Oliver Heesakkers).
3183 PP/02 Restore caching of spamd results with expanded spamd_address.
3184 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3186 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3187 Improves build reliability. Fix from: Frank Elsner
3189 NM/02 Fix wide character breakage in the rfc2047 coding
3190 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3192 NM/03 Allow underscore in dnslist lookups
3193 Fixes bug 1026. Patch from Graeme Fowler
3195 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3196 Code patches from Adam Ciarcinski of NetBSD.
3198 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3201 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3202 is logged, to avoid truncation. Patch from John Horne.
3204 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3205 Patch from Jakob Hirsch.
3207 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3208 SQL string expansion failure details.
3209 Patch from Andrey Oktyabrski.
3211 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3212 Patch from Simon Arlott.
3214 PP/09 New lookups functionality failed to compile on old gcc which rejects
3215 extern declarations in function scope.
3216 Patch from Oliver Fleischmann
3218 PP/10 Use sig_atomic_t for flags set from signal handlers.
3219 Check getgroups() return and improve debugging.
3220 Fixed developed for diagnosis in bug 927 (which turned out to be
3223 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3224 Patch from Mark Zealey.
3226 PP/12 Bugzilla 1056: Improved spamd server selection.
3227 Patch from Mark Zealey.
3229 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3230 Based on patch from Heiko Schlittermann.
3232 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3233 Patch from Uwe Doering, sign-off by Michael Haardt.
3235 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3236 variable declaration deep within a block. Bug and patch from
3239 PP/15 lookups-Makefile IRIX compatibility coercion.
3241 PP/16 Make DISABLE_DKIM build knob functional.
3243 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3244 Patch by Simon Arlott
3246 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3247 variable argument macros. Our copy now differs from upstream.
3253 TF/01 Failure to get a lock on a hints database can have serious
3254 consequences so log it to the panic log.
3256 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3257 controlled using the smtp_confirmation log selector.
3259 TF/03 Include the error message when we fail to unlink a spool file.
3261 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3262 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3263 for maintaining out-of-tree patches for some time.
3265 PP/01 Bugzilla 139: Documentation and portability issues.
3266 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3267 Handle per-OS dynamic-module compilation flags.
3269 PP/02 Let /dev/null have normal permissions.
3270 The 4.73 fixes were a little too stringent and complained about the
3271 permissions on /dev/null. Exempt it from some checks.
3272 Reported by Andreas M. Kirchwitz.
3274 PP/03 Report version information for many libraries, including
3275 Exim version information for dynamically loaded libraries. Created
3276 version.h, now support a version extension string for distributors
3277 who patch heavily. Dynamic module ABI change.
3279 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3280 privilege escalation vulnerability whereby the Exim run-time user
3281 can cause root to append content of the attacker's choosing to
3284 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3287 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3288 If dropping privileges for untrusted macros, we disabled normal logging
3289 on the basis that it would fail; for the Exim run-time user, this is not
3290 the case, and it resulted in successful deliveries going unlogged.
3291 Fixed. Reported by Andreas Metzler.
3297 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3298 only prepend for the Resent-* case. Fixes regression introduced in
3299 Exim 4.70 by NM/22 for Bugzilla 607.
3301 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3302 increasing numbers of administrators be bitten by this.
3304 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3306 PP/03 Bugzilla 994: added openssl_options main configuration option.
3308 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3310 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3312 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3314 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3315 without a peer certificate, leading to a segfault because of an
3316 assumption that peers always have certificates. Be a little more
3317 paranoid. Problem reported by Martin Tscholak.
3319 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3320 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3321 NB: ClamAV planning to remove STREAM in "middle of 2010".
3322 CL also introduces -bmalware, various -d+acl logging additions and
3323 more caution in buffer sizes.
3325 PP/09 Implemented reverse_ip expansion operator.
3327 PP/10 Bugzilla 937: provide a "debug" ACL control.
3329 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3331 PP/12 Bugzilla 973: Implement --version.
3333 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3335 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3337 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3339 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3340 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3341 ignore trailing whitespace.
3343 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3345 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3348 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3349 Notification from Dr Andrew Aitchison.
3351 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3352 ExtendedDetectionInfo response format.
3353 Notification from John Horne.
3355 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3358 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3359 XSL and documented dependency on system catalogs, with examples of how
3362 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3365 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3366 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3367 configuration file which is writeable by the Exim user or group.
3369 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3370 of configuration files to cover files specified with the -C option if
3371 they are going to be used with root privileges, not just the default
3374 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3375 option (effectively making it always true).
3377 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3378 files to be used while preserving root privileges.
3380 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3381 that rogue child processes cannot use them.
3383 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3384 run-time user, instead of root.
3386 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3387 Exim run-time user without dropping privileges.
3389 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3390 result string, instead of calling string_vformat() twice with the same
3393 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3394 for other users. Others should always drop root privileges if they use
3395 -C on the command line, even for a whitelisted configure file.
3397 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3399 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3405 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3406 $data_path, and $header_path variables; fixed documentation bugs and
3409 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3410 exipick to access non-standard spools, including the "frozen" queue
3413 NM/01 Bugzilla 965: Support mysql stored procedures.
3414 Patch from Alain Williams
3416 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3418 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3419 Patch from Andreas Metzler
3421 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3422 Patch from Kirill Miazine
3424 NM/05 Bugzilla 671: Added umask to procmail example.
3426 JJ/03 installed exipick 20100323.0, fixing doc bug
3428 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3429 directory. Notification and patch from Dan Rosenberg.
3431 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3433 TK/02 Improve log output when DKIM signing operation fails.
3435 MH/01 Treat the transport option dkim_domain as a colon separated
3436 list, not as a single string, and sign the message with each element,
3437 omitting multiple occurences of the same signer.
3439 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3440 Bugzilla 985, 986. Patch by Simon Arlott
3442 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3443 Patch by Simon Arlott
3445 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3446 MBX locking. Notification from Dan Rosenberg.
3452 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3454 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3456 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3458 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3460 NM/04 Bugzilla 915: Flag broken perl installation during build.
3466 TK/01 Added patch by Johannes Berg that expands the main option
3467 "spamd_address" if it starts with a dollar sign.
3469 TK/02 Write list of recipients to X-Envelope-Sender header when building
3470 the mbox-format spool file for content scanning (suggested by Jakob
3473 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3474 (http://www.dcc-servers.net/) support via dccifd. Activated by
3475 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3477 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3478 by Mark Daniel Reidel <mr@df.eu>.
3480 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3481 When building exim an external PCRE library is now needed -
3482 PCRE is a system library on the majority of modern systems.
3483 See entry on PCRE_LIBS in EDITME file.
3485 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3486 conversation. Added nologin parameter to request.
3487 Patch contributed by Kirill Miazine.
3489 TF/01 Do not log submission mode rewrites if they do not change the address.
3491 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3493 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3494 log files in place. Contributed by Roberto Lima.
3496 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3498 TF/03 Bugzilla 615: When checking the local_parts router precondition
3499 after a local_part_suffix or local_part_prefix option, Exim now
3500 does not use the address's named list lookup cache, since this
3501 contains cached lookups for the whole local part.
3503 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3504 Robert Millan. Documentation is in experimental-spec.txt.
3506 TF/04 Bugzilla 668: Fix parallel build (make -j).
3508 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3510 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3511 Patch provided by Jan Srzednicki.
3513 TF/05 Leading white space used to be stripped from $spam_report which
3514 wrecked the formatting. Now it is preserved.
3516 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3517 that they are available at delivery time.
3519 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3521 TF/08 TLS error reporting now respects the incoming_interface and
3522 incoming_port log selectors.
3524 TF/09 Produce a more useful error message if an SMTP transport's hosts
3525 setting expands to an empty string.
3527 NM/06 Bugzilla 744: EXPN did not work under TLS.
3528 Patch provided by Phil Pennock.
3530 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3531 Patch provided by Richard Godbee.
3533 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3534 acl_smtp_notquit, added index entry.
3536 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3537 Patch provided by Eugene Bujak.
3539 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3540 accept(). Patch provided by Maxim Dounin.
3542 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3543 Patch provided by Phil Pennock.
3545 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3547 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3548 Patch provided by Brad "anomie" Jorsch.
3550 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3551 Patch provided by Dean Brooks.
3553 TK/05 Add native DKIM support (does not depend on external libraries).
3555 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3556 Patch provided by Graeme Fowler.
3558 NM/16 Bugzilla 851: Documentation example syntax fix.
3560 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3562 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3565 NM/19 Bugzilla 745: TLS version reporting.
3566 Patch provided by Phil Pennock.
3568 NM/20 Bugzilla 167: bool: condition support.
3569 Patch provided by Phil Pennock.
3571 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3572 clients. Patch provided by Phil Pennock.
3574 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3575 Patch provided by Brad "anomie" Jorsch.
3577 NM/23 Bugzilla 687: Fix misparses in eximstats.
3578 Patch provided by Heiko Schlittermann.
3580 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3581 Patch provided by Heiko Schlittermann.
3583 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3584 plus update to original patch.
3586 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3588 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3589 Patch provided by David Brownlee.
3591 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3593 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3595 NM/30 Bugzilla 888: TLS documentation bugfixes.
3597 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3599 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3600 Unlike the original bugzilla I have changed all shell scripts in src tree.
3602 NM/33 Bugzilla 898: Transport filter timeout fix.
3603 Patch by Todd Rinaldo.
3605 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3606 Patch by Serge Demonchaux.
3608 NM/35 Bugzilla 39: Base64 decode bug fixes.
3609 Patch by Jakob Hirsch.
3611 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3613 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3615 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3617 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3623 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3624 ALT-N's libdkim that I have put here:
3625 http://duncanthrax.net/exim-experimental/
3627 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3628 were called 'true' and it seems that C99 defines that as a reserved
3629 keyword to be used with 'bool' variable types. That means you could
3630 not include C99-style headers which use bools without triggering
3631 build errors in sieve.c.
3633 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3634 as mailq or other aliases. Changed the --help handling significantly
3635 to do whats expected. exim_usage() emits usage/help information.
3637 SC/01 Added the -bylocaldomain option to eximstats.
3639 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3641 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3643 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3649 PH/01 Another patch from the Sieve maintainer.
3651 PH/02 When an IPv6 address is converted to a string for single-key lookup
3652 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3653 dots are used instead of colons so that keys in lsearch files need not
3654 contain colons. This was done some time before quoting was made available
3655 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3656 (notated using the quote facility) so as to distinguish them from IPv4
3657 keys. This meant that lookups for IP addresses in host lists did not work
3658 for iplsearch lookups.
3660 This has been fixed by arranging for IPv6 addresses to be expressed with
3661 colons if the lookup type is iplsearch. This is not incompatible, because
3662 previously such lookups could never work.
3664 The situation is now rather anomalous, since one *can* have colons in
3665 ordinary lsearch keys. However, making the change in all cases is
3666 incompatible and would probably break a number of configurations.
3668 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3671 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3672 conversion specification without a maximum field width, thereby enabling
3673 a rogue spamd server to cause a buffer overflow. While nobody in their
3674 right mind would setup Exim to query an untrusted spamd server, an
3675 attacker that gains access to a server running spamd could potentially
3676 exploit this vulnerability to run arbitrary code as the Exim user.
3678 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3679 $primary_hostname instead of what libspf2 thinks the hosts name is.
3681 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3682 a directory entry by the name of the lookup key. Previously, if a
3683 symlink pointed to a non-existing file or a file in a directory that
3684 Exim lacked permissions to read, a lookup for a key matching that
3685 symlink would fail. Now it is enough that a matching directory entry
3686 exists, symlink or not. (Bugzilla 503.)
3688 PH/03 The body_linecount and body_zerocount variables are now exported in the
3691 PH/04 Added the $dnslist_matched variable.
3693 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3694 This means they are set thereafter only if the connection becomes
3697 PH/06 Added the client_condition to authenticators so that some can be skipped
3698 by clients under certain conditions.
3700 PH/07 The error message for a badly-placed control=no_multiline_responses left
3701 "_responses" off the end of the name.
3703 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3705 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3706 (without spaces) instead of just copying the configuration text.
3708 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3710 PH/11 Added $max_received_linelength.
3712 PH/12 Added +ignore_defer and +include_defer to host lists.
3714 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3715 way in which PCRE > 7.0 is built.
3717 PH/14 Implemented queue_only_load_latch.
3719 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3720 MAIL command. The effect was to mangle the value on 64-bit systems.
3722 PH/16 Another patch from the Sieve maintainer.
3724 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3726 PH/18 If a system quota error occurred while trying to create the file for
3727 a maildir delivery, the message "Mailbox is full" was not appended to the
3728 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3729 only to a quota excession during the actual writing of the file.
3731 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3732 characters?) which causes problems in log lines. The DN values are now
3733 passed through string_printing() before being added to log lines.
3735 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3736 and InterBase are left for another time.)
3738 PH/21 Added message_body_newlines option.
3740 PH/22 Guard against possible overflow in moan_check_errorcopy().
3742 PH/23 POSIX allows open() to be a macro; guard against that.
3744 PH/24 If the recipient of an error message contained an @ in the local part
3745 (suitably quoted, of course), incorrect values were put in $domain and
3746 $local_part during the evaluation of errors_copy.
3752 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3753 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3756 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3757 issue a MAIL command.
3759 PH/02 In an ACL statement such as
3761 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3763 if a client was not listed at all, or was listed with a value other than
3764 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3765 the condition was not true (as it should be), so access was not denied.
3766 The bug was that the ! inversion was incorrectly passed on to the second
3767 item. This has been fixed.
3769 PH/03 Added additional dnslists conditions == and =& which are different from
3770 = and & when the dns lookup returns more than one IP address.
3772 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3773 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3775 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3776 FSYNC, which compiles an option called disable_fsync that allows for
3777 bypassing fsync(). The documentation is heavily laced with warnings.
3779 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3781 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3782 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3783 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3784 including adding "make clean"; (3) Added -fPIC when compiling the test
3785 dynamically loaded module, to get rid of a warning.
3787 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3788 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3789 = 0s. The bug is otherwise harmless.
3791 PH/07 There was a bug in the dovecot authenticator such that the value of
3792 $auth1 could be overwritten, and so not correctly preserved, after a
3793 successful authentication. This usually meant that the value preserved by
3794 the server_setid option was incorrect.
3796 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3798 PH/09 Installed PCRE release 7.0.
3800 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3801 run for batched SMTP input. It is now run at the start of every message
3802 in the batch. While fixing this I discovered that the process information
3803 (output by running exiwhat) was not always getting set for -bs and -bS
3804 input. This is fixed, and it now also says "batched" for BSMTP.
3806 PH/11 Added control=no_pipelining.
3808 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3809 patch, slightly modified), and move the expansion of helo_data till after
3810 the connection is made in the smtp transport (so it can use these
3813 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3815 PH/14 Added log_selector = +pid.
3817 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3819 PH/16 Add ${if forany and ${if forall.
3821 PH/17 Added dsn_from option to vary the From: line in DSNs.
3823 PH/18 Flush SMTP output before performing a callout, unless control =
3824 no_callout_flush is set.
3826 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3827 was true (the default) a successful delivery failed to delete the retry
3828 item, thus causing premature timeout of the address. The bug is now
3831 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3833 PH/21 Long custom messages for fakedefer and fakereject are now split up
3834 into multiline responses in the same way that messages for "deny" and
3835 other ACL rejections are.
3837 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3838 with slight modification.
3840 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3841 draft, changing the syntax and factoring some duplicate code".
3843 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3844 for deliveries of the second and subsequent messages over the same SMTP
3847 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3848 ${reduce, with only minor "tidies".
3850 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3852 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3853 expansion side effects.
3855 PH/27 When a message times out after an over-quota error from an Exim-imposed
3856 quota, the bounce message says "mailbox is full". This message was not
3857 being given when it was a system quota that was exceeded. It now should
3860 MH/03 Made $recipients available in local_scan(). local_scan() already has
3861 better access to the recipient list through recipients_list[], but
3862 $recipients can be useful in postmaster-provided expansion strings.
3864 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3865 in the case of a MAIL command with additional options following the
3866 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3867 were accidentally chopped off.
3869 PH/29 SMTP synchronization checks are implemented when a command is read -
3870 there is a check that no more input is waiting when there shouldn't be
3871 any. However, for some commands, a delay in an ACL can mean that it is
3872 some time before the response is written. In this time, more input might
3873 arrive, invalidly. So now there are extra checks after an ACL has run for
3874 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3875 pipelining has not been advertised.
3877 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3879 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3880 returned in the SMTP response when smtp_return_error_details was false.
3881 This has been fixed.
3883 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3884 instead of the C I/O that was originally supplied, because problems were
3885 reported on Solaris.
3887 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3888 Exim which did not show up earlier: it was assuming that a call to
3889 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3890 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3891 was a macro that became an assignment, so it seemed to work. This has
3892 changed to a proper function call with a void return, hence the compile
3893 error. Exim's code has been fixed.
3895 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3898 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3900 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3902 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3903 because it turns out to be too big or there is a timeout) there is no
3904 "Completed" line in the log. When some messages of this type were
3905 selected by exigrep, they were listed as "not completed". Others were
3906 picked up by some special patterns. I have improved the selection
3907 criteria to be more general.
3909 PH/38 The host_find_failed option in the manualroute router can now be set
3910 to "ignore", to completely ignore a host whose IP address cannot be
3911 found. If all hosts are ignored, the behaviour is controlled by the new
3912 host_all_ignored option.
3914 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3915 homing or because of multiple MX records with /mx) generated more than
3916 one IP address, and the following item turned out to be the local host,
3917 all the secondary addresses of the first item were incorrectly removed
3918 from the list, along with the local host and any following hosts (which
3919 is what is supposed to happen).
3921 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3922 whoever called Exim into the -H file. In the case of the daemon it was
3923 behaving confusingly. When first started, it used values for whoever
3924 started the daemon, but after a SIGHUP it used the Exim user (because it
3925 calls itself on a restart). I have changed the code so that it now always
3928 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3929 message are rejected with the same error (e.g. no authentication or bad
3930 sender address), and a DATA command is nevertheless sent (as can happen
3931 with PIPELINING or a stupid MUA), the error message that was given to the
3932 RCPT commands is included in the rejection of the DATA command. This is
3933 intended to be helpful for MUAs that show only the final error to their
3936 PH/42 Another patch from the Sieve maintainer.
3938 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3939 Eximstats - Fixed some broken HTML links and added missing column headers
3941 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3942 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3944 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3946 PH/43 Yet another patch from the Sieve maintainer.
3948 PH/44 I found a way to check for a TCP/IP connection going away before sending
3949 the response to the final '.' that terminates a message, but only in the
3950 case where the client has not sent further data following the '.'
3951 (unfortunately, this is allowed). However, in many cases there won't be
3952 any further data because there won't be any more messages to send. A call
3953 to select() can be used: if it shows that the input is "ready", there is
3954 either input waiting, or the socket has been closed. An attempt to read
3955 the next input character can distinguish the two cases. Previously, Exim
3956 would have sent an OK response which the client would never have see.
3957 This could lead to message repetition. This fix should cure that, at
3958 least in a lot of common cases.
3960 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3961 advertised in response to EHLO.
3967 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3968 fixed by 4.65/MH/01 (is this a record?) are fixed:
3970 (i) An empty string was always treated as zero by the numeric comparison
3971 operators. This behaviour has been restored.
3973 (ii) It is documented that the numeric comparison operators always treat
3974 their arguments as decimal numbers. This was broken in that numbers
3975 starting with 0 were being interpreted as octal.
3977 While fixing these problems I realized that there was another issue that
3978 hadn't been noticed. Values of message_size_limit (both the global option
3979 and the transport option) were treated as octal if they started with 0.
3980 The documentation was vague. These values are now always treated as
3981 decimal, and I will make that clear in the documentation.
3987 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3988 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3991 MH/01 Don't check that the operands of numeric comparison operators are
3992 integers when their expansion is in "skipping" mode (fixes bug
3993 introduced by 4.64-PH/07).
3995 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3996 child addresses, Exim now panics and dies. Previously, because the count
3997 is held in a short int, deliveries were likely to be lost. As such a
3998 large number of recipients for a single message is ridiculous
3999 (performance will be very, very poor), I have chosen to impose a limit
4000 rather than extend the field.
4006 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4007 leftover -K file (the existence of which was triggered by #402).
4008 While we were at it, introduced process PID as part of the -K
4009 filename. This should rule out race conditions when creating
4012 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4013 processing considerably. Previous code took too long for large mails,
4014 triggering a timeout which in turn triggers #401.
4016 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4017 in the DK code in transports.c. sendfile() is not really portable,
4018 hence the _LINUX specificness.
4020 TF/01 In the add_headers option to the mail command in an Exim filter,
4021 there was a bug that Exim would claim a syntax error in any
4022 header after the first one which had an odd number of characters
4025 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4026 callout verification, Exim cached a "reject" for the entire domain. This
4027 is correct for most verifications, but it is not correct for a recipient
4028 verification with use_sender or use_postmaster set, because in that case
4029 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4030 case of MAIL FROM:<> rejection from other early rejections (e.g.
4031 rejection of HELO). When verifying a recipient using a non-null MAIL
4032 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4033 Whatever the result of the callout, the value of the domain cache is
4034 left unchanged (for any other kind of callout, getting as far as trying
4035 RCPT means that the domain itself is ok).
4037 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4040 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4041 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4042 session, the daemon ignores ECONNECT errors and logs others; it now
4043 ignores EPIPE as well.
4045 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4046 (quoted-printable decoding).
4048 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4049 later the small subsequent patch to fix an introduced bug.
4051 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4053 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4055 PH/08 An error is now given if message_size_limit is specified negative.
4057 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4058 to be given (somewhat) arbitrary names.
4060 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4063 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4064 miscellaneous code fixes
4066 PH/10 Added the log_reject_target ACL modifier to specify where to log
4069 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4070 hostname. This is wrong, because it relates to the incoming message (and
4071 probably the interface on which it is arriving) and not to the outgoing
4072 callout (which could be using a different interface). This has been
4073 changed to use the value of the helo_data option from the smtp transport
4074 instead - this is what is used when a message is actually being sent. If
4075 there is no remote transport (possible with a router that sets up host
4076 addresses), $smtp_active_hostname is used.
4078 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4079 tweaks were necessary in order to get it to work (see also 21 below):
4080 (a) The code assumed that strncpy() returns a negative number on buffer
4081 overflow, which isn't the case. Replaced with Exim's string_format()
4083 (b) There were several signed/unsigned issues. I just did the minimum
4084 hacking in of casts. There is scope for a larger refactoring.
4085 (c) The code used strcasecmp() which is not a standard C function.
4086 Replaced with Exim's strcmpic() function.
4087 (d) The code set only $1; it now sets $auth1 as well.
4088 (e) A simple test gave the error "authentication client didn't specify
4089 service in request". It would seem that Dovecot has changed its
4090 interface. Fortunately there's a specification; I followed it and
4091 changed what the client sends and it appears to be working now.
4093 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4096 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4097 address A is aliased to B and C, where B exists and C does not. Without
4098 -v the output is "A verified" because verification stops after a
4099 successful redirection if more than one address is generated. However,
4100 with -v the child addresses are also verified. Exim was outputting "A
4101 failed to verify" and then showing the successful verification for C,
4102 with its parentage. It now outputs "B failed to verify", showing B's
4103 parentage before showing the successful verification of C.
4105 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4106 look up a TXT record in a specific list after matching in a combined
4109 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4110 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4111 they consult the DNS. I had assumed they would set it the way they
4112 wanted; and indeed my experiments on Linux seem to show that in some
4113 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4114 To be on the safe side, however, I have now made the interface to
4115 host_find_byname() similar to host_find_bydns(), with an argument
4116 containing the DNS resolver options. The host_find_byname() function now
4117 sets these options at its start, just as host_find_bydns() does. The smtp
4118 transport options dns_qualify_single and dns_search_parents are passed to
4119 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4120 of host_find_byname() use the default settings of RES_DEFNAMES
4121 (qualify_single) but not RES_DNSRCH (search_parents).
4123 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4124 spool_read_header() do less string testing, by means of a preliminary
4125 switch on the second character of optional "-foo" lines. (This is
4126 overdue, caused by the large number of possibilities that now exist.
4127 Originally there were few.) While I was there, I also converted the
4128 str(n)cmp tests so they don't re-test the leading "-" and the first
4129 character, in the hope this might squeeze out yet more improvement.
4131 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4132 flag allowing group syntax was set by the header_syntax check but not
4133 turned off, possible causing trouble later; (2) The flag was not being
4134 set at all for the header_verify test, causing "group"-style headers to
4135 be rejected. I have now set it in this case, and also caused header_
4136 verify to ignore an empty address taken from a group. While doing this, I
4137 came across some other cases where the code for allowing group syntax
4138 while scanning a header line wasn't quite right (mostly, not resetting
4139 the flag correctly in the right place). These bugs could have caused
4140 trouble for malformed header lines. I hope it is now all correct.
4142 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4143 with the "reply" argument non-NULL. The code, however (which originally
4144 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4145 but it didn't always do it. This confused somebody who was copying the
4146 code for some other use. I have removed all the tests.
4148 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4149 feature that was used to support insecure browsers during the U.S. crypto
4150 embargo. It requires special client support, and Exim is probably the
4151 only MTA that supported it -- and would never use it because real RSA is
4152 always available. This code has been removed, because it had the bad
4153 effect of slowing Exim down by computing (never used) parameters for the
4154 RSA_EXPORT functionality.
4156 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4157 authenticator to fail if there's a tab character in the incoming data
4158 (there should never be unless someone is messing about, as it's supposed
4159 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4160 if the connection is using TLS or if the remote IP is the same as the
4161 local IP, and the "valid-client-cert option" if a client certificate has
4164 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4165 authenticators. This can be used for authorization after authentication
4166 succeeds. (In the case of plaintext, it servers for both authentication
4169 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4170 if any retry times were supplied.
4172 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4173 connection, where there is no client IP address to check. In this
4174 situation, the verify now always succeeds.
4176 PH/25 Applied John Jetmore's -Mset patch.
4178 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4180 PH/27 In a string expansion for a processed (not raw) header when multiple
4181 headers of the same name were present, leading whitespace was being
4182 removed from all of them, but trailing whitespace was being removed only
4183 from the last one. Now trailing whitespace is removed from each header
4184 before concatenation. Completely empty headers in a concatenation (as
4185 before) are ignored.
4187 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4188 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4190 PH/29 [Removed. This was a change that I later backed out, and forgot to
4191 correct the ChangeLog entry (that I had efficiently created) before
4192 committing the later change.]
4194 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4195 address errors (4xx response to RCPT) over the same connection as other
4196 messages routed to the same hosts. Such deliveries are always "forced",
4197 so retry times are not inspected. This resulted in far too many retries
4198 for the affected addresses. The effect occurred only when there were more
4199 hosts than the hosts_max_try setting in the smtp transport when it had
4200 the 4xx errors. Those hosts that it had tried were not added to the list
4201 of hosts for which the message was waiting, so if all were tried, there
4202 was no problem. Two fixes have been applied:
4204 (i) If there are any address or message errors in an SMTP delivery, none
4205 of the hosts (tried or untried) are now added to the list of hosts
4206 for which the message is waiting, so the message should not be a
4207 candidate for sending over the same connection that was used for a
4208 successful delivery of some other message. This seems entirely
4209 reasonable: after all the message is NOT "waiting for some host".
4210 This is so "obvious" that I'm not sure why it wasn't done
4211 previously. Hope I haven't missed anything, but it can't do any
4212 harm, as the worst effect is to miss an optimization.
4214 (ii) If, despite (i), such a delivery is accidentally attempted, the
4215 routing retry time is respected, so at least it doesn't keep
4216 hammering the server.
4218 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4219 in ${readsocket because some servers need this prod.
4221 PH/32 Added some extra debug output when updating a wait-xxx database.
4223 PH/33 The hint "could be header name not terminated by colon", which has been
4224 given for certain expansion errors for a long time, was not being given
4225 for the ${if def:h_colon_omitted{... case.
4227 PH/34 The spec says: "With one important exception, whenever a domain list is
4228 being scanned, $domain contains the subject domain." There was at least
4229 one case where this was not true.
4231 PH/35 The error "getsockname() failed: connection reset by peer" was being
4232 written to the panic log as well as the main log, but it isn't really
4233 panic-worthy as it just means the connection died rather early on. I have
4234 removed the panic log writing for the ECONNRESET error when getsockname()
4237 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4238 runs only) independently of the message's sender address. This meant
4239 that, if the 4xx error was in fact related to the sender, a different
4240 message to the same recipient with a different sender could confuse
4241 things. In particular, this can happen when sending to a greylisting
4242 server, but other circumstances could also provoke similar problems.
4243 I have changed the default so that the retry time for these errors is now
4244 based a combination of the sender and recipient addresses. This change
4245 can be overridden by setting address_retry_include_sender=false in the
4248 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4249 remote server are returned as part of bounce messages. This was not
4250 happening for LMTP over a pipe (the lmtp transport), but now it is the
4251 same for both kinds of LMTP.
4253 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4254 in header lines that were in fact CNAMEs. This is no longer the case.
4256 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4257 and queue runs started by the daemon processed all messages. This has
4258 been fixed so that -R and -S can now usefully be given with -q<time>.
4260 PH/40 Import PCRE release 6.7 (fixes some bugs).
4262 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4264 PH/42 Give an error if -q is specified more than once.
4266 PH/43 Renamed the variables $interface_address and $interface_port as
4267 $received_ip_address and $received_port, to make it clear that these
4268 values apply to message reception, and not to the outgoing interface when
4269 a message is delivered. (The old names remain recognized, of course.)
4271 PH/44 There was no timeout on the connect() call when using a Unix domain
4272 socket in the ${readsocket expansion. There now is.
4274 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4275 be meaningful with "accept".
4277 SC/01 Eximstats V1.43
4278 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4280 SC/02 Eximstats V1.44
4281 Use a glob alias rather than an array ref in the generated
4282 parser. This improves both readability and performance.
4284 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4285 Collect SpamAssassin and rejection statistics.
4286 Don't display local sender or destination tables unless
4287 there is data to show.
4288 Added average volumes into the top table text output.
4290 SC/04 Eximstats V1.46
4291 Collect data on the number of addresses (recipients)
4292 as well as the number of messages.
4294 SC/05 Eximstats V1.47
4295 Added 'Message too big' to the list of mail rejection
4296 reasons (thanks to Marco Gaiarin).
4298 SC/06 Eximstats V1.48
4299 Mainlog lines which have GMT offsets and are too short to
4300 have a flag are now skipped.
4302 SC/07 Eximstats V1.49 (Alain Williams)
4303 Added the -emptyok flag.
4305 SC/08 Eximstats V1.50
4306 Fixes for obtaining the IP address from reject messages.
4308 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4309 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4310 whitespace changes from 4.64-PH/27
4312 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4315 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4316 are found, allow negative numbers in numeric criteria)
4318 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4320 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4321 to match changes made in 4.64-PH/43
4323 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4325 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4326 "domain" is actually an IP address, and force a failure. This locks out
4327 those revolvers/nameservers that support "A-for-A" lookups, in
4328 contravention of the specifications.
4330 PH/48 When a host name was looked up from an IP address, and the subsequent
4331 forward lookup of the name timed out, the host name was left in
4332 $sender_host_name, contrary to the specification.
4334 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4335 restricted to single-key lookups, Exim was not diagnosing an error if
4336 * or *@ was used with a query-style lookup.
4338 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4340 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4341 long ago, but noone interested enough thought of it. Let's just say that
4342 the "1.1" means that there are some new functions that weren't there at
4343 some point in the past.
4345 PH/51 Error processing for expansion failure of helo_data from an smtp
4346 transport during callout processing was broken.
4348 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4349 tested/used via the -bh/-bhc/-bs options.
4351 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4352 bug, fixed in subsequent PCRE releases).
4354 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4355 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4357 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4363 SC/01 Use a glob alias rather than an array ref in eximstats generated
4364 parser. This improves both readability and performance.
4366 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4367 Don't display local sender or destination tables in eximstats unless
4368 there is data to show.
4369 Added average volumes into the eximstats top table text output.
4371 SC/03 Collect data on the number of addresses (recipients) as well
4372 as the number of messages in eximstats.
4374 TF/01 Correct an error in the documentation for the redirect router. Exim
4375 does (usually) call initgroups() when daemonizing.
4377 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4378 with consistent privilege compared to when running as a daemon.
4380 TF/03 Note in the spec that $authenticated_id is not set for local
4381 submissions from trusted users.
4383 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4384 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4386 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4387 by adding some example configuration directives to the default
4388 configuration file. A little bit of work is required to uncomment the
4389 directives and define how usernames and passwords are checked, but
4390 there is now a framework to start from.
4392 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4393 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4394 without this. I don't know how relevant this is to other LDAP libraries.
4396 PH/02 Add the verb name to the "unknown ACL verb" error.
4398 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4400 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4402 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4403 directory not expanded when it should be if an expanded home directory
4404 was set for the address (which is overridden by the transport).
4406 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4409 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4410 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4411 because it is too late at that time, and has no effect.
4413 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4414 security issue with \' (bugzilla #107). I could not use the
4415 PQescapeStringConn() function, because it needs a PGconn value as one of
4418 PH/08 When testing addresses using -bt, indicate those final addresses that
4419 are duplicates that would not cause an additional delivery. At least one
4420 person was confused, thinking that -bt output corresponded to deliveries.
4421 (Suppressing duplicates isn't a good idea as you lose the information
4422 about possibly different redirections that led to the duplicates.)
4424 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4425 systems where poll() doesn't work, in particular OS X.
4427 PH/10 Added more information to debugging output for retry time not reached.
4429 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4430 operations in malware.c.
4432 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4435 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4436 syslog_duplication set false, log lines that would normally be written
4437 both the the main log and to the reject log were not written to syslog at
4440 PH/14 In the default configuration, change the use of "message" in ACL warn
4441 statements to "add_header".
4443 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4444 not followed by a command (e.g. "seen endif").
4446 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4447 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4450 PH/17 Added extra conditions to the default value of delay_warning_condition
4454 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4455 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4456 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4459 The Auto-Submitted: and various List- headers are standardised, whereas I
4460 don't think Precedence: ever was.
4462 PH/18 Refactored debugging code in route_finduser() to show more information,
4463 in particular, the error code if getpwnam() issues one.
4465 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4466 This is apparently needed in addition to the PH/07 change above to avoid
4467 any possible encoding problems.
4469 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4470 but not after initializing Perl.
4472 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4473 output them only if debugging. By default they are written stderr,
4474 apparently, which is not desirable.
4476 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4479 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4482 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4484 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4485 authenticated or an ident call has been made. Suppress the default
4486 values for $authenticated_id and $authenticated_sender (but permit -oMai
4487 and -oMas) when testing with -bh.
4489 PH/24 Re-jigged the order of the tests in the default configuration so that the
4490 tests for valid domains and recipients precede the DNS black list and CSA
4491 tests, on the grounds that those ones are more expensive.
4493 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4494 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4495 This bug exists in every version of Exim that I still have, right back to
4498 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4499 However, an attempt to turn on case-sensitivity in a regex key by
4500 including (?-i) didn't work because the subject string was already
4501 lowercased, and the effects were non-intuitive. It turns out that a
4502 one-line patch can be used to allow (?-i) to work as expected.
4508 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4509 other effects) broke the use of negated acl sub-conditions.
4511 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4514 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4515 "Deny" causes Exim to reject the incoming connection with a 554 error.
4516 Unfortunately, if there is a major crisis, such as a disk failure,
4517 tcp-wrappers gives "deny", whereas what one would like would be some
4518 kind of temporary error. A kludge has been added to help with this.
4519 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4520 554 error is used if errno is still zero or contains ENOENT (which occurs
4521 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4524 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4526 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4527 errors. Otherwise a message that provokes a temporary error (when other
4528 messages do not) can cause a whole host to time out.
4530 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4531 addresses were routed directly to files or pipes from a redirect router.
4532 File deliveries just didn't batch; pipe deliveries might have suffered
4535 PH/06 A failure to get a lock for a hints database would erroneously always say
4536 "Failed to get write lock", even when it was really a read lock.
4538 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4539 of 0600. This has been changed to use the value of the lockfile_mode
4540 option (which defaults to 0600).
4542 PH/08 Applied small patch from the Sieve maintainer.
4544 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4545 folder from quota calculations, a direct delivery into this folder messed
4546 up the contents of the maildirsize file. This was because the regex was
4547 used only to exclude .Trash (or whatever) when the size of the mailbox
4548 was calculated. There was no check that a delivery was happening into an
4549 excluded directory. This bug has been fixed by ignoring all quota
4550 processing for deliveries into excluded directories.
4552 PH/10 Added the maildirfolder_create_regex option to appendfile.
4558 PH/01 The code for finding all the local interface addresses on a FreeBSD
4559 system running IPv6 was broken. This may well have applied to all BSD
4560 systems, as well as to others that have similar system calls. The broken
4561 code found IPv4 interfaces correctly, but gave incorrect values for the
4562 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4563 that it would not match correctly against @[] and not recognize the IPv6
4566 PH/02 The ipliteral router was not recognizing addresses of the form user@
4567 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4569 PH/03 Added disable_ipv6.
4571 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4572 decoded form, because it is most often used to construct To: headers
4573 lines in autoreplies, and the decoded form may well be syntactically
4574 invalid. However, $reply_address has leading white space removed, and all
4575 newlines turned into spaces so that the autoreply transport does not
4578 PH/05 If group was specified without a user on a router, and no group or user
4579 was specified on a transport, the group from the router was ignored.
4581 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4582 for visible compile-time settings that can be used to change these
4583 numbers, for those that want even more. Backwards compatibility with old
4584 spool files has been maintained. However, going back to a previous Exim
4585 release will lost any variables that are in spool files.
4587 PH/07 Two small changes when running in the test harness: increase delay when
4588 passing a TCP/IP connection to a new process, in case the original
4589 process has to generate a bounce, and remove special handling of
4590 127.0.0.2 (sic), which is no longer necessary.
4592 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4593 be the same on different OS.
4595 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4598 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4599 whether --show-vars was specified or not
4601 JJ/02 exipick: Added support for new ACL variable spool format introduced
4604 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4605 syntactically invalid From: or Reply-to: line, and a filter used this to
4606 generate an autoreply, and therefore failed to obtain an address for the
4607 autoreply, Exim could try to deliver to a non-existent relative file
4608 name, causing unrelated and misleading errors. What now happens is that
4609 it logs this as a hard delivery error, but does not attempt to create a
4612 PH/11 The exinext utility has a -C option for testing purposes, but although
4613 the given file was scanned by exinext itself; it wasn't being passed on
4614 when Exim was called.
4616 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4617 an end-of-file indication when reading a command response.
4619 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4620 compiled. In many other places in Exim, IPv6 addresses are always
4621 recognized, so I have changed this. It also means that IPv4 domain
4622 literals of the form [IPV4:n.n.n.n] are now always recognized.
4624 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4625 used if the router is not running as root, for example, when verifying at
4626 ACL time, or when using -bh. The debugging output from this situation was
4627 non-existent - all you got was a failure to exec. I have made two
4630 (a) Failures to set uid/gid, the current directory, or a process leader
4631 in a subprocess such as that created by queryprogram now generate
4632 suitable debugging output when -d is set.
4634 (b) The queryprogram router detects when it is not running as root,
4635 outputs suitable debugging information if -d is set, and then runs
4636 the subprocess without attempting to change uid/gid.
4638 PH/15 Minor change to Makefile for building test_host (undocumented testing
4641 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4642 additional section of a DNS packet that returns MX or SRV records.
4643 Instead, it always explicitly searches for A/AAAA records. This avoids
4644 major problems that occur when a DNS server includes only records of one
4645 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4646 fixed another bug: if SRV records were looked up and the corresponding
4647 address records were *not* found in the additional section, the port
4648 values from the SRV records were lost.
4650 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4651 using the correct key (the original address) when searching the retry
4652 rules in order to find which one to use for generating the retry hint.
4654 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4655 adding the default one. Similarly, if it contains a Reply-To: header, the
4656 errors_reply_to option, if set, is not used.
4658 PH/19 When calculating a retry time, Exim used to measure the "time since
4659 failure" by looking at the "first failed" field in the retry record. Now
4660 it does not use this if it is later than than the arrival time of the
4661 message. Instead it uses the arrival time. This makes for better
4662 behaviour in cases where some deliveries succeed, thus re-setting the
4663 "first failed" field. An example is a quota failure for a huge message
4664 when small messages continue to be delivered. Without this change, the
4665 "time since failure" will always be short, possible causing more frequent
4666 delivery attempts for the huge message than are intended.
4667 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4669 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4670 $1, $2, $3) because the numerical variables can be reset during some
4671 expansion items (e.g. "match"), thereby losing the authentication data.
4673 PH/21 Make -bV show the size of off_t variables so that the test suite can
4674 decide whether to run tests for quotas > 2G.
4676 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4677 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4678 filecount value is greater than 2G or if a quota value is greater than 2G
4679 on a system where the size of off_t is not greater than 4, a panic error
4682 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4683 never match. The debug and -bh output now contains an explicit error
4684 message indicating a malformed IPv4 address or mask.
4686 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4687 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4688 PH/23 above applies.
4690 PH/25 Do not write to syslog when running in the test harness. The only
4691 occasion when this arises is a failure to open the main or panic logs
4692 (for which there is an explicit test).
4694 PH/26 Added the /no_tell option to "control=freeze".
4696 PH/27 If a host name lookup failed very early in a connection, for example, if
4697 the IP address matched host_lookup and the reverse lookup yielded a name
4698 that did not have a forward lookup, an error message of the form "no IP
4699 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4700 could be logged. Now it outputs the IP address instead of "NULL".
4702 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4703 allows the sender and the authenticated sender to be set when
4704 submitting a message from within Exim. Since child_open_exim() is
4705 documented for local_scan(), the new function should be too.
4707 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4708 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4709 results in an empty string is now treated as unset.
4711 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4713 PH/31 Added sender_verify_fail logging option.
4715 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4716 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4717 tidied the source and removed it altogether.
4719 PH/33 When a queue run was abandoned because the load average was too high, a
4720 log line was always written; now it is written only if the queue_run log
4721 selector is set. In addition, the log line for abandonment now contains
4722 information about the queue run such as the pid. This is always present
4723 in "start" and "stop" lines but was omitted from the "abandon" line.
4725 PH/34 Omit spaces between a header name and the colon in the error message that
4726 is given when verify = headers_syntax fails (if there are lots of them,
4727 the message gets confusing).
4729 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4730 names, as there are now some PTR records that contain slashes. This check
4731 is only to protect against broken name servers that fall over on strange
4732 characters, so the fact that it applies to all lookups doesn't matter.
4734 PH/36 Now that the new test suite is complete, we can remove some of the
4735 special code in Exim that was needed for the old test suite. For example,
4736 sorting DNS records because real resolvers return them in an arbitrary
4737 order. The new test suite's fake resolver always returns records in the
4740 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4741 bounces) except when queue_only is set, to avoid logging races between
4742 the different processes.
4744 PH/38 Panic-die if .include specifies a non-absolute path.
4746 PH/39 A tweak to the "H" retry rule from its user.
4748 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4749 a label. They prevented compilation on older perls.
4751 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4752 a warning to be raised on newish perls.
4754 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4755 on queue. Changes to match documented behaviour of showing count of
4756 messages matching specified criteria.
4758 PH/40 Changed the default ident timeout from 30s to 5s.
4760 PH/41 Added support for the use of login_cap features, on those BSD systems
4761 that have them, for controlling the resources used by pipe deliveries.
4763 PH/42 The content-scanning code uses fopen() to create files in which to put
4764 message data. Previously it was not paying any attention to the mode of
4765 the files. Exim runs with umask(0) because the rest of the code creates
4766 files with open(), and sets the required mode explicitly. Thus, these
4767 files were ending up world-writeable. This was not a big issue, because,
4768 being within the spool directory, they were not world-accessible. I have
4769 created a function called modefopen, which takes an additional mode
4770 argument. It sets umask(777), creates the file, chmods it to the required
4771 mode, then resets the umask. All the relevant calls to fopen() in the
4772 content scanning code have been changed to use this function.
4774 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4775 to 24 hours. This avoids potential overflow problems when processing G
4776 and H retry rules. I suspect nobody ever tinkers with this value.
4778 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4780 PH/45 When the plaintext authenticator is running as a client, the server's
4781 challenges are checked to ensure they are valid base64 strings. By
4782 default, the authentication attempt is cancelled if an invalid string is
4783 received. Setting client_ignore_invalid_base64 true ignores these errors.
4784 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4785 they are received. Thus, the responses can be made to depend on the
4786 challenges. If an invalid string is ignored, an empty string is placed in
4789 PH/46 Messages that are created by the autoreply transport now contains a
4790 References: header, in accordance with RFCs 2822 and 3834.
4792 PH/47 Added authenticated_sender_force to the smtp transport.
4794 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4796 PH/49 Installed latest patch from the Sieve maintainer.
4798 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4799 was also set, the appendfile transport was unnecessarily scanning a
4800 directory of message files (e.g. for maildir delivery) to find the count
4801 of files (along with the size), even though it did not need this
4802 information. It now does the scan only if it needs to find either the
4803 size of the count of files.
4805 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4807 PH/52 Two bugs concerned with error handling when the smtp transport is
4810 (i) Exim was not creating retry information for temporary errors given
4811 for individual recipients after the DATA command when the smtp transport
4812 was used in LMTP mode. This meant that they could be retried too
4813 frequently, and not timed out correctly.
4815 (ii) Exim was setting the flag that allows error details to be returned
4816 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4817 recipients that were returned after the DATA command.
4819 PH/53 This is related to PH/52, but is more general: for any failing address,
4820 when detailed error information was permitted to be returned to the
4821 sender, but the error was temporary, then after the final timeout, only
4822 "retry timeout exceeded" was returned. Now it returns the full error as
4823 well as "retry timeout exceeded".
4825 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4826 do this, and (what is worse) MTAs that accept it.
4828 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4829 will now be deprecated.
4831 PH/56 New os.c-cygwin from the Cygwin maintainer.
4833 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4834 formats (previously only available in exim formats via -bpr, -bpru,
4835 and -bpra. Now also available in native and exiqgrep formats)
4837 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4838 with very large, slow to parse queues
4840 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4842 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4844 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4845 responses to authentication challenges, though it was showing the
4846 challenges; (ii) I've removed the CR characters from the debug output for
4849 PH/58 Allow for the insertion of a newline as well as a space when a string
4850 is turned into more than one encoded-word during RFC 2047 encoding. The
4851 Sieve code now uses this.
4853 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4854 data_4xx, lost_connection, tls_required.
4856 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4857 message was being sent as an SMTP response.
4859 PH/61 Add -l and -k options to exicyclog.
4861 PH/62 When verifying, if an address was redirected to one new address, so that
4862 verification continued, and the new address failed or deferred after
4863 having set something in $address_data, the value of $address_data was not
4864 passed back to the ACL. This was different to the case when no
4865 redirection occurred. The value is now passed back in both cases.
4867 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4868 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4869 use login_cap.h, so on its own it isn't the distinguishing feature. The
4870 new name refers directly to the setclassresources() function.
4872 PH/65 Added configuration files for NetBSD3.
4874 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4876 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4877 is preferred over IPv4.
4879 PH/68 The bounce_return_message and bounce_return_body options were not being
4880 honoured for bounces generated during the reception of non-SMTP messages.
4881 In particular, this applied to messages rejected by the ACL. This bug has
4882 been fixed. However, if bounce_return_message is true and bounce_return_
4883 body is false, the headers that are returned for a non-SMTP message
4884 include only those that have been read before the error was detected.
4885 (In the case of an ACL rejection, they have all been read.)
4887 PH/69 The HTML version of the specification is now built in a directory called
4888 spec_html instead of spec.html, because the latter looks like a path with
4889 a MIME-type, and this confuses some software.
4891 PH/70 Catch two compiler warnings in sieve.c.
4893 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4894 function verify_get_ident() calls ip_connect() to connect a socket, but
4895 if the "connect()" function timed out, ip_connect() used to close the
4896 socket. However, verify_get_ident() also closes the socket later, and in
4897 between Exim writes to the log, which may get opened at this point. When
4898 the socket was closed in ip_connect(), the log could get the same file
4899 descriptor number as the socket. This naturally causes chaos. The fix is
4900 not to close the socket in ip_connect(); the socket should be closed by
4901 the function that creates it. There was only one place in the code where
4902 this was missing, in the iplookup router, which I don't think anybody now
4903 uses, but I've fixed it anyway.
4905 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4906 well as to direct DNS lookups. Otherwise the handling of names in host
4907 lists is inconsistent and therefore confusing.
4913 PH/01 Two changes to the default runtime configuration:
4915 (1) Move the checks for relay_from_hosts and authenticated clients from
4916 after to before the (commented out) DNS black list checks.
4918 (2) Add control=submission to the relay_from_hosts and authenticated
4919 clients checks, on the grounds that messages accepted by these
4920 statements are most likely to be submissions.
4922 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4924 (1) Generate an error if the third argument for the ${prvs expansion is
4927 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4930 (3) Reset the variables that are obtained from the first argument of
4931 ${prvscheck and used in the second argument before leaving the code,
4932 because their memory is reclaimed, so using them afterwards may do
4935 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4936 one (it's much easier than Tom thought :-).
4938 (5) Because of (4), we can now allow for the use of $prvscheck_result
4939 inside the third argument.
4941 PH/03 For some reason, the default setting of PATH when running a command from
4942 a pipe transport was just "/usr/bin". I have changed it to
4945 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4946 anything to be listed in the output from -bV.
4948 PH/05 When a filter generated an autoreply, the entire To: header line was
4949 quoted in the delivery log line, like this:
4951 => >A.N.Other <ano@some.domain> <original@ddress> ...
4953 This has been changed so that it extracts the operative address. There
4954 may be more than one such address. If so, they are comma-separated, like
4957 => >ano@some.domain,ona@other.domain <original@ddress> ...
4959 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4960 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4961 not being looked up in the rDNS to get a host name, Exim was showing the
4962 IP address twice in Received: lines, even though the IP addresses were
4963 identical. For example:
4965 Received: from [1.2.3.4] (helo=[1.2.3.4])
4967 However, if the real host name was known, it was omitting the HELO data
4968 if it matched the actual IP address. This has been tidied up so that it
4969 doesn't show the same IP address twice.
4971 PH/07 When both +timestamp and +memory debugging was on, the value given by
4972 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4973 called by the debug printing, thereby overwriting the timestamp buffer.
4974 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4976 PH/08 When the original message was included in an autoreply transport, it
4977 always said "this is a copy of the message, including all the headers",
4978 even if body_only or headers_only was set. It now gives an appropriate
4981 PH/09 Applied a patch from the Sieve maintainer which:
4983 o fixes some comments
4984 o adds the (disabled) notify extension core
4985 o adds some debug output for the result of if/elsif tests
4986 o points to the current vacation draft in the documentation
4987 and documents the missing references header update
4991 o fixes a bug in processing the envelope test (when testing
4992 multiple envelope elements, the last element determined the
4995 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4996 Electronic Mail") by including:
4998 Auto-submitted: auto-generated
5000 in the messages that it generates (bounce messages and others, such as
5001 warnings). In the case of bounce messages for non-SMTP messages, there was
5002 also a typo: it was using "Auto_submitted" (underscore instead of
5003 hyphen). Since every message generated by Exim is necessarily in response
5004 to another message, thes have all been changed to:
5006 Auto-Submitted: auto-replied
5008 in accordance with these statements in the RFC:
5010 The auto-replied keyword:
5012 - SHOULD be used on messages sent in direct response to another
5013 message by an automatic process,
5015 - MUST NOT be used on manually-generated messages,
5017 - MAY be used on Delivery Status Notifications (DSNs) and Message
5018 Disposition Notifications (MDNs),
5020 - MUST NOT be used on messages generated by automatic or periodic
5021 processes, except for messages which are automatic responses to
5024 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5025 to the default Received: header definition.
5027 PH/12 Added log selector acl_warn_skipped (default on).
5029 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5030 variables because (a) they are in the wrong storage pool and (b) even if
5031 they were copied, it wouldn't work properly because of the caching.
5033 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5034 checking when decoding. Apparently there are clients that generate
5035 overlong encoded strings. Why am I not surprised?
5037 PH/15 If the first argument of "${if match_address" was not empty, but did not
5038 contain an "@" character, Exim crashed. Now it writes a panic log message
5039 and treats the condition as false.
5041 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5043 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5044 extension "envelope-auth". The code is finished and in agreement with
5045 other implementations, but there is no documentation so far and in fact,
5046 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5047 not changing the active code.
5049 Print executed "if" and "elsif" statements when debugging is used. This
5050 helps a great deal to understand what a filter does.
5052 Document more things not specified clearly in RFC3028. I had all this
5053 sorted out, when out of a sudden new issues came to my mind. Oops."
5055 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5058 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5059 canonical form (as documented). However, after a host name lookup from
5060 the IP address, check_host() was doing a simple string comparison with
5061 addresses acquired from the DNS when checking that the found name did
5062 have the original IP as one of its addresses. Since any found IPv6
5063 addresses are likely to be in abbreviated form, the comparison could
5064 fail. Luckily, there already exists a function for doing the comparison
5065 by converting both addresses to binary, so now that is used instead of
5066 the text comparison.
5068 PH/20 There was another similar case to PH/19, when a complete host name was
5069 given in a host list; looking up its IP address could give an abbreviated
5070 form, whereas the current host's name might or might not be abbreviated.
5071 The same fix has been applied.
5077 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5078 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5081 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5082 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5084 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5086 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5087 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5088 Also, for "Auto-Submitted", treat anything other than "no" as
5089 non-personal, in accordance with RFC 3834. (Previously it treated
5090 anything starting "auto-" as non-personal.)
5092 TF/01 The control=submission/name=... option had a problem with syntax
5093 errors if the name included a slash character. The /name= option
5094 now slurps the rest of the string, so it can include any characters
5095 but it must come last in the list of options (after /sender_retain
5098 PH/05 Some modifications to the interface to the fake nameserver for the new
5106 TK/01 Added the "success_on_redirect" address verification option. See
5107 NewStuff for rationale and an example.
5109 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5111 PH/02 Patch to exigrep to allow it to work on syslog lines.
5113 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5114 fread() to skip over the body file's header line, because in Cygwin the
5115 header line is locked and is inaccessible.
5117 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5118 co-exist for some time) to make it clear that it is the Exim ID that is
5119 referenced, not the Message-ID: header line.
5121 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5122 string_format() function, because snprintf() does not exist on all
5125 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5126 ${stat: expansion item.
5128 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5129 protocol synchronization error", to keep the pedants happy.
5131 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5132 well as for IRIX systems, when gcc is being used. See the host.c source
5135 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5137 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5140 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5141 transport and to the smtp transport in LMTP mode.
5143 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5145 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5146 run a filter in a subprocess. This could lead to confusion in subsequent
5147 lookups in the parent process. There should also be a search_tidyup() at
5148 the end of the subprocess.
5150 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5151 only if the host matched helo_try_verify_hosts, which caused the
5152 verification to occur when the EHLO/HELO command was issued. The ACL just
5153 tested the remembered result. Now, if a previous verification attempt has
5154 not happened, "verify = helo" does it there and then.
5156 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5158 TK/03 Fix log output including CR from clamd.
5160 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5161 exist provoked a memory error which could cause a segfault.
5163 PH/15 Installed PCRE 6.2
5165 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5167 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5168 of the problem. Specifically, suggested +O2 rather than +O1 for the
5171 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5173 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5174 record, the port was not correctly logged when the outgoing_port log
5175 selector was set (it logged the transort's default port).
5177 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5178 fallback_hosts, and "hosts" in the smtp transport.
5180 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5181 host errors such as "Connection refused".
5183 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5184 authentication with radiusclient 0.4.9:
5186 - Error returned from rc_read_config was caught wrongly
5187 - Username/password not passed on to radius server due to wrong length.
5189 The presumption is that some radiusclient API changes for 4.51/PH/17
5190 were not taken care of correctly. The code is still untested by me (my
5191 Linux distribution still has 0.3.2 of radiusclient), but it was
5192 contributed by a Radius user.
5194 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5195 expanding the "port" option of the smtp transport.
5197 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5198 while reading a MIME header. Thanks to Tom Hughes for a patch.
5200 PH/24 Include config.h inside local_scan.h so that configuration settings are
5203 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5204 that in an ACL for RCPT (for example), you can examine exactly what was
5207 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5208 commands, but it was not correctly comparing the address with the actual
5209 client host address. Thus, it would show the EHLO address in Received:
5210 header lines when this was not necessary.
5212 PH/27 Added the % operator to ${eval:}.
5214 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5215 it should be ignoring failures (because with -C, for example, it has lost
5216 privilege). It wasn't ignoring creation failures other than "already
5219 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5222 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5223 because an input line was too long, either on its own, or by virtue of
5224 too many continuations, the temporary file was not being removed, and the
5225 return code was incorrect.
5227 PH/31 Missing "BOOL" in function definition in filtertest.c.
5229 PH/32 Applied Sieve patches from the maintainer.
5231 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5233 PH/33 Added "verify = not_blind".
5235 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5236 Local/Makefile (with some defaults set). These are used in built scripts
5237 such as exicyclog, but they have never been used in the exim_install
5238 script (though there are many overriding facilities there). I have
5239 arranged that the exim_install script now takes note of these two
5242 PH/35 Installed configuration files for Dragonfly.
5244 PH/36 When a locally submitted message by a trusted user did not contain a
5245 From: header, and the sender address was obtained from -f or from an SMTP
5246 MAIL command, and the trusted user did not use -F to supply a sender
5247 name, $originator_name was incorrectly used when constructing a From:
5248 header. Furthermore, $originator_name was used for submission mode
5249 messages from external hosts without From: headers in a similar way,
5250 which is clearly wrong.
5252 PH/37 Added control=suppress_local_fixups.
5254 PH/38 When log_selector = +received_sender was set, and the addition of the
5255 sender made the log line's construction buffer exactly full, or one byte
5256 less than full, an overflow happened when the terminating "\n" was
5259 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5260 when the result of a list match is failure because a DNS lookup failed.
5262 PH/40 RM_COMMAND is now used in the building process.
5264 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5265 the "build-* directories that it finds.
5267 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5268 address in a domain literal was a prefix of an interface address.
5270 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5271 when verifying a sender address, unless rewrite_headers is false.
5273 PH/44 Wrote a long comment about why errors_to addresses are verified as
5274 recipients, not senders.
5276 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5277 the ratelimit ACL was added.
5279 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5281 PH/46 Added extra information about PostgreSQL errors to the error string.
5283 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5284 avoiding the need to install special zones in a real server. This is
5285 backwards compatible; if it can't find the fake resolver, it drops back.
5286 Thus, both old and new test suites can be run.
5288 TF/02 Added util/ratelimit.pl
5290 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5291 clock is set back in time.
5293 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5294 Candler <B.Candler@pobox.com>.
5296 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5297 OK for addresses that are the result of redirections.
5299 PH/48 A number of further additions for the benefit of the new test suite,
5300 including a fake gethostbyname() that interfaces to the fake DNS resolver
5303 TF/06 The fix for widen_domains has also been applied to qualify_single and
5304 search_parents which are the other dnslookup options that can cause
5307 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5310 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5312 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5313 service" if it encountered an MX record with an empty target hostname.
5314 The message is now "an MX or SRV record indicated no SMTP service".
5316 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5317 if the DNS of the sending site is misconfigured. This is quite a
5318 common situation. This change restores the behaviour of treating a
5319 helo verification defer as a failure.
5321 PH/51 If self=fail was set on a router, the bounce message did not include the
5322 actual error message.
5328 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5330 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5331 command itself ended in error, the underlying message about the transport
5332 filter timeout was being overwritten with the pipe command error. Now the
5333 underlying error message should be appended to the second error message.
5335 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5337 PH/02 Reduce the amount of output that "make" produces by default. Full output
5338 can still be requested.
5340 PH/03 The warning log line about a condition test deferring for a "warn" verb
5341 was being output only once per connection, rather than after each
5342 occurrence (because it was using the same function as for successful
5343 "warn" verbs). This seems wrong, so I have changed it.
5345 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5346 it should not have, which might have caused a crash in the right
5347 circumstances, but probably never did.
5349 PH/04 Installed a modified version of Tony Finch's patch to make submission
5350 mode fix the return path as well as the Sender: header line, and to
5351 add a /name= option so that you can make the user's friendly name appear
5354 TF/03 Added the control = fakedefer ACL modifier.
5356 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5357 Mark Lowes for thorough testing.
5359 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5361 TK/03 Merged latest SRS patch from Miles Wilton.
5363 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5364 with the definition in sysexits.h (which is #included earlier).
5365 Fortunately, Exim does not actually use EX_OK. The code used to try to
5366 preserve the sysexits.h value, by assuming that macro definitions were
5367 scanned for macro replacements. I have been disabused of this notion,
5368 so now the code just undefines EX_OK before #including unistd.h.
5370 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5371 in the smtp transport. When a block could not be written in a single
5372 write() function, the timeout was being re-applied to each part-write.
5373 This seems wrong - if the receiver was accepting one byte at a time it
5374 would take for ever. The timeout is now adjusted when this happens. It
5375 doesn't have to be particularly precise.
5377 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5378 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5380 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5381 without a domain if the check to <postmaster@domain> fails.
5383 SC/01 Eximstats: added -xls and the ability to specify output files
5384 (patch written by Frank Heydlauf).
5386 SC/02 Eximstats: use FileHandles for outputting results.
5388 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5390 SC/04 Eximstats: fixed display of large numbers with -nvr option
5392 SC/05 Eximstats: fixed merging of reports with empty tables.
5394 SC/06 Eximstats: added the -include_original_destination flag
5396 SC/07 Eximstats: removed tabs and trailing whitespace.
5398 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5400 TK/06 MBOX spool code: Add real "From " MBOX separator line
5401 so the .eml file is really in mbox format (even though
5402 most programs do not really care). Patch from Alex Miller.
5404 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5405 The latter is generated from $received_to and is only set if the
5406 message has one envelope recipient. SA can use these headers,
5407 obviously out-of-the-box. Patch from Alex Miller.
5409 PH/08 The ${def test on a variable was returning false if the variable's
5410 value was "0", contrary to what the specification has always said!
5411 The result should be true unless the variable is empty.
5413 PH/09 The syntax error of a character other than { following "${if
5414 def:variable_name" (after optional whitespace) was not being diagnosed.
5415 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5416 accidental colon was present, for example, could give incorrect results.
5418 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5419 result is used (not including appendfile, where other changes are about
5422 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5423 This involved changing a lot of size variables from int to off_t. It
5424 should work with maildirs and everything.
5426 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5427 spamd dying while we are connected to it.
5429 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5432 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5433 function for BDB 4.3.
5435 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5437 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5438 the sources for PCRE have been re-arranged and more files are now
5441 PH/15 The code I had for printing potentially long long variables in PH/11
5442 above was not the best (it lost precision). The length of off_t variables
5443 is now inspected at build time, and an appropriate printing format (%ld
5444 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5445 to be "long long int" or "long int". This is needed for the internal
5446 formatting function string_vformat().
5448 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5449 the configuration file to be ":syslog", then the script "guesses" where
5450 the logs files are, rather than using the compiled in default. In our
5451 case the guess is not the same as the compiled default, so the script
5452 suddenly stopped working when I started to use syslog. The patch checks
5453 to see if log_file_path is "". If so, it attempts to read it from exim
5454 with no configuration file to get the compiled in version, before it
5455 falls back to the previous guessing code."
5457 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5458 implementing BATV in an Exim configuration. See NewStuff for the gory
5461 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5462 Makefile that are specific to HP-UX.
5464 PH/18 If the "use_postmaster" option was set for a recipient callout together
5465 with the "random" option, the postmaster address was used as the MAIL
5466 FROM address for the random test, but not for the subsequent recipient
5467 test. It is now used for both.
5469 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5470 patch removes a few documentation additions to RFC 3028, because the
5471 latest draft now contains them. It adds the new en;ascii-case comparator
5472 and a new error check for 8bit text in MIME parts. Comparator and
5473 require names are now matched exactly. I enabled the subaddress
5474 extension, but it is not well tested yet (read: it works for me)."
5476 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5477 rework some of the code of TK/09 above to avoid the hardwired use of
5478 "%lld" and "long long". Replaced the call to snprintf() with a call to
5481 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5482 records point to non-existent hosts", "retry timeout exceeded", and
5483 "retry time not reached for any host after a long failure period".
5485 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5486 experimental DomainKeys support:
5488 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5489 (2) On an error such as an illegally used "control", the wrong name for
5490 the control was given.
5492 These problems did NOT occur unless DomainKeys support was compiled.
5494 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5496 PH/24 Added ${if match_ip condition.
5498 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5499 databases so that it will be absolutely obvious if a crash occurs in the
5500 DB library. This is a regular occurrence (often caused by mis-matched
5503 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5504 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5505 picked up on a user's system that detects such things. There doesn't seem
5506 to be a gcc warning option for this - only an attribute that has to be
5507 put on the function's prototype. It seems that in Fedora Core 4 they have
5508 set this on a number of new functions. No doubt there will be more in due
5511 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5512 specify a transport. However, if an address that was verified by such a
5513 router was the subject of a callout, Exim crashed because it tried to
5514 read the rcpt_include_affixes from the non-existent transport. Now it
5515 just assumes that the setting of that option is false. This bug was
5516 introduced by 4.51/PH/31.
5518 PH/28 Changed -d+all to exclude +memory, because that information is very
5519 rarely of interest, but it makes the output a lot bigger. People tend to
5520 do -d+all out of habit.
5522 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5523 code in os-type was giving problems when libc.so lives in lib64, like on
5526 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5527 aren't the modern standard, and it seems that some systems' include files
5528 don't always have them. Exim was already checking for some of the newer
5529 ones like T_AAAA, and defining it itself. I've added checks for all the
5530 record types that Exim uses.
5532 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5533 not automatically generating a new one, as it is supposed to. This
5534 prevented TLS from working. If the file did exist, but contained invalid
5535 data, a new version was generated, as expected. It was only the case of a
5536 non-existent file that was broken.
5538 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5539 with a change in libdomainkeys > 0.64.
5541 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5542 from DNS. If the selector record carries the flag, it now has
5543 precedence over the domain-wide flag.
5545 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5547 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5548 the use of a port name that isn't defined in /etc/services) occurred, the
5549 message was deferred as in a normal delivery, and thus remained on the
5550 spool, instead of being failed because of the mua_wrapper setting. This
5551 is now fixed, and I tidied up some of the mua_wrapper messages at the
5554 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5555 the messages in a hash of arrays rather than using individual hashes.
5556 This is a bit cleaner and results in dramatic memory savings, albeit
5557 at a slight CPU cost.
5559 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5560 as requested by Marc Sherman.
5562 SC/10 Eximstats: added histograms for user specified patterns as requested
5565 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5567 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5568 fopen() in the content-scanning modules that did not already have it.
5574 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5575 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5577 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5579 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5581 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5582 to test Sieve filters that use "vacation".
5584 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5585 that changes the way the GnuTLS parameters are stored in the cache file.
5586 The new format can be generated externally. For backward compatibility,
5587 if the data in the cache doesn't make sense, Exim assumes it has read an
5588 old-format file, and it generates new data and writes a new file. This
5589 means that you can't go back to an older release without removing the
5592 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5593 work if there are any delivery delays because "one_time" forces the
5594 parent to be marked "delivered", so its unseen clone is never tried
5595 again. For this reason, Exim now forbids the simultaneous setting of
5598 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5599 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5600 entry if you want to know the details. The fix, however, neglected to
5601 consider the case when local delivery batching is involved. The test for
5602 "previously delivered" was not happening when checking to see if an
5603 address could be batched with a previous (undelivered) one; under
5604 certain circumstances this could lead to multiple deliveries to the same
5607 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5608 in its include files, and this causes problems building Exim.
5610 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5611 header_syntax) but Exim was just ignoring anything given after a slash.
5612 In particular, this caused confusion with an attempt to use "verify =
5613 reverse_host_lookup/defer_ok". An error is now given when options are
5614 supplied for verify items that do not have them. (Maybe reverse_host_
5615 lookup should have a defer_ok option, but that's a different point.)
5617 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5618 defined by RFC 821) to 2048, because there were problems with some AUTH
5619 commands, and RFC 1869 says the size should be increased for extended
5620 SMTP commands that take arguments.
5622 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5625 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5626 "unknown" error; now it says that the functionality isn't in the binary.
5628 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5629 an address' error message when a string expansion fails (syntax or
5630 whatever). Otherwise the password may appear in the log. Following change
5631 PH/42 below, there is no longer a chance of it appearing in a bounce
5634 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5636 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5637 of its addresses was ever tried. (Bugzilla bug #2.)
5639 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5640 the result incorrectly in the debug output. (It correctly added a newline
5641 to what was transported.)
5643 TF/01 Added $received_time.
5645 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5646 commented out examples of how to interface to a virus scanner and to
5647 SpamAssassin. Also added commented examples of av_scanner and
5648 spamd_address settings.
5650 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5651 and controls are allowed in which ACLs. There were a couple of minor
5652 errors. Some of the entries in the conditions table (which is a table of
5653 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5654 negation of where the condition IS allowed.
5656 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5658 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5659 header file does not have a version number, so I've had to invent a new
5660 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5661 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5662 radiusclient), but it was contributed by a Radius user.
5664 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5665 files or directories, for OpenSSL.
5667 PH/19 When an Exim process that is running as root has to create an Exim log
5668 file, it does so in a subprocess that runs as exim:exim so as to get the
5669 ownership right at creation (otherwise, other Exim processes might see
5670 the file with the wrong ownership). There was no test for failure of this
5671 fork() call, which would lead to the process getting stuck as it waited
5672 for a non-existent subprocess. Forks do occasionally fail when resources
5673 run out. I reviewed all the other calls to fork(); they all seem to check
5676 PH/20 When checking for unexpected SMTP input at connect time (before writing
5677 the banner), Exim was not dealing correctly with a non-positive return
5678 from the read() function. If the client had disconnected by this time,
5679 the result was a log entry for a synchronization error with an empty
5680 string after "input=" when read() returned zero. If read() returned -1
5681 (an event I could not check), uninitialized data bytes were printed.
5682 There were reports of junk text (parts of files, etc) appearing after
5685 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5687 PH/22 Added support for macro redefinition, and (re)definition in between
5688 driver and ACL definitions.
5690 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5691 forgetting to use the resulting value; it was using the unexpanded value.
5693 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5694 hadn't been configured. The fix is from Juergen Kreileder, who
5695 understands it better than I do:
5697 "Here's what I see happening with three configured cyrus_sasl
5698 authenticators configured (plain, login, cram-md5):
5700 On startup auth_cyrus_sasl_init() gets called for each of these.
5701 This means three calls to sasl_listmech() without a specified mech_list.
5702 => SASL tests which mechs of all available mechs actually work
5703 => three warnings about OTP not working
5704 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5706 With the patch, sasl_listmech() also gets called three times. But now
5707 SASL's mech_list option is set to the server_mech specified in the the
5708 authenticator. Or in other words, the answer from sasl_listmech()
5709 gets limited to just the mech you're testing for (which is different
5711 => the return list contains just 'plain' or 'login', 'cram-md5' or
5712 nothing depending on the value of ob->server_mech.
5714 I've just tested the patch: Authentication still works fine,
5715 unavailable mechs specified in the exim configuration are still
5716 caught, and the auth.log warnings about OTP are gone."
5718 PH/25 When debugging is enabled, the contents of the command line are added
5719 to the debugging output, even when log_selector=+arguments is not
5722 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5723 answer is "GNU", and only if the return is "GNU/something" is the answer
5726 PH/27 $acl_verify_message is now set immediately after the failure of a
5727 verification in an ACL, and so is available in subsequent modifiers. In
5728 particular, the message can be preserved by coding like this:
5730 warn !verify = sender
5731 set acl_m0 = $acl_verify_message
5733 Previously, $acl_verify_message was set only while expanding "message"
5734 and "log_message" when a very denied access.
5736 PH/28 Modified OS/os.c-Linux with
5738 -#ifndef OS_LOAD_AVERAGE
5739 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5741 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5742 nomenclature these days.)
5744 PH/29 Installed patch from the Sieve maintainer that adds the options
5745 sieve_useraddress and sieve_subaddress to the redirect router.
5747 PH/30 In these circumstances:
5748 . Two addresses routed to the same list of hosts;
5749 . First host does not offer TLS;
5750 . First host accepts first address;
5751 . First host gives temporary error to second address;
5752 . Second host offers TLS and a TLS session is established;
5753 . Second host accepts second address.
5754 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5755 and peerdn, if requested) that were in fact used only for the second
5758 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5759 attention to any local part prefix or suffix that was matched by the
5760 router that accepted the address. It now behaves in the same way as it
5761 does for delivery: the affixes are removed from the local part unless
5762 rcpt_include_affixes is set on the transport.
5764 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5765 timeout during the DATA phase of an incoming message.
5767 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5768 applied a patch sanctioned by the Sieve maintainer.
5770 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5771 the uid or gid is negative. A case of a negative gid caused this to be
5772 noticed. The fix allows for either to be negative.
5774 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5775 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5778 PH/36 The change PH/12 above was broken. Fixed it.
5780 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5781 the grounds that routing the same address twice would always produce the
5782 same answer. This might have been true once, but it is certainly no
5783 longer true now. Routing a child address may depend on the previous
5784 routing that produced that child. Some complicated redirection strategies
5785 went wrong when messages had multiple recipients, and made Exim's
5786 behaviour dependent on the order in which the addresses were given.
5788 I have moved the duplicate checking until after the routing is complete.
5789 Exim scans the addresses that are assigned to local and remote
5790 transports, and removes any duplicates. This means that more work will be
5791 done, as duplicates will always all be routed, but duplicates are
5792 presumably rare, so I don't expect this is of any significance.
5794 For deliveries to pipes, files, and autoreplies, the duplicate checking
5795 still happens during the routing process, since they are not going to be
5798 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5799 It corrects a timeout issue with spamd. This is Ian's comment: "The
5800 background is that sometimes spamd either never reads data from a
5801 connection it has accepted, or it never writes response data. The exiscan
5802 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5803 blindly assumes that writes won't block so it may never time out."
5805 PH/39 Allow G after quota size as well as K and M.
5807 PH/40 The value set for $authenticated_id in an authenticator may not contain
5808 binary zeroes or newlines because the value is written to log lines and
5809 to spool files. There was no check on this. Now the value is run through
5810 the string_printing() function so that such characters are converted to
5811 printable escape sequences.
5813 PH/41 $message_linecount is a new variable that contains the total number of
5814 lines in the message. Compare $body_linecount, which is the count for the
5817 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5818 bounce and delay warning messages, except in certain special cases, which
5821 (a) An SMTP error message from a remote host;
5822 (b) A message specified in a :fail: redirection;
5823 (c) A message specified in a "fail" command in a system filter;
5824 (d) A message specified in a FAIL return from the queryprogram router;
5825 (e) A message specified by the cannot_route_message router option.
5827 In these cases only, Exim does include the error details in bounce and
5828 warning messages. There are also a few cases where bland messages such
5829 as "unrouteable address" or "local delivery error" are given.
5831 PH/43 $value is now also set for the "else" part of a ${run expansion.
5833 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5834 being worked on, but at least Exim now implements the latest version to
5837 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5838 process to complete was treated as a delivery failure, a timeout while
5839 writing the message to the pipe was logged, but erroneously treated as a
5840 successful delivery. Such timeouts include transport filter timeouts. For
5841 consistency with the overall process timeout, these timeouts are now
5842 treated as errors, giving rise to delivery failures by default. However,
5843 there is now a new Boolean option for the pipe transport called
5844 timeout_defer, which, if set TRUE, converts the failures into defers for
5845 both kinds of timeout. A transport filter timeout is now identified in
5848 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5849 systems where "make" and "gmake" are different, calling "gmake" at top
5850 level broke things. I've arranged for the value of $(MAKE) to be passed
5851 from the Makefile to this script so that it can call the same version of
5855 A note about Exim versions 4.44 and 4.50
5856 ----------------------------------------
5858 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5859 changes of various kinds. As a consequence, a big documentation update was
5860 needed. This delayed the release for rather longer than seemed good, especially
5861 in the light of a couple of (minor) security issues. Therefore, the changes
5862 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5863 release. So 4.44 and 4.50 are in effect two different branches that both start
5866 I have left the 4.50 change log unchanged; it contains all the changes since
5867 4.43. The change log for 4.44 is below; many of its items are identical to
5868 those for 4.50. This seems to be the most sensible way to preserve the
5869 historical information.
5875 1. Minor wording change to the doc/README.SIEVE file.
5877 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5878 computation of the current number of files was incorrect.
5880 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5881 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5884 4. Give more explanation in the error message when the command for a transport
5885 filter fails to execute.
5887 5. There are several places where Exim runs a non-Exim command in a
5888 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5889 was being done only for the command run by the queryprogram router. It is
5890 now done for all such subprocesses. The other cases are: ${run, transport
5891 filters, and the commands run by the lmtp and pipe transports.
5893 6. Added CONFIGURE_GROUP build-time option.
5895 7. Some older OS have a limit of 256 on the maximum number of file
5896 descriptors. Exim was using setrlimit() to set 1000 as a large value
5897 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5898 systems. I've change it so that if it can't get 1000, it tries for 256.
5900 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5901 was an oversight, and furthermore, ever since the addition of extra
5902 controls (e.g. 4.43/32), the checks on when to allow different forms of
5903 "control" were broken. There should now be diagnostics for all cases when a
5904 control that does not make sense is encountered.
5906 9. Added the /retain_sender option to "control=submission".
5908 10. $recipients is now available in the predata ACL (oversight).
5910 11. Tidy the search cache before the fork to do a delivery from a message
5911 received from the command line. Otherwise the child will trigger a lookup
5912 failure and thereby defer the delivery if it tries to use (for example) a
5913 cached ldap connection that the parent has called unbind on.
5915 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5916 of $address_data from the recipient verification was clobbered by the
5917 sender verification.
5919 13. The value of address_data from a sender verification is now available in
5920 $sender_address_data in subsequent conditions in the ACL statement.
5922 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5924 15. Added a new option "connect=<time>" to callout options, to set a different
5927 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5928 was its contents. (It was OK if the option was not defined at all.)
5930 17. A "Completed" log line is now written for messages that are removed from
5931 the spool by the -Mrm option.
5933 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5934 information about exactly what failed.
5936 19. Added -dd to debug only the daemon process.
5938 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5939 handles timeouts, both on the server side and network timeouts. Renamed the
5940 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5942 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5943 It is now set to "smtps".
5945 22. $host_address is now set to the target address during the checking of
5946 ignore_target_hosts.
5948 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5949 being passed; this would have caused $sender_host_name to have been used if
5950 matching the list had actually called for a host name (not very likely,
5951 since this list is usually IP addresses). A host name is now passed as
5954 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5955 code that specifies a non-restarting handler (typically sigaction() in
5956 modern systems) in an attempt to fix a rare and obscure crash bug.
5958 25. Narrowed the window for a race in the daemon that could cause it to ignore
5959 SIGCHLD signals. This is not a major problem, because they are used only to
5960 wake it up if nothing else does.
5962 26. A malformed maildirsize file could cause Exim to calculate negative values
5963 for the mailbox size or file count. Odd effects could occur as a result.
5964 The maildirsize information is now recalculated if the size or filecount
5967 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5968 support for a long time. Removed HAVE_SYS_VFS_H.
5970 28. Installed the latest version of exipick from John Jetmore.
5972 29. In an address list, if the pattern was not a regular expression, an empty
5973 subject address (from a bounce message) matched only if the pattern was an
5974 empty string. Non-empty patterns were not even tested. This was the wrong
5975 because it is perfectly reasonable to use an empty address as part of a
5976 database query. An empty address is now tested by patterns that are
5977 lookups. However, all the other forms of pattern expect the subject to
5978 contain a local part and a domain, and therefore, for them, an empty
5979 address still always fails if the pattern is not itself empty.
5981 30. Exim went into a mad DNS loop when attempting to do a callout where the
5982 host was specified on an smtp transport, and looking it up yielded more
5983 than one IP address.
5985 31. Re-factored the code for checking spool and log partition space into a
5986 function that finds that data and another that does the check. The former
5987 is then used to implement four new variables: $spool_space, $log_space,
5988 $spool_inodes, and $log_inodes.
5990 32. The RFC2047 encoding function was originally intended for short strings
5991 such as real names; it was not keeping to the 75-character limit for
5992 encoded words that the RFC imposes. It now respects the limit, and
5993 generates multiple encoded words if necessary. To be on the safe side, I
5994 have increased the buffer size for the ${rfc2047: expansion operator from
5997 33. It is now permitted to omit both strings after an "if" condition; if the
5998 condition is true, the result is "true". As before, when the second string
5999 is omitted, a false condition yields an empty string. This makes it less
6000 cumbersome to write custom ACL and router conditions.
6002 34. Failure to deliver a bounce message always caused it to be frozen, even if
6003 there was an errors_to setting on the router. The errors_to setting is now
6006 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6007 canonical form (fully expanded) before being placed in
6008 $sender_host_address.
6010 36. The table in the code that translates DNS record types into text (T_A to
6011 "A" for instance) was missing entries for NS and CNAME. It is just possible
6012 that this could have caused confusion if both these types were looked up
6013 for the same domain, because the text type is used as part of Exim's
6014 per-process caching. But the chance of anyone hitting this buglet seems
6017 37. The dnsdb lookup has been extended in a number of ways.
6019 (1) There is a new type, "zns", which walks up the domain tree until it
6020 finds some nameserver records. It should be used with care.
6022 (2) There is a new type, "mxh", which is like "mx" except that it returns
6023 just the host names, not the priorities.
6025 (3) It is now possible to give a list of domains (or IP addresses) to be
6026 looked up. The behaviour when one of the lookups defers can be
6027 controlled by a keyword.
6029 (4) It is now possible to specify the separator character for use when
6030 multiple records are returned.
6032 38. The dnslists ACL condition has been extended: it is now possible to supply
6033 a list of IP addresses and/or domains to be looked up in a particular DNS
6036 39. Added log_selector=+queue_time_overall.
6038 40. When running the queue in the test harness, wait just a tad after forking a
6039 delivery process, to get repeatability of debugging output.
6041 41. Include certificate and key file names in error message when GnuTLS fails
6042 to set them up, because the GnuTLS error message doesn't include the name
6043 of the failing file when there is a problem reading it.
6045 42. Allow both -bf and -bF in the same test run.
6047 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6049 44. The "Exiscan patch" is now merged into the mainline Exim source.
6051 45. Sometimes the final signoff response after QUIT could fail to get
6052 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6053 < 0 before doing a fflush(). This bug looks as though it goes back to the
6054 introduction of TLS in release 3.20, but "sometimes" must have been rare
6055 because the tests only now provoked it.
6057 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6058 (this can affect the format of dates).
6060 47. exim_tidydb, when checking for the continued existence of a message for
6061 which it has found a message-specific retry record, was not finding
6062 messages that were in split spool directories. Consequently, it was
6063 deleting retry records that should have stayed in existence.
6065 48. Steve fixed some bugs in eximstats.
6067 49. The SPA authentication driver was not abandoning authentication and moving
6068 on to the next authenticator when an expansion was forced to fail,
6069 contradicting the general specification for all authenticators. Instead it
6070 was generating a temporary error. It now behaves as specified.
6072 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6073 (the order specifies the preference for clients). The order is now AES256,
6074 AES128, 3DES, ARCFOUR128.
6076 51. Small patch to Sieve code - explicitly set From: when generating an
6079 52. Exim crashed if a remote delivery caused a very long error message to be
6080 recorded - for instance if somebody sent an entire SpamAssassin report back
6081 as a large number of 550 error lines. This bug was coincidentally fixed by
6082 increasing the size of one of Exim's internal buffers (big_buffer) that
6083 happened as part of the Exiscan merge. However, to be on the safe side, I
6084 have made the code more robust (and fixed the comments that describe what
6087 53. Now that there can be additional text after "Completed" in log lines (if
6088 the queue_time_overall log selector is set), a one-byte patch to exigrep
6089 was needed to allow it to recognize "Completed" as not the last thing in
6092 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6093 patch that reportedly fixes this has been added. I am not expert enough to
6094 create a test for it. This is what the patch creator wrote:
6096 "I found a little strange behaviour of ldap code when working with
6097 Windows 2003 AD Domain, where users was placed in more than one
6098 Organization Units. When I tried to give exim partial DN, the exit code
6099 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6100 But simultaneously result of request was absolutely normal ldap result,
6101 so I produce this patch..."
6103 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6104 REFERENCE, so I have modified the code to exclude the patch when that macro
6107 55. Some experimental protocols are using DNS PTR records for new purposes. The
6108 keys for these records are domain names, not reversed IP addresses. The
6109 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6110 leaves it alone. Component reversal etc. now happens only for IP addresses.
6113 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6115 57. Double the size of the debug message buffer (to 2048) so that more of very
6116 long debug lines gets shown.
6118 58. The exicyclog utility now does better if the number of log files to keep
6119 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6121 59. Two changes related to the smtp_active_hostname option:
6123 (1) $smtp_active_hostname is now available as a variable.
6124 (2) The default for smtp_banner uses $smtp_active_hostname instead
6125 of $primary_hostname.
6127 60. The host_aton() function is supposed to be passed a string that is known
6128 to be a valid IP address. However, in the case of IPv6 addresses, it was
6129 not checking this. This is a hostage to fortune. Exim now panics and dies
6130 if the condition is not met. A case was found where this could be provoked
6131 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6132 components; fortuitously, this particular loophole had already been fixed
6133 by change 4.50/55 above.
6135 If there are any other similar loopholes, the new check in host_aton()
6136 itself should stop them being exploited. The report I received stated that
6137 data on the command line could provoke the exploit when Exim was running as
6138 exim, but did not say which command line option was involved. All I could
6139 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6140 running as the user.
6143 61. There was a buffer overflow vulnerability in the SPA authentication code
6144 (which came originally from the Samba project). I have added a test to the
6145 spa_base64_to_bits() function which I hope fixes it.
6148 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6149 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6151 63. The daemon start-up calls getloadavg() while still root for those OS that
6152 need the first call to be done as root, but it missed one case: when
6153 deliver_queue_load_max is set with deliver_drop_privilege. This is
6154 necessary for the benefit of the queue runner, because there is no re-exec
6155 when deliver_drop_privilege is set.
6157 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6158 This has been fixed.
6160 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6161 was in use, was not putting the data itself into the right store pool;
6162 consequently, it could be overwritten for a subsequent message in the same
6163 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6166 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6168 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6169 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6170 the function were treating the return as a boolean value, which happened to
6171 work because 0=false and not-0=true, but is not correct code.
6173 68. The host_aton() function was not handling scoped IPv6 addresses (those
6174 with, for example, "%eth0" on the end) correctly.
6176 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6177 negated items (that is, ~something) in unsigned ints. Some compilers
6178 apparently mutter when there is no cast.
6180 70. If an address verification called from an ACL failed, and did not produce a
6181 user-specific message (i.e. there was only a "system" message), nothing was
6182 put in $acl_verify_message. In this situation, it now puts the system
6185 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6186 session; change 4.31/43 added the unwanted input to the log line - except
6187 that it did not do this in the start of session case. It now does.
6189 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6190 This is wrong and can cause the other end to generate a synchronization
6191 error if it is another Exim or anything else that does the synchronization
6192 check. A QUIT command is no longer sent after a timeout.
6194 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6195 during host lookups.
6197 74. The defer_ok option of callout verification was not working if it was used
6198 when verifying addresses in header lines, that is, for this case:
6200 verify = header_sender/callout=defer_ok
6202 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6203 those file descriptors could be used for SMTP connections. If anything
6204 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6205 could be sent to the SMTP client, causing chaos. The daemon now opens
6206 stdin, stdout, and stderr to /dev/null when it puts itself into the
6209 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6210 log by default. The user can override this with suitable Perl magic.
6212 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6213 the log message when discard triggers, was not working for the DATA ACL or
6214 for the non-SMTP ACL.
6216 78. Error message wording change in sieve.c.
6218 79. If smtp_accept_max_per_host was set, the number of connections could be
6219 restricted to fewer than expected, because the daemon was trying to set up
6220 a new connection before checking whether the processes handling previous
6221 connections had finished. The check for completed processes is now done
6222 earlier. On busy systems, this bug wouldn't be noticed because something
6223 else would have woken the daemon, and it would have reaped the completed
6226 80. If a message was submitted locally by a user whose login name contained one
6227 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6228 It caused a spool format error. I have fixed the spool reading code. A
6229 related problem was that the "from" clause in the Received: line became
6230 illegal because of the space(s). It is now covered by ${quote_local_part.
6232 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6235 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6237 83. Patch from Sieve maintainer to fix unterminated string problem in
6238 "vacation" handling.
6240 84. Some minor changes to the Linux configuration files to help with other
6241 OS variants using glibc.
6243 85. One more patch for Sieve to update vacation handling to latest spec.
6246 ----------------------------------------------------
6247 See the note above about the 4.44 and 4.50 releases.
6248 ----------------------------------------------------
6254 1. Change 4.43/35 introduced a bug that caused file counts to be
6255 incorrectly computed when quota_filecount was set in an appendfile
6258 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6259 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6262 3. Give more explanation in the error message when the command for a transport
6263 filter fails to execute.
6265 4. There are several places where Exim runs a non-Exim command in a
6266 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6267 was being done only for the command run by the queryprogram router. It is
6268 now done for all such subprocesses. The other cases are: ${run, transport
6269 filters, and the commands run by the lmtp and pipe transports.
6271 5. Some older OS have a limit of 256 on the maximum number of file
6272 descriptors. Exim was using setrlimit() to set 1000 as a large value
6273 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6274 systems. I've change it so that if it can't get 1000, it tries for 256.
6276 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6277 was an oversight, and furthermore, ever since the addition of extra
6278 controls (e.g. 4.43/32), the checks on when to allow different forms of
6279 "control" were broken. There should now be diagnostics for all cases when a
6280 control that does not make sense is encountered.
6282 7. $recipients is now available in the predata ACL (oversight).
6284 8. Tidy the search cache before the fork to do a delivery from a message
6285 received from the command line. Otherwise the child will trigger a lookup
6286 failure and thereby defer the delivery if it tries to use (for example) a
6287 cached ldap connection that the parent has called unbind on.
6289 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6290 of $address_data from the recipient verification was clobbered by the
6291 sender verification.
6293 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6294 was its contents. (It was OK if the option was not defined at all.)
6296 11. A "Completed" log line is now written for messages that are removed from
6297 the spool by the -Mrm option.
6299 12. $host_address is now set to the target address during the checking of
6300 ignore_target_hosts.
6302 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6303 being passed; this would have caused $sender_host_name to have been used if
6304 matching the list had actually called for a host name (not very likely,
6305 since this list is usually IP addresses). A host name is now passed as
6308 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6309 code that specifies a non-restarting handler (typically sigaction() in
6310 modern systems) in an attempt to fix a rare and obscure crash bug.
6312 15. Narrowed the window for a race in the daemon that could cause it to ignore
6313 SIGCHLD signals. This is not a major problem, because they are used only to
6314 wake it up if nothing else does.
6316 16. A malformed maildirsize file could cause Exim to calculate negative values
6317 for the mailbox size or file count. Odd effects could occur as a result.
6318 The maildirsize information is now recalculated if the size or filecount
6321 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6322 support for a long time. Removed HAVE_SYS_VFS_H.
6324 18. Updated exipick to current release from John Jetmore.
6326 19. Allow an empty sender to be matched against a lookup in an address list.
6327 Previously the only cases considered were a regular expression, or an
6330 20. Exim went into a mad DNS lookup loop when doing a callout where the
6331 host was specified on the transport, if the DNS lookup yielded more than
6334 21. The RFC2047 encoding function was originally intended for short strings
6335 such as real names; it was not keeping to the 75-character limit for
6336 encoded words that the RFC imposes. It now respects the limit, and
6337 generates multiple encoded words if necessary. To be on the safe side, I
6338 have increased the buffer size for the ${rfc2047: expansion operator from
6341 22. Failure to deliver a bounce message always caused it to be frozen, even if
6342 there was an errors_to setting on the router. The errors_to setting is now
6345 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6346 canonical form (fully expanded) before being placed in
6347 $sender_host_address.
6349 24. Updated eximstats to version 1.33
6351 25. Include certificate and key file names in error message when GnuTLS fails
6352 to set them up, because the GnuTLS error message doesn't include the name
6353 of the failing file when there is a problem reading it.
6355 26. Expand error message when OpenSSL has problems setting up cert/key files.
6358 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6359 (this can affect the format of dates).
6361 28. exim_tidydb, when checking for the continued existence of a message for
6362 which it has found a message-specific retry record, was not finding
6363 messages that were in split spool directories. Consequently, it was
6364 deleting retry records that should have stayed in existence.
6366 29. eximstats updated to version 1.35
6367 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6368 1.35 - bugfix such that pie charts by volume are generated correctly
6370 30. The SPA authentication driver was not abandoning authentication and moving
6371 on to the next authenticator when an expansion was forced to fail,
6372 contradicting the general specification for all authenticators. Instead it
6373 was generating a temporary error. It now behaves as specified.
6375 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6376 (the order specifies the preference for clients). The order is now AES256,
6377 AES128, 3DES, ARCFOUR128.
6379 31. Small patch to Sieve code - explicitly set From: when generating an
6382 32. Exim crashed if a remote delivery caused a very long error message to be
6383 recorded - for instance if somebody sent an entire SpamAssassin report back
6384 as a large number of 550 error lines. This bug was coincidentally fixed by
6385 increasing the size of one of Exim's internal buffers (big_buffer) that
6386 happened as part of the Exiscan merge. However, to be on the safe side, I
6387 have made the code more robust (and fixed the comments that describe what
6390 33. Some experimental protocols are using DNS PTR records for new purposes. The
6391 keys for these records are domain names, not reversed IP addresses. The
6392 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6393 leaves it alone. Component reversal etc. now happens only for IP addresses.
6396 34. The host_aton() function is supposed to be passed a string that is known
6397 to be a valid IP address. However, in the case of IPv6 addresses, it was
6398 not checking this. This is a hostage to fortune. Exim now panics and dies
6399 if the condition is not met. A case was found where this could be provoked
6400 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6401 components; fortuitously, this particular loophole had already been fixed
6402 by change 4.50/55 or 4.44/33 above.
6404 If there are any other similar loopholes, the new check in host_aton()
6405 itself should stop them being exploited. The report I received stated that
6406 data on the command line could provoke the exploit when Exim was running as
6407 exim, but did not say which command line option was involved. All I could
6408 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6409 running as the user.
6412 35. There was a buffer overflow vulnerability in the SPA authentication code
6413 (which came originally from the Samba project). I have added a test to the
6414 spa_base64_to_bits() function which I hope fixes it.
6417 36. The daemon start-up calls getloadavg() while still root for those OS that
6418 need the first call to be done as root, but it missed one case: when
6419 deliver_queue_load_max is set with deliver_drop_privilege. This is
6420 necessary for the benefit of the queue runner, because there is no re-exec
6421 when deliver_drop_privilege is set.
6423 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6424 was in use, was not putting the data itself into the right store pool;
6425 consequently, it could be overwritten for a subsequent message in the same
6426 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6429 38. Sometimes the final signoff response after QUIT could fail to get
6430 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6431 < 0 before doing a fflush(). This bug looks as though it goes back to the
6432 introduction of TLS in release 3.20, but "sometimes" must have been rare
6433 because the tests only now provoked it.
6439 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6440 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6441 Now it returns an integer. A number of calls were still expecting a T/F
6442 return. Fortuitously, in all cases, the tests worked in OK situations,
6443 which is the norm. However, things would have gone wrong on any write
6444 failures on the smtp file descriptor. This function is used when sending
6445 messages over SMTP and also when doing verify callouts.
6447 2. When Exim is called to do synchronous delivery of a locally submitted
6448 message (the -odf or -odi options), it no longer closes stderr before doing
6451 3. Implemented the mua_wrapper option.
6453 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6455 5. Implemented the functions header_remove(), header_testname(),
6456 header_add_at_position(), and receive_remove_recipient(), and exported them
6459 6. If an ACL "warn" statement specified the addition of headers, Exim already
6460 inserted X-ACL-Warn: at the start if there was no header name. However, it
6461 was not making this test for the second and subsequent header lines if
6462 there were newlines in the string. This meant that an invalid header could
6463 be inserted if Exim was badly configured.
6465 7. Allow an ACL "warn" statement to add header lines at the start or after all
6466 the Received: headers, as well as at the end.
6468 8. Added the rcpt_4xx retry error code.
6470 9. Added postmaster_mailfrom=xxx to callout verification option.
6472 10. Added mailfrom=xxxx to the callout verification option, for verify=
6475 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6476 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6478 12. Inserted some casts to stop certain compilers warning when using pointer
6479 differences as field lengths or precisions in printf-type calls (mostly
6480 affecting debugging statements).
6482 13. Added optional readline() support for -be (dynamically loaded).
6484 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6485 same clock tick as a message's arrival, so that its received time was the
6486 same as the "first fail" time on the retry record, and that message
6487 remained on the queue past the ultimate address timeout, every queue runner
6488 would try a delivery (because it was past the ultimate address timeout) but
6489 after another failure, the ultimate address timeout, which should have then
6490 bounced the address, did not kick in. This was a "< instead of <=" error;
6491 in most cases the first failure would have been in the next clock tick
6492 after the received time, and all would be well.
6494 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6495 being recognized when the domain list was tested by the match_domain
6496 condition in an expansion string.
6498 16. Added the ${str2b64: operator.
6500 17. Exim was always calling setrlimit() to set a large limit for the number of
6501 processes, without checking whether the existing limit was already
6502 adequate. (It did check for the limit on file descriptors.) Furthermore,
6503 errors from getrlimit() and setrlimit() were being ignored. Now they are
6504 logged to the main and panic logs, but Exim does carry on, to try to do its
6505 job under whatever limits there are.
6507 18. Imported PCRE 5.0.
6509 19. Trivial typo in log message " temporarily refused connection" (the leading
6512 20. If the log selector return_path_on_delivery was set and an address was
6513 redirected to /dev/null, the delivery process crashed because it assumed
6514 that a return path would always be set for a "successful" delivery. In this
6515 case, the whole delivery is bypassed as an optimization, and therefore no
6518 21. Internal re-arrangement: the function for sending a challenge and reading
6519 a response while authentication was assuming a zero-terminated challenge
6520 string. It's now changed to take a pointer and a length, to allow for
6521 binary data in such strings.
6523 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6525 23. Exim was not respecting finduser_retries when seeking the login of the
6526 uid under which it was called; it was always trying 10 times. (The default
6527 setting of finduser_retries is zero.) Also, it was sleeping after the final
6528 failure, which is pointless.
6530 24. Implemented tls_on_connect_ports.
6532 25. Implemented acl_smtp_predata.
6534 26. If the domain in control=submission is set empty, Exim assumes that the
6535 authenticated id is a complete email address when it generates From: or
6536 Sender: header lines.
6538 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6539 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6540 chown and chgrp in /bin and hostname in /usr/bin.
6542 28. Exim was keeping the "process log" file open after each use, just as it
6543 does for the main log. This opens the possibility of it remaining open for
6544 long periods when the USR1 signal hits a daemon. Occasional processlog
6545 errors were reported, that could have been caused by this. Anyway, it seems
6546 much more sensible not to leave this file open at all, so that is what now
6549 29. The long-running daemon process does not normally write to the log once it
6550 has entered its main loop, and it closes the log before doing so. This is
6551 so that log files can straightforwardly be renamed and moved. However,
6552 there are a couple of unusual error situations where the daemon does write
6553 log entries, and I had neglected to close the log afterwards.
6555 30. The text of an SMTP error response that was received during a remote
6556 delivery was being truncated at 512 bytes. This is too short for some of
6557 the long messages that one sometimes sees. I've increased the limit to
6560 31. It is now possible to make retry rules that apply only when a message has a
6561 specific sender, in particular, an empty sender.
6563 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6564 it possible to be selective about when SMTP synchronization is enforced.
6566 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6568 32. Implemented hosts_connection_nolog.
6570 33. Added an ACL for QUIT.
6572 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6575 35. Added mailbox_size and mailbox_filecount to appendfile.
6577 36. Added control = no_multiline_responses to ACLs.
6579 37. There was a bug in the logic of the code that waits for the clock to tick
6580 in the case where the clock went backwards by a substantial amount such
6581 that the microsecond fraction of "now" was more than the microsecond
6582 fraction of "then" (but the whole seconds number was less).
6584 38. Added support for the libradius Radius client library this is found on
6585 FreeBSD (previously only the radiusclient library was supported).
6591 1. When certain lookups returned multiple values in the form name=value, the
6592 quoting of the values was not always being done properly. Specifically:
6593 (a) If the value started with a double quote, but contained no whitespace,
6595 (b) If the value contained whitespace other than a space character (i.e.
6596 tabs or newlines or carriage returns) it was not quoted.
6597 This fix has been applied to the mysql and pgsql lookups by writing a
6598 separate quoting function and calling it from the lookup code. The fix
6599 should probably also be applied to nisplus, ibase and oracle lookups, but
6600 since I cannot test any of those, I have not disturbed their existing code.
6602 2. A hit in the callout cache for a specific address caused a log line with no
6603 reason for rejecting RCPT. Now it says "Previous (cached) callout
6604 verification failure".
6606 3. There was an off-by-one bug in the queryprogram router. An over-long
6607 return line was truncated at 256 instead of 255 characters, thereby
6608 overflowing its buffer with the terminating zero. As well as fixing this, I
6609 have increased the buffer size to 1024 (and made a note to document this).
6611 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6612 when Exim is waiting for an SMTP response from a remote server, Exim
6613 restarts its select() call on the socket, thereby resetting its timeout.
6614 This is not a problem when such interrupts are rare. Somebody set up a cron
6615 job to run exiwhat every 2 minutes, which is less than the normal select()
6616 timeout (5 or 10 minutes). This meant that the select() timeout never
6617 kicked in because it was always reset. I have fixed this by comparing the
6618 time when an interrupt arrives with the time at the start of the first call
6619 to select(). If more time than the timeout has elapsed, the interrupt is
6620 treated as a timeout.
6622 5. Some internal re-factoring in preparation for the addition of Sieve
6623 extensions (by MH). In particular, the "personal" test is moved to a
6624 separate function, and given an option for scanning Cc: and Bcc: (which is
6625 not set for Exim filters).
6627 6. When Exim created an email address using the login of the caller as the
6628 local part (e.g. when creating a From: or Sender: header line), it was not
6629 quoting the local part when it contained special characters such as @.
6631 7. Installed new OpenBSD configuration files.
6633 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6634 try to make them clearer.
6636 9. Callout options, other than the timeout value, were being ignored when
6637 verifying sender addresses in header lines. For example, when using
6639 verify = header_sender/callout=no_cache
6641 the cache was (incorrectly) being used.
6643 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6644 only the Cygwin environment.
6646 11. When return_path_on_delivery was set as a log selector, if different remote
6647 addresses in the same message used different return paths and parallel
6648 remote delivery occurred, the wrong values would sometimes be logged.
6649 (Whenever a remote delivery process finished, the return path value from
6650 the most recently started remote delivery process was logged.)
6652 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6653 lines when AUTH and/or TLS are in use. This is the "received protocol"
6654 field. Exim used to use "asmtp" for authenticated SMTP, without any
6655 indication (in the protocol name) for TLS use. Now it follows the RFC and
6656 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6657 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6658 names appear in log lines as well as in Received: header lines.
6660 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6661 extensions, and comparison tests, and to fix some bugs.
6663 14. Changes to the "personal" filter test:
6665 (1) The test was buggy in that it was just doing the equivalent of
6666 "contains" tests on header lines. For example, if a user's address was
6667 anne@some.where, the "personal" test would incorrectly be true for
6669 To: susanne@some.where
6671 This test is now done by extracting each address from the header in turn,
6672 and checking the entire address. Other tests that are part of "personal"
6673 are now done using regular expressions (for example, to check local parts
6674 of addresses in From: header lines).
6676 (2) The list of non-personal local parts in From: addresses has been
6677 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6678 taken from the Sieve specification recommendations.
6680 (3) If the message contains any header line starting with "List-" it is
6681 treated as non-personal.
6683 (4) The test for "circular" in the Subject: header line has been removed
6684 because it now seems ill-conceived.
6686 15. Minor typos in src/EDITME comments corrected.
6688 16. Installed latest exipick from John Jetmore.
6690 17. If headers_add on a router specified a text string that was too long for
6691 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6692 of string_sprintf() is now avoided.
6694 18. $message_body_size was not set (it was always zero) when running the DATA
6695 ACL and the local_scan() function.
6697 19. For the "mail" command in an Exim filter, no default was being set for
6698 the once_repeat time, causing a random time value to be used if "once" was
6699 specified. (If the value happened to be <= 0, no repeat happened.) The
6700 default is now 0s, meaning "never repeat". The "vacation" command was OK
6701 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6702 (I found it when inspecting the code).
6704 20. There is now an overall timeout for performing a callout verification. It
6705 defaults to 4 times the callout timeout, which applies to individual SMTP
6706 commands during the callout. The overall timeout applies when there is more
6707 than one host that can be tried. The timeout is checked before trying the
6708 next host. This prevents very long delays if there are a large number of
6709 hosts and all are timing out (e.g. when the network connections are timing
6710 out). The value of the overall timeout can be changed by specifying an
6711 additional sub-option for "callout", called "maxwait". For example:
6713 verify = sender/callout=5s,maxwait=20s
6715 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6716 to the end before writing, but this should make it even safer).
6718 22. Exim was forgetting that it had advertised PIPELINING for the second and
6719 subsequent messages on an SMTP connection. It was also not resetting its
6720 memory on STARTTLS and an internal HELO.
6722 23. When Exim logs an SMTP synchronization error within a session, it now
6723 records whether PIPELINING has been advertised or not.
6725 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6726 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6727 rather than long int.
6729 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6731 26. Added the never_mail option to autoreply.
6737 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6738 crash if the getsockname() call failed; this can happen if a connection is
6739 closed very soon after it is established. The problem was simply in the
6740 order in which certain operations were done, causing Exim to try to write
6741 to the SMTP stream before it had set up the file descriptor. The bug has
6742 been fixed by making things happen in the correct order.
6748 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6749 before the connection was closed, thus losing the rejection response.
6751 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6752 some early Solaris releases, but causes trouble in current releases where
6753 socklen_t is defined.
6755 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6758 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6761 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6762 that releasing the top of it at the end releases what was used for sub-
6763 expansions (unless the block got too big). However, discard this block if
6764 the first thing is a variable or header, so that we can use its block when
6765 it is dynamic (useful for very large $message_headers, for example).
6767 6. Lookups now cache *every* query, not just the most recent. A new, separate
6768 store pool is used for this. It can be recovered when all lookup caches are
6769 flushed. Lookups now release memory at the end of their result strings.
6770 This has involved some general refactoring of the lookup sources.
6772 7. Some code has been added to the store_xxx() functions to reduce the amount
6773 of flapping under certain conditions.
6775 8. log_incoming_interface used to affect only the <= reception log lines. Now
6776 it causes the local interface and port to be added to several more SMTP log
6777 lines, for example "SMTP connection from", and rejection lines.
6779 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6781 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6783 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6784 could be overwritten at the end of the current message (or the start of a
6785 new message if it was set in a HELO ACL). The value is now preserved for
6786 the duration of the SMTP connection.
6788 12. If a transport had a headers_rewrite setting, and a matching header line
6789 contained an unqualified address, that address was qualified, even if it
6790 did not match any rewriting rules. The underlying bug was that the values
6791 of the flags that permit the existence of unqualified sender and recipient
6792 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6793 non-local messages, and by -bnq for local messages) were not being
6794 preserved with the message after it was received.
6796 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6797 "next input=" as part of the text comprising the host identity instead of
6798 the correct text. The code was using the same buffer for two different
6799 strings. However, depending on which order the printing function evaluated
6800 its arguments, the bug did not always show up. Under Linux, for example, my
6801 test suite worked just fine.
6803 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6804 doesn't work with some older versions of Perl. It has been changed to "my",
6805 which in any case is probably the better facility to use.
6807 15. A really picky compiler found some instances of statements for creating
6808 error messages that either had too many or two few arguments for the format
6811 16. The size of the buffer for calls to the DNS resolver has been increased
6812 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6813 for addresses that have a lot of PTR records. This alleviates a problem; it
6814 does not fully solve it.
6816 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6817 buffer now truncates the list and logs the incident, which is the same
6818 action as happens when Exim is looking up a host name and its aliases.
6819 Previously in this situation something unpredictable would happen;
6820 sometimes it was "internal error: store_reset failed".
6822 18. If a server dropped the connection unexpectedly when an Exim client was
6823 using GnuTLS and trying to read a response, the client delivery process
6824 crashed while trying to generate an error log message.
6826 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6827 string, for example:
6829 warn message = H1: something\nH2: something
6831 the text was added as a single header line from Exim's point of view
6832 though it ended up OK in the delivered message. However, searching for the
6833 second and subsequent header lines using $h_h2: did not work. This has been
6834 fixed. Similarly, if a system filter added multiple headers in this way,
6835 the routers could not see them.
6837 20. Expanded the error message when iplsearch is called with an invalid key to
6838 suggest using net-iplsearch in a host list.
6840 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6841 ACLs are no longer actually imposed (and a message to that effect is
6844 22. If a "gecos" field in a passwd entry contained escaped characters, in
6845 particular, if it contained a \" sequence, Exim got it wrong when building
6846 a From: or a Sender: header from that name. A second bug also caused
6847 incorrect handling when an unquoted " was present following a character
6848 that needed quoting.
6850 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6851 was not being matched caselessly.
6853 24. Arranged for all hyphens in the exim.8 source to be escaped with
6856 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6857 properly. Recipient callout cache records were still being keyed to include
6858 the sender, even when use_sender was set false. This led to far more
6859 callouts that were necessary. The sender is no longer included in the key
6860 when use_sender is false.
6862 26. Added "control = submission" modifier to ACLs.
6864 27. Added the ${base62d: operator to decode base 62 numbers.
6866 28. dnsdb lookups can now access SRV records.
6868 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6869 the configuration file.
6871 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6872 (-v) mode. This makes the output for a verbose queue run more intelligible.
6874 31. Added a use_postmaster feature to recipient callouts.
6876 32. Added the $body_zerocount variable, containing the number of binary zero
6877 bytes in the message body.
6879 33. The time of last modification of the "new" subdirectory is now used as the
6880 "mailbox time last read" when there is a quota error for a maildir
6883 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6885 35. Added +ignore_unknown as a special item in host lists.
6887 36. Code for decoding IPv6 addresses in host lists is now included, even if
6888 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6889 address was recognized as an IP address, but was then not correctly decoded
6890 into binary, causing unexpected and incorrect effects when compared with
6897 1. Very minor rewording of debugging text in manualroute to say "list of
6898 hosts" instead of "hostlist".
6900 2. If verify=header_syntax was set, and a header line with an unqualified
6901 address (no domain) and a large number of spaces between the end of the
6902 name and the colon was received, the reception process suffered a buffer
6903 overflow, and (when I tested it) crashed. This was caused by some obsolete
6904 code that should have been removed. The fix is to remove it!
6906 3. When running in the test harness, delay a bit after writing a bounce
6907 message to get a bit more predictability in the log output.
6909 4. Added a call to search_tidyup() just before forking a reception process. In
6910 theory, someone could use a lookup in the expansion of smtp_accept_max_
6911 per_host which, without the tidyup, could leave open a database connection.
6913 5. Added the variables $recipient_data and $sender_data which get set from a
6914 lookup success in an ACL "recipients" or "senders" condition, or a router
6915 "senders" option, similar to $domain_data and $local_part_data.
6917 6. Moved the writing of debug_print from before to after the "senders" test
6920 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6921 problems for message scanning, either using a data ACL, or using
6922 local_scan() because the Received: header was not generated till after they
6923 were called (in order to set the time as the time of reception completion).
6924 I have revised the way this works. The header is now generated after the
6925 body is received, but before the ACL or local_scan() are called. After they
6926 are run, the timestamp in the header is updated.
6932 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6933 before starting a queue runner without re-exec. This happened only when
6934 deliver_drop_privilege was set or when the Exim user was set to root. The
6935 effect of the bug was that timeouts during subsequent deliveries caused
6936 crashes instead of being properly handled. The handler is now left at its
6937 default (and expected) setting.
6939 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6940 message, again when deliver_drop_privilege is set or Exim is run as root.
6941 The bug described in (1) was not present in this case, but the tidying up
6942 of the other signals was missing. I have made the two cases consistent.
6944 3. The ignore_target_hosts setting on a manualroute router was being ignored
6945 for hosts that were looked up using the /MX notation.
6947 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6950 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6951 operated on the sender address. After changing the $sender_address to <>
6952 for the sender address verify, Exim was re-instated it as the original
6953 (before rewriting) address, but remembering that it had rewritten it, so it
6954 wasn't rewriting it again. This bug also had the effect of breaking the
6955 sender address verification caching when the sender address was rewritten.
6957 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6958 This has been changed so that if the ip literal address matches
6959 ignore_target_hosts, the router declines.
6961 7. Added expansion conditions match_domain, match_address, and match_local_
6962 part (NOT match_host).
6964 8. The placeholder for the Received: header didn't have a length field set.
6966 9. Added code to Exim itself and to exim_lock to test for a specific race
6967 condition that could lead to file corruption when using MBX delivery. The
6968 issue is with the lockfile that is created in /tmp. If this file is removed
6969 after a process has opened it but before that process has acquired a lock,
6970 there is the potential for a second process to recreate the file and also
6971 acquire a lock. This could lead to two Exim processes writing to the file
6972 at the same time. The added code performs the same test as UW imapd; it
6973 checks after acquiring the lock that its file descriptor still refers to
6974 the same named file.
6976 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6977 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6980 11. Added the smtp_active_hostname option. If used, this will typically be made
6981 to depend on the incoming interface address. Because $interface_address is
6982 not set up until the daemon has forked a reception process, error responses
6983 that can happen earlier (such as "too many connections") no longer contain
6986 12. If an expansion in a condition on a "warn" statement fails because a lookup
6987 defers, the "warn" statement is abandoned, and the next ACL statement is
6988 processed. Previously this caused the whole ACL to be aborted.
6990 13. Added the iplsearch lookup type.
6992 14. Added ident_timeout as a log selector.
6994 15. Added tls_certificate_verified as a log selector.
6996 16. Added a global option tls_require_ciphers (compare the smtp transport
6997 option of the same name). This controls incoming TLS connections.
6999 17. I finally figured out how to make tls_require_ciphers do a similar thing
7000 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7001 before starting the TLS session.
7003 18. Tabs are now shown as \t in -bP output.
7005 19. If the log selector return_path_on_delivery was set, Exim crashed when
7006 bouncing a message because it had too many Received: header lines.
7008 20. If two routers both had headers_remove settings, and the first one included
7009 a superfluous trailing colon, the final name in the first list and the
7010 first name in the second list were incorrectly joined into one item (with a
7011 colon in the middle).
7017 1. Added -C and -D options to the exinext utility, mainly to make it easier
7018 to include in the automated testing, but these could be helpful when
7019 multiple configurations are in use.
7021 2. The exinext utility was not formatting the output nicely when there was
7022 an alternate port involved in the retry record key, nor when there was a
7023 message id as well (for retries that were specific to a specific message
7024 and a specific host). It was also confused by IPv6 addresses, because of
7025 the additional colons they contain. I have fixed the IPv4 problem, and
7026 patched it up to do a reasonable job for IPv6.
7028 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7029 delivery, the log line now contains "pipelined" if PIPELINING was used.
7031 4. An SMTP transport process used to panic and die if the bind() call to set
7032 an explicit outgoing interface failed. This has been changed; it is now
7033 treated in the same way as a connect() failure.
7035 5. A reference to $sender_host_name in the part of a conditional expansion
7036 that was being skipped was still causing a DNS lookup. This no longer
7039 6. The def: expansion condition was not recognizing references to header lines
7040 that used bh_ and bheader_.
7042 7. Added the _cache feature to named lists.
7044 8. The code for checking quota_filecount in the appendfile transport was
7045 allowing one more file than it should have been.
7047 9. For compatibility with Sendmail, the command line option
7055 and sets the incoming protocol and host name (for trusted callers). The
7056 host name and its colon can be omitted when only the protocol is to be set.
7057 Note the Exim already has two private options, -pd and -ps, that refer to
7058 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7059 "s", but I don't think that's a major issue.
7061 10. A number of refactoring changes to the code, none of which should affect
7064 (a) The number of logging options was getting close to filling up the
7065 32-bit word that was used as a bit map. I have split them into two classes:
7066 those that are passed in the argument to log_write(), and those that are
7067 only ever tested independently outside of that function. These are now in
7068 separate 32-bit words, so there is plenty of room for expansion again.
7069 There is no change in the user interface or the logging behaviour.
7071 (b) When building, for example, log lines, the code previously used a
7072 macro that called string_cat() twice, in order to add two strings. This is
7073 not really sufficiently general. Furthermore, there was one instance where
7074 it was actually wrong because one of the argument was used twice, and in
7075 one call a function was used. (As it happened, calling the function twice
7076 did not affect the overall behaviour.) The macro has been replaced by a
7077 function that can join an arbitrary number of extra strings onto a growing
7080 (c) The code for expansion conditions now uses a table and a binary chop
7081 instead of a serial search (which was left over from when there were very
7082 few conditions). Also, it now recognizes conditions like "pam" even when
7083 the relevant support is not compiled in: a suitably worded error message is
7084 given if an attempt is made to use such a condition.
7086 11. Added ${time_interval:xxxxx}.
7088 12. A bug was causing one of the ddress fields not to be passed back correctly
7089 from remote delivery subprocesses. The field in question was not being
7090 subsequently used, so this caused to problems in practice.
7092 13. Added new log selectors queue_time and deliver_time.
7094 14. Might have fixed a bug in maildirsizefile handling that threw up
7095 "unexpected character" debug warnings, and recalculated the data
7096 unnecessarily. In any case, I expanded the warning message to give more
7099 15. Added the message "Restricted characters in address" to the statements in
7100 the default ACL that block characters like @ and % in local parts.
7102 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7103 Three changes have been made:
7105 (a) There was a serious bug; a negative response to MAIL caused the whole
7106 recipient domain to be cached as invalid, thereby blocking all messages
7107 to all local parts at the same domain, from all senders. This bug has
7108 been fixed. The domain is no longer cached after a negative response to
7109 MAIL if the sender used is not empty.
7111 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7114 (c) A new callout option, "use_sender" has been added for people who want
7115 the modified behaviour.
7121 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7124 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7125 indeed breaks things for older releases.
7127 3. Added additional logging to the case where there is a problem reading data
7128 from a filter that is running in a subprocess using a pipe, in order to
7129 try to track down a specific problem.
7131 4. Testing facility fudge: when running in the test harness and attempting
7132 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7133 getting "No route to host". Convert this to a timeout.
7135 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7138 6. Some OS don't have socklen_t but use size_t instead. This affects the
7139 fifth argument of getsockopt() amongst other things. This is now
7140 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7141 can be set for individual OS. I have set it for SunOS5, OSF1, and
7142 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7143 some earlier ones do not.
7145 7. Change 4.30/15 was not doing the test caselessly.
7147 8. The standard form for an IPv6 address literal was being rejected by address
7148 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7149 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7150 this, as well as the form without the "IPv6" on the front (but only when
7151 address literals are enabled, of course).
7153 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7155 10. Exim crashed if a message with an empty sender address specified by -f
7156 encountered a router with an errors_to setting. This could be provoked only
7157 by a command such as
7161 where an empty string was supplied; "<>" did not hit this bug.
7163 11. Installed PCRE release 4.5.
7165 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7166 remained set. It is now erased.
7168 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7169 times from message ids (which are base 36 rather than the normal 62).
7171 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7172 were being counted as actual protocol errors, and logged if the log
7173 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7174 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7175 and DATA following a set of rejected RCPTs do not count as protocol errors.
7176 In other words, Exim assumes they were pipelined, though this may not
7177 actually be the case. Of course, in all cases the client gets an
7178 appropriate error code.
7180 15. If a lookup fails in an ACL condition, a message about the failure may
7181 be available; it is used if testing the ACL cannot continue, because most
7182 such messages specify what the cause of the deferral is. However, some
7183 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7184 that caused an old message to be retained and used if a later statement
7185 caused a defer, replacing the real cause of the deferral.
7187 16. If an IP address had so many PTR records that the DNS lookup buffer
7188 was not large enough to hold them, Exim could crash while trying to process
7189 the truncated data. It now detects and logs this case.
7191 17. Further to 4.21/58, another change has been made: if (and only if) the
7192 first line of a message (the first header line) ends with CRLF, a bare LF
7193 in a subsequent header line has a space inserted after it, so as not to
7194 terminate the header.
7196 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7197 unnecessarily, used atoi() instead of strtol(), and didn't check the
7198 termination when getting file sizes from file names by regex.
7200 19. Completely re-implemented the support for maildirsize files, in the light
7201 of a number of problems with the previous contributed implementation
7202 (4.30/29). In particular:
7204 . If the quota is zero, the maildirsize file is maintained, but no quota is
7207 . If the maildir directory does not exist, it is created before any attempt
7208 to write a maildirsize file.
7210 . The quota value in the file is just a cache; if the quota is changed in
7211 the transport, the new value overrides.
7213 . A regular expression is available for excluding directories from the
7216 20. The autoreply transport checks the characters in options that define the
7217 message's headers; it allows continued headers, but it was checking with
7218 isspace() after an embedded newline instead of explicitly looking for a
7221 21. If all the "regular" hosts to which an address was routed had passed their
7222 expiry times, and had not reached their retry times, the address was
7223 bounced, even if fallback hosts were defined. Now Exim should go on to try
7226 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7227 equivalent code in the SMTP transport. Some hosts send humungous responses
7228 to HELO/EHLO, more than 1024 it seems.
7230 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7231 gives compiler warnings in some environments. I've now done it "properly",
7234 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7235 (because of problems with the built-in one) was declared to return uschar *
7236 instead of char *, causing compiler failure.
7238 25. Fixed a file descriptor leak when processing alias/forward files.
7240 26. Fixed a minor format string issue in dbfn.c.
7242 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7244 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7245 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7248 29. When a sender address is verified, it is cached, to save repeating the test
7249 when there is more than one recipient in a message. However, when the
7250 verification involves a callout, it is possible for different callout
7251 options to be set for different recipients. It is too complicated to keep
7252 track of this in the cache, so now Exim always runs a verification when a
7253 callout is required, relying on the callout cache for the optimization.
7254 The overhead is duplication of the address routing, but this should not be
7257 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7258 to be verified with callout=postmaster, and the main callout worked but the
7259 postmaster check failed, the verification correctly failed. However, if a
7260 subsequent RCPT command asked for sender verification *without* the
7261 postmaster check, incorrect caching caused this verification also to fail,
7264 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7265 it was not caching the DNS options (qualify_single, search_parents) that
7266 were used when the lookup failed. A subsequent lookup with different
7267 options therefore always gave the same answer, though there were cases
7268 where it should not have. (Example: a "domains = !$mx_any" option on a
7269 dnslookup router: the "domains" option is always processed without any
7270 widening, but the router might have qualify_single set.) Now Exim uses the
7271 cached value only when the same options are set.
7273 32. Added John Jetmore's "exipick" utility to the distribution.
7275 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7276 than a timeout (e.g. a certificate is required, and is not provided), an
7277 Exim server now closes the connection immediately. Previously it waited for
7278 the client to close - but if the client is SSL, it seems that they each
7279 wait for each other, leading to a delay before one of them times out.
7281 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7282 maintained 0.8.x compatibility because I don't think many are using it, and
7283 it is clearly obsolete.
7285 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7288 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7289 client certificate was expired. A simple patch fixes this, though I don't
7290 understand the full logic of why the verify callback is called multiple
7293 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7294 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7295 which causes problems with some clients (such as the Certicom SSL Plus
7296 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7297 disables the coutermeasure allowing Eudora to connect."
7299 38. Exim was not checking that a write() to a log file succeeded. This could
7300 lead to Bad Things if a log got too big, in particular if it hit a file
7301 size limit. Exim now panics and dies if it cannot write to a log file, just
7302 as it does if it cannot open a log file.
7304 39. Modified OS/Makefile-Linux so that it now contains
7306 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7308 The two -D definitions ensure that Exim is compiled with large file
7309 support, which makes it possible to handle log files that are bigger than
7312 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7313 instance) a domain was checked against a named list that involved a lookup,
7314 causing $domain_data to be set, then another domain was checked against the
7315 same list, then the first domain was re-checked, the value of $domain_data
7316 after the final check could be wrong. In particular, if the second check
7317 failed, it could be set empty. This bug probably also applied to
7320 41. The strip_trailing_dot option was not being applied to the address given
7321 with the -f command-line option.
7323 42. The code for reading a message's header from the spool was incrementing
7324 $received_count, but never initializing it. This meant that the value was
7325 incorrect (doubled) while delivering a message in the same process in which
7326 it was received. In the most common configuration of Exim, this never
7327 happens - a fresh exec is done - but it can happen when
7328 deliver_drop_privilege is set.
7330 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7331 it now includes up to 150 characters of the unexpected data in the log
7334 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7335 and building data strings. The size of both of these buffers was 10 000
7336 bytes - far larger than anybody would *ever* want, thought I. Needless to
7337 say, somebody hit the limit. I have increased the maximum line length to
7338 20 000 and the maximum data length of concatenated lines to 100 000. I have
7339 also fixed two bugs, because there was no checking on these buffers. Tsk,
7340 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7341 buffer is too small.
7343 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7344 lsearch lookups. Now it does.
7346 46. When parsing a route_list item in a manualroute router, a fixed-length
7347 buffer was used for the list of hosts. I made this 1024 bytes long,
7348 thinking that nobody would ever have a list of hosts that long. Wrong.
7349 Somebody had a whole pile of complicated expansion conditions, and the
7350 string was silently truncated, leading to an expansion error. It turns out
7351 that it is easier to change to an unlimited length (owing to other changes
7352 that have happened since this code was originally written) than to build
7353 structure for giving a limitation error. The length of the item that
7354 expands into the list of hosts is now unlimited.
7356 47. The lsearch lookup could not handle data where the length of text line was
7357 more than 4095 characters. Such lines were truncated, leading to shortened
7358 data being returned. It should now handle lines of any length.
7360 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7361 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7364 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7365 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7366 no longer happens in comment lines. A list of replacements is now placed
7367 at the head of all of the source files, except those whose only change is
7368 to replace PERL_COMMAND in the very first #! line.
7370 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7371 messages on the queue, with a bottom-up merge sort, using code contributed
7372 by Michael Haardt. This should make operations like -bp somewhat faster on
7373 large queues. It won't affect queue runners, except when queue_run_in_order
7376 51. Installed eximstats 1.31 in the distribution.
7378 52. Added support for SRV lookups to the dnslookup router.
7380 53. If an ACL referred to $message_body or $message_body_end, the value was not
7381 reset for any messages that followed in the same SMTP session.
7383 54. The store-handling optimization for building very long strings was not
7384 differentiating between the different store pools. I don't think this
7385 actually made any difference in practice, but I've tidied it.
7387 55. While running the routers to verify a sender address, $sender_address
7388 was still set to the sender address. This is wrong, because when routing to
7389 send a bounce to the sender, it would be empty. Therefore, I have changed
7390 it so that, while verifying a sender address, $sender_address is set to <>.
7391 (There is no change to what happens when verifying a recipient address.)
7393 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7394 target A or AAAA records (if not already returned) without resetting the
7395 qualify_single or search_parents options of the DNS resolver. These are
7396 inappropriate in this case because the targets of MX and SRV records must
7397 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7398 target that, when qualified, matched something in the local domain. These
7399 two options are now turned off when doing these lookups.
7401 57. It seems that at least some releases of Reiserfs (which does not have the
7402 concept of a fixed number of inodes) returns zero and not -1 for the
7403 number of available inodes. This interacted badly with check_spool_inodes,
7404 which assumed that -1 was the "no such thing" setting. What I have done is
7405 to check that the total number of inodes is greater than zero before doing
7406 the test of how many are available.
7408 58. When a "warn" ACL statement has a log_message modifier, the message is
7409 remembered, and not repeated. This is to avoid a lot of repetition when a
7410 message has many recipients that cause the same warning to be written.
7411 However, Exim was preserving the list of already written lines for an
7412 entire SMTP session, which doesn't seem right. The memory is now reset if a
7413 new message is started.
7415 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7416 debugging output unless log_rewrite was also set.
7418 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7419 of (int)(handle) when we know that handle contains (void *)(-1).
7421 61. The Exim daemon panic-logs an error return when it closes the incoming
7422 connection. However "connection reset by peer" seems to be common, and
7423 isn't really an error worthy of noting specially, so that particular error
7426 62. When Exim is trying to find all the local interfaces, it used to panic and
7427 die if the ioctl to get the interface flags failed. However, it seems that
7428 on at least one OS (Solaris 9) it is possible to have an interface that is
7429 included in the list of interfaces, but for which you get a failure error
7430 for this call. This happens when the interface is not "plumbed" into a
7431 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7432 failure of the "get flags" call assumes that the interface is down.
7434 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7435 makes life easier for people who are doing arithmetic on fields extracted
7436 from dates, where you often get leading zeros that should not be
7437 interpreted as octal.
7439 64. Added qualify_domain to the redirect router, to override the global
7442 65. If a pathologically long header line contained very many addresses (the
7443 report of this problem mentioned 10 000) and each of them was rewritten,
7444 Exim could use up a very large amount of memory. (It kept on making new
7445 copies of the header line as it rewrote, and never released the old ones.)
7446 At the expense of a bit more processing, the header rewriting function has
7447 been changed so that it no longer eats memory in this way.
7449 66. The generation of the Received: header has been moved from the time that a
7450 message starts to be received, to the time that it finishes. The timestamp
7451 in the Received: header should now be very close to that of the <= log
7452 line. There are two side-effects of this change:
7454 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7455 logged header lines no longer include the local Received: line, because
7456 it has not yet been created. The same applies to a copy of the message
7457 that is returned to a non-SMTP sender when a message is rejected.
7459 (b) When a filter file is tested using -bf, no additional Received: header
7460 is added to the test message. After some thought, I decided that this
7463 This change does not affect the value of $received_for. It is still set
7464 after address rewriting, but before local_scan() is called.
7466 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7468 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7469 gave an unhelpful panic error message, and a defer error. I have managed to
7470 change this behaviour so that it now rejects any supplied certificate,
7471 which seems right, as the list of acceptable certificates is empty.
7473 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7474 gave an unhelpful defer error. I have not managed to make this reject any
7475 supplied certificates, but the error message it gives is "no certificate
7476 supplied", which is not helpful.
7478 70. exigrep's output now also includes lines that are not associated with any
7479 message, but which match the given pattern. Implemented by a patch from
7480 Martin Sluka, which also tidied up the Perl a bit.
7482 71. Recipient callout verification, like sender verification, was using <> in
7483 the MAIL FROM command. This isn't really the right thing, since the actual
7484 sender may affect whether the remote host accepts the recipient or not. I
7485 have changed it to use the actual sender in the callout; this means that
7486 the cache record is now keyed on a recipient/sender pair, not just the
7487 recipient address. There doesn't seem to be a real danger of callout loops,
7488 since a callout by the remote host to check the sender would use <>.
7489 [SEE ABOVE: changed after hitting problems.]
7491 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7492 temporary errors. However, in the case of such a code being given after
7493 the end of a data transmission (i.e. after ".") Exim was failing to write
7494 a retry record for the message. (Yes, there was some broken host that was
7495 actually sending 8xx at this point.)
7497 73. An unknown lookup type in a host list could cause Exim to panic-die when
7498 the list was checked. (An example that provoked this was putting <; in the
7499 middle of a list instead of at the start.) If this happened during a DATA
7500 ACL check, a -D file could be left lying around. This kind of configuration
7501 error no longer causes Exim to die; instead it causes a defer error. The
7502 incident is still logged to the main and panic logs.
7504 74. Buglet left over from Exim 3 conversion. The message "too many messages
7505 in one connection" was written to the rejectlog but not the mainlog, except
7506 when address rewriting (yes!) was being logged.
7508 75. Added write_rejectlog option.
7510 76. When a system filter was run not as root (that is, when system_filter_user
7511 was set), the values of the $n variables were not being returned to the
7512 main process; thus, they were not subsequently available in the $sn
7515 77. Added +return_path_on_delivery log selector.
7517 78. A connection timeout was being treated differently from recipients deferred
7518 when testing hosts_max_try with a message that was older than the host's
7519 retry timeout. (The host should not be counted, thus allowing all hosts to
7520 be tried at least once before bouncing.) This may have been the cause of an
7521 occasionally reported bug whereby a message would remain on the queue
7522 longer than the retry timeout, but would be bounced if a delivery was
7523 forced. I say "may" because I never totally pinned down the problem;
7524 setting up timeout/retry tests is difficult. See also the next item.
7526 79. The ultimate address timeout was not being applied to errors that involved
7527 a combination of host plus message (for example, a timeout on a MAIL
7528 command). When an address resolved to a number of possible hosts, and they
7529 were not all tried for each delivery (e.g. because of hosts_max_try), a
7530 message could remain on the queue longer than the retry timeout.
7532 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7535 81. Fixed an obscure SMTP outgoing bug which required at least the following
7536 conditions: (a) there was another message waiting for the same server;
7537 (b) the server returned 5xx to all RCPT commands in the first message so
7538 that the message was not completed; (c) the server dropped the connection
7539 or gave a negative response to the RSET that Exim sends to abort the
7540 transaction. The observed case was a dropped connection after DATA that had
7541 been sent in pipelining mode. That is, the server had advertised PIPELINING
7542 but was not implementing it correctly. The effect of the bug was incorrect
7543 behaviour, such as trying another host, and this could lead to a crash.
7549 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7550 and daemon.c were passed as pointers to ints; they should have been
7551 pointers to socklen_t variables (which are typically unsigned ints).
7553 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7556 3. Fixed a really odd bug that affected only the testing scheme; patching a
7557 certain fixed string in the binary changed the value of another string that
7558 happened to be identical to the end of the original first string.
7560 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7561 name", it returns that address as the IP address. On some operating
7562 systems (e.g. Solaris), it also passes back the IP address string as the
7563 "host name". However, on others (e.g. Linux), it passes back an empty
7564 string. Exim wasn't checking for this, and was changing the host name to an
7565 empty string, assuming it had been canonicalized.
7567 5. Although rare, it is permitted to have more than one PTR record for a given
7568 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7569 all the names associated with an address, because they do in Solaris.
7570 However, it seems that they do not in Linux for data that comes from the
7571 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7572 I found this out when I moved to a new Linux workstation and tried to run
7573 the Exim test suite.
7575 To get round this problem I have changed the code so that it now does its
7576 own call to the DNS to look up PTR records when searching for a host name.
7577 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7578 addresses that are only in /etc/hosts are still found.
7580 This behaviour is, however, controlled by an option called host_lookup_
7581 order, which defaults to "bydns:byaddr". If people want to use the other
7582 order, or indeed, just use one or the other means of lookup, they can
7583 specify it in this variable.
7585 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7586 some operating systems, this comes back from gethostbyaddr() as an empty
7587 string, and this is what Exim used to test for. However, it seems that in
7588 other systems, "." is yielded. Exim now tests for this case too.
7590 7. The values of check_spool_space and check_log_space are now held internally
7591 as a number of kilobytes instead of an absolute number of bytes. If a
7592 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7593 kilobyte. This means that much larger values can be stored.
7595 8. Exim monitor: an attempt to get the action menu when not actually pointing
7596 at a message produces an empty menu entitled "No message selected". This
7597 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7598 no entries in it ("Shell widget menu has zero width and/or height"). So I
7599 have added a single, blank menu entry in this case.
7601 9. Added ${quote_local_part.
7603 10. MIME decoding is now applied to the contents of Subject: header lines when
7606 11. Now that a reference to $sender_host_address automatically causes a reverse
7607 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7608 host lookup before query-style lookups in lists that might use this
7609 variable. This has therefore been abolished, and the "net-" prefix is no
7610 longer necessary for query-style lookups.
7612 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7613 have been a typo for LFLAGS, so it has been changed.
7615 13. The install script calls Exim with "-C /dev/null" in order to find the
7616 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7617 to be output. However, since Exim outputs its version number before the
7618 error, it didn't break the script. It just looked ugly. I fixed this by
7619 always allowing "-C /dev/null" if the caller is root.
7621 14. Ignore overlarge ACL variable number when reading spool file - insurance
7622 against a later release with more variables having written the file.
7624 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7625 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7626 this, as well as the form without the "IPv6" on the front.
7628 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7629 OS/Makefile-Darwin file.
7631 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7632 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7634 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7635 variable. However, some OS use ints for inodes. I've added cast to long int
7636 to get rid of the compiler warning.
7638 19. I had forgotten to lock out "/../" in configuration file names when
7639 ALT_CONFIG_PREFIX was set.
7641 20. Routers used for verification do not need to specify transports. However,
7642 if such a router generated a host list, and callout was configured, Exim
7643 crashed, because it could not find a port number from the (non-existent)
7644 transport. It now assumes port 25 in this circumstance.
7646 21. Added the -t option to exigrep.
7648 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7649 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7650 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7651 them is set, LOOKUP_LSEARCH is forced.
7653 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7655 24. Added sender and host information to the "rejected by local_scan()" log
7656 line; previously there was no indication of these.
7658 25. Added .include_if_exists.
7660 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7661 Linux. It turns out that not all file systems support this. Apparently some
7662 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7663 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7664 not supported on the file descriptor, is now ignored when Exim is trying to
7665 sync a directory. This applies only to Linux.
7667 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7669 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7670 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7673 29. Added support for maildirsize files from supplied patch (modified a bit).
7675 30. The use of :fail: followed by an empty string could lead Exim to respond to
7676 sender verification failures with (e.g.):
7678 550 Verification failed for <xxx>
7679 550 Sender verify failed
7681 where the first response line was missing the '-' that indicates it is not
7682 the final line of the response.
7684 31. The loop for finding the name of the user that called Exim had a hardwired
7685 limit of 10; it now uses the value of finduser_retries, which is used for
7686 all other user lookups.
7688 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7691 33. Exim was neglecting to zero errno before one call of strtol() when
7692 expanding a string and expecting an integer value. On some systems this
7693 resulted in spurious "integer overflow" errors. Also, it was casting the
7694 result into an int without checking.
7696 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7697 did not work. The code looks as if it has *never* worked, though it appears
7698 to have been documented since at least release 1.62. I have made it work.
7700 35. The "timeout_DNS" error in retry rules, also documented since at least
7701 1.62, also never worked. As it isn't clear exactly what this means, and
7702 clearly it isn't a major issue, I have abolished the feature by treating it
7703 as "timeout", and writing a warning to the main and panic logs.
7705 36. The display of retry rules for -brt wasn't always showing the error code
7708 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7709 timeout_connect_A, timeout_connect_MX.
7711 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7712 to the empty sender.
7714 39. The daemon was not analysing the content of -oX till after it had closed
7715 stderr and disconnected from the controlling terminal. This meant that any
7716 syntax errors were only noted on the panic log, and the return code from
7717 the command was 0. By re-arranging the code a little, I've made the
7718 decoding happen first, so such errors now appear on stderr, and the return
7719 code is 1. However, the actual setting up of the sockets still happens in
7720 the disconnected process, so errors there are still only recorded on the
7723 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7724 connections (as happens on some IP stacks) was logged at start up time as
7725 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7726 it from "IPv6 and IPv4", which means that two separate sockets are being
7729 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7730 says whether AF_INET or AF_INET6 was passed as an argument.
7732 42. Exiwhat output was messed up when time zones were included in log
7735 43. Exiwhat now gives more information about the daemon's listening ports,
7736 and whether -tls-on-connect was used.
7738 44. The "port" option of the smtp transport is now expanded.
7740 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7741 silently ignored. Now an error message is written to the main and panic
7744 46. There's a new ACL modifier called "logwrite" which writes to a log file
7745 as soon as it is encountered.
7747 47. Added $local_user_uid and $local_user_gid at routing time.
7749 48. Exim crashed when trying to verify a sender address that was being
7752 49. Exim was recognizing only a space character after ".include". It now also
7753 recognizes a tab character.
7755 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7756 extracting the relevant information from the specification. The man page no
7757 longer contains scrambled data for the -d option, and I've added a section
7758 at the front about calling Exim under different names.
7760 51. Added "extra_headers" argument to the "mail" command in filter files.
7762 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7765 53. Installed eximstats 1.29.
7767 54. Added transport_filter_timeout as a generic transport option.
7769 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7770 Cc: header lines. This was required by RFC 822, but it not required by RFC
7773 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7774 incoming messages that did not have them. Now it does so only if the
7775 message originates locally, that is, if there is no associated remote host
7776 address. When Resent- header lines are present, this applies to the Resent-
7777 lines rather than the non-Resent- lines.
7779 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7780 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7782 58. Messages for configuration errors now include the name of the main
7783 configuration file - useful now that there may be more than one file in a
7784 list (.included file names were always shown).
7786 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7787 for those rare installations that do not start the daemon as root or run it
7788 setuid root. I've cut out the call to initgroups() if the daemon is not
7791 60. The Exim user and group can now be bound into the binary as text strings
7792 that are looked up at the start of Exim's processing.
7794 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7796 62. Added $mailstore_basename variable.
7798 63. Installed patch to sieve.c from Michael Haardt.
7800 64. When Exim failed to open the panic log after failing to open the main log,
7801 the original message it was trying to log was written to stderr and debug
7802 output, but if they were not available (the usual case in production), it
7803 was lost. Now it is written to syslog before the two lines that record the
7804 failures to open the logs.
7806 65. Users' Exim filters run in subprocesses under the user's uid. It is
7807 possible for a "deliver" command or an alias in a "personal" command to
7808 provoke an address rewrite. If logging of address rewriting is configured,
7809 this fails because the process is not running as root or exim. There may be
7810 a better way of dealing with this, but for the moment (because 4.30 needs
7811 to be released), I have disabled address rewrite logging when running a
7812 filter in a non-root, non-exim process.
7818 1. The buildconfig auxiliary program wasn't quoting the value set for
7819 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7820 not defined. This bug was masked in 4.22 by the effect that was fixed in
7823 2. Some messages that were rejected after a message id was allocated were
7824 shown as "incomplete" by exigrep. It no longer does this for messages that
7825 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7827 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7828 have allow_domain_literals set, the ID did not get logged in the <= line.
7829 Domain literals are now always recognized in Message-ID: header lines.
7831 4. The first argument for a ${extract expansion item is the key name or field
7832 number. Leading and trailing spaces in this item were not being ignored,
7833 causing some misleading effects.
7835 5. When deliver_drop_privilege was set, single queue runner processes started
7836 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7837 same command in the process it spins off) were not dropping privilege.
7839 6. When the daemon running as "exim" started a queue runner, it always
7840 re-executed Exim in the spun-off process. This is a waste of effort when
7841 deliver_drop_privilege is set. The new process now just calls the
7842 queue-runner function directly.
7848 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7851 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7852 ignored. Though the use of -oP was forcing the writing of a pid file, it
7853 was always written to the default place.
7855 3. If the message "no IP address found for host xxxx" is generated during
7856 incoming verification, it is now followed by identification of the incoming
7857 connection (so you can more easily find what provoked it).
7859 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7861 5. Added some features to "harden" Exim a bit more against certain attacks:
7863 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7864 be put in Local/Makefile. This is like the never_users runtime option,
7865 but it cannot be overridden. The default setting is "root".
7867 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7868 prefix string with which any file named in a -C command line option
7871 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7872 is retained for -C and -D only if the caller of Exim is root. Without
7873 it, the exim user may also use -C and -D and retain privilege.
7875 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7876 command line option is disabled.
7878 6. Macro names set by the -D option must start with an upper case letter, just
7879 like macro names defined in the configuration file.
7881 7. Added "dereference=" facility to LDAP.
7883 8. Two instances of the typo "uknown" in the source files are fixed.
7885 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7886 the Configure-Makefile script screwed up while processing it.
7888 10. Incorporated PCRE 4.4.
7890 11. The SMTP synchronization check was not operating right at the start of an
7891 SMTP session. For example, it could not catch a HELO sent before the client
7892 waited for the greeting. There is now a check for outstanding input at the
7893 point when the greeting is written. Because of the duplex, asynchronous
7894 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7895 way, but not yet received, when the check is performed.
7897 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7898 on TCP/IP sockets, because this apparently causes some broken clients to
7901 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7902 unchanged) from the Cygwin maintainer.
7904 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7905 was supported instead of testing for mbx. Effectively a typo.
7907 15. The spa authenticator server code was not checking that the input it
7908 received was valid base64.
7910 16. The debug output line for the "set" modifier in ACLs was not showing the
7911 name of the variable that was being set.
7913 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7915 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7916 lookup on its own. Something else was needed to trigger the lookup. For
7917 example, a match in host_lookup or the need for a host name in a host list.
7918 Now, if $sender_host_name is referenced and the host name has not yet been
7919 looked up, a lookup is performed. If the lookup fails, the variable remains
7920 empty, and $host_lookup_failed is set to "1".
7922 19. Added "eqi" as a case-independent comparison operator.
7924 20. The saslauthd authentication condition could segfault if neither service
7925 nor realm was specified.
7927 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7928 error message that was logged was misleading, and incoming SMTP
7929 connections were dropped. The message is now more accurate, and temporary
7930 errors are given to SMTP connections.
7932 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7933 (except RSET) with a 421 error, until QUIT is received. However, it was
7934 failing to send a response to QUIT.
7936 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7937 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7938 ACL is now run after the helo_try_verify_hosts code.
7940 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7941 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7942 case-independent, so other case variants are also recognized). Apparently
7943 some systems use these upper case variants.
7945 25. If more than two messages were waiting for the same host, and a transport
7946 filter was specified for the transport, Exim sent two messages over the
7947 same TCP/IP connection, and then failed with "socket operation on non-
7948 socket" when it tried to send the third.
7950 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7952 27. The extern definition of crypt16() in expand.c was not being excluded when
7953 the OS had its own crypt16() function.
7955 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7956 as a preferred synonym for return_size_limit, both as an option and as an
7959 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7961 30. Changed the default configuration ACL to relax the local part checking rule
7962 for addresses that are not in any local domains. For these addresses,
7963 slashes and pipe symbols are allowed within local parts, but the sequence
7964 /../ is explicitly forbidden.
7966 31. SPA server authentication was not clearing the challenge buffer before
7969 32. log_message in a "warn" ACL statement was writing to the reject log as
7970 well as to the main log, which contradicts the documentation and doesn't
7971 seem right (because no rejection is happening). So I have stopped it.
7973 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7974 However, I am unable to do any testing of this.
7976 34. Fixed an infelicity in the appendfile transport. When checking directories
7977 for a mailbox, to see if any needed to be created, it was accidentally
7978 using path names with one or more superfluous leading slashes; tracing
7979 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7981 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7982 contents are added to the log line that is written for every discarded
7983 recipient. (Previously a log_message setting was ignored.)
7985 36. The ${quote: operator now quotes the string if it is empty.
7987 37. The install script runs exim in order to find its version number. If for
7988 some reason other than non-existence or emptiness, which it checks, it
7989 could not run './exim', it was installing it with an empty version number,
7990 i.e. as "exim-". This error state is now caught, and the installation is
7993 38. An argument was missing from the function that creates an error message
7994 when Exim fails to connect to the socket for saslauthd authentication.
7995 This could cause Exim to crash, or give a corrupted message.
7997 39. Added isip, isip4, and isip6 to ${if conditions.
7999 40. The ACL variables $acl_xx are now saved with the message, and can be
8000 accessed later in routers, transports, and filters.
8002 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8003 strings in the file are not string-expanded.
8005 42. If a MAIL command specified a SIZE value that was too large to fit into an
8006 int variable, the check against message_size_limit failed. Such values are
8007 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8008 day this will have to be increased, but I don't think I want to be around
8009 when emails are that large.
8016 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8017 iconv() is not standard in FreeBSD.
8019 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8020 IPv6 enabled. The observed symptom was a segmentation fault on return from
8021 the function os_common_find_running_interfaces() in src/os.c.
8023 3. In the check_special_case() function in daemon.c I had used "errno" as an
8024 argument name, which causes warnings on some systems. This was basically a
8025 typo, since it was named "eno" in the comments!
8027 4. The code that waits for the clock to tick (at a resolution of some fraction
8028 of a second) so as to ensure message-id uniqueness was always waiting for
8029 at least one whole tick, when it could have waited for less. [This is
8030 almost certainly not relevant at current processor speeds, where it is
8031 unlikely to ever wait at all. But we try to future-proof.]
8033 5. The function that sleeps for a time interval that includes fractions of a
8034 second contained a race. It did not block SIGALRM between setting the
8035 timer, and suspending (a couple of lines later). If the interval was short
8036 and the sigsuspend() was delayed until after it had expired, the suspension
8037 never ended. On busy systems this could lead to processes getting stuck for
8040 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8041 process, before it forks any delivery processes. The open lookup caching
8042 mechanism meant that the open file or database connection was passed into
8043 the delivery process. The problem was that delivery processes always tidy
8044 up cached lookup data. This could cause a problem for the next delivery
8045 process started by the queue runner, because the external queue runner
8046 process does not know about the closure. So the next delivery process
8047 still has data in the lookup cache. In the case of a file lookup, there was
8048 no problem because closing a file descriptor in a subprocess doesn't affect
8049 the parent. However, if the lookup was caching a connection to a database,
8050 the connection was closed, and the second delivery process was likely to
8051 see errors such as "PGSQL: query failed: server closed the connection
8052 unexpectedly". The problem has been fixed by closing all cached lookups
8053 in a queue runner before running a delivery process.
8055 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8056 seem to have the "const" qualifier which it has on other OS. I've
8059 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8060 *of the same type* (client or server) with the same public name that an
8061 error should be diagnosed.
8063 9. When Exim looked up a host name for an IP address, but failed to find the
8064 original IP address when looking up the host name (a safety check), it
8065 output the message "<ip address> does not match any IP for NULL", which was
8066 confusing, to say the least. The bug was that the host name should have
8067 appeared instead of "NULL".
8069 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8070 user that is built into the binary, and the -C or -D options is used, root
8071 privilege is dropped before the configuration file is read. In addition,
8072 logging is switched to stderr instead of the normal log files. If the
8073 configuration then re-defines the Exim user, the unprivileged environment
8074 is probably not what is expected, so Exim logs a panic warning message (but
8077 However, if deliver_drop_privilege is set, the unprivileged state may well
8078 be exactly what is intended, so the warning has been cut out in that case,
8079 and Exim is allowed to try to write to its normal log files.
8085 1. smtp_return_error_details was not giving details for temporary sender
8086 or receiver verification errors.
8088 2. Diagnose a configuration error if two authenticators have the same public
8091 3. Exim used not to create the message log file for a message until the first
8092 delivery attempt. This could be confusing when incoming messages were held
8093 for policy or load reasons. The message log file is now created at the time
8094 the message is received, and an initial "Received" line is written to it.
8096 4. The automatically generated man page for command line options had a minor
8097 bug that caused no ill effects; however, a more serious problem was that
8098 the procedure for building the man page automatically didn't always
8099 operate. Consequently, release 4.20 contains an out-of-date version. This
8100 shouldn't happen again.
8102 5. When building Exim with embedded Perl support, the script that builds the
8103 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8104 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8106 6. The freeze_tell option was not being used for messages that were frozen on
8107 arrival, either by an ACL or by local_scan().
8109 7. Added the smtp_incomplete_transaction log selector.
8111 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8112 was accepting AUTH without a new EHLO.
8114 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8115 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8118 10. Logging of TCP/IP connections (when configured) now happens in the main
8119 daemon process instead of the child process, so that the TCP/IP connection
8120 count is more accurate (but it can never be perfect).
8122 11. The use of "drop" in a nested ACL was not being handled correctly in the
8123 outer ACL. Now, if condition failure induced by the nested "drop" causes
8124 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8125 or "require"), the connection is dropped.
8127 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8128 that yield "discard" can now be used with an "accept" or a "discard" verb,
8129 but an error is generated for any others (because I can't see a useful way
8130 to define what should happen).
8132 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8133 are now processed in the same way as lines in the Exim configuration file.
8134 In particular, continuation lines are supported.
8136 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8138 15. Added -ti meaning -t -i.
8140 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8141 domains, and warn by logging if others are found.
8143 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8144 returned from GIFCONF when Exim is trying to find the list of interfaces on
8145 a host. The code in os.c has been modified to copy each ifreq to an aligned
8146 structure in all cases.
8148 Also, in some cases, the returned ifreq's were being copied to a 'struct
8149 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8150 means the last couple of bytes of an IPv6 address could be chopped if the
8151 ifreq contained only a normal sockaddr (14 bytes storage).
8153 18. Named domain lists were not supported in the hosts_treat_as_local option.
8154 An entry such as +xxxx was not recognized, and was treated as a literal
8157 19. Ensure that header lines added by a DATA ACL are included in the reject log
8158 if the ACL subsequently rejects the message.
8160 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8161 MD5 (which is deprecated).
8163 21. When testing a filter file using -bf, Exim was writing a message when it
8164 took the sender from a "From " line in the message, but it was not doing so
8165 when it took $return_path from a Return-Path: header line. It now does.
8167 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8168 with a valid header line field name (a series of printing characters
8169 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8171 23. Changed "disc" in the source to "disk" to conform to the documentation and
8172 the book and for uniformity.
8174 24. Ignore Sendmail's -Ooption=value command line item.
8176 25. When execve() failed while trying to run a command in a pipe transport,
8177 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8178 could be confused with a return value of 69 from the command itself. This
8179 has been changed to 127, the value the shell returns if it is asked to run
8180 a non-existent command. The wording for the related log line suggests a
8181 non-existent command as the problem.
8183 26. If received_header_text expands to an empty string, do not add a Received:
8184 header line to the message. (Well, it adds a token one on the spool, but
8185 marks it "old" so that it doesn't get used or transmitted.)
8187 27. Installed eximstats 1.28 (addition of -nt option).
8189 28. There was no check for failure on the call to getsockname() in the daemon
8190 code. This can fail if there is a shortage of resources on the system, with
8191 ENOMEM, for example. A temporary error is now given on failure.
8193 29. Contrary to the C standard, it seems that in some environments, the
8194 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8195 program. Exim now does this explicitly; it affects the formatting of
8196 timestamps using strftime().
8198 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8199 complaints. I've now initialized all the variables, to avoid this.
8201 32. Header lines added by a system filter were not being "seen" during
8202 transport-time rewrites.
8204 33. The info_callback() function passed to OpenSSL is set up with type void
8205 (*)(SSL *, int, int), as described somewhere. However, when calling the
8206 function (actually a macro) that sets it up, the type void(*)() is
8207 expected. I've put in a cast to prevent warnings from picky compilers.
8209 34. If a DNS black list lookup found a CNAME record, but there were no A
8210 records associated with the domain it pointed at, Exim crashed.
8212 35. If a DNS black list lookup returned more than one A record, Exim ignored
8213 all but the first. It now scans all returned addresses if a particular IP
8214 value is being sought. In this situation, the contents of the
8215 $dnslist_value variable are a list of all the addresses, separated by a
8218 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8219 to accept a host name and all its aliases if the forward lookup for any of
8220 them yielded the IP address of the incoming connection. Now it accepts only
8221 those names whose forward lookup yields the correct IP address. Any other
8222 names are discarded. This closes a loophole whereby a rogue DNS
8223 administrator could create reverse DNS records to break through a
8224 wildcarded host restriction in an ACL.
8226 37. If a user filter or a system filter that ran in a subprocess used any of
8227 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8228 the wrong values were passed to the pipe command ($thisaddress had the
8229 value of $0, $0 had the value of $1, etc). This bug was introduced by
8230 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8232 38. Improved the line breaking for long SMTP error messages from ACLs.
8233 Previously, if there was no break point between 40 and 75 characters, Exim
8234 left the rest of the message alone. Two changes have been made: (a) I've
8235 reduced the minimum length to 35 characters; (b) if it can't find a break
8236 point between 35 and 75 characters, it looks ahead and uses the first one
8237 that it finds. This may give the occasional overlong line, but at least the
8238 remaining text gets split now.
8240 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8241 file descriptors might be low, and that setting 1000 would always raise it.
8242 It turns out that in some environments, the limit is already over 1000 and
8243 that lowering it causes trouble. So now Exim takes care not to decrease it.
8245 40. When delivering a message, the value of $return_path is set to $sender_
8246 address at the start of routing (routers may change the value). By an
8247 oversight, this default was not being set up when an address was tested by
8248 -bt or -bv, which affected the outcome if any router or filter referred to
8251 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8252 log only when "message" or "log_message" are set. However, if one of the
8253 conditions was an address verification, or a call to a nested ACL, the
8254 messages generated by the underlying test were being passed through. This
8255 no longer happens. The underlying message is available in $acl_verify_
8256 message for both "message" and "log_message" expansions, so it can be
8257 passed through if needed.
8259 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8260 new expansion $bh_ to give the encoded byte string without charset
8261 translation. Translation happens only if iconv() is available; HAVE_ICONV
8262 indicates this at build time. HEADERS_CHARSET gives the charset to
8263 translate to; headers_charset can change it in the configuration, and
8264 "headers charset" can change it in an individual filter file.
8266 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8267 that creates RFC 2047 encoded "words" labels them as that charset instead
8268 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8269 expansion operator; (ii) when Exim creates a From: line for a local
8270 message; (iii) when a header line is rewritten to include a "phrase" part.
8272 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8273 buggy, causing it to skip the first lines of messages whose message ID
8274 ended in 'D'. This would not have bitten before Exim release 4.14, because
8275 message IDs were unlikely to end in 'D' before then. The effect was to have
8276 incorrect size information for certain domains.
8278 45. #include "config.h" was missing at the start of the crypt16.c module. This
8279 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8282 46. If there was a timeout during a "random" callout check, Exim treated it as
8283 a failure of the random address, and carried on sending RSET and the real
8284 address. If the delay was just some slowness somewhere, the response to the
8285 original RCPT would be taken as a response to RSET and so on, causing
8286 mayhem of various kinds.
8288 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8289 when I implemented it. It didn't allow for the fact that some option values
8290 may legitimately be negative (e.g. size_addition), and it didn't even do
8291 the right test for positive values.
8293 48. Domain names in DNS records are case-independent. Exim always looks them up
8294 in lower case. Some resolvers return domain names in exactly the case they
8295 appear in the zone file, that is, they may contain uppercase letters. Not
8296 all resolvers do this - some return always lower case. Exim was treating a
8297 change of case by a resolver as a change of domain, similar to a widening
8298 of a domain abbreviation. This triggered its re-routing code and so it was
8299 trying to route what was effectively the same domain again. This normally
8300 caused routing to fail (because the router wouldn't handle the domain
8301 twice). Now Exim checks for this case specially, and just changes the
8302 casing of the domain that it ultimately uses when it transmits the message
8305 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8308 50. If a filter generated a file delivery with a non-absolute name (possible if
8309 no home directory exists for the router), the forbid_file option was not
8312 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8313 the existing equality matching.
8315 52. Exim was using ints instead of ino_t variables in some places where it was
8316 dealing with inode numbers.
8318 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8319 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8320 called TMPDIR, and if it finds it is different, it changes its value.
8322 54. The smtp_printf() function is now made available to local_scan() so
8323 additional output lines can be written before returning. There is also an
8324 smtp_fflush() function to enable the detection of a dropped connection.
8325 The variables smtp_input and smtp_batched_input are exported to
8328 55. Changed the default runtime configuration: the message "Unknown user"
8329 has been removed from the ACL, and instead placed on the localuser router,
8330 using the cannot_route_message feature. This means that any verification
8331 failures that generate their own messages won't get overridden. Similarly,
8332 the "Unrouteable address" message that was in the ACL for unverifiable
8333 relay addresses has also been removed.
8335 56. Added hosts_avoid_esmtp to the smtp transport.
8337 57. The exicyclog script was not checking for the esoteric option
8338 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8339 will work only if exicyclog is run under the appropriate euid.
8341 58. Following a discussion on the list, the rules by which Exim recognises line
8342 endings on incoming messages have been changed. The -dropcr and drop_cr
8343 options are now no-ops, retained only for backwards compatibility. The
8344 following line terminators are recognized: LF CRLF CR. However, special
8345 processing applies to CR:
8347 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8348 nor a local message in the state where . is a terminator.
8350 (ii) If a bare CR is encountered in a header line, an extra space is added
8351 after the line terminator so as not to end the header. The reasoning
8352 behind this is that bare CRs in header lines are most likely either
8353 to be mistakes, or people trying to play silly games.
8355 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8356 was being incorrectly given as 18 bytes larger than it should have been.
8357 This is a VOB (very old bug).
8359 60. This may never have affected anything current, but just in case it has:
8360 When the local host is found other than at the start of a list of hosts,
8361 the local host, those with the same MX, and any that follow, are discarded.
8362 When the list in question was part of a longer list of hosts, the following
8363 hosts (not currently being processed) were also being discarded. This no
8364 longer happens. I'm not sure if this situation could ever has previously
8367 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8370 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8371 header. This is something that is recommended in a new Internet Draft, and
8372 is something that is documented as being done by Sendmail. There are two
8373 possible values. For messages generated by the autoreply transport, Exim
8376 Auto-Submitted: auto-replied
8378 whereas for all other generated messages (e.g. bounces) it adds
8380 Auto-Submitted: auto-generated
8382 63. The "personal" condition in filters now includes a test for the
8383 Auto-Submitted: header. If it contains the string "auto-" the message it
8384 not considered personal.
8386 64. Added rcpt_include_affixes as a generic transport option.
8388 65. Added queue_only_override (default true).
8390 66. Added the syslog_duplication option.
8392 67. If what should have been the first header line of a message consisted of
8393 a space followed by a colon, Exim was mis-interpreting it as a header line.
8394 It isn't of course - it is syntactically invalid and should therefore be
8395 treated as the start of the message body. The misbehaviour could have
8396 caused a number of strange effects, including loss of data in subsequent
8397 header lines, and spool format errors.
8399 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8400 client host had authenticated. This control can now be exercised by an ACL
8401 for more flexibility.
8403 69. By default, callouts do not happen when testing with -bh. There is now a
8404 variant, -bhc, which does actually run the callout code, including
8405 consulting and updating the callout cache.
8407 70. Added support for saslauthd authentication, courtesy of Alexander
8410 71. If statvfs() failed on the spool or log directories while checking their
8411 size for availability, Exim confusingly gave the error "space shortage".
8412 Furthermore, in debugging mode it crashed with a floating point exception.
8413 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8414 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8415 serious problem, Exim now writes to the main and panic logs when this
8416 happens, with details of the failure. It then refuses to accept the
8417 incoming message, giving the message "spool directory problem" or "log
8418 directory problem" with a 421 code for SMTP messages.
8420 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8421 0, 1, and 2 exist, because some OS complain for execs without them (see
8422 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8423 descriptors. However, the code omitted to check that the open succeeded,
8424 causing mysterious errors if for some reason the permissions on /dev/null
8425 got screwed. Now Exim writes a message to the main and panic logs, and
8426 bombs out if it can't open /dev/null.
8428 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8429 interact so that it is all more flexible. It is supposed to remain
8430 backwards compatible. Also added extra_local_interfaces.
8432 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8433 to bomb out with an assertion failure - to the client this appears as a
8434 connection drop. This problem occurs in the part of the code that was taken
8435 from the Samba project. Fortunately, the assertion is in a very simple
8436 function, so I have fixed this by reproducing the function inline in the
8437 one place where it is called, and arranging for authentication to fail
8438 instead of killing the process with assert().
8440 75. The SPA client code was not working when the server requested OEM rather
8441 than Unicode encoding.
8443 76. Added code to make require_files with a specific uid setting more usable in
8444 the case where statting the file as root fails - usually a non-root-mounted
8445 NFS file system. When this happens and the failure is EACCES, Exim now
8446 forks a subprocess and does the per-uid checking as the relevant uid.
8448 77. Added process_log_path.
8450 78. If log_file_path was not explicitly set, a setting of check_log_space or
8451 check_log_inodes was ignored.
8453 79. If a space check for the spool or log partitions fails, the incident is now
8454 logged. Of course, in the latter case the data may get lost...
8456 80. Added the %p formatting code to string_format() so that it can be used to
8457 print addresses in debug_print(). Adjusted all the address printing in the
8458 debugging in store.c to use %p rather than %d.
8460 81. There was a concern that a line of code in smtp_in.c could overflow a
8461 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8462 initially expressed, the concern was not well-founded, because trailing
8463 spaces are removed early. However, if the trailing spaces were followed by
8464 a NULL, they did not get removed, so the overflow was possible. Two fixes
8467 (a) I re-wrote the offending code in a cleaner fashion.
8468 (b) If an incoming SMTP command contains a NULL character, it is rejected
8471 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8472 runs initgroups(), so that if the Exim user is in any additional groups,
8473 they will be used during message reception.
8479 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob