1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
21 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
22 (under Linux) we watch the dir containing the final file; previously
23 it would be the dir with the first symlink. We still do not monitor
26 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
27 dn_expand() is happy to pass them through.
29 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
30 the file with the proof had an unchanged name, the new proof(s) were
31 loaded on top of the old ones (and nover used; the old ones were stapled).
33 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
34 more than one message arrived in a single connection a reference from
35 the earlier message could be re-used. Often a sigsegv resulted.
36 These variables were introduced in Exim 4.87.
37 Debug help from Graeme Fowler.
39 JH/09 Fix ${filter } for conditions that modify $value. Previously the
40 modified version would be used in construction the result, and a memory
43 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
44 Find and fix by Jasen Betts.
46 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
47 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
48 the systemwide configuration to override the Exim config.
50 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
51 API changes in libopendmarc.
53 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
54 pid 1, in the normal "background daemon" mode, having to drop process-
55 group leadership also lost track of needing to create listener sockets.
57 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
58 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
60 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
61 a capture group which obtained no text (eg. "(abc)*" matching zero
62 occurrences) could cause a segfault if the corresponding $<n> was
65 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
66 included a close-brace character (eg. it itself used an expansion) an
69 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
70 starting TLS. Previously it was after, meaning that attackers on such
71 ports had to be screened using the host_reject_connection main config
72 option. The new sequence aligns better with the STARTTLS behaviour, and
73 permits defences against crypto-processing load attacks, even though it
74 is strictly an incompatible change.
75 Also, avoid sending any SMTP fail response for either the connect ACL
76 or host_reject_connection, for TLS-on-connect ports.
78 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
79 Previously this was not permitted, but it makes reasonable sense.
80 While there, restore a restriction on using it from a connect ACL; given
81 the change JH/16 it could only return false (and before 4.91 was not
84 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
85 was exactly sized compared to the log buffer, a crash occurred with the
86 misleading message "bad memory reference; pool not found".
87 Found and traced by Jasen Betts.
89 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
90 dns_again_means_nonexist included an element causing a DNS lookup which
91 iteslf returned DNS_AGAIN, unbounded recursion occurred. Possible results
92 included (though probably not limited to) a process crash from stack
93 memory limit, or from excessive open files. Replace this with a paniclog
94 whine (as this is likely a configuration error), and returning
97 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
98 this always failed, probably leading to the usual downgrade to in-clear
101 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
102 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
103 to not checking that list for these lookups.
105 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
108 JH/23 Fix crash in string expansions. Previously, if an empty variable was
109 immediately followed by an expansion operator, a null-indirection read
110 was done, killing the process.
112 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
113 include an SMTP response string which is longer than that supported
114 by the delivering transport. Alleviate by wrapping such lines before
117 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
118 chars (RFC limit). Previously a limit of 12 items was made, which with
119 a not-impossible References: in the message being bounced could still
120 be over-large and get stopped in the transport.
122 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
123 close. Previously a bare socket close was done.
125 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
128 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
129 message being considered for delivery was already being handled by
130 another process, and having an SMTP connection already open, the function
131 to close it tried to use an uninitialized variable. This would afftect
132 high-volume sites more, especially when running mailing-list-style loads.
133 Pollution of logs was the major effect, as the other process delivered
134 the message. Found and partly investigated by Graeme Fowler.
136 JH/29 Change format of the internal ID used for message identification. The old
137 version only supported 31 bits for a PID element; the new 64 (on systems
138 which can use Base-62 encoding, which is all currently supported ones
139 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
140 and must use Base-36). The new ID is 23 characters rather than 16, and is
141 visible in various places - notably logs, message headers, and spool file
142 names. Various of the ancillary utilities also have to know the format.
143 As well as the expanded PID portion, the sub-second part of the time
144 recorded in the ID is expanded to support finer precision. Theoretically
145 this permits a receive rate from a single comms channel of better than the
147 The major timestamp part of the ID is not changed; at 6 characters it is
148 usable until about year 3700.
149 Updating from previously releases is fully supported: old-format spool
150 files are still usable, and the utilities support both formats. New
151 message will use the new format. The one hints-DB file type which uses
152 message-IDs (the transport wait- DB) will be discarded if an old-format ID
153 is seen; new ones will be built with only new-format IDs.
154 Optionally, a utility can be used to convert spool files from old to new,
155 but this is only an efficiency measure not a requirement for operation
156 Downgrading from new to old requires running a provided utility, having
157 first stopped all operations. This will convert any spool files from new
158 back to old (losing time-precision and PID information) and remove any
159 wait- hints databases.
161 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
162 we treated them as item separators when parsing for a list item, but they
163 need to be protected by the doublequotes. While there, add handling for
166 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
167 Found and fixed by Jasen Betts. No testcase for this as my usual text
168 editor insists on emitting only valid UTF-8.
170 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
171 a null-indirection SIGSEGV for the receive process.
173 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
174 -bh use. Internal checking would spot this and take a panic.
176 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
177 In 4.96 this would expand to empty.
179 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
180 certificate. Find and fix by Andreas Metzler.
182 JH/36 Add ARC info to DMARC hostory records.
184 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
185 or fakedefer. Previously the sender could discover that the message
186 had in fact been accepted.
188 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
189 sequences. Previously the input was not noted as being tainted; notably
190 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
191 bad coding of authenticators.
193 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
194 and ${tr...}. Found and diagnosed by Heiko Schlichting.
196 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
197 could be triggered by externally-supplied input. Found by Trend Micro.
199 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
200 be triggered by externally-controlled input. Found by Trend Micro.
202 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
203 be triggered by externally-controlled input. Found by Trend Micro.
209 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
210 after reception to before a subsequent reception. This should
211 mean slightly faster delivery, and also confirmation of reception
214 JH/02 Move from using the pcre library to pcre2. The former is no longer
215 being developed or supported (by the original developer).
217 JH/03 Constification work in the filters module required a major version
218 bump for the local-scan API. Specifically, the "headers_charset"
219 global which is visible via the API is now const and may therefore
220 not be modified by local-scan code.
222 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
223 sendfile() didi not account for the way the ClamAV driver code called it.
225 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
226 time check was being done for every character; replace that with one
229 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
230 sent was prefixed with a length byte.
232 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
233 RFC 5321. Previously Dovecot (at least) would log errors during
236 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
237 in 4.61 without a reason logged. Binaries will be bigger, which might
238 matter on diskspace-constrained systems, but debug is easier.
240 JH/09 Fix macro-definition during "-be" expansion testing. The move to
241 write-protected store for macros had not accounted for these runtime
242 additions; fix by removing this protection for "-be" mode.
244 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
245 handing out large-numbered file descriptors, violating the usual Unix
246 assumption (and required by Posix) that the lowest possible number will be
247 allocated by the kernel when a new one is needed. In the daemon, and any
248 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
249 are not useable for FD_SET() [and hence select()] and overwrite the stack.
250 Assorted crashes happen.
252 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
253 main-section options or in a connect ACL, the value from the first ever
254 connection was never replaced for subsequent connections. Found by
257 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
258 though only once PCRE2 was introduced: the memory accounting used under
259 debug offset allocations by an int, giving a hard trap in early startup.
260 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
262 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
263 with underbars is given. The write-protection of configuration introduced
264 in 4.95 trapped when normalisation was applied to an option not needing
267 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
269 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
270 failing to close the certificates directory, every hour or any time it
273 JH/16 Debugging initiated by an ACL control now continues through into routing
274 and transport processes. Previously debugging stopped any time Exim
275 re-execs, or for processing a queued message.
277 JH/17 The "expand" debug selector now gives more detail, specifically on the
278 result of expansion operators and items.
280 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
281 bad comparison between the option value and the name of the file to
282 be included was done, and a mismatch was wrongly identified.
283 4.88 to 4.95 are affected.
285 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
287 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
288 supplied as the db file-pair basename. Previously, if a directory
289 path was given, for example via the autoreply "once" option, the DB
290 file.pag and file.dir files would be created in that directory's
293 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
294 "taint" log_selector. These were previously deprecated.
296 JH/22 Fix static address-list lookups to properly return the matched item.
297 Previously only the domain part was returned.
299 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
300 the call into OpenSSL to send a TLS Close was being repeated; this
301 resulted in the library waiting for the peer's Close. If that was never
302 sent we waited forever. Fix by tracking send calls.
304 JH/24 The ${run} expansion item now expands its command string elements after
305 splitting. Previously it was before; the new ordering makes handling
306 zero-length arguments simpler. The old ordering can be obtained by
307 appending a new option "preexpand", after a comma, to the "run".
309 JH/25 Taint-check exec arguments for transport-initiated external processes.
310 Previously, tainted values could be used. This affects "pipe", "lmtp" and
311 "queryprogram" transport, transport-filter, and ETRN commands.
312 The ${run} expansion is also affected: in "preexpand" mode no part of
313 the command line may be tainted, in default mode the executable name
316 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
317 the the facility was not passed across execs, and only the first message
318 passed over a connection could use BDAT; any further ones using DATA.
320 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
321 uses $sending_ip_address and an interface is specified.
322 Previously any use of the local address in the EHLO name disabled
323 PIPECONNECT, the common case being to use the rDNS of it.
325 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
326 resumption. Previously verify failed because no certificate status is
327 passed on the wire for the restarted session. Fix by using the recorded
328 ocsp status of the stored session for the new connection.
330 JH/29 TLS resumption: the key for session lookup in the client now includes
331 more info that a server could potentially use in configuring a TLS
332 session, avoiding oferring mismatching sessions to such a server.
333 Previously only the server IP was used.
335 JH/30 Fix string_copyn() for limit greater than actual string length.
336 Previously the copied amount was the limit, which could result in a
337 overlapping memcpy for newly allocated destination soon after a
338 source string shorter than the limit. Found/investigated by KM.
340 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
341 close; it may be needed for a subsequent connection. This caused a
342 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
344 JH/32 Fix CHUNKING for a second message on a connection when the first was
345 rejected. Previously we did not reset the chunking-offered state, and
346 erroneously rejected the BDAT command. Investigation help from
349 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
350 an empty address. Previously the expansion returned an error.
352 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
353 proxy. Previously these were misparsed, leading to paniclog entries.
359 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
360 related applications. Previously an "H" was used where available info
361 says that "M" should be, so change to match.
363 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
364 as arguments, so an implementation trying to copy these into a local
365 buffer was taking a taint-enforcement trap. Fix by using dynamically
366 created buffers. Similar fix for radius expansion condition.
368 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
369 reasonable, eg. to count headers. Fix by using dynamically created
370 buffers rather than a local. Do similar fixes for ACL actions "dcc",
371 "log_reject_target", "malware" and "spam"; the arguments are expanded
372 so could be handling tainted values.
374 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
375 broken the (no-op) support for this sendmail command. Restore it
376 to doing nothing, silently, and returning good status.
378 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
379 record path was given (or the default used) without a leading directory
380 path, an error occurred on trying to open it. Use the transport's working
383 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
384 transport. Previously it was the name on the DNS A-record; use instead
385 the head of the CNAME chain leading there (if there is one). This seems
386 to align better with RFC 6125.
388 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
389 smtp_accept_max_per_host allocated resources which were not released
390 when the limit was exceeded. This eventually crashed the daemon. Fix
391 by adding a release action in that path.
393 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
394 expanded; previously using tainted values was rejected. Fix by using
395 dynamically-created buffers.
397 JH/09 Relax restrictions on ACL verify condition needing access to message
398 headers. Previously they were only permitted in data and non-smtp ACLs;
399 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
400 not_blind, header_sender and header_names_ascii verification.
402 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
403 Previously a macro used one argument twice; when called with the
404 argument as an expression having side-effects, incorrect operation
405 resulted. Use an inlineable function.
407 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
408 held open for a verify callout. Previously this wan not accounted for
409 and a corrupt onward SMTP conversation resulted.
411 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
412 Previously they were embedded naively in the constructed address; when
413 needed, strip the quoting and quote the entire local_part.
414 Also make the inbound_srs expansion condition handle quoting.
416 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
417 excluded, not matching the documentation.
419 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
420 was given for the sqlite_dbfile a trap resulted.
422 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
423 "name" argument resulted in a trap. There is no reason to disallow such;
424 this was a coding error.
426 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
427 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
428 spent suspended, ignoring the POSIX definition. Previously we assumed
429 it did and a constant offset from real time could be used as a correction.
430 Change to using the same clock source for the start-of-message and the
431 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
432 exists, just to get a clock slightly more aligned to reality.
434 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
435 RFC says it is optional some validators care. The missing char was not
436 intended but triggered by a line-wrap alignment. Discovery and fix by
437 Guillaume Outters, hacked on by JH.
439 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
440 name being quoted was tainted a trap would be taken. Fix by using
441 dynamically created buffers. The routine could have been called by a
442 rewrite with the "h" flag, by using the "-F" command-line option, or
443 by using a "name=" option on a control=submission ACL modifier.
445 JH/19 SPF: change the Authentication-Results expansion component to give
446 smtp.helo when the sender domain is empty. Previously it gave
449 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
450 not in 127.0.0.0/8 to help in spotting list domains taken over by a
451 domain-parking registrar.
453 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
454 Previously when a whitespace character was specified it was not inserted
455 after removing the newline.
457 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
458 the domain part of the recipient address. This overrides any tls_sni
459 option set, which was previously used.
461 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
464 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
465 is_tainted() had an off-by-one error in the overenthusiastic direction.
466 Find and fix by Gavan. Although NetBSD is not a supported platform for
467 4.94 this bug could affect other platforms.
469 PP/01 Fix default prime selection to be consistent.
470 One path used ike23 still, instead of exim.dev.20160529.3; now both
471 execution flows will use the same DH primes (currently
472 exim.dev.20160529.3).
474 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
475 option in smtp transport, to match the documentation. Previously
476 verification was not being done in some cases where it should have been.
478 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
479 than one server was defined and depending on the platform memory layout
480 details, an internal consistency trap could be hit while walking the list
483 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
484 files. The value is used by the authresults expansion item. Previously
485 if this was used in a router or transport, a crash could result.
487 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
488 have a sufficient incidence of EINTR returns from select that an
489 interaction with other operations done by the main daemon loop exposed
490 a bug in the error-handling. This was benign apart from the log
493 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
494 for consistency with delivered "=>" and failed "**" lines. While we're
495 there, handle PRX and TFO.
497 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
498 applied. This resulted, if any header-line rewrite rules were configured,
499 in a panic-log triggerable by sending a message with a long address in
500 a header. Fix by increasing the arbitrary limit to larger than a single
501 (dewrapped) 5322 header line maximum size.
503 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
504 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
505 RFC 6648 which deprecates X- options in protocols as a general practice.
506 Changeover between the implementations is handled by the mechanisms
509 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
510 a remote which uses callout/hold. Previously the local was queued.
512 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
515 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
516 log line. The asterisk is now consistently appended to the remote IP
517 (and port, if given), and will also be provided on defer and fail log
518 lines. Previously it could be placed on the local IP if that was being
519 logged, and was only provided on delivery lines.
521 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
523 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
524 authenticator client_send option. Previously the next char, after a pair
525 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
526 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
527 documentation. There is still no way to get a leading ^ immediately
528 after a NUL (ie. for the password of a PLAIN method authenticator.
530 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
531 files. For bad sizes read, delete the record and whine to paniclog.
533 JH/38 When logging an AUTH failure, as server, do not include sensitive
534 information. Previously, the credentials would be included if given
535 as part of the AUTH command line and an ACL denied authentication.
537 JH/39 Bug 2691: fix $local_part_data. When the matching list element
538 referred to a file, bad data was returned. This likely also affected
541 JH/40 The gsasl authenticator now supports caching of the salted password
542 generated by the client-side implementation. This required the addition
543 of a new variable: $auth4.
545 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
546 left undeleted; the attempt to re-create it then failed - resulting in
547 the usual "SIGHUP tp have daemon reload configuration" to not work.
548 This affected any platform not supporting "abstract" Unix-domain
549 sockets (i.e. not Linux).
551 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
552 recipients" response to RCPT in a later response, with a 250. The
553 previous coding assumed this would not happen, and under PIPELINING
554 would result in both lost and duplicate recipients for a message.
556 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
557 Previously the weighting was incorrectly applied. Similar fix for socks
558 proxies. Found and fixed by Heiko Schlichting.
560 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
561 not handle sub-lists included using the +namedlist syntax. While
562 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
563 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
566 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
567 Previously the permanent pool was used, so the sore could not be freed.
568 This meant a connection with many messages would use continually-growing
571 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
572 per-pool so as not to waste too much space. Previously a constant size
573 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
574 DOS attacks harder. The cost is wasted memory use in the larger blocks.
576 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
577 same space repeatedly, and a smaller process footprint.
579 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
580 for GnuTLS. Previously the name started "%s" which, while not a bug,
581 looked as if if might be one.
583 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
584 "continued connection") make the $proxy_* variables available. Previously
585 the information was not passed across the exec() call for subsequent
586 transport executions. This also mean that the log lines for the
587 messages can show the proxy information.
589 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
590 receive time. With modern systems the difference is significant.
591 The historical behaviour can be restored by disabling (a new) log_selector
592 "queue_time_exclusive".
594 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
595 content) were specifically tested for. Now, also cover expansions
596 resulting in ACL names and inline ACL content.
598 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
599 making it unusable in complex expressions.
601 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
602 fail with a taint-check on the spoolfile name, and leave the message
605 HS/01 Enforce absolute PID file path name.
607 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
609 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
611 PP/02 Bug 2643: Correct TLS DH constants.
612 A missing NUL termination in our code-generation tool had led to some
613 incorrect Diffie-Hellman constants in the Exim source.
614 Reported by kylon94, code-gen tool fix by Simon Arlott.
616 PP/03 Impose security length checks on various command-line options.
617 Fixes CVE-2020-SPRSS reported by Qualys.
619 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
620 better. Reported by Qualys.
622 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
623 providing a particularly obnoxious sender full name.
626 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
628 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
631 PP/08 Change default for recipients_max from unlimited to 50,000.
633 PP/09 Fix security issue with too many recipients on a message (to remove a
634 known security problem if someone does set recipients_max to unlimited,
635 or if local additions add to the recipient list).
636 Fixes CVE-2020-RCPTL reported by Qualys.
638 PP/10 Fix security issue in SMTP verb option parsing
639 Fixes CVE-2020-EXOPT reported by Qualys.
641 PP/11 Fix security issue in BDAT state confusion.
642 Ensure we reset known-good where we know we need to not be reading BDAT
643 data, as a general case fix, and move the places where we switch to BDAT
644 mode until after various protocol state checks.
645 Fixes CVE-2020-BDATA reported by Qualys.
647 HS/03 Die on "/../" in msglog file names
649 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
650 the Exim runtime user are allowed to create files.
652 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
655 QS/03 When reading the output from interpreted forward files we do not
656 pass the pipe between the parent and the interpreting process to
657 executed child processes (if any).
659 QS/04 Always die if requested from internal logging, even is logging is
662 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
663 the API; compilation noo longer completes with DMARC support included.
664 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
665 been reported on other platforms.
667 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
669 JH/56 Make the majority of info read from config files readonly, for defence-in-
670 depth against exploits. Suggestion by Qualys.
671 Not supported on Solaris 10.
673 JH/57 Fix control=fakreject for a custom message containing tainted data.
674 Previously this resulted in a log complaint, due to a re-expansion present
675 since fakereject was originally introduced.
677 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
678 modifier was given, a loop resulted.
680 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
681 pipelined SMTP command followed the BDAT LAST then it would be
682 incorrectly treated as part of the message body, causing a verification
685 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
686 require looser validation rules than those for 821-level addresses,
687 which only permit IP addresses.
693 JH/01 Avoid costly startup code when not strictly needed. This reduces time
694 for some exim process initialisations. It does mean that the logging
695 of TLS configuration problems is only done for the daemon startup.
697 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
699 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
700 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
703 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
704 DKIM signing is being done. Previously a transport_filter would always
705 disable CHUNKING, falling back to traditional DATA.
707 JH/05 Regard command-line recipients as tainted.
709 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
711 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
712 PAM library frees one of the arguments given to it, despite the
713 documentation. Therefore a plain malloc must be used.
715 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
716 on-stack buffers were used, resulting in a taint trap when DSN information
717 copied from a received message was written into the buffer.
719 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
720 the ordering of its ARC headers. This caused a crash.
722 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
723 a new record was being constructed with information from the peer, a trap
726 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
727 installation would get error messages from DMARC verify, when it hit the
728 nonexistent file indicated by the default. Distros wanting DMARC enabled
729 should both provide the file and set the option.
730 Also enforce no DMARC verification for command-line sourced messages.
732 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
733 could, depending on the platform, hang at the STARTTLS response.
735 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
736 message on a connection. Previously if one message had ARC headers and
737 the following one did not, a crash could result when adding an
738 Authentication-Results: header.
740 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
741 Exim main code and Exim-related utities. The introduction of taint
742 tracking also did many adjustments to string handling. Since then, eximon
743 frequently terminated with an assert failure.
745 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
746 check for 452 responses. This slightly helps the inefficieny of doing
747 a large alias-expansion into a recipient-limited target. The max_rcpt
748 transport option still applies (and at the current default, will override
749 the new feature). The check is done for either cause of synch, and forces
750 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
751 connection. The new facility is not tunable at this time.
753 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
754 library live data was being used, so the results became garbage. Make
755 copies while it is still usable.
757 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
758 on delivery deferred (==) and failed (**) lines (if a delivery was
759 attemtped). Previously it was only on completion (=>) lines.
761 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
762 for the expansion of the server_scram_iter and server_scram_salt options.
764 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
765 are now specifically given a NO_DATA response without hitting the system
766 resolver. The library goes on to do the now-standard TXT lookup.
767 Use of dnsdb lookups is not affected.
769 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
770 only retrieve the errormessage once. Previously two calls to dlerror()
771 were used, and the second one (for mainlog/paniclog) retrieved null
774 JH/20 Taint checking: disallow use of tainted data for
775 - the appendfile transport file and directory options
776 - the pipe transport command
777 - the autoreply transport file, log and once options
778 - file names used by the redirect router (including filter files)
780 - paths used by single-key lookups
781 Previously this was permitted.
783 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
784 adjusted the size of a major service buffer; this failed because the
785 buffer was in use at the time. Change to a compile-time increase in the
786 buffer size, when this authenticator is compiled into exim.
788 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
789 previous fast-mode was untenable in the face of glibs using mmap to
790 support larger malloc requests.
792 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
793 New values supported, if defined on system where compiled:
794 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
795 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
797 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
798 running a limited number of proceses in parallel, a benefit is gained. The
799 amount varies with the platform hardware and load. The use of the option
800 queue_run_in_order means we cannot do this, as ordering becomes
803 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
804 had introduced a string-copy (for ensuring NUL-termination) which was not
805 appropriate for that case, which can include embedded NUL bytes in the
806 block of data. Investigation showed the copy to actually be needless, the
807 data being length-specified.
809 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
810 done during a receiving connection, and both used TLS, global info was
811 used rather than per-connection info for tracking the state of data
812 queued for transmission. This could result in a connection hang.
814 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
815 Previously, when delivering serveral messages down a single connection
816 only the first would provide a SIZE. This was due to the size information
817 not being properly tracked.
819 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
820 TAI (at 37 seconds currently), pretend to be in UTC for time-related
821 expansion and logging. Previously, spurious values such as a future
822 minute could be seen.
824 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
825 it could crash from a null-deref. This could also affect the
826 ${addresses: } operator and ${readsock } item.
828 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
829 message following a mime one, the variable was not reset.
831 JH/30 When an pipelined-connect fails at the first response, assume incorrect
832 cached capability (perhaps the peer reneged?) and immediately retry in
835 JH/31 Fix spurious detection of timeout while writing to transport filter.
837 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
838 an attempt to copy the string was made before checking it.
840 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
841 taint of the lookup key was maintained; we now regard the presence in the
842 filesystem as sufficient validation.
844 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
845 argument is supplied.
847 JH/35 The dsearch lookup now requires that the directory is an absolute path.
848 Previously this was not checked, and nonempty relative paths made an
849 access under Exim's current working directory.
851 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
852 Previously no event was raised.
854 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
855 parameter supplied by the sender MAIL FROM command. Previously it was
856 ignored, and only the check_spool_space option value for the required
859 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
860 the size of the signing public-key. Previously it was instead giving
861 the size of the signature hash.
863 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
864 the default. See the (new) dkim_verify_min_keysizes option.
866 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
867 and an ACL use a lookup for checking either the local_part or domain,
868 stale data could be accessed. Ensure that variable references are
869 dropped between messages.
871 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
872 by the client was not checked as pointing within response data before
873 being used. A malicious client could thus cause an out-of-bounds read and
874 possibly gain authentication. Fix by adding the check.
876 JH/42 Internationalisation: change the default for downconversion in the smtp
877 transport to be "if needed". Previously it was "as previously set" for
878 the message, which usually meant "if needed" for message-submission but
879 "no" for everything else. However, MTAs have been seen using SMTPUTF8
880 even when the envelope addresses did not need it, resulting in forwarding
881 failures to non-supporting MTAs. A downconvert in such cases will be
882 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
883 transport. The change does mean that addresses needing conversion will
884 be converted when previously a delivery failure would occur.
886 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
887 response was received it would be used unchecked in a fail-DSN, violating
888 standards on line-length limits. Truncate if needed.
890 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
891 doesn't work. (Additionally add a new main config option to configure the
898 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
899 side, to decode a TLS 1.3 packet capture.
901 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
902 Previously the default library behaviour applied, sending two, each in
905 JH/03 Debug output for ACL now gives the config file name and line number for
908 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
910 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
912 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
913 buffer overrun for (non-chunking) other transports.
915 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
916 TLS1.3, means that a server rejecting a client certificate is not visible
917 to the client until the first read of encrypted data (typically the
918 response to EHLO). Add detection for that case and treat it as a failed
919 TLS connection attempt, so that the normal retry-in-clear can work (if
920 suitably configured).
922 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
923 and/or domain. Found and fixed by Jason Betts.
925 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
926 configuration). If a CNAME target was not a wellformed name pattern, a
929 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
930 the OS reports them interleaved with other addresses.
932 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
933 used both for input and for a verify callout, both encrypted, SMTP
934 responses being sent by the server could be lost. This resulted in
935 dropped connections and sometimes bounces generated by a peer sending
938 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
939 string. Previously it was possible to cause undefined behaviour in a
940 library routine (usually a crash). Found by "zerons".
942 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
945 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
946 API was removed, so update to use the newer ones.
948 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
949 any timeout set, is taking a long time. Previously we would hang on to a
950 rotated logfile "forever" if the input was arriving with long gaps
951 (a previous attempt to fix addressed lack, for a long time, of initial
954 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
955 shared (NFS) environment. The length of the tempfile name is now
956 4 + 16 ("hdr.$message_exim_id") which might break on file
957 systems which restrict the file name length to lower values.
960 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
961 shared (NFS) environment.
963 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
964 did for all versions <4.90). Notably -M, -m, --invert, -I may be
967 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
968 on some platforms for bit 31.
970 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
971 to changes apparently associated with TLS1.3 handling some of the APIs
972 previously used were either nonfunctional or inappropriate. Strings
973 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
974 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
975 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
976 This affects log line X= elements, the $tls_{in,out}_cipher variables,
977 and the use of specific cipher names in the encrypted= ACL condition.
979 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
981 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
982 verification result was not updated unless hosts_require_ocsp applied.
984 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
985 queue_list_requires_admin set to false, non-admin users were denied the
988 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
989 directory-of-certs mode. Previously they were advertised despite the
992 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
993 A single TCP connection by a client will now hold a TLS connection open
994 for multiple message deliveries, by default. Previously the default was to
997 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
998 default. If built with the facility, DANE will be used. The facility
999 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1001 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1002 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1003 must be defined and you must still, unless you define DISABLE_TLS, manage
1004 the the include-dir and library-file requirements that go with that
1005 choice. Non-TLS builds are still supported.
1007 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1010 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1011 default. If the platform supports and has the facility enabled, it will
1012 be requested on all coneections.
1014 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1015 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1017 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1019 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1020 success-DSN messages. Previously the From: header was always the default
1021 one for these; the option was ignored.
1023 JH/28 Fix the timeout on smtp response to apply to the whole response.
1024 Previously it was reset for every read, so a teergrubing peer sending
1025 single bytes within the time limit could extend the connection for a
1026 long time. Credit to Qualsys Security Advisory Team for the discovery.
1028 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1029 delivery address, which leaked information of the results of local
1030 forwarding. Change to the original envelope recipient address, per
1033 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1034 requested. Previously not bounce was generated and a log entry of
1035 error ignored was made.
1037 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1039 JH/32 Introduce a general tainting mechanism for values read from the input
1040 channel, and values derived from them. Refuse to expand any tainted
1041 values, to catch one form of exploit.
1043 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1044 was unused and the unexpanded text used for the test. Found and
1045 fixed by Ruben Jenster.
1047 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1048 an attempt to use a TLS library read routine dereffed a nul pointer,
1051 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1052 them in our smtp response.
1054 JH/36 Have the general router option retry_use_local_part default to true when
1055 any of the restrictive preconditions are set (to anything). Previously it
1056 was only for check_local user. The change removes one item of manual
1057 configuration which is required for proper retries when a remote router
1058 handles a subset of addresses for a domain.
1060 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1061 link count into consideration.
1063 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1064 caused the extension of big_buffer, the following lines were ignored.
1066 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1067 accordance with RFC 2308. Previously there was no expiry, so a longlived
1068 receive process (eg. due to ACL delays) versus a short SOA value could
1071 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1073 JH/39 Promote DMARC support to mainline.
1075 JH/40 Bug 2452: Add a References: header to DSNs.
1077 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1078 parameters. The relevant library call is documented as "Deprecated: This
1079 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1080 3.6.0, DH parameters are negotiated following RFC7919."
1082 HS/06 Change the default of dnssec_request_domains to "*"
1084 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1085 carried on and emitted a BDAT command, even when PIPELINING was not
1088 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1089 buffer was used for the filename, resulting in a trap when tainted
1090 arguments (eg. $domain) were used.
1092 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1093 recommended to avoid a possible server-load attack. The feature can be
1094 re-enabled via the openssl_options main cofiguration option.
1096 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1097 for version 4.90 - adding a "more data" boolean to the arguments.
1098 Bumped the ABI version number also, this having been missed previously;
1099 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1100 respect of usage of smtp_printf() by either local_scan code or libraries
1101 accessed via the ${dlfunc } expansion item. Both will need coding
1102 adjustment for any calls to smtp_printf() to match the new function
1103 signature; a FALSE value for the new argument is always safe.
1105 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1106 the file-offset (which the Linux syscall does, and exim expects); this
1107 resulted in an indefinite loop.
1109 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1110 to do ARC verification. The Authentication-Results: header line added
1111 by the configuration then had no ARC item.
1113 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1114 an incorrect mode was used for reading the result, resulting in it being
1121 JH/01 Remove code calling the customisable local_scan function, unless a new
1122 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1124 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1125 non-signal-safe functions being used.
1127 JH/03 Bug 2269: When presented with a received message having a stupidly large
1128 number of DKIM-Signature headers, disable DKIM verification to avoid
1129 a resource-consumption attack. The limit is set at twenty.
1131 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1132 report of oldest_pass in ${authres } in consequence, and separate out
1133 some descriptions of reasons for verification fail.
1135 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1136 files in the spool were present and unlocked. A queue-runner could spot
1137 them, resulting in a duplicate delivery. Fix that by doing the unlock
1138 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1139 add more error-checking on spoolfile handling while that code is being
1142 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1143 No known attacks, no CVE, this is defensive hardening.
1145 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1146 a queue-runner could start a delivery while other operations were ongoing.
1147 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1148 Found and investigated by Tim Stewart. Fix by using the open message data
1149 file handle rather than opening another, and not locally closing it (which
1150 releases a lock) for that case, while creating the temporary .eml format
1151 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1153 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1154 $sender_verify_failure/$recipient_verify_failure to "random".
1156 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1159 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1160 Previously this would segfault.
1162 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1165 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1166 like zero, since the resolver should be doing this for us, But we need one
1167 as a CNAME but no MX presence gets the CNAME returned; we need to check
1168 that doesn't point to an MX to declare it "no MX returned" rather than
1169 "error, loop". A new main option is added so the older capability of
1170 following some limited number of chain links is maintained.
1172 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1174 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1175 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1176 now-deprecated earlier definitions used only specified the range up to TLS
1177 1.2 (in the older-version library docs).
1179 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1181 JH/15 Rework TLS client-side context management. Stop using a global, and
1182 explicitly pass a context around. This enables future use of TLS for
1183 connections to service-daemons (eg. malware scanning) while a client smtp
1184 connection is using TLS; with cutthrough connections this is quite likely.
1186 JH/16 Fix ARC verification to do AS checks in reverse order.
1188 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1190 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1191 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1192 Previously the "utf8" would be re-prepended for every additional message.
1194 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1195 Previously thery were accepted, resulting in issues when attempting to
1196 forward messages to a non-supporting MTA.
1198 PP/02 Let -n work with printing macros too, not just options.
1200 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1201 one parent address was copied, and bogus data was used at delivery-logging
1202 time. Either a crash (after delivery) or bogus log data could result.
1203 Discovery and analysis by Tim Stewart.
1205 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1206 Previously if the string ended mid-character, we did not insert the
1207 promised '?' replacement.
1209 PP/04 Documentation: current string operators work on bytes, not codepoints.
1211 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1212 should pack well giving a smaller memory footprint so better caching and
1213 therefore performance. Group the declarations where this can't be done so
1214 that the byte-sized flag variables are not interspersed among pointer
1215 variables, giving a better chance of good packing by the compiler.
1217 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1218 non-null, to avoid issues with sites running BATV. Previously reports were
1219 sent with an empty envelope sender so looked like bounces.
1221 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1222 The ignore_error flag wasn't being returned from the filter subprocess so
1223 was not set for later routers. Investigation and fix by Matthias Kurz.
1225 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1226 and a msg:complete for the whole, when a message is manually removed using
1227 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1229 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1230 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1231 hope that is portable enough.
1233 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1234 requiring a known-CA anchor certificate; make it now rely entirely on the
1235 TLSA as an anchor. Checking the name on the leaf cert against the name
1236 on the A-record for the host is still done for TA (but not for EE mode).
1238 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1239 included in delivery lines for non-proxied connections, when compiled with
1240 SUPPORT_SOCKS and running with proxy logging enabled.
1242 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1243 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1244 move the existing event to fire before the normal logging of message
1245 failure so that custom logging is bracketed by normal logging.
1247 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1248 msg:fail:internal event. Developement by Matthias Kurz.
1250 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1251 far too small for todays use of crypto signatures stored there. Go all
1252 the way to the max DNS message size of 64kB, even though this might be
1253 overmuch for IOT constrained device use.
1255 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1256 copy a string over itself. The library routine is documented as not
1257 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1259 JH/32 For main options check_spool_space and check_inode_space, where the
1260 platform supports 64b integers, support more than the previous 2^31 kB
1261 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1262 the previous G, M, k.
1264 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1265 $authenticated_fail_id variable on authentication failure. Previously
1268 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1269 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1270 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1271 GNUTLS_SEC_PARAM_MEDIUM.
1273 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1274 side. Previously we would continue as if no SNI had been received.
1276 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1277 "<" character, which should be a list-separator specification, we walked
1278 off past the nul-terimation.
1280 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1281 causes) even when the retry time is not yet met. Previously they were
1282 not, meaning that when (say) an account was over-quota and temp-rejecting,
1283 and multiple senders' messages were queued, only one sender would get
1284 notified on each configured delay_warning cycle.
1286 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1288 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1289 cipher-suites, an error can be left on the stack even for a succeeding
1290 accept; this results in impossible error messages when a later operation
1293 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1294 return error codes indicating retry. Under TLS1.3 this becomes required.
1296 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1297 it only wrote the new authenticators, resulting in a lack of tracking of
1298 peer changes of ESMTP extensions until the next cache flush.
1300 JH/41 Fix the loop reading a message header line to check for integer overflow,
1301 and more-often against header_maxsize. Previously a crafted message could
1302 induce a crash of the recive process; now the message is cleanly rejected.
1304 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1305 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1311 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1312 When redis_servers is set to a list of > 1 element, and the Redis servers
1313 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1314 case of MOVED into a DEFER case instead, thus moving the query onto the
1315 next server in the list. For a cluster of N elements, all N servers must
1316 be defined in redis_servers.
1318 GF/02 Catch and remove uninitialized value warning in exiqsumm
1319 Check for existence of @ARGV before looking at $ARGV[0]
1321 JH/01 Replace the store_release() internal interface with store_newblock(),
1322 which internalises the check required to safely use the old one, plus
1323 the allocate and data copy operations duplicated in both (!) of the
1324 extant use locations.
1326 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1327 modifier. This matches the restriction on the commandline.
1329 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1330 Previously only the last row was returned.
1332 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1333 we assumed that tags in the header were well-formed, and parsed the
1334 element content after inspecting only the first char of the tag.
1335 Assumptions at that stage could crash the receive process on malformed
1338 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1339 While running the DKIM ACL we operate on the Permanent memory pool so that
1340 variables created with "set" persist to the DATA ACL. Also (at any time)
1341 DNS lookups that fail create cache records using the Permanent pool. But
1342 expansions release any allocations made on the current pool - so a dnsdb
1343 lookup expansion done in the DKIM ACL releases the memory used for the
1344 DNS negative-cache, and bad things result. Solution is to switch to the
1345 Main pool for expansions.
1346 While we're in that code, add checks on the DNS cache during store_reset,
1347 active in the testsuite.
1348 Problem spotted, and debugging aided, by Wolfgang Breyha.
1350 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1351 When none of the hosts presented to a transport match an already-open
1352 connection, close it and proceed with the list. Previously we would
1353 queue the message. Spotted by Lena with Yahoo, probably involving
1356 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1357 Previously a spurious "250 OK id=" response was appended to the proper
1360 JH/08 The "support for" informational output now, which built with Content
1361 Scanning support, has a line for the malware scanner interfaces compiled
1362 in. Interface can be individually included or not at build time.
1364 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1365 by the template makefile "src/EDITME". The "STREAM" support for an older
1366 ClamAV interface method is removed.
1368 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1369 rows affected is given instead).
1371 JH/11 The runtime Berkeley DB library version is now additionally output by
1372 "exim -d -bV". Previously only the compile-time version was shown.
1374 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1375 SMTP connection. Previously, when one had more recipients than the
1376 first, an abortive onward connection was made. Move to full support for
1377 multiple onward connections in sequence, handling cutthrough connection
1378 for all multi-message initiating connections.
1380 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1381 routers. Previously, a multi-recipient message would fail to match the
1382 onward-connection opened for the first recipient, and cause its closure.
1384 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1385 a timeout on read on a GnuTLS initiating connection, resulting in the
1386 initiating connection being dropped. This mattered most when the callout
1387 was marked defer_ok. Fix to keep the two timeout-detection methods
1390 JH/15 Relax results from ACL control request to enable cutthrough, in
1391 unsupported situations, from error to silently (except under debug)
1392 ignoring. This covers use with PRDR, frozen messages, queue-only and
1395 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1397 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1398 metadata, resulting in a crash in free().
1400 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1401 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1402 Broken also in d185889f4, with init system revamp.
1404 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1405 Previously we abruptly closed the connection after reading a malware-
1406 found indication; now we go on to read the "scan ok" response line,
1409 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1410 ACL. Previously, a crash would result.
1412 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1413 macro text after a replacement (previously it was only once per line) and
1414 by skipping builtin macros when searching for an uppercase lead character.
1416 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1417 for the build is renamed.
1419 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1420 was allocated for every new TLS startup, meaning one per message. Fix
1421 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1423 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1424 reported the original. Fix to report (as far as possible) the ACL
1425 result replacing the original.
1427 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1428 OpenSSL. Certificate information is loaded for every new TLS startup,
1429 and the resources needed to be freed.
1431 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1433 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1436 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1437 DATA response info to the (existing) per-recipient response info for
1438 the "C=" log element. It can have useful tracking info from the
1439 destination system. Patch from Simon Arlott.
1441 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1442 length value. Previously this would segfault.
1444 HS/02 Support Avast multiline protoocol, this allows passing flags to
1445 newer versions of the scanner.
1447 JH/28 Ensure that variables possibly set during message acceptance are marked
1448 dead before release of memory in the daemon loop. This stops complaints
1449 about them when the debug_store option is enabled. Discovered specifically
1450 for sender_rate_period, but applies to a whole set of variables.
1451 Do the same for the queue-runner and queue-list loops, for variables set
1452 from spool message files. Do the same for the SMTP per-message loop, for
1453 certain variables indirectly set in ACL operations.
1455 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1456 as a multi-recipient message from a mailinglist manager). The coding had
1457 an arbitrary cutoff number of characters while checking for more input;
1458 enforced by writing a NUL into the buffer. This corrupted long / fast
1459 input. The problem was exposed more widely when more pipelineing of SMTP
1460 responses was introduced, and one Exim system was feeding another.
1461 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1462 receiving system, and refused recipients seen by the sending system
1463 (propating to people being dropped from mailing lists).
1464 Discovered and pinpointed by David Carter.
1466 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1467 replaced by the ${authresults } expansion.
1469 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1471 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1472 allows proper process termination in container environments.
1474 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1475 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1477 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1478 and "err_perm", deprecated since 4.83 when the RFC-defined words
1479 "temperror" and "permerror" were introduced.
1481 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1482 transport-filters or DKIM-signing. The restriction was lost in the
1483 consolidation of verify-callout and delivery SMTP handling.
1484 Extend the restriction to also cover ARC-signing.
1486 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1487 in defer=pass mode supply a 450 to the initiator. Previously the message
1490 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1491 tls_require_ciphers is used as before.
1493 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1494 "pass_unscanned". Only tmpfails from the scanner are written to
1495 the paniclog, as they may require admin intervention (permission
1496 denied, license issues). Other scanner errors (like decompression
1497 bombs) do not cause a paniclog entry.
1499 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1500 Previously it was possible to log spurious information in receive log
1503 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1504 triggered odd behaviour from Outlook Express clients.
1506 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1509 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1510 since the IETF WG has not yet settled on that versus the original
1511 "bare" representation.
1513 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1514 Previously the millisecond value corrupted the output.
1515 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1516 corrupted the output.
1522 JH/01 Rework error string handling in TLS interface so that the caller in
1523 more cases is responsible for logging. This permits library-sourced
1524 string to be attached to addresses during delivery, and collapses
1525 pairs of long lines into single ones.
1527 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1528 during configuration. Wildcards are allowed and expanded.
1530 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1531 This permits better logging.
1533 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1534 do not close it down and have the child transport start it up again on
1535 the passed-on TCP connection. Instead, proxy the child (and any
1536 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1537 affected: the continued delivery log lines do not have any DNSSEC, TLS
1538 Certificate or OCSP information. TLS cipher information is still logged.
1540 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1541 identical IP addresses on different listening ports. Will also affect
1544 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1545 add noisy ifdef guards to special-case this sillyness.
1546 Patch from Bernd Kuhls.
1548 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1549 than 255 are no longer allowed.
1551 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1552 Disable the session-cache too, which might reduce our load. Since we
1553 currrectly use a new context for every connection, both as server and
1554 client, there is no benefit for these.
1555 GnuTLS appears to not support tickets server-side by default (we don't
1556 call gnutls_session_ticket_enable_server()) but client side is enabled
1557 by default on recent versions (3.1.3 +) unless the PFS priority string
1560 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1561 <https://reproducible-builds.org/specs/source-date-epoch/>.
1563 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1564 the check for any unsuccessful recipients did not notice the limit, and
1565 erroneously found still-pending ones.
1567 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1568 MSG_MORE. Only in-clear (not on TLS connections).
1570 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1571 transport-filter is involved we can buffer the headers in memory for
1572 creating the signature, and read the spool data file once for the
1573 signature and again for transmission.
1575 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1576 4.77 as the kernel support then wasn't solid, having issues in 64bit
1577 mode. Now, it's been long enough. Add support for FreeBSD also.
1579 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1580 case where the routing stage had gathered several addresses to send to
1581 a host before calling the transport for the first, we previously failed
1582 to close down TLS in the old transport process before passing the TCP
1583 connection to the new process. The new one sent a STARTTLS command
1584 which naturally failed, giving a failed delivery and bloating the retry
1585 database. Investigation and fix prototype from Wolfgang Breyha.
1587 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1588 false-negatives in the check that the sender had not preempted a response
1589 or prompt from Exim (running as a server), due to that code's lack of
1590 awareness of the SMTP input buffering.
1592 PP/04 Add commandline_checks_require_admin option.
1593 Exim drops privileges sanely, various checks such as -be aren't a
1594 security problem, as long as you trust local users with access to their
1595 own account. When invoked by services which pass untrusted data to
1596 Exim, this might be an issue. Set this option in main configuration
1597 AND make fixes to the calling application, such as using `--` to stop
1600 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1601 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1602 responses to those, into a single TLS record each way (this usually means
1603 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1606 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1607 affects you only if you're dancing at the edge of the param size limits.
1608 If you are, and this message makes sense to you, then: raise the
1609 configured limit or use OpenSSL 1.1. Nothing we can do for older
1612 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1613 cmdline element to get the documented default one. Previously it was
1616 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1617 are unsuable for non-message-specific reasons - eg. connection timeout,
1618 banner-time rejection.
1620 JH/16 Fix logging of delivery remote port, when specified by router, under
1623 PP/06 Repair manualroute's ability to take options in any order, even if one
1624 is the name of a transport.
1627 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1629 JH/17 Change the list-building routines interface to use the expanding-string
1630 triplet model, for better allocation and copying behaviour.
1632 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1633 Previously it was constructed the first time a possibly-matching string
1634 was met in the configuration file input during startup; now it is done
1637 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1638 compatible one, to avoid the (poorly documented) possibility of a config
1639 file in the working directory redirecting the DB files, possibly correpting
1640 some existing file. CVE-2017-10140 assigned for BDB.
1642 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1643 cache-hot. Previously, although the result was properly cached, the
1644 initial verify call returned a defer.
1646 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1647 the main verify for receipient in uncached-mode.
1649 JH/22 Retire historical build files to an "unsupported" subdir. These are
1650 defined as "ones for which we have no current evidence of testing".
1652 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1653 if present. Previously it was ignored.
1655 JH/24 Start using specified-initialisers in C structure init coding. This is
1656 a C99 feature (it's 2017, so now considered safe).
1658 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1659 if was a fixed-sized field and bitmask ops via macros; it is now more
1662 PP/07 GitHub PR 56: Apply MariaDB build fix.
1663 Patch provided by Jaroslav Škarvada.
1665 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1666 during Coverity cleanups [4.87 JH/47]
1667 Diagnosis and fix provided by Michael Fischer v. Mollard.
1669 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1670 the right size to place the terminating semicolon on its own folded
1671 line, the header hash was calculated to an incorrect value thanks to
1672 the (relaxed) space the fold became.
1674 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1675 and confused the parent.
1677 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1678 which could crash as a result. This could lead to undeliverable messages.
1680 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1683 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1684 matters on fast-turnover and PID-randomising systems, which were getting
1685 out-of-order delivery.
1687 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1688 a possibly-overlapping copy. The symptom was that "Remote host closed
1689 connection in response to HELO" was logged instead of the actual 4xx
1692 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1693 Previously only that bufferd was discarded, resulting in SYMTP command
1696 JH/32 DKIM: when a message has multiple signatures matching an identity given
1697 in dkim_verify_signers, run the dkim acl once for each. Previously only
1698 one run was done. Bug 2189.
1700 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1701 "panic the current process" to "deliberately defer". The panic log is
1702 still written with the problem list name; the mail and reject logs now
1703 get a temp-reject line for the message that was being handled, saying
1704 something like "domains check lookup or other defer". The SMTP 451
1705 message is still "Temporary local problem".
1707 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1708 A crafted sequence of BDAT commands could result in in-use memory beeing
1709 freed. CVE-2017-16943.
1711 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1712 from SMTP input. Previously it was always done; now only done for DATA
1713 and not BDAT commands. CVE-2017-16944.
1715 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1716 to the message (such as an overlong header line). Previously this was
1717 not done and we did not exit BDAT mode. Followon from the previous item
1718 though a different problem.
1724 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1725 than -2003 did; needs libidn2 in addition to libidn.
1727 JH/02 The path option on a pipe transport is now expanded before use.
1729 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1730 Patch provided by "Björn", documentation fix added too.
1732 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1733 missing a wire-to-host endian conversion.
1735 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1736 close after a BDAT command line could be taken as a following command,
1737 giving a synch failure. Fix by only checking for synch immediately
1738 before acknowledging the chunk.
1740 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1741 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1742 macro. Patches provided by Josh Soref.
1744 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1745 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1746 by IANA as a service extension. However, John Klensin suggests that we
1749 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1750 the dkim code may be unix-mode line endings rather than smtp wire-format
1751 CRLF, so prepend a CR to any bare LF.
1753 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1754 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1756 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1757 the first were themselves being wrongly included in the feed into dkim
1758 processing; with most chunk sizes in use this resulted in an incorrect
1759 body hash calculated value.
1761 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1762 DKIM signature block, for verification. Although advised against by
1763 standards it is specifically not ruled illegal.
1765 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1767 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1768 missing a body hash (the bh= tag).
1770 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1771 It seems that HAProxy sends the Proxy Protocol information in clear and
1772 only then does a TLS startup, so do the same.
1774 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1775 TCP connections (such as for Spamd) unless the daemon successfully set
1776 Fast Open mode on its listening sockets. This fixes breakage seen on
1777 too-old kernels or those not configured for Fast Open, at the cost of
1778 requiring both directions being enabled for TFO, and TFO never being used
1779 by non-daemon-related Exim processes.
1781 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1782 endings, at least on the first header line. Try to canonify any that get
1783 past that check, despite the cost.
1785 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1786 now limited to an arbitrary five deep, while parsing addresses with the
1787 strip_excess_angle_brackets option enabled.
1789 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1790 instead leave the unprompted TLS handshake in socket buffer for the
1791 TLS library to consume.
1793 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1795 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1797 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1798 region in one operation, for speed, and this leaves assigned pointers
1799 dangling. Add checks run only under the testsuite which checks all
1800 variables at a store-reset and panics on a dangling pointer; add code
1801 explicitly nulling out all the variables discovered. Fixes one known
1802 bug: a transport crash, where a dangling pointer for $sending_ip_address
1803 originally assigned in a verify callout, is re-used.
1805 PP/06 Drop '.' from @INC in various Perl scripts.
1807 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1809 PP/08 Reduce a number of compilation warnings under clang; building with
1810 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1811 should be warning-free.
1813 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1815 HS/01 Fix portability problems introduced by PP/08 for platforms where
1816 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1818 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1819 chunk. This allows us to accept broken chunked messages. We need a more
1820 general solution here.
1822 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1823 already-broken messages in the queue.
1825 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1827 JH/19 Fix reference counting bug in routing-generated-address tracking.
1833 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1834 supports it and a size is available (ie. the sending peer gave us one).
1836 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1837 years of being deprecated). The replacements are the ACLs
1838 acl_smtp_mime and acl_not_smtp_mime.
1840 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1841 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1842 or even an in-clear connection were permitted. Now, if the host lookup
1843 was dnssec and dane was requested then the host is only used if the
1844 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1845 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1846 if one fails this test.
1847 This means that a poorly-configured remote DNS will make it incommunicado;
1848 but it protects against a DNS-interception attack on it.
1850 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1851 non-noisy when a race steals the message being considered.
1853 JH/05 If main configuration option tls_certificate is unset, generate a
1854 self-signed certificate for inbound TLS connections.
1856 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1857 in rewrites and routers.
1859 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1860 and logged a warning sing 4.83; now they are a configuration file error.
1862 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1863 (lacking @domain). Apply the same qualification processing as RCPT.
1865 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1867 JH/10 Support ${sha256:} applied to a string (as well as the previous
1870 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1871 a cutthrough deliver is pending, as we always want to make a connection.
1872 This also avoids re-routing the message when later placing the cutthrough
1873 connection after a verify cache hit.
1874 Do not update it with the verify result either.
1876 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1877 when routing results in more than one destination address.
1879 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1880 signing (which inhibits the cutthrough capability). Previously only
1881 the presence of an option was tested; now an expansion evaluating as
1882 empty is permissible (obviously it should depend only on data available
1883 when the cutthrough connection is made).
1885 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1886 the relevant preceding SMTP command did not note the pipelining mode.
1888 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1889 Previously they were not counted.
1891 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1892 as one having no matching records. Previously we deferred the message
1893 that needed the lookup.
1895 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1896 distinguished as "(=".
1898 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1899 for missing MX records. Previously it only worked for missing A records.
1901 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1903 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1904 after the data-go-ahead and data-ack. Patch from Jason Betts.
1906 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1907 even for a "none" policy. Patch from Tony Meyer.
1909 JH/22 Fix continued use of a connection for further deliveries. If a port was
1910 specified by a router, it must also match for the delivery to be
1913 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1914 When one of the recipients of a message was unsuitable for the connection
1915 (has no matching addresses), we lost track of needing to mark it
1916 deferred. As a result mail would be lost.
1918 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1920 JH/25 Decoding ACL controls is now done using a binary search; the source code
1921 takes up less space and should be simpler to maintain. Merge the ACL
1922 condition decode tables also, with similar effect.
1924 JH/26 Fix problem with one_time used on a redirect router which returned the
1925 parent address unchanged. A retry would see the parent address marked as
1926 delivered, so not attempt the (identical) child. As a result mail would
1929 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1930 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1931 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1934 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1935 Main config options check_{log,spool}_{inodes,space} are now
1936 100 inodes, 10MB unless set otherwise in the configuration.
1938 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1939 Previously it only applied to the main-section connection policy
1942 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1944 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1945 by me. Added RFC7919 DH primes as an alternative.
1947 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1948 are not in the system include path.
1950 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1951 GnuTLS, when a session startup failed (eg because the client disconnected)
1952 Exim did stdio operations after fclose. This was exposed by a recent
1953 change which nulled out the file handle after the fclose.
1955 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1956 signed directly by the cert-signing cert, rather than an intermediate
1957 OCSP-signing cert. This is the model used by LetsEncrypt.
1959 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1961 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1962 an incoming connection.
1964 HS/02 Bug 1802: Do not half-close the connection after sending a request
1967 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1968 fallback to "prime256v1".
1970 JH/34 SECURITY: Use proper copy of DATA command in error message.
1971 Could leak key material. Remotely exploitable. CVE-2016-9963.
1977 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1978 and 3.4.4 - once the server is enabled to respond to an OCSP request
1979 it does even when not requested, resulting in a stapling non-aware
1980 client dropping the TLS connection.
1982 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1983 support variable-length bit vectors. No functional change.
1985 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1986 The I= interface field on outgoing lines is now after the H= remote
1987 host field, same as incoming lines. There is a separate
1988 outgoing_interface log selector which allows you to disable the
1991 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1992 If not running log_selector +smtp_connection the mainlog would be held
1993 open indefinitely after a "too many connections" event, including to a
1994 deleted file after a log rotate. Leave the per net connection logging
1995 leaving it open for efficiency as that will be quickly detected by the
1996 check on the next write.
1998 HS/01 Bug 1671: Fix post transport crash.
1999 Processing the wait-<transport> messages could crash the delivery
2000 process if the message IDs didn't exist for some reason. When
2001 using 'split_spool_directory=yes' the construction of the spool
2002 file name failed already, exposing the same netto behaviour.
2004 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2005 mime_regex ACL conditions.
2007 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2008 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2009 to HELO, local diagnostic string.
2011 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2012 log line to debug. Even when configured with a tls authenticator many
2013 client connections are expected to not authenticate in this way, so
2014 an authenticate fail is not an error.
2016 HS/02 Add the Exim version string to the process info. This way exiwhat
2017 gives some more detail about the running daemon.
2019 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2020 matter for fast-change records such as DNSBLs.
2022 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2023 retry record, even if constant. There may be multiple transports with
2024 different interface settings and the retry behaviour needs to be kept
2027 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2029 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2031 JH/10 Bug 840: fix log_defer_output option of pipe transport
2033 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2034 hosts_randomize. This matters a lot when combined with mua_wrapper.
2036 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2037 ${quote_pgsql:<string>} operator.
2039 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2041 JH/14 Change header file name for freeradius-client. Relevant if compiling
2042 with Radius support; from the Gentoo tree and checked under Fedora.
2044 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2046 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2047 option transport_filter, meaning no filtering.
2049 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2051 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2052 defaults to "*" (all hosts). The variable is now available when not built
2053 with TLS, default unset, mainly to enable keeping the testsuite sane.
2054 If a server certificate is not supplied (via tls_certificate) an error is
2055 logged, and clients will find TLS connections fail on startup. Presumably
2056 they will retry in-clear.
2057 Packagers of Exim are strongly encouraged to create a server certificate
2058 at installation time.
2060 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2061 with the $config_file variable.
2063 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2064 in transport context, after the attempt, and per-recipient. The latter type
2065 is per host attempted. The event data is the error message, and the errno
2066 information encodes the lookup type (A vs. MX) used for the (first) host,
2067 and the trailing two digits of the smtp 4xx response.
2069 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2070 to write to mainlog (or rejectlog, paniclog) in the window between file
2071 creation and permissions/ownership being changed. Particularly affects
2072 installations where exicyclog is run as root, rather than exim user;
2073 result is that the running daemon panics and dies.
2075 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2077 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2078 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2079 "pri" and "weight". Note that the previous implicit priority given by the
2080 list order is no longer honoured.
2082 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2083 for DKIM processing.
2085 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2086 by defining SUPPORT_SOCKS.
2088 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2089 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2090 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2091 variables are renamed to proxy_{local,external}_{address,port}.
2093 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2094 for a build by defining SUPPORT_I18N
2096 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2097 of the query string, and make ${quote_redis:} do that quoting.
2099 JH/29 Move Events support from Experimental to mainline, enabled by default
2100 and removable for a build by defining DISABLE_EVENT.
2102 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2104 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2105 cached by the daemon.
2107 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2108 by defining LOOKUP_REDIS. The libhiredis library is required.
2110 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2111 keys are given for lookup.
2113 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2114 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2115 only supported when built with TLS support. The PolarSSL SHA routines
2116 are still used when the TLS library is too old for convenient support.
2118 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2119 openssl_options), for security. OpenSSL forces this from version 1.1.0
2120 server-side so match that on older versions.
2122 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2123 allocation for $value could be released as the expansion processing
2124 concluded, but leaving the global pointer active for it.
2126 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2127 and to use the domains and local_parts ACL conditions.
2129 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2130 incorrectly not doubled on cutthrough transmission, hence seen as a
2131 body-termination at the receiving system - resulting in truncated mails.
2132 Commonly the sender saw a TCP-level error, and retransmitted the message
2133 via the normal store-and-forward channel. This could result in duplicates
2134 received - but deduplicating mailstores were liable to retain only the
2135 initial truncated version.
2137 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2139 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2141 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2142 we're in there, support oversigning also; bug 1309.
2144 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2146 HS/04 Add support for keep_environment and add_environment options.
2148 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2149 either intentional arithmetic overflow during PRNG, or testing config-
2152 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2153 delivery resulted in actual delivery. Cancel cutthrough before DATA
2156 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2157 rejects a recipient: pass the reject to the originator.
2159 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2160 Many were false-positives and ignorable, but it's worth fixing the
2163 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2164 for the new environment-manipulation done at startup. Move the routines
2165 from being local to tls.c to being global via the os.c file.
2167 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2168 an extract embedded as result-arg for a map, the first arg for extract
2169 is unavailable so we cannot tell if this is a numbered or keyed
2170 extraction. Accept either.
2176 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2179 JH/02 The smtp transport option "multi_domain" is now expanded.
2181 JH/03 The smtp transport now requests PRDR by default, if the server offers
2184 JH/04 Certificate name checking on server certificates, when exim is a client,
2185 is now done by default. The transport option tls_verify_cert_hostnames
2186 can be used to disable this per-host. The build option
2187 EXPERIMENTAL_CERTNAMES is withdrawn.
2189 JH/05 The value of the tls_verify_certificates smtp transport and main options
2190 default to the word "system" to access the system default CA bundle.
2191 For GnuTLS, only version 3.0.20 or later.
2193 JH/06 Verification of the server certificate for a TLS connection is now tried
2194 (but not required) by default. The verification status is now logged by
2195 default, for both outbound TLS and client-certificate supplying inbound
2198 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2201 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2202 Status Notification (bounce) messages are now MIME format per RFC 3464.
2203 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2204 under the control of the dsn_advertise_hosts option, and routers may
2205 have a dsn_lasthop option.
2207 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2208 default, modifiable by a malware= option. The list separator for
2209 the options can now be changed in the usual way. Bug 68.
2211 JH/10 The smtp_receive_timeout main option is now expanded before use.
2213 JH/11 The incoming_interface log option now also enables logging of the
2214 local interface on delivery outgoing connections.
2216 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2217 if the interface and destination host and port all match.
2219 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2222 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2223 Patch from Andrew Lewis.
2225 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2226 now supports optional time-restrictions, weighting, and priority
2227 modifiers per server. Patch originally by <rommer@active.by>.
2229 JH/16 The spamd_address main option now supports a mixed list of local
2230 and remote servers. Remote servers can be IPv6 addresses, and
2231 specify a port-range.
2233 JH/17 Bug 68: The spamd_address main option now supports an optional
2234 timeout value per server.
2236 JH/18 Bug 1581: Router and transport options headers_add/remove can
2237 now have the list separator specified.
2239 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2242 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2245 JH/21 Support for the A6 type of dns record is withdrawn.
2247 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2248 rather than the verbs used.
2250 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2251 from 255 to 1024 chars.
2253 JH/24 Verification callouts now attempt to use TLS by default.
2255 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2256 are generic router options now. The defaults didn't change.
2258 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2259 Original patch from Alexander Shikoff, worked over by JH.
2261 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2262 files if ZCAT_COMMAND is not executable.
2264 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2266 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2268 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2269 Normally benign, it bites when the pair was led to by a CNAME;
2270 modern usage is to not canonicalize the domain to a CNAME target
2271 (and we were inconsistent anyway for A-only vs AAAA+A).
2273 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2275 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2276 when evaluating $sender_host_dnssec.
2278 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2279 $sender_helo_dnssec variable.
2281 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2283 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2285 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2287 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2288 documented as working, but never had. Support all but $spam_report.
2290 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2291 added for tls authenticator.
2293 HS/03 Add perl_taintmode main config option
2299 TL/01 When running the test suite, the README says that variables such as
2300 no_msglog_check are global and can be placed anywhere in a specific
2301 test's script, however it was observed that placement needed to be near
2302 the beginning for it to behave that way. Changed the runtest perl
2303 script to read through the entire script once to detect and set these
2304 variables, reset to the beginning of the script, and then run through
2305 the script parsing/test process like normal.
2307 TL/02 The BSD's have an arc4random API. One of the functions to induce
2308 adding randomness was arc4random_stir(), but it has been removed in
2309 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2310 function when detected.
2312 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2313 cause callback expansion.
2315 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2316 syntax errors in an expansion can be treated as a string instead of
2317 logging or causing an error, due to the internal use of bool_lax
2318 instead of bool when processing it.
2320 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2321 server certificates when making smtp deliveries.
2323 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2325 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2327 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2329 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2330 Merged patch from Sebastian Wiedenroth.
2332 JH/05 Fix results-pipe from transport process. Several recipients, combined
2333 with certificate use, exposed issues where response data items split
2334 over buffer boundaries were not parsed properly. This eventually
2335 resulted in duplicates being sent. This issue only became common enough
2336 to notice due to the introduction of connection certificate information,
2337 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2339 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2340 size buffer was used, resulting in syntax errors when an expansion
2343 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2344 version 3.3.6 or later.
2346 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2347 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2348 both become "event_action", the variables become $event_name, $event_data
2349 and $event_defer_errno. There is a new variable $verify_mode, usable in
2350 routers, transports and related events. The tls:cert event is now also
2351 raised for inbound connections, if the main configuration event_action
2354 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2355 early OCSP support, but no stapling (appears to be less than 1.0.0).
2357 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2358 server certificate names available under the smtp transport option
2359 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2362 JH/10 Time-related extraction expansions from certificates now use the main
2363 option "timezone" setting for output formatting, and are consistent
2364 between OpenSSL and GnuTLS compilations. Bug 1541.
2366 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2367 encoded parameter in the incoming message. Bug 1558.
2369 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2370 include certificate info, eximon was claiming there were spoolfile
2373 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2375 JH/14 Log delivery-related information more consistently, using the sequence
2376 "H=<name> [<ip>]" wherever possible.
2378 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2379 are problematic for Debian distribution, omit them from the release
2382 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2384 JH/16 Fix string representation of time values on 64bit time_t architectures.
2387 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2388 output list separator was used.
2393 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2394 checkers that were complaining about end of non-void function with no
2397 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2398 This was a regression introduced in 4.83 by another bugfix.
2400 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2402 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2403 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2409 TF/01 Correctly close the server side of TLS when forking for delivery.
2411 When a message was received over SMTP with TLS, Exim failed to clear up
2412 the incoming connection properly after forking off the child process to
2413 deliver the message. In some situations the subsequent outgoing
2414 delivery connection happened to have the same fd number as the incoming
2415 connection previously had. Exim would try to use TLS and fail, logging
2416 a "Bad file descriptor" error.
2418 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2419 utilities have not been installed.
2421 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2422 temporary space as the ACL may create new global variables.
2424 TL/01 LDAP support uses per connection or global context settings, depending
2425 upon the detected version of the libraries at build time.
2427 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2428 to extract and use the src ip:port in logging and expansions as if it
2429 were a direct connection from the outside internet. PPv2 support was
2430 updated based on HAProxy spec change in May 2014.
2432 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2434 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2435 Properly escape header and check for NULL return.
2437 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2440 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2442 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2443 characters in header names, implemented as a verify condition.
2444 Contributed by Michael Fischer v. Mollard.
2446 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2447 results permerror and temperror. Previous values are deprecated but
2448 still accepted. In a future release, err_perm and err_temp will be
2449 completely removed, which will be a backward incompatibility if the
2450 ACL tests for either of these two old results. Patch contributed by
2451 user bes-internal on the mailing list.
2453 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2455 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2456 selectors, in both main and reject logs.
2458 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2461 JH/07 Add malware type "sock" for talking to simple daemon.
2463 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2465 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2466 routers/transports under cutthrough routing.
2468 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2469 numbers. Touch up "bool" conditional to keep the same definition.
2471 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2473 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2475 JH/12 Expand items in router/transport headers_add or headers_remove lists
2476 individually rather than the list as a whole. Bug 1452.
2478 Required for reasonable handling of multiple headers_ options when
2479 they may be empty; requires that headers_remove items with embedded
2480 colons must have them doubled (or the list-separator changed).
2482 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2483 view the policy declared in the DMARC record. Currently, $dmarc_status
2484 is a combined value of both the record presence and the result of the
2487 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2489 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2490 dnslookup router and the smtp transport (applying to the forward
2493 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2494 of ldap servers used for a specific lookup. Patch provided by Heiko
2497 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2498 New variable $lookup_dnssec_authenticated for observability.
2500 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2501 Patch submitted by Lars Timman.
2503 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2505 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2506 Requires trusted mode and valid format message id, aborts otherwise.
2507 Patch contributed by Heiko Schlichting.
2509 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2510 certextract with support for various fields. Bug 1358.
2512 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2513 is requested by default, modifiable by smtp transport option
2516 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2517 operate on certificate variables to give certificate fingerprints
2518 Also new ${sha256:cert_variable}.
2520 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2522 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2523 Christian Aistleitner.
2525 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2527 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2528 file. Patch from Wolfgang Breyha.
2530 JH/25 Expand the coverage of the delivery $host and $host_address to
2531 client authenticators run in verify callout. Bug 1476.
2533 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2534 align with daemon_smtp_ports. Bug 72.
2536 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2537 support and error reporting did not work properly.
2539 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2540 and is readable. Patch from Andrew Colin Kissa.
2542 TL/14 Enhance documentation of ${run expansion and how it parses the
2543 commandline after expansion, particularly in the case when an
2544 unquoted variable expansion results in an empty value.
2546 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2548 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2549 by renaming away from it. Addresses GH issue 15, reported by
2552 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2554 TL/15 SECURITY: prevent double expansion in math comparison functions
2555 (can expand unsanitized data). Not remotely exploitable.
2562 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2564 PP/02 Make -n do something, by making it not do something.
2565 When combined with -bP, the name of an option is not output.
2567 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2570 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2571 $sender_host_name and config options to manage this, and basic check
2574 PP/05 DSCP support for outbound connections and control modifier for inbound.
2576 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2577 (Only plugin which currently uses this is kerberos4, which nobody should
2578 be using, but we should make it available and other future plugins might
2579 conceivably use it, even though it would break NAT; stuff *should* be
2580 using channel bindings instead).
2582 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2583 name; added for Sendmail compatibility; requires admin caller.
2584 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2585 just ignore it); requires trusted caller.
2586 Also parse but ignore: -Ac -Am -X<logfile>
2589 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2591 TL/02 Add +smtp_confirmation as a default logging option.
2593 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2594 Patch by Magnus Holmgren from 2007-02-20.
2596 TL/04 Bugzilla 1281 - Spec typo.
2597 Bugzilla 1283 - Spec typo.
2598 Bugzilla 1290 - Spec grammar fixes.
2600 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2602 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2604 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2605 mailing list by by Dmitry Isaikin.
2607 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2609 JH/02 Support "G" suffix to numbers in ${if comparisons.
2611 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2613 NM/01 Bugzilla 1197 - Spec typo
2614 Bugzilla 1196 - Spec examples corrections
2616 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2618 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2619 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2622 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2623 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2625 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2626 "acl {{name}{arg}...}", and optional args on acl condition
2629 JH/05 Permit multiple router/transport headers_add/remove lines.
2631 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2633 JH/07 Avoid using a waiting database for a single-message-only transport.
2634 Performance patch from Paul Fisher. Bugzilla 1262.
2636 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2639 JH/09 Add $headers_added variable, with content from use of ACL modifier
2640 add_header (but not yet added to the message). Bugzilla 199.
2642 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2643 Pulled from Bugzilla 817 by Wolfgang Breyha.
2645 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2647 (nb: this is the same fix as in Exim 4.80.1)
2649 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2652 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2655 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2657 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2658 advertise SMTP AUTH mechanism to us, instead of a generic
2659 protocol violation error. Also, make Exim more robust to bad
2660 data from the Dovecot auth socket.
2662 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2664 When a queue runner is handling a message, Exim first routes the
2665 recipient addresses, during which it prunes them based on the retry
2666 hints database. After that it attempts to deliver the message to
2667 any remaining recipients. It then updates the hints database using
2670 So if a recipient address works intermittently, it can get repeatedly
2671 deferred at routing time. The retry hints record remains fresh so the
2672 address never reaches the final cutoff time.
2674 This is a fairly common occurrence when a user is bumping up against
2675 their storage quota. Exim had some logic in its local delivery code
2676 to deal with this. However it did not apply to per-recipient defers
2677 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2679 This change adds a proper retry rule check during routing so that the
2680 final cutoff time is checked against the message's age. We only do
2681 this check if there is an address retry record and there is not a
2682 domain retry record; this implies that previous attempts to handle
2683 the address had the retry_use_local_parts option turned on. We use
2684 this as an approximation for the destination being like a local
2685 delivery, as in LMTP.
2687 I suspect this new check makes the old local delivery cutoff check
2688 redundant, but I have not verified this so I left the code in place.
2690 TF/02 Correct gecos expansion when From: is a prefix of the username.
2692 Test 0254 submits a message to Exim with the header
2696 When I ran the test suite under the user fanf2, Exim expanded
2697 the header to contain my full name, whereas it should have added
2698 a Resent-Sender: header. It erroneously treats any prefix of the
2699 username as equal to the username.
2701 This change corrects that bug.
2703 GF/01 DCC debug and logging tidyup
2704 Error conditions log to paniclog rather than rejectlog.
2705 Debug lines prefixed by "DCC: " to remove any ambiguity.
2707 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2709 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2710 Bug spotted by Jeremy Harris; was flawed since initial commit.
2711 Would have resulted in OCSP responses post-SNI triggering an Exim
2712 NULL dereference and crash.
2714 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2716 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2717 Bug detection, analysis and fix by Samuel Thibault.
2718 Bugzilla 1331, Debian bug #698092.
2720 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2722 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2723 Server implementation by Todd Lyons, client by JH.
2724 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2725 config variable "prdr_enable" controls whether the server
2726 advertises the facility. If the client requests PRDR a new
2727 acl_data_smtp_prdr ACL is called once for each recipient, after
2728 the body content is received and before the acl_smtp_data ACL.
2729 The client is controlled by both of: a hosts_try_prdr option
2730 on the smtp transport, and the server advertisement.
2731 Default client logging of deliveries and rejections involving
2732 PRDR are flagged with the string "PRDR".
2734 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2735 fclose(). Diagnosis by Todd Lyons.
2737 PP/17 Update configure.default to handle IPv6 localhost better.
2738 Patch by Alain Williams (plus minor tweaks).
2741 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2742 This is now consistent with GnuTLS, and is now documented: the
2743 previous undocumented portable approach to treating the option as
2744 unset was to force an expansion failure. That still works, and
2745 an empty string is now equivalent.
2747 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2748 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2749 not performing validation itself.
2751 PP/20 Added force_command boolean option to pipe transport.
2752 Patch from Nick Koston, of cPanel Inc.
2754 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2757 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2759 PP/21 Fix eximon continuous updating with timestamped log-files.
2760 Broken in a format-string cleanup in 4.80, missed when I repaired the
2761 other false fix of the same issue.
2762 Report and fix from Heiko Schlichting.
2765 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2766 Report from Prashanth Katuri.
2768 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2769 It's SecureTransport, so affects any MacOS clients which use the
2770 system-integrated TLS libraries, including email clients.
2772 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2773 using a MIME ACL for non-SMTP local injection.
2774 Report and assistance in diagnosis by Warren Baker.
2776 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2778 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2780 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2781 last id that failed so it may be referenced in subsequent ACL's.
2783 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2786 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2787 ldap library initialization, allowing self-signed CA's to be
2788 used. Also properly sets require_cert option later in code by
2789 using NULL (global ldap config) instead of ldap handle (per
2790 session). Bug diagnosis and testing by alxgomz.
2792 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2793 the src/util/ subdirectory.
2795 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2796 renamed to Transport Post Delivery Action by Jeremy Harris, as
2799 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2800 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2801 redis_servers = needs to be configured which will be used by the redis
2802 lookup. Patch from Warren Baker, of The Packet Hub.
2804 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2806 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2807 hostname or reverse DNS when processing a host list. Used suggestions
2808 from multiple comments on this bug.
2810 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2812 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2813 Missed a few lines, added it to make the runtest require no keyboard
2816 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2817 contains upper case chars. Make router use caseful_local_part.
2819 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2820 support when GnuTLS has been built with p11-kit.
2826 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2828 This, or similar/improved, will also be change PP/11 of 4.82.
2834 PP/01 Handle short writes when writing local log-files.
2835 In practice, only affects FreeBSD (8 onwards).
2836 Bugzilla 1053, with thanks to Dmitry Isaikin.
2838 NM/01 Bugzilla 949 - Documentation tweak
2840 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2843 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2845 PP/02 Implemented gsasl authenticator.
2847 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2849 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2850 `pkg-config foo` for cflags/libs.
2852 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2853 with rest of GSASL and with heimdal_gssapi.
2855 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2856 `pkg-config foo` for cflags/libs for the TLS implementation.
2858 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2859 properties get this fed in as external SSF. A number of robustness
2860 and debugging improvements to the cyrus_sasl authenticator.
2862 PP/08 cyrus_sasl server now expands the server_realm option.
2864 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2865 Patch by Jeremy Harris.
2867 PP/10 Added dbmjz lookup type.
2869 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2871 PP/12 MAIL args handles TAB as well as SP, for better interop with
2872 non-compliant senders.
2873 Analysis and variant patch by Todd Lyons.
2875 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2876 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2877 Patch from Dirk Mueller <dmueller@suse.com>
2879 PP/13 tls_peerdn now print-escaped for spool files.
2880 Observed some $tls_peerdn in wild which contained \n, which resulted
2881 in spool file corruption.
2883 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2884 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2885 or write after TLS renegotiation, which otherwise led to messages
2888 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2889 as a tracking header (ie: a signed header comes before the signature).
2890 Patch from Wolfgang Breyha.
2892 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2893 comma-sep list; embedded commas doubled.
2895 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2897 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2899 Report and patch from Dmitry Banschikov.
2901 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2902 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2903 failures. We appear to now support TLS1.1+ with Exim.
2905 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2906 lets Exim select keys and certificates based upon TLS SNI from client.
2907 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2908 before an outbound SMTP session. New log_selector, +tls_sni.
2910 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2911 NULL dereference. Report and patch from Alun Jones.
2913 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2914 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2915 Not seeing resolver debug output on NetBSD, but suspect this is a
2916 resolver implementation change.
2918 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2919 Left warnings. Added "eximon gdb" invocation mode.
2921 PP/21 Defaulting "accept_8bitmime" to true, not false.
2923 PP/22 Added -bw for inetd wait mode support.
2925 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2926 locate the relevant includes and libraries. Made this the default.
2928 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2929 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2931 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2932 This may cause build issues on older platforms.
2934 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2935 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2936 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2937 Added SNI support via GnuTLS too.
2938 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2940 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2942 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2943 Applied second patch from Janne, implementing suggestion to default
2944 multiple-strings-in-record handling to match SPF spec.
2946 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2948 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2949 read-only, out of scope).
2950 Patch from Wolfgang Breyha, report from Stuart Northfield.
2952 PP/29 Fix three issues highlighted by clang analyser static analysis.
2953 Only crash-plausible issue would require the Cambridge-specific
2954 iplookup router and a misconfiguration.
2955 Report from Marcin Mirosław.
2957 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2959 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2960 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2961 As part of this, removing so much warning spew let me fix some minor
2962 real issues in debug logging.
2964 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2965 assignment on my part. Fixed.
2967 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2968 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2969 Janne Snabb (who went above and beyond: thank you).
2971 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2972 string otherwise requires a connection and a bunch more work and it's
2973 relatively easy to get wrong. Should also expose TLS library linkage
2976 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2977 64-bit ${eval} (JH/03).
2979 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2980 GNU libc to support some of the 64-bit stuff, should not lead to
2981 conflicts. Defined before os.h is pulled in, so if a given platform
2982 needs to override this, it can.
2984 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2985 protection layer was required, which is not implemented.
2986 Bugzilla 1254, patch from Wolfgang Breyha.
2988 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2989 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2990 tls_dhparam take prime identifiers. Also unbreak combination of
2991 OpenSSL+DH_params+TLSSNI.
2993 PP/39 Disable SSLv2 by default in OpenSSL support.
2999 PP/01 Solaris build fix for Oracle's LDAP libraries.
3000 Bugzilla 1109, patch from Stephen Usher.
3002 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3004 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3007 TF/02 Fix a couple more cases where we did not log the error message
3008 when unlink() failed. See also change 4.74-TF/03.
3010 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3011 lock up or crash if it happened to be inside a call to libc when it
3012 got a SIGUSR1 from exiwhat.
3014 The SIGUSR1 handler appends the current process status to the process
3015 log which is later printed by exiwhat. It used to use the general
3016 purpose logging code to do this, but several functions it calls are
3017 not safe for signals.
3019 The new output code in the SIGUSR1 handler is specific to the process
3020 log, and simple enough that it's easy to inspect for signal safety.
3021 Removing some special cases also simplifies the general logging code.
3022 Removing the spurious timestamps from the process log simplifies
3025 TF/04 Improved ratelimit ACL condition.
3027 The /noupdate option has been deprecated in favour of /readonly which
3028 has clearer semantics. The /leaky, /strict, and /readonly update modes
3029 are mutually exclusive. The update mode is no longer included in the
3030 database key; it just determines when the database is updated. (This
3031 means that when you upgrade Exim will forget old rate measurements.)
3033 Exim now checks that the per_* options are used with an update mode that
3034 makes sense for the current ACL. For example, when Exim is processing a
3035 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3036 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3037 must specify per_mail/readonly. If you omit the update mode it defaults to
3038 /leaky where that makes sense (as before) or /readonly where required.
3040 The /noupdate option is now undocumented but still supported for
3041 backwards compatibility. It is equivalent to /readonly except that in
3042 ACLs where /readonly is required you may specify /leaky/noupdate or
3043 /strict/noupdate which are treated the same as /readonly.
3045 A useful new feature is the /count= option. This is a generalization
3046 of the per_byte option, so that you can measure the throughput of other
3047 aggregate values. For example, the per_byte option is now equivalent
3048 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3050 The per_rcpt option has been generalized using the /count= mechanism
3051 (though it's more complicated than the per_byte equivalence). When it is
3052 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3053 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3054 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3055 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3056 Note that using per_rcpt with a non-readonly update mode in more than
3057 one ACL will cause the recipients to be double-counted. (The per_mail
3058 and per_byte options don't have this problem.)
3060 The handling of very low rates has changed slightly. If the computed rate
3061 is less than the event's count (usually one) then this event is the first
3062 after a long gap. In this case the rate is set to the same as this event's
3063 count, so that the first message of a spam run is counted properly.
3065 The major new feature is a mechanism for counting the rate of unique
3066 events. The new per_addr option counts the number of different
3067 recipients that someone has sent messages to in the last time period. It
3068 behaves like per_rcpt if all the recipient addresses are different, but
3069 duplicate recipient addresses do not increase the measured rate. Like
3070 the /count= option this is a general mechanism, so the per_addr option
3071 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3072 example, measure the rate that a client uses different sender addresses
3073 with the options per_mail/unique=$sender_address. There are further
3074 details in the main documentation.
3076 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3078 TF/06 Removed a few PCRE remnants.
3080 TF/07 Automatically extract Exim's version number from tags in the git
3081 repository when doing development or release builds.
3083 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3084 Bugzilla 879. Patch from Paul Fisher.
3086 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3087 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3090 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3092 PP/05 Variable $av_failed, true if the AV scanner deferred.
3093 Bugzilla 1078. Patch from John Horne.
3095 PP/06 Stop make process more reliably on build failure.
3096 Bugzilla 1087. Patch from Heiko Schlittermann.
3098 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3099 Bugzilla 1089. Patch from Heiko Schlittermann.
3101 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3102 Bugzilla 1131. Patch from Holger Weiß.
3104 PP/09 Handle IPv6 addresses with SPF.
3105 Bugzilla 860. Patch from Wolfgang Breyha.
3107 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3109 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3112 PP/11 match_* no longer expand right-hand-side by default.
3113 New compile-time build option, EXPAND_LISTMATCH_RHS.
3114 New expansion conditions, "inlist", "inlisti".
3116 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3118 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3120 PP/14 fix log_write() format string regression from TF/03.
3121 Bugzilla 1152. Patch from Dmitry Isaikin.
3127 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3129 PP/02 Harmonised TLS library version reporting; only show if debugging.
3130 Layout now matches that introduced for other libraries in 4.74 PP/03.
3132 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3134 PP/04 New "dns_use_edns0" global option.
3136 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3139 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3140 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3142 TK/01 Updated PolarSSL code to 0.14.2.
3143 Bugzilla 1097. Patch from Andreas Metzler.
3145 PP/07 Catch divide-by-zero in ${eval:...}.
3146 Fixes bugzilla 1102.
3148 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3151 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3152 format-string attack -- SECURITY: remote arbitrary code execution.
3154 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3155 time unintentionally subject to list matching rules, letting the header
3156 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3157 arbitrary string expansion). This allowed for information disclosure.
3159 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3160 INT_MIN/-1 -- value coerced to INT_MAX.
3166 NM/01 Workaround for PCRE version dependency in version reporting
3169 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3170 This fixes portability to compilers other than gcc, notably
3171 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3173 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3174 makefiles for portability to HP-UX and POSIX correctness.
3176 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3177 Also via indented variable definition in the Makefile.
3178 (Debugging by Oliver Heesakkers).
3180 PP/02 Restore caching of spamd results with expanded spamd_address.
3181 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3183 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3184 Improves build reliability. Fix from: Frank Elsner
3186 NM/02 Fix wide character breakage in the rfc2047 coding
3187 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3189 NM/03 Allow underscore in dnslist lookups
3190 Fixes bug 1026. Patch from Graeme Fowler
3192 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3193 Code patches from Adam Ciarcinski of NetBSD.
3195 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3198 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3199 is logged, to avoid truncation. Patch from John Horne.
3201 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3202 Patch from Jakob Hirsch.
3204 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3205 SQL string expansion failure details.
3206 Patch from Andrey Oktyabrski.
3208 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3209 Patch from Simon Arlott.
3211 PP/09 New lookups functionality failed to compile on old gcc which rejects
3212 extern declarations in function scope.
3213 Patch from Oliver Fleischmann
3215 PP/10 Use sig_atomic_t for flags set from signal handlers.
3216 Check getgroups() return and improve debugging.
3217 Fixed developed for diagnosis in bug 927 (which turned out to be
3220 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3221 Patch from Mark Zealey.
3223 PP/12 Bugzilla 1056: Improved spamd server selection.
3224 Patch from Mark Zealey.
3226 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3227 Based on patch from Heiko Schlittermann.
3229 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3230 Patch from Uwe Doering, sign-off by Michael Haardt.
3232 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3233 variable declaration deep within a block. Bug and patch from
3236 PP/15 lookups-Makefile IRIX compatibility coercion.
3238 PP/16 Make DISABLE_DKIM build knob functional.
3240 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3241 Patch by Simon Arlott
3243 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3244 variable argument macros. Our copy now differs from upstream.
3250 TF/01 Failure to get a lock on a hints database can have serious
3251 consequences so log it to the panic log.
3253 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3254 controlled using the smtp_confirmation log selector.
3256 TF/03 Include the error message when we fail to unlink a spool file.
3258 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3259 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3260 for maintaining out-of-tree patches for some time.
3262 PP/01 Bugzilla 139: Documentation and portability issues.
3263 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3264 Handle per-OS dynamic-module compilation flags.
3266 PP/02 Let /dev/null have normal permissions.
3267 The 4.73 fixes were a little too stringent and complained about the
3268 permissions on /dev/null. Exempt it from some checks.
3269 Reported by Andreas M. Kirchwitz.
3271 PP/03 Report version information for many libraries, including
3272 Exim version information for dynamically loaded libraries. Created
3273 version.h, now support a version extension string for distributors
3274 who patch heavily. Dynamic module ABI change.
3276 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3277 privilege escalation vulnerability whereby the Exim run-time user
3278 can cause root to append content of the attacker's choosing to
3281 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3284 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3285 If dropping privileges for untrusted macros, we disabled normal logging
3286 on the basis that it would fail; for the Exim run-time user, this is not
3287 the case, and it resulted in successful deliveries going unlogged.
3288 Fixed. Reported by Andreas Metzler.
3294 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3295 only prepend for the Resent-* case. Fixes regression introduced in
3296 Exim 4.70 by NM/22 for Bugzilla 607.
3298 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3299 increasing numbers of administrators be bitten by this.
3301 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3303 PP/03 Bugzilla 994: added openssl_options main configuration option.
3305 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3307 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3309 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3311 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3312 without a peer certificate, leading to a segfault because of an
3313 assumption that peers always have certificates. Be a little more
3314 paranoid. Problem reported by Martin Tscholak.
3316 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3317 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3318 NB: ClamAV planning to remove STREAM in "middle of 2010".
3319 CL also introduces -bmalware, various -d+acl logging additions and
3320 more caution in buffer sizes.
3322 PP/09 Implemented reverse_ip expansion operator.
3324 PP/10 Bugzilla 937: provide a "debug" ACL control.
3326 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3328 PP/12 Bugzilla 973: Implement --version.
3330 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3332 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3334 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3336 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3337 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3338 ignore trailing whitespace.
3340 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3342 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3345 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3346 Notification from Dr Andrew Aitchison.
3348 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3349 ExtendedDetectionInfo response format.
3350 Notification from John Horne.
3352 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3355 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3356 XSL and documented dependency on system catalogs, with examples of how
3359 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3362 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3363 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3364 configuration file which is writeable by the Exim user or group.
3366 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3367 of configuration files to cover files specified with the -C option if
3368 they are going to be used with root privileges, not just the default
3371 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3372 option (effectively making it always true).
3374 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3375 files to be used while preserving root privileges.
3377 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3378 that rogue child processes cannot use them.
3380 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3381 run-time user, instead of root.
3383 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3384 Exim run-time user without dropping privileges.
3386 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3387 result string, instead of calling string_vformat() twice with the same
3390 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3391 for other users. Others should always drop root privileges if they use
3392 -C on the command line, even for a whitelisted configure file.
3394 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3396 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3402 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3403 $data_path, and $header_path variables; fixed documentation bugs and
3406 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3407 exipick to access non-standard spools, including the "frozen" queue
3410 NM/01 Bugzilla 965: Support mysql stored procedures.
3411 Patch from Alain Williams
3413 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3415 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3416 Patch from Andreas Metzler
3418 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3419 Patch from Kirill Miazine
3421 NM/05 Bugzilla 671: Added umask to procmail example.
3423 JJ/03 installed exipick 20100323.0, fixing doc bug
3425 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3426 directory. Notification and patch from Dan Rosenberg.
3428 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3430 TK/02 Improve log output when DKIM signing operation fails.
3432 MH/01 Treat the transport option dkim_domain as a colon separated
3433 list, not as a single string, and sign the message with each element,
3434 omitting multiple occurences of the same signer.
3436 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3437 Bugzilla 985, 986. Patch by Simon Arlott
3439 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3440 Patch by Simon Arlott
3442 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3443 MBX locking. Notification from Dan Rosenberg.
3449 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3451 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3453 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3455 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3457 NM/04 Bugzilla 915: Flag broken perl installation during build.
3463 TK/01 Added patch by Johannes Berg that expands the main option
3464 "spamd_address" if it starts with a dollar sign.
3466 TK/02 Write list of recipients to X-Envelope-Sender header when building
3467 the mbox-format spool file for content scanning (suggested by Jakob
3470 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3471 (http://www.dcc-servers.net/) support via dccifd. Activated by
3472 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3474 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3475 by Mark Daniel Reidel <mr@df.eu>.
3477 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3478 When building exim an external PCRE library is now needed -
3479 PCRE is a system library on the majority of modern systems.
3480 See entry on PCRE_LIBS in EDITME file.
3482 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3483 conversation. Added nologin parameter to request.
3484 Patch contributed by Kirill Miazine.
3486 TF/01 Do not log submission mode rewrites if they do not change the address.
3488 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3490 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3491 log files in place. Contributed by Roberto Lima.
3493 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3495 TF/03 Bugzilla 615: When checking the local_parts router precondition
3496 after a local_part_suffix or local_part_prefix option, Exim now
3497 does not use the address's named list lookup cache, since this
3498 contains cached lookups for the whole local part.
3500 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3501 Robert Millan. Documentation is in experimental-spec.txt.
3503 TF/04 Bugzilla 668: Fix parallel build (make -j).
3505 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3507 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3508 Patch provided by Jan Srzednicki.
3510 TF/05 Leading white space used to be stripped from $spam_report which
3511 wrecked the formatting. Now it is preserved.
3513 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3514 that they are available at delivery time.
3516 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3518 TF/08 TLS error reporting now respects the incoming_interface and
3519 incoming_port log selectors.
3521 TF/09 Produce a more useful error message if an SMTP transport's hosts
3522 setting expands to an empty string.
3524 NM/06 Bugzilla 744: EXPN did not work under TLS.
3525 Patch provided by Phil Pennock.
3527 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3528 Patch provided by Richard Godbee.
3530 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3531 acl_smtp_notquit, added index entry.
3533 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3534 Patch provided by Eugene Bujak.
3536 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3537 accept(). Patch provided by Maxim Dounin.
3539 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3540 Patch provided by Phil Pennock.
3542 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3544 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3545 Patch provided by Brad "anomie" Jorsch.
3547 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3548 Patch provided by Dean Brooks.
3550 TK/05 Add native DKIM support (does not depend on external libraries).
3552 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3553 Patch provided by Graeme Fowler.
3555 NM/16 Bugzilla 851: Documentation example syntax fix.
3557 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3559 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3562 NM/19 Bugzilla 745: TLS version reporting.
3563 Patch provided by Phil Pennock.
3565 NM/20 Bugzilla 167: bool: condition support.
3566 Patch provided by Phil Pennock.
3568 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3569 clients. Patch provided by Phil Pennock.
3571 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3572 Patch provided by Brad "anomie" Jorsch.
3574 NM/23 Bugzilla 687: Fix misparses in eximstats.
3575 Patch provided by Heiko Schlittermann.
3577 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3578 Patch provided by Heiko Schlittermann.
3580 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3581 plus update to original patch.
3583 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3585 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3586 Patch provided by David Brownlee.
3588 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3590 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3592 NM/30 Bugzilla 888: TLS documentation bugfixes.
3594 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3596 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3597 Unlike the original bugzilla I have changed all shell scripts in src tree.
3599 NM/33 Bugzilla 898: Transport filter timeout fix.
3600 Patch by Todd Rinaldo.
3602 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3603 Patch by Serge Demonchaux.
3605 NM/35 Bugzilla 39: Base64 decode bug fixes.
3606 Patch by Jakob Hirsch.
3608 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3610 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3612 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3614 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3620 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3621 ALT-N's libdkim that I have put here:
3622 http://duncanthrax.net/exim-experimental/
3624 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3625 were called 'true' and it seems that C99 defines that as a reserved
3626 keyword to be used with 'bool' variable types. That means you could
3627 not include C99-style headers which use bools without triggering
3628 build errors in sieve.c.
3630 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3631 as mailq or other aliases. Changed the --help handling significantly
3632 to do whats expected. exim_usage() emits usage/help information.
3634 SC/01 Added the -bylocaldomain option to eximstats.
3636 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3638 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3640 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3646 PH/01 Another patch from the Sieve maintainer.
3648 PH/02 When an IPv6 address is converted to a string for single-key lookup
3649 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3650 dots are used instead of colons so that keys in lsearch files need not
3651 contain colons. This was done some time before quoting was made available
3652 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3653 (notated using the quote facility) so as to distinguish them from IPv4
3654 keys. This meant that lookups for IP addresses in host lists did not work
3655 for iplsearch lookups.
3657 This has been fixed by arranging for IPv6 addresses to be expressed with
3658 colons if the lookup type is iplsearch. This is not incompatible, because
3659 previously such lookups could never work.
3661 The situation is now rather anomalous, since one *can* have colons in
3662 ordinary lsearch keys. However, making the change in all cases is
3663 incompatible and would probably break a number of configurations.
3665 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3668 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3669 conversion specification without a maximum field width, thereby enabling
3670 a rogue spamd server to cause a buffer overflow. While nobody in their
3671 right mind would setup Exim to query an untrusted spamd server, an
3672 attacker that gains access to a server running spamd could potentially
3673 exploit this vulnerability to run arbitrary code as the Exim user.
3675 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3676 $primary_hostname instead of what libspf2 thinks the hosts name is.
3678 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3679 a directory entry by the name of the lookup key. Previously, if a
3680 symlink pointed to a non-existing file or a file in a directory that
3681 Exim lacked permissions to read, a lookup for a key matching that
3682 symlink would fail. Now it is enough that a matching directory entry
3683 exists, symlink or not. (Bugzilla 503.)
3685 PH/03 The body_linecount and body_zerocount variables are now exported in the
3688 PH/04 Added the $dnslist_matched variable.
3690 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3691 This means they are set thereafter only if the connection becomes
3694 PH/06 Added the client_condition to authenticators so that some can be skipped
3695 by clients under certain conditions.
3697 PH/07 The error message for a badly-placed control=no_multiline_responses left
3698 "_responses" off the end of the name.
3700 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3702 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3703 (without spaces) instead of just copying the configuration text.
3705 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3707 PH/11 Added $max_received_linelength.
3709 PH/12 Added +ignore_defer and +include_defer to host lists.
3711 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3712 way in which PCRE > 7.0 is built.
3714 PH/14 Implemented queue_only_load_latch.
3716 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3717 MAIL command. The effect was to mangle the value on 64-bit systems.
3719 PH/16 Another patch from the Sieve maintainer.
3721 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3723 PH/18 If a system quota error occurred while trying to create the file for
3724 a maildir delivery, the message "Mailbox is full" was not appended to the
3725 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3726 only to a quota excession during the actual writing of the file.
3728 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3729 characters?) which causes problems in log lines. The DN values are now
3730 passed through string_printing() before being added to log lines.
3732 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3733 and InterBase are left for another time.)
3735 PH/21 Added message_body_newlines option.
3737 PH/22 Guard against possible overflow in moan_check_errorcopy().
3739 PH/23 POSIX allows open() to be a macro; guard against that.
3741 PH/24 If the recipient of an error message contained an @ in the local part
3742 (suitably quoted, of course), incorrect values were put in $domain and
3743 $local_part during the evaluation of errors_copy.
3749 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3750 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3753 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3754 issue a MAIL command.
3756 PH/02 In an ACL statement such as
3758 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3760 if a client was not listed at all, or was listed with a value other than
3761 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3762 the condition was not true (as it should be), so access was not denied.
3763 The bug was that the ! inversion was incorrectly passed on to the second
3764 item. This has been fixed.
3766 PH/03 Added additional dnslists conditions == and =& which are different from
3767 = and & when the dns lookup returns more than one IP address.
3769 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3770 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3772 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3773 FSYNC, which compiles an option called disable_fsync that allows for
3774 bypassing fsync(). The documentation is heavily laced with warnings.
3776 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3778 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3779 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3780 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3781 including adding "make clean"; (3) Added -fPIC when compiling the test
3782 dynamically loaded module, to get rid of a warning.
3784 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3785 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3786 = 0s. The bug is otherwise harmless.
3788 PH/07 There was a bug in the dovecot authenticator such that the value of
3789 $auth1 could be overwritten, and so not correctly preserved, after a
3790 successful authentication. This usually meant that the value preserved by
3791 the server_setid option was incorrect.
3793 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3795 PH/09 Installed PCRE release 7.0.
3797 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3798 run for batched SMTP input. It is now run at the start of every message
3799 in the batch. While fixing this I discovered that the process information
3800 (output by running exiwhat) was not always getting set for -bs and -bS
3801 input. This is fixed, and it now also says "batched" for BSMTP.
3803 PH/11 Added control=no_pipelining.
3805 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3806 patch, slightly modified), and move the expansion of helo_data till after
3807 the connection is made in the smtp transport (so it can use these
3810 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3812 PH/14 Added log_selector = +pid.
3814 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3816 PH/16 Add ${if forany and ${if forall.
3818 PH/17 Added dsn_from option to vary the From: line in DSNs.
3820 PH/18 Flush SMTP output before performing a callout, unless control =
3821 no_callout_flush is set.
3823 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3824 was true (the default) a successful delivery failed to delete the retry
3825 item, thus causing premature timeout of the address. The bug is now
3828 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3830 PH/21 Long custom messages for fakedefer and fakereject are now split up
3831 into multiline responses in the same way that messages for "deny" and
3832 other ACL rejections are.
3834 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3835 with slight modification.
3837 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3838 draft, changing the syntax and factoring some duplicate code".
3840 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3841 for deliveries of the second and subsequent messages over the same SMTP
3844 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3845 ${reduce, with only minor "tidies".
3847 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3849 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3850 expansion side effects.
3852 PH/27 When a message times out after an over-quota error from an Exim-imposed
3853 quota, the bounce message says "mailbox is full". This message was not
3854 being given when it was a system quota that was exceeded. It now should
3857 MH/03 Made $recipients available in local_scan(). local_scan() already has
3858 better access to the recipient list through recipients_list[], but
3859 $recipients can be useful in postmaster-provided expansion strings.
3861 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3862 in the case of a MAIL command with additional options following the
3863 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3864 were accidentally chopped off.
3866 PH/29 SMTP synchronization checks are implemented when a command is read -
3867 there is a check that no more input is waiting when there shouldn't be
3868 any. However, for some commands, a delay in an ACL can mean that it is
3869 some time before the response is written. In this time, more input might
3870 arrive, invalidly. So now there are extra checks after an ACL has run for
3871 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3872 pipelining has not been advertised.
3874 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3876 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3877 returned in the SMTP response when smtp_return_error_details was false.
3878 This has been fixed.
3880 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3881 instead of the C I/O that was originally supplied, because problems were
3882 reported on Solaris.
3884 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3885 Exim which did not show up earlier: it was assuming that a call to
3886 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3887 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3888 was a macro that became an assignment, so it seemed to work. This has
3889 changed to a proper function call with a void return, hence the compile
3890 error. Exim's code has been fixed.
3892 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3895 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3897 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3899 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3900 because it turns out to be too big or there is a timeout) there is no
3901 "Completed" line in the log. When some messages of this type were
3902 selected by exigrep, they were listed as "not completed". Others were
3903 picked up by some special patterns. I have improved the selection
3904 criteria to be more general.
3906 PH/38 The host_find_failed option in the manualroute router can now be set
3907 to "ignore", to completely ignore a host whose IP address cannot be
3908 found. If all hosts are ignored, the behaviour is controlled by the new
3909 host_all_ignored option.
3911 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3912 homing or because of multiple MX records with /mx) generated more than
3913 one IP address, and the following item turned out to be the local host,
3914 all the secondary addresses of the first item were incorrectly removed
3915 from the list, along with the local host and any following hosts (which
3916 is what is supposed to happen).
3918 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3919 whoever called Exim into the -H file. In the case of the daemon it was
3920 behaving confusingly. When first started, it used values for whoever
3921 started the daemon, but after a SIGHUP it used the Exim user (because it
3922 calls itself on a restart). I have changed the code so that it now always
3925 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3926 message are rejected with the same error (e.g. no authentication or bad
3927 sender address), and a DATA command is nevertheless sent (as can happen
3928 with PIPELINING or a stupid MUA), the error message that was given to the
3929 RCPT commands is included in the rejection of the DATA command. This is
3930 intended to be helpful for MUAs that show only the final error to their
3933 PH/42 Another patch from the Sieve maintainer.
3935 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3936 Eximstats - Fixed some broken HTML links and added missing column headers
3938 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3939 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3941 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3943 PH/43 Yet another patch from the Sieve maintainer.
3945 PH/44 I found a way to check for a TCP/IP connection going away before sending
3946 the response to the final '.' that terminates a message, but only in the
3947 case where the client has not sent further data following the '.'
3948 (unfortunately, this is allowed). However, in many cases there won't be
3949 any further data because there won't be any more messages to send. A call
3950 to select() can be used: if it shows that the input is "ready", there is
3951 either input waiting, or the socket has been closed. An attempt to read
3952 the next input character can distinguish the two cases. Previously, Exim
3953 would have sent an OK response which the client would never have see.
3954 This could lead to message repetition. This fix should cure that, at
3955 least in a lot of common cases.
3957 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3958 advertised in response to EHLO.
3964 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3965 fixed by 4.65/MH/01 (is this a record?) are fixed:
3967 (i) An empty string was always treated as zero by the numeric comparison
3968 operators. This behaviour has been restored.
3970 (ii) It is documented that the numeric comparison operators always treat
3971 their arguments as decimal numbers. This was broken in that numbers
3972 starting with 0 were being interpreted as octal.
3974 While fixing these problems I realized that there was another issue that
3975 hadn't been noticed. Values of message_size_limit (both the global option
3976 and the transport option) were treated as octal if they started with 0.
3977 The documentation was vague. These values are now always treated as
3978 decimal, and I will make that clear in the documentation.
3984 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3985 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3988 MH/01 Don't check that the operands of numeric comparison operators are
3989 integers when their expansion is in "skipping" mode (fixes bug
3990 introduced by 4.64-PH/07).
3992 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3993 child addresses, Exim now panics and dies. Previously, because the count
3994 is held in a short int, deliveries were likely to be lost. As such a
3995 large number of recipients for a single message is ridiculous
3996 (performance will be very, very poor), I have chosen to impose a limit
3997 rather than extend the field.
4003 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4004 leftover -K file (the existence of which was triggered by #402).
4005 While we were at it, introduced process PID as part of the -K
4006 filename. This should rule out race conditions when creating
4009 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4010 processing considerably. Previous code took too long for large mails,
4011 triggering a timeout which in turn triggers #401.
4013 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4014 in the DK code in transports.c. sendfile() is not really portable,
4015 hence the _LINUX specificness.
4017 TF/01 In the add_headers option to the mail command in an Exim filter,
4018 there was a bug that Exim would claim a syntax error in any
4019 header after the first one which had an odd number of characters
4022 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4023 callout verification, Exim cached a "reject" for the entire domain. This
4024 is correct for most verifications, but it is not correct for a recipient
4025 verification with use_sender or use_postmaster set, because in that case
4026 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4027 case of MAIL FROM:<> rejection from other early rejections (e.g.
4028 rejection of HELO). When verifying a recipient using a non-null MAIL
4029 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4030 Whatever the result of the callout, the value of the domain cache is
4031 left unchanged (for any other kind of callout, getting as far as trying
4032 RCPT means that the domain itself is ok).
4034 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4037 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4038 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4039 session, the daemon ignores ECONNECT errors and logs others; it now
4040 ignores EPIPE as well.
4042 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4043 (quoted-printable decoding).
4045 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4046 later the small subsequent patch to fix an introduced bug.
4048 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4050 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4052 PH/08 An error is now given if message_size_limit is specified negative.
4054 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4055 to be given (somewhat) arbitrary names.
4057 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4060 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4061 miscellaneous code fixes
4063 PH/10 Added the log_reject_target ACL modifier to specify where to log
4066 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4067 hostname. This is wrong, because it relates to the incoming message (and
4068 probably the interface on which it is arriving) and not to the outgoing
4069 callout (which could be using a different interface). This has been
4070 changed to use the value of the helo_data option from the smtp transport
4071 instead - this is what is used when a message is actually being sent. If
4072 there is no remote transport (possible with a router that sets up host
4073 addresses), $smtp_active_hostname is used.
4075 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4076 tweaks were necessary in order to get it to work (see also 21 below):
4077 (a) The code assumed that strncpy() returns a negative number on buffer
4078 overflow, which isn't the case. Replaced with Exim's string_format()
4080 (b) There were several signed/unsigned issues. I just did the minimum
4081 hacking in of casts. There is scope for a larger refactoring.
4082 (c) The code used strcasecmp() which is not a standard C function.
4083 Replaced with Exim's strcmpic() function.
4084 (d) The code set only $1; it now sets $auth1 as well.
4085 (e) A simple test gave the error "authentication client didn't specify
4086 service in request". It would seem that Dovecot has changed its
4087 interface. Fortunately there's a specification; I followed it and
4088 changed what the client sends and it appears to be working now.
4090 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4093 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4094 address A is aliased to B and C, where B exists and C does not. Without
4095 -v the output is "A verified" because verification stops after a
4096 successful redirection if more than one address is generated. However,
4097 with -v the child addresses are also verified. Exim was outputting "A
4098 failed to verify" and then showing the successful verification for C,
4099 with its parentage. It now outputs "B failed to verify", showing B's
4100 parentage before showing the successful verification of C.
4102 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4103 look up a TXT record in a specific list after matching in a combined
4106 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4107 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4108 they consult the DNS. I had assumed they would set it the way they
4109 wanted; and indeed my experiments on Linux seem to show that in some
4110 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4111 To be on the safe side, however, I have now made the interface to
4112 host_find_byname() similar to host_find_bydns(), with an argument
4113 containing the DNS resolver options. The host_find_byname() function now
4114 sets these options at its start, just as host_find_bydns() does. The smtp
4115 transport options dns_qualify_single and dns_search_parents are passed to
4116 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4117 of host_find_byname() use the default settings of RES_DEFNAMES
4118 (qualify_single) but not RES_DNSRCH (search_parents).
4120 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4121 spool_read_header() do less string testing, by means of a preliminary
4122 switch on the second character of optional "-foo" lines. (This is
4123 overdue, caused by the large number of possibilities that now exist.
4124 Originally there were few.) While I was there, I also converted the
4125 str(n)cmp tests so they don't re-test the leading "-" and the first
4126 character, in the hope this might squeeze out yet more improvement.
4128 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4129 flag allowing group syntax was set by the header_syntax check but not
4130 turned off, possible causing trouble later; (2) The flag was not being
4131 set at all for the header_verify test, causing "group"-style headers to
4132 be rejected. I have now set it in this case, and also caused header_
4133 verify to ignore an empty address taken from a group. While doing this, I
4134 came across some other cases where the code for allowing group syntax
4135 while scanning a header line wasn't quite right (mostly, not resetting
4136 the flag correctly in the right place). These bugs could have caused
4137 trouble for malformed header lines. I hope it is now all correct.
4139 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4140 with the "reply" argument non-NULL. The code, however (which originally
4141 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4142 but it didn't always do it. This confused somebody who was copying the
4143 code for some other use. I have removed all the tests.
4145 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4146 feature that was used to support insecure browsers during the U.S. crypto
4147 embargo. It requires special client support, and Exim is probably the
4148 only MTA that supported it -- and would never use it because real RSA is
4149 always available. This code has been removed, because it had the bad
4150 effect of slowing Exim down by computing (never used) parameters for the
4151 RSA_EXPORT functionality.
4153 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4154 authenticator to fail if there's a tab character in the incoming data
4155 (there should never be unless someone is messing about, as it's supposed
4156 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4157 if the connection is using TLS or if the remote IP is the same as the
4158 local IP, and the "valid-client-cert option" if a client certificate has
4161 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4162 authenticators. This can be used for authorization after authentication
4163 succeeds. (In the case of plaintext, it servers for both authentication
4166 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4167 if any retry times were supplied.
4169 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4170 connection, where there is no client IP address to check. In this
4171 situation, the verify now always succeeds.
4173 PH/25 Applied John Jetmore's -Mset patch.
4175 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4177 PH/27 In a string expansion for a processed (not raw) header when multiple
4178 headers of the same name were present, leading whitespace was being
4179 removed from all of them, but trailing whitespace was being removed only
4180 from the last one. Now trailing whitespace is removed from each header
4181 before concatenation. Completely empty headers in a concatenation (as
4182 before) are ignored.
4184 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4185 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4187 PH/29 [Removed. This was a change that I later backed out, and forgot to
4188 correct the ChangeLog entry (that I had efficiently created) before
4189 committing the later change.]
4191 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4192 address errors (4xx response to RCPT) over the same connection as other
4193 messages routed to the same hosts. Such deliveries are always "forced",
4194 so retry times are not inspected. This resulted in far too many retries
4195 for the affected addresses. The effect occurred only when there were more
4196 hosts than the hosts_max_try setting in the smtp transport when it had
4197 the 4xx errors. Those hosts that it had tried were not added to the list
4198 of hosts for which the message was waiting, so if all were tried, there
4199 was no problem. Two fixes have been applied:
4201 (i) If there are any address or message errors in an SMTP delivery, none
4202 of the hosts (tried or untried) are now added to the list of hosts
4203 for which the message is waiting, so the message should not be a
4204 candidate for sending over the same connection that was used for a
4205 successful delivery of some other message. This seems entirely
4206 reasonable: after all the message is NOT "waiting for some host".
4207 This is so "obvious" that I'm not sure why it wasn't done
4208 previously. Hope I haven't missed anything, but it can't do any
4209 harm, as the worst effect is to miss an optimization.
4211 (ii) If, despite (i), such a delivery is accidentally attempted, the
4212 routing retry time is respected, so at least it doesn't keep
4213 hammering the server.
4215 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4216 in ${readsocket because some servers need this prod.
4218 PH/32 Added some extra debug output when updating a wait-xxx database.
4220 PH/33 The hint "could be header name not terminated by colon", which has been
4221 given for certain expansion errors for a long time, was not being given
4222 for the ${if def:h_colon_omitted{... case.
4224 PH/34 The spec says: "With one important exception, whenever a domain list is
4225 being scanned, $domain contains the subject domain." There was at least
4226 one case where this was not true.
4228 PH/35 The error "getsockname() failed: connection reset by peer" was being
4229 written to the panic log as well as the main log, but it isn't really
4230 panic-worthy as it just means the connection died rather early on. I have
4231 removed the panic log writing for the ECONNRESET error when getsockname()
4234 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4235 runs only) independently of the message's sender address. This meant
4236 that, if the 4xx error was in fact related to the sender, a different
4237 message to the same recipient with a different sender could confuse
4238 things. In particular, this can happen when sending to a greylisting
4239 server, but other circumstances could also provoke similar problems.
4240 I have changed the default so that the retry time for these errors is now
4241 based a combination of the sender and recipient addresses. This change
4242 can be overridden by setting address_retry_include_sender=false in the
4245 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4246 remote server are returned as part of bounce messages. This was not
4247 happening for LMTP over a pipe (the lmtp transport), but now it is the
4248 same for both kinds of LMTP.
4250 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4251 in header lines that were in fact CNAMEs. This is no longer the case.
4253 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4254 and queue runs started by the daemon processed all messages. This has
4255 been fixed so that -R and -S can now usefully be given with -q<time>.
4257 PH/40 Import PCRE release 6.7 (fixes some bugs).
4259 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4261 PH/42 Give an error if -q is specified more than once.
4263 PH/43 Renamed the variables $interface_address and $interface_port as
4264 $received_ip_address and $received_port, to make it clear that these
4265 values apply to message reception, and not to the outgoing interface when
4266 a message is delivered. (The old names remain recognized, of course.)
4268 PH/44 There was no timeout on the connect() call when using a Unix domain
4269 socket in the ${readsocket expansion. There now is.
4271 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4272 be meaningful with "accept".
4274 SC/01 Eximstats V1.43
4275 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4277 SC/02 Eximstats V1.44
4278 Use a glob alias rather than an array ref in the generated
4279 parser. This improves both readability and performance.
4281 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4282 Collect SpamAssassin and rejection statistics.
4283 Don't display local sender or destination tables unless
4284 there is data to show.
4285 Added average volumes into the top table text output.
4287 SC/04 Eximstats V1.46
4288 Collect data on the number of addresses (recipients)
4289 as well as the number of messages.
4291 SC/05 Eximstats V1.47
4292 Added 'Message too big' to the list of mail rejection
4293 reasons (thanks to Marco Gaiarin).
4295 SC/06 Eximstats V1.48
4296 Mainlog lines which have GMT offsets and are too short to
4297 have a flag are now skipped.
4299 SC/07 Eximstats V1.49 (Alain Williams)
4300 Added the -emptyok flag.
4302 SC/08 Eximstats V1.50
4303 Fixes for obtaining the IP address from reject messages.
4305 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4306 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4307 whitespace changes from 4.64-PH/27
4309 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4312 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4313 are found, allow negative numbers in numeric criteria)
4315 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4317 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4318 to match changes made in 4.64-PH/43
4320 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4322 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4323 "domain" is actually an IP address, and force a failure. This locks out
4324 those revolvers/nameservers that support "A-for-A" lookups, in
4325 contravention of the specifications.
4327 PH/48 When a host name was looked up from an IP address, and the subsequent
4328 forward lookup of the name timed out, the host name was left in
4329 $sender_host_name, contrary to the specification.
4331 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4332 restricted to single-key lookups, Exim was not diagnosing an error if
4333 * or *@ was used with a query-style lookup.
4335 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4337 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4338 long ago, but noone interested enough thought of it. Let's just say that
4339 the "1.1" means that there are some new functions that weren't there at
4340 some point in the past.
4342 PH/51 Error processing for expansion failure of helo_data from an smtp
4343 transport during callout processing was broken.
4345 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4346 tested/used via the -bh/-bhc/-bs options.
4348 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4349 bug, fixed in subsequent PCRE releases).
4351 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4352 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4354 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4360 SC/01 Use a glob alias rather than an array ref in eximstats generated
4361 parser. This improves both readability and performance.
4363 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4364 Don't display local sender or destination tables in eximstats unless
4365 there is data to show.
4366 Added average volumes into the eximstats top table text output.
4368 SC/03 Collect data on the number of addresses (recipients) as well
4369 as the number of messages in eximstats.
4371 TF/01 Correct an error in the documentation for the redirect router. Exim
4372 does (usually) call initgroups() when daemonizing.
4374 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4375 with consistent privilege compared to when running as a daemon.
4377 TF/03 Note in the spec that $authenticated_id is not set for local
4378 submissions from trusted users.
4380 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4381 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4383 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4384 by adding some example configuration directives to the default
4385 configuration file. A little bit of work is required to uncomment the
4386 directives and define how usernames and passwords are checked, but
4387 there is now a framework to start from.
4389 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4390 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4391 without this. I don't know how relevant this is to other LDAP libraries.
4393 PH/02 Add the verb name to the "unknown ACL verb" error.
4395 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4397 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4399 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4400 directory not expanded when it should be if an expanded home directory
4401 was set for the address (which is overridden by the transport).
4403 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4406 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4407 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4408 because it is too late at that time, and has no effect.
4410 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4411 security issue with \' (bugzilla #107). I could not use the
4412 PQescapeStringConn() function, because it needs a PGconn value as one of
4415 PH/08 When testing addresses using -bt, indicate those final addresses that
4416 are duplicates that would not cause an additional delivery. At least one
4417 person was confused, thinking that -bt output corresponded to deliveries.
4418 (Suppressing duplicates isn't a good idea as you lose the information
4419 about possibly different redirections that led to the duplicates.)
4421 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4422 systems where poll() doesn't work, in particular OS X.
4424 PH/10 Added more information to debugging output for retry time not reached.
4426 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4427 operations in malware.c.
4429 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4432 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4433 syslog_duplication set false, log lines that would normally be written
4434 both the the main log and to the reject log were not written to syslog at
4437 PH/14 In the default configuration, change the use of "message" in ACL warn
4438 statements to "add_header".
4440 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4441 not followed by a command (e.g. "seen endif").
4443 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4444 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4447 PH/17 Added extra conditions to the default value of delay_warning_condition
4451 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4452 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4453 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4456 The Auto-Submitted: and various List- headers are standardised, whereas I
4457 don't think Precedence: ever was.
4459 PH/18 Refactored debugging code in route_finduser() to show more information,
4460 in particular, the error code if getpwnam() issues one.
4462 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4463 This is apparently needed in addition to the PH/07 change above to avoid
4464 any possible encoding problems.
4466 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4467 but not after initializing Perl.
4469 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4470 output them only if debugging. By default they are written stderr,
4471 apparently, which is not desirable.
4473 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4476 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4479 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4481 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4482 authenticated or an ident call has been made. Suppress the default
4483 values for $authenticated_id and $authenticated_sender (but permit -oMai
4484 and -oMas) when testing with -bh.
4486 PH/24 Re-jigged the order of the tests in the default configuration so that the
4487 tests for valid domains and recipients precede the DNS black list and CSA
4488 tests, on the grounds that those ones are more expensive.
4490 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4491 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4492 This bug exists in every version of Exim that I still have, right back to
4495 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4496 However, an attempt to turn on case-sensitivity in a regex key by
4497 including (?-i) didn't work because the subject string was already
4498 lowercased, and the effects were non-intuitive. It turns out that a
4499 one-line patch can be used to allow (?-i) to work as expected.
4505 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4506 other effects) broke the use of negated acl sub-conditions.
4508 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4511 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4512 "Deny" causes Exim to reject the incoming connection with a 554 error.
4513 Unfortunately, if there is a major crisis, such as a disk failure,
4514 tcp-wrappers gives "deny", whereas what one would like would be some
4515 kind of temporary error. A kludge has been added to help with this.
4516 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4517 554 error is used if errno is still zero or contains ENOENT (which occurs
4518 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4521 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4523 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4524 errors. Otherwise a message that provokes a temporary error (when other
4525 messages do not) can cause a whole host to time out.
4527 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4528 addresses were routed directly to files or pipes from a redirect router.
4529 File deliveries just didn't batch; pipe deliveries might have suffered
4532 PH/06 A failure to get a lock for a hints database would erroneously always say
4533 "Failed to get write lock", even when it was really a read lock.
4535 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4536 of 0600. This has been changed to use the value of the lockfile_mode
4537 option (which defaults to 0600).
4539 PH/08 Applied small patch from the Sieve maintainer.
4541 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4542 folder from quota calculations, a direct delivery into this folder messed
4543 up the contents of the maildirsize file. This was because the regex was
4544 used only to exclude .Trash (or whatever) when the size of the mailbox
4545 was calculated. There was no check that a delivery was happening into an
4546 excluded directory. This bug has been fixed by ignoring all quota
4547 processing for deliveries into excluded directories.
4549 PH/10 Added the maildirfolder_create_regex option to appendfile.
4555 PH/01 The code for finding all the local interface addresses on a FreeBSD
4556 system running IPv6 was broken. This may well have applied to all BSD
4557 systems, as well as to others that have similar system calls. The broken
4558 code found IPv4 interfaces correctly, but gave incorrect values for the
4559 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4560 that it would not match correctly against @[] and not recognize the IPv6
4563 PH/02 The ipliteral router was not recognizing addresses of the form user@
4564 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4566 PH/03 Added disable_ipv6.
4568 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4569 decoded form, because it is most often used to construct To: headers
4570 lines in autoreplies, and the decoded form may well be syntactically
4571 invalid. However, $reply_address has leading white space removed, and all
4572 newlines turned into spaces so that the autoreply transport does not
4575 PH/05 If group was specified without a user on a router, and no group or user
4576 was specified on a transport, the group from the router was ignored.
4578 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4579 for visible compile-time settings that can be used to change these
4580 numbers, for those that want even more. Backwards compatibility with old
4581 spool files has been maintained. However, going back to a previous Exim
4582 release will lost any variables that are in spool files.
4584 PH/07 Two small changes when running in the test harness: increase delay when
4585 passing a TCP/IP connection to a new process, in case the original
4586 process has to generate a bounce, and remove special handling of
4587 127.0.0.2 (sic), which is no longer necessary.
4589 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4590 be the same on different OS.
4592 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4595 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4596 whether --show-vars was specified or not
4598 JJ/02 exipick: Added support for new ACL variable spool format introduced
4601 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4602 syntactically invalid From: or Reply-to: line, and a filter used this to
4603 generate an autoreply, and therefore failed to obtain an address for the
4604 autoreply, Exim could try to deliver to a non-existent relative file
4605 name, causing unrelated and misleading errors. What now happens is that
4606 it logs this as a hard delivery error, but does not attempt to create a
4609 PH/11 The exinext utility has a -C option for testing purposes, but although
4610 the given file was scanned by exinext itself; it wasn't being passed on
4611 when Exim was called.
4613 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4614 an end-of-file indication when reading a command response.
4616 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4617 compiled. In many other places in Exim, IPv6 addresses are always
4618 recognized, so I have changed this. It also means that IPv4 domain
4619 literals of the form [IPV4:n.n.n.n] are now always recognized.
4621 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4622 used if the router is not running as root, for example, when verifying at
4623 ACL time, or when using -bh. The debugging output from this situation was
4624 non-existent - all you got was a failure to exec. I have made two
4627 (a) Failures to set uid/gid, the current directory, or a process leader
4628 in a subprocess such as that created by queryprogram now generate
4629 suitable debugging output when -d is set.
4631 (b) The queryprogram router detects when it is not running as root,
4632 outputs suitable debugging information if -d is set, and then runs
4633 the subprocess without attempting to change uid/gid.
4635 PH/15 Minor change to Makefile for building test_host (undocumented testing
4638 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4639 additional section of a DNS packet that returns MX or SRV records.
4640 Instead, it always explicitly searches for A/AAAA records. This avoids
4641 major problems that occur when a DNS server includes only records of one
4642 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4643 fixed another bug: if SRV records were looked up and the corresponding
4644 address records were *not* found in the additional section, the port
4645 values from the SRV records were lost.
4647 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4648 using the correct key (the original address) when searching the retry
4649 rules in order to find which one to use for generating the retry hint.
4651 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4652 adding the default one. Similarly, if it contains a Reply-To: header, the
4653 errors_reply_to option, if set, is not used.
4655 PH/19 When calculating a retry time, Exim used to measure the "time since
4656 failure" by looking at the "first failed" field in the retry record. Now
4657 it does not use this if it is later than than the arrival time of the
4658 message. Instead it uses the arrival time. This makes for better
4659 behaviour in cases where some deliveries succeed, thus re-setting the
4660 "first failed" field. An example is a quota failure for a huge message
4661 when small messages continue to be delivered. Without this change, the
4662 "time since failure" will always be short, possible causing more frequent
4663 delivery attempts for the huge message than are intended.
4664 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4666 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4667 $1, $2, $3) because the numerical variables can be reset during some
4668 expansion items (e.g. "match"), thereby losing the authentication data.
4670 PH/21 Make -bV show the size of off_t variables so that the test suite can
4671 decide whether to run tests for quotas > 2G.
4673 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4674 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4675 filecount value is greater than 2G or if a quota value is greater than 2G
4676 on a system where the size of off_t is not greater than 4, a panic error
4679 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4680 never match. The debug and -bh output now contains an explicit error
4681 message indicating a malformed IPv4 address or mask.
4683 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4684 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4685 PH/23 above applies.
4687 PH/25 Do not write to syslog when running in the test harness. The only
4688 occasion when this arises is a failure to open the main or panic logs
4689 (for which there is an explicit test).
4691 PH/26 Added the /no_tell option to "control=freeze".
4693 PH/27 If a host name lookup failed very early in a connection, for example, if
4694 the IP address matched host_lookup and the reverse lookup yielded a name
4695 that did not have a forward lookup, an error message of the form "no IP
4696 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4697 could be logged. Now it outputs the IP address instead of "NULL".
4699 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4700 allows the sender and the authenticated sender to be set when
4701 submitting a message from within Exim. Since child_open_exim() is
4702 documented for local_scan(), the new function should be too.
4704 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4705 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4706 results in an empty string is now treated as unset.
4708 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4710 PH/31 Added sender_verify_fail logging option.
4712 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4713 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4714 tidied the source and removed it altogether.
4716 PH/33 When a queue run was abandoned because the load average was too high, a
4717 log line was always written; now it is written only if the queue_run log
4718 selector is set. In addition, the log line for abandonment now contains
4719 information about the queue run such as the pid. This is always present
4720 in "start" and "stop" lines but was omitted from the "abandon" line.
4722 PH/34 Omit spaces between a header name and the colon in the error message that
4723 is given when verify = headers_syntax fails (if there are lots of them,
4724 the message gets confusing).
4726 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4727 names, as there are now some PTR records that contain slashes. This check
4728 is only to protect against broken name servers that fall over on strange
4729 characters, so the fact that it applies to all lookups doesn't matter.
4731 PH/36 Now that the new test suite is complete, we can remove some of the
4732 special code in Exim that was needed for the old test suite. For example,
4733 sorting DNS records because real resolvers return them in an arbitrary
4734 order. The new test suite's fake resolver always returns records in the
4737 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4738 bounces) except when queue_only is set, to avoid logging races between
4739 the different processes.
4741 PH/38 Panic-die if .include specifies a non-absolute path.
4743 PH/39 A tweak to the "H" retry rule from its user.
4745 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4746 a label. They prevented compilation on older perls.
4748 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4749 a warning to be raised on newish perls.
4751 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4752 on queue. Changes to match documented behaviour of showing count of
4753 messages matching specified criteria.
4755 PH/40 Changed the default ident timeout from 30s to 5s.
4757 PH/41 Added support for the use of login_cap features, on those BSD systems
4758 that have them, for controlling the resources used by pipe deliveries.
4760 PH/42 The content-scanning code uses fopen() to create files in which to put
4761 message data. Previously it was not paying any attention to the mode of
4762 the files. Exim runs with umask(0) because the rest of the code creates
4763 files with open(), and sets the required mode explicitly. Thus, these
4764 files were ending up world-writeable. This was not a big issue, because,
4765 being within the spool directory, they were not world-accessible. I have
4766 created a function called modefopen, which takes an additional mode
4767 argument. It sets umask(777), creates the file, chmods it to the required
4768 mode, then resets the umask. All the relevant calls to fopen() in the
4769 content scanning code have been changed to use this function.
4771 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4772 to 24 hours. This avoids potential overflow problems when processing G
4773 and H retry rules. I suspect nobody ever tinkers with this value.
4775 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4777 PH/45 When the plaintext authenticator is running as a client, the server's
4778 challenges are checked to ensure they are valid base64 strings. By
4779 default, the authentication attempt is cancelled if an invalid string is
4780 received. Setting client_ignore_invalid_base64 true ignores these errors.
4781 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4782 they are received. Thus, the responses can be made to depend on the
4783 challenges. If an invalid string is ignored, an empty string is placed in
4786 PH/46 Messages that are created by the autoreply transport now contains a
4787 References: header, in accordance with RFCs 2822 and 3834.
4789 PH/47 Added authenticated_sender_force to the smtp transport.
4791 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4793 PH/49 Installed latest patch from the Sieve maintainer.
4795 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4796 was also set, the appendfile transport was unnecessarily scanning a
4797 directory of message files (e.g. for maildir delivery) to find the count
4798 of files (along with the size), even though it did not need this
4799 information. It now does the scan only if it needs to find either the
4800 size of the count of files.
4802 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4804 PH/52 Two bugs concerned with error handling when the smtp transport is
4807 (i) Exim was not creating retry information for temporary errors given
4808 for individual recipients after the DATA command when the smtp transport
4809 was used in LMTP mode. This meant that they could be retried too
4810 frequently, and not timed out correctly.
4812 (ii) Exim was setting the flag that allows error details to be returned
4813 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4814 recipients that were returned after the DATA command.
4816 PH/53 This is related to PH/52, but is more general: for any failing address,
4817 when detailed error information was permitted to be returned to the
4818 sender, but the error was temporary, then after the final timeout, only
4819 "retry timeout exceeded" was returned. Now it returns the full error as
4820 well as "retry timeout exceeded".
4822 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4823 do this, and (what is worse) MTAs that accept it.
4825 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4826 will now be deprecated.
4828 PH/56 New os.c-cygwin from the Cygwin maintainer.
4830 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4831 formats (previously only available in exim formats via -bpr, -bpru,
4832 and -bpra. Now also available in native and exiqgrep formats)
4834 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4835 with very large, slow to parse queues
4837 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4839 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4841 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4842 responses to authentication challenges, though it was showing the
4843 challenges; (ii) I've removed the CR characters from the debug output for
4846 PH/58 Allow for the insertion of a newline as well as a space when a string
4847 is turned into more than one encoded-word during RFC 2047 encoding. The
4848 Sieve code now uses this.
4850 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4851 data_4xx, lost_connection, tls_required.
4853 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4854 message was being sent as an SMTP response.
4856 PH/61 Add -l and -k options to exicyclog.
4858 PH/62 When verifying, if an address was redirected to one new address, so that
4859 verification continued, and the new address failed or deferred after
4860 having set something in $address_data, the value of $address_data was not
4861 passed back to the ACL. This was different to the case when no
4862 redirection occurred. The value is now passed back in both cases.
4864 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4865 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4866 use login_cap.h, so on its own it isn't the distinguishing feature. The
4867 new name refers directly to the setclassresources() function.
4869 PH/65 Added configuration files for NetBSD3.
4871 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4873 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4874 is preferred over IPv4.
4876 PH/68 The bounce_return_message and bounce_return_body options were not being
4877 honoured for bounces generated during the reception of non-SMTP messages.
4878 In particular, this applied to messages rejected by the ACL. This bug has
4879 been fixed. However, if bounce_return_message is true and bounce_return_
4880 body is false, the headers that are returned for a non-SMTP message
4881 include only those that have been read before the error was detected.
4882 (In the case of an ACL rejection, they have all been read.)
4884 PH/69 The HTML version of the specification is now built in a directory called
4885 spec_html instead of spec.html, because the latter looks like a path with
4886 a MIME-type, and this confuses some software.
4888 PH/70 Catch two compiler warnings in sieve.c.
4890 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4891 function verify_get_ident() calls ip_connect() to connect a socket, but
4892 if the "connect()" function timed out, ip_connect() used to close the
4893 socket. However, verify_get_ident() also closes the socket later, and in
4894 between Exim writes to the log, which may get opened at this point. When
4895 the socket was closed in ip_connect(), the log could get the same file
4896 descriptor number as the socket. This naturally causes chaos. The fix is
4897 not to close the socket in ip_connect(); the socket should be closed by
4898 the function that creates it. There was only one place in the code where
4899 this was missing, in the iplookup router, which I don't think anybody now
4900 uses, but I've fixed it anyway.
4902 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4903 well as to direct DNS lookups. Otherwise the handling of names in host
4904 lists is inconsistent and therefore confusing.
4910 PH/01 Two changes to the default runtime configuration:
4912 (1) Move the checks for relay_from_hosts and authenticated clients from
4913 after to before the (commented out) DNS black list checks.
4915 (2) Add control=submission to the relay_from_hosts and authenticated
4916 clients checks, on the grounds that messages accepted by these
4917 statements are most likely to be submissions.
4919 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4921 (1) Generate an error if the third argument for the ${prvs expansion is
4924 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4927 (3) Reset the variables that are obtained from the first argument of
4928 ${prvscheck and used in the second argument before leaving the code,
4929 because their memory is reclaimed, so using them afterwards may do
4932 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4933 one (it's much easier than Tom thought :-).
4935 (5) Because of (4), we can now allow for the use of $prvscheck_result
4936 inside the third argument.
4938 PH/03 For some reason, the default setting of PATH when running a command from
4939 a pipe transport was just "/usr/bin". I have changed it to
4942 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4943 anything to be listed in the output from -bV.
4945 PH/05 When a filter generated an autoreply, the entire To: header line was
4946 quoted in the delivery log line, like this:
4948 => >A.N.Other <ano@some.domain> <original@ddress> ...
4950 This has been changed so that it extracts the operative address. There
4951 may be more than one such address. If so, they are comma-separated, like
4954 => >ano@some.domain,ona@other.domain <original@ddress> ...
4956 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4957 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4958 not being looked up in the rDNS to get a host name, Exim was showing the
4959 IP address twice in Received: lines, even though the IP addresses were
4960 identical. For example:
4962 Received: from [1.2.3.4] (helo=[1.2.3.4])
4964 However, if the real host name was known, it was omitting the HELO data
4965 if it matched the actual IP address. This has been tidied up so that it
4966 doesn't show the same IP address twice.
4968 PH/07 When both +timestamp and +memory debugging was on, the value given by
4969 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4970 called by the debug printing, thereby overwriting the timestamp buffer.
4971 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4973 PH/08 When the original message was included in an autoreply transport, it
4974 always said "this is a copy of the message, including all the headers",
4975 even if body_only or headers_only was set. It now gives an appropriate
4978 PH/09 Applied a patch from the Sieve maintainer which:
4980 o fixes some comments
4981 o adds the (disabled) notify extension core
4982 o adds some debug output for the result of if/elsif tests
4983 o points to the current vacation draft in the documentation
4984 and documents the missing references header update
4988 o fixes a bug in processing the envelope test (when testing
4989 multiple envelope elements, the last element determined the
4992 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4993 Electronic Mail") by including:
4995 Auto-submitted: auto-generated
4997 in the messages that it generates (bounce messages and others, such as
4998 warnings). In the case of bounce messages for non-SMTP messages, there was
4999 also a typo: it was using "Auto_submitted" (underscore instead of
5000 hyphen). Since every message generated by Exim is necessarily in response
5001 to another message, thes have all been changed to:
5003 Auto-Submitted: auto-replied
5005 in accordance with these statements in the RFC:
5007 The auto-replied keyword:
5009 - SHOULD be used on messages sent in direct response to another
5010 message by an automatic process,
5012 - MUST NOT be used on manually-generated messages,
5014 - MAY be used on Delivery Status Notifications (DSNs) and Message
5015 Disposition Notifications (MDNs),
5017 - MUST NOT be used on messages generated by automatic or periodic
5018 processes, except for messages which are automatic responses to
5021 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5022 to the default Received: header definition.
5024 PH/12 Added log selector acl_warn_skipped (default on).
5026 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5027 variables because (a) they are in the wrong storage pool and (b) even if
5028 they were copied, it wouldn't work properly because of the caching.
5030 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5031 checking when decoding. Apparently there are clients that generate
5032 overlong encoded strings. Why am I not surprised?
5034 PH/15 If the first argument of "${if match_address" was not empty, but did not
5035 contain an "@" character, Exim crashed. Now it writes a panic log message
5036 and treats the condition as false.
5038 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5040 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5041 extension "envelope-auth". The code is finished and in agreement with
5042 other implementations, but there is no documentation so far and in fact,
5043 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5044 not changing the active code.
5046 Print executed "if" and "elsif" statements when debugging is used. This
5047 helps a great deal to understand what a filter does.
5049 Document more things not specified clearly in RFC3028. I had all this
5050 sorted out, when out of a sudden new issues came to my mind. Oops."
5052 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5055 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5056 canonical form (as documented). However, after a host name lookup from
5057 the IP address, check_host() was doing a simple string comparison with
5058 addresses acquired from the DNS when checking that the found name did
5059 have the original IP as one of its addresses. Since any found IPv6
5060 addresses are likely to be in abbreviated form, the comparison could
5061 fail. Luckily, there already exists a function for doing the comparison
5062 by converting both addresses to binary, so now that is used instead of
5063 the text comparison.
5065 PH/20 There was another similar case to PH/19, when a complete host name was
5066 given in a host list; looking up its IP address could give an abbreviated
5067 form, whereas the current host's name might or might not be abbreviated.
5068 The same fix has been applied.
5074 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5075 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5078 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5079 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5081 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5083 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5084 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5085 Also, for "Auto-Submitted", treat anything other than "no" as
5086 non-personal, in accordance with RFC 3834. (Previously it treated
5087 anything starting "auto-" as non-personal.)
5089 TF/01 The control=submission/name=... option had a problem with syntax
5090 errors if the name included a slash character. The /name= option
5091 now slurps the rest of the string, so it can include any characters
5092 but it must come last in the list of options (after /sender_retain
5095 PH/05 Some modifications to the interface to the fake nameserver for the new
5103 TK/01 Added the "success_on_redirect" address verification option. See
5104 NewStuff for rationale and an example.
5106 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5108 PH/02 Patch to exigrep to allow it to work on syslog lines.
5110 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5111 fread() to skip over the body file's header line, because in Cygwin the
5112 header line is locked and is inaccessible.
5114 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5115 co-exist for some time) to make it clear that it is the Exim ID that is
5116 referenced, not the Message-ID: header line.
5118 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5119 string_format() function, because snprintf() does not exist on all
5122 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5123 ${stat: expansion item.
5125 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5126 protocol synchronization error", to keep the pedants happy.
5128 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5129 well as for IRIX systems, when gcc is being used. See the host.c source
5132 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5134 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5137 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5138 transport and to the smtp transport in LMTP mode.
5140 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5142 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5143 run a filter in a subprocess. This could lead to confusion in subsequent
5144 lookups in the parent process. There should also be a search_tidyup() at
5145 the end of the subprocess.
5147 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5148 only if the host matched helo_try_verify_hosts, which caused the
5149 verification to occur when the EHLO/HELO command was issued. The ACL just
5150 tested the remembered result. Now, if a previous verification attempt has
5151 not happened, "verify = helo" does it there and then.
5153 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5155 TK/03 Fix log output including CR from clamd.
5157 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5158 exist provoked a memory error which could cause a segfault.
5160 PH/15 Installed PCRE 6.2
5162 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5164 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5165 of the problem. Specifically, suggested +O2 rather than +O1 for the
5168 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5170 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5171 record, the port was not correctly logged when the outgoing_port log
5172 selector was set (it logged the transort's default port).
5174 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5175 fallback_hosts, and "hosts" in the smtp transport.
5177 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5178 host errors such as "Connection refused".
5180 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5181 authentication with radiusclient 0.4.9:
5183 - Error returned from rc_read_config was caught wrongly
5184 - Username/password not passed on to radius server due to wrong length.
5186 The presumption is that some radiusclient API changes for 4.51/PH/17
5187 were not taken care of correctly. The code is still untested by me (my
5188 Linux distribution still has 0.3.2 of radiusclient), but it was
5189 contributed by a Radius user.
5191 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5192 expanding the "port" option of the smtp transport.
5194 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5195 while reading a MIME header. Thanks to Tom Hughes for a patch.
5197 PH/24 Include config.h inside local_scan.h so that configuration settings are
5200 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5201 that in an ACL for RCPT (for example), you can examine exactly what was
5204 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5205 commands, but it was not correctly comparing the address with the actual
5206 client host address. Thus, it would show the EHLO address in Received:
5207 header lines when this was not necessary.
5209 PH/27 Added the % operator to ${eval:}.
5211 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5212 it should be ignoring failures (because with -C, for example, it has lost
5213 privilege). It wasn't ignoring creation failures other than "already
5216 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5219 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5220 because an input line was too long, either on its own, or by virtue of
5221 too many continuations, the temporary file was not being removed, and the
5222 return code was incorrect.
5224 PH/31 Missing "BOOL" in function definition in filtertest.c.
5226 PH/32 Applied Sieve patches from the maintainer.
5228 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5230 PH/33 Added "verify = not_blind".
5232 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5233 Local/Makefile (with some defaults set). These are used in built scripts
5234 such as exicyclog, but they have never been used in the exim_install
5235 script (though there are many overriding facilities there). I have
5236 arranged that the exim_install script now takes note of these two
5239 PH/35 Installed configuration files for Dragonfly.
5241 PH/36 When a locally submitted message by a trusted user did not contain a
5242 From: header, and the sender address was obtained from -f or from an SMTP
5243 MAIL command, and the trusted user did not use -F to supply a sender
5244 name, $originator_name was incorrectly used when constructing a From:
5245 header. Furthermore, $originator_name was used for submission mode
5246 messages from external hosts without From: headers in a similar way,
5247 which is clearly wrong.
5249 PH/37 Added control=suppress_local_fixups.
5251 PH/38 When log_selector = +received_sender was set, and the addition of the
5252 sender made the log line's construction buffer exactly full, or one byte
5253 less than full, an overflow happened when the terminating "\n" was
5256 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5257 when the result of a list match is failure because a DNS lookup failed.
5259 PH/40 RM_COMMAND is now used in the building process.
5261 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5262 the "build-* directories that it finds.
5264 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5265 address in a domain literal was a prefix of an interface address.
5267 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5268 when verifying a sender address, unless rewrite_headers is false.
5270 PH/44 Wrote a long comment about why errors_to addresses are verified as
5271 recipients, not senders.
5273 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5274 the ratelimit ACL was added.
5276 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5278 PH/46 Added extra information about PostgreSQL errors to the error string.
5280 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5281 avoiding the need to install special zones in a real server. This is
5282 backwards compatible; if it can't find the fake resolver, it drops back.
5283 Thus, both old and new test suites can be run.
5285 TF/02 Added util/ratelimit.pl
5287 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5288 clock is set back in time.
5290 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5291 Candler <B.Candler@pobox.com>.
5293 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5294 OK for addresses that are the result of redirections.
5296 PH/48 A number of further additions for the benefit of the new test suite,
5297 including a fake gethostbyname() that interfaces to the fake DNS resolver
5300 TF/06 The fix for widen_domains has also been applied to qualify_single and
5301 search_parents which are the other dnslookup options that can cause
5304 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5307 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5309 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5310 service" if it encountered an MX record with an empty target hostname.
5311 The message is now "an MX or SRV record indicated no SMTP service".
5313 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5314 if the DNS of the sending site is misconfigured. This is quite a
5315 common situation. This change restores the behaviour of treating a
5316 helo verification defer as a failure.
5318 PH/51 If self=fail was set on a router, the bounce message did not include the
5319 actual error message.
5325 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5327 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5328 command itself ended in error, the underlying message about the transport
5329 filter timeout was being overwritten with the pipe command error. Now the
5330 underlying error message should be appended to the second error message.
5332 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5334 PH/02 Reduce the amount of output that "make" produces by default. Full output
5335 can still be requested.
5337 PH/03 The warning log line about a condition test deferring for a "warn" verb
5338 was being output only once per connection, rather than after each
5339 occurrence (because it was using the same function as for successful
5340 "warn" verbs). This seems wrong, so I have changed it.
5342 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5343 it should not have, which might have caused a crash in the right
5344 circumstances, but probably never did.
5346 PH/04 Installed a modified version of Tony Finch's patch to make submission
5347 mode fix the return path as well as the Sender: header line, and to
5348 add a /name= option so that you can make the user's friendly name appear
5351 TF/03 Added the control = fakedefer ACL modifier.
5353 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5354 Mark Lowes for thorough testing.
5356 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5358 TK/03 Merged latest SRS patch from Miles Wilton.
5360 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5361 with the definition in sysexits.h (which is #included earlier).
5362 Fortunately, Exim does not actually use EX_OK. The code used to try to
5363 preserve the sysexits.h value, by assuming that macro definitions were
5364 scanned for macro replacements. I have been disabused of this notion,
5365 so now the code just undefines EX_OK before #including unistd.h.
5367 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5368 in the smtp transport. When a block could not be written in a single
5369 write() function, the timeout was being re-applied to each part-write.
5370 This seems wrong - if the receiver was accepting one byte at a time it
5371 would take for ever. The timeout is now adjusted when this happens. It
5372 doesn't have to be particularly precise.
5374 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5375 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5377 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5378 without a domain if the check to <postmaster@domain> fails.
5380 SC/01 Eximstats: added -xls and the ability to specify output files
5381 (patch written by Frank Heydlauf).
5383 SC/02 Eximstats: use FileHandles for outputting results.
5385 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5387 SC/04 Eximstats: fixed display of large numbers with -nvr option
5389 SC/05 Eximstats: fixed merging of reports with empty tables.
5391 SC/06 Eximstats: added the -include_original_destination flag
5393 SC/07 Eximstats: removed tabs and trailing whitespace.
5395 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5397 TK/06 MBOX spool code: Add real "From " MBOX separator line
5398 so the .eml file is really in mbox format (even though
5399 most programs do not really care). Patch from Alex Miller.
5401 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5402 The latter is generated from $received_to and is only set if the
5403 message has one envelope recipient. SA can use these headers,
5404 obviously out-of-the-box. Patch from Alex Miller.
5406 PH/08 The ${def test on a variable was returning false if the variable's
5407 value was "0", contrary to what the specification has always said!
5408 The result should be true unless the variable is empty.
5410 PH/09 The syntax error of a character other than { following "${if
5411 def:variable_name" (after optional whitespace) was not being diagnosed.
5412 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5413 accidental colon was present, for example, could give incorrect results.
5415 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5416 result is used (not including appendfile, where other changes are about
5419 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5420 This involved changing a lot of size variables from int to off_t. It
5421 should work with maildirs and everything.
5423 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5424 spamd dying while we are connected to it.
5426 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5429 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5430 function for BDB 4.3.
5432 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5434 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5435 the sources for PCRE have been re-arranged and more files are now
5438 PH/15 The code I had for printing potentially long long variables in PH/11
5439 above was not the best (it lost precision). The length of off_t variables
5440 is now inspected at build time, and an appropriate printing format (%ld
5441 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5442 to be "long long int" or "long int". This is needed for the internal
5443 formatting function string_vformat().
5445 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5446 the configuration file to be ":syslog", then the script "guesses" where
5447 the logs files are, rather than using the compiled in default. In our
5448 case the guess is not the same as the compiled default, so the script
5449 suddenly stopped working when I started to use syslog. The patch checks
5450 to see if log_file_path is "". If so, it attempts to read it from exim
5451 with no configuration file to get the compiled in version, before it
5452 falls back to the previous guessing code."
5454 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5455 implementing BATV in an Exim configuration. See NewStuff for the gory
5458 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5459 Makefile that are specific to HP-UX.
5461 PH/18 If the "use_postmaster" option was set for a recipient callout together
5462 with the "random" option, the postmaster address was used as the MAIL
5463 FROM address for the random test, but not for the subsequent recipient
5464 test. It is now used for both.
5466 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5467 patch removes a few documentation additions to RFC 3028, because the
5468 latest draft now contains them. It adds the new en;ascii-case comparator
5469 and a new error check for 8bit text in MIME parts. Comparator and
5470 require names are now matched exactly. I enabled the subaddress
5471 extension, but it is not well tested yet (read: it works for me)."
5473 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5474 rework some of the code of TK/09 above to avoid the hardwired use of
5475 "%lld" and "long long". Replaced the call to snprintf() with a call to
5478 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5479 records point to non-existent hosts", "retry timeout exceeded", and
5480 "retry time not reached for any host after a long failure period".
5482 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5483 experimental DomainKeys support:
5485 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5486 (2) On an error such as an illegally used "control", the wrong name for
5487 the control was given.
5489 These problems did NOT occur unless DomainKeys support was compiled.
5491 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5493 PH/24 Added ${if match_ip condition.
5495 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5496 databases so that it will be absolutely obvious if a crash occurs in the
5497 DB library. This is a regular occurrence (often caused by mis-matched
5500 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5501 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5502 picked up on a user's system that detects such things. There doesn't seem
5503 to be a gcc warning option for this - only an attribute that has to be
5504 put on the function's prototype. It seems that in Fedora Core 4 they have
5505 set this on a number of new functions. No doubt there will be more in due
5508 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5509 specify a transport. However, if an address that was verified by such a
5510 router was the subject of a callout, Exim crashed because it tried to
5511 read the rcpt_include_affixes from the non-existent transport. Now it
5512 just assumes that the setting of that option is false. This bug was
5513 introduced by 4.51/PH/31.
5515 PH/28 Changed -d+all to exclude +memory, because that information is very
5516 rarely of interest, but it makes the output a lot bigger. People tend to
5517 do -d+all out of habit.
5519 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5520 code in os-type was giving problems when libc.so lives in lib64, like on
5523 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5524 aren't the modern standard, and it seems that some systems' include files
5525 don't always have them. Exim was already checking for some of the newer
5526 ones like T_AAAA, and defining it itself. I've added checks for all the
5527 record types that Exim uses.
5529 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5530 not automatically generating a new one, as it is supposed to. This
5531 prevented TLS from working. If the file did exist, but contained invalid
5532 data, a new version was generated, as expected. It was only the case of a
5533 non-existent file that was broken.
5535 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5536 with a change in libdomainkeys > 0.64.
5538 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5539 from DNS. If the selector record carries the flag, it now has
5540 precedence over the domain-wide flag.
5542 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5544 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5545 the use of a port name that isn't defined in /etc/services) occurred, the
5546 message was deferred as in a normal delivery, and thus remained on the
5547 spool, instead of being failed because of the mua_wrapper setting. This
5548 is now fixed, and I tidied up some of the mua_wrapper messages at the
5551 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5552 the messages in a hash of arrays rather than using individual hashes.
5553 This is a bit cleaner and results in dramatic memory savings, albeit
5554 at a slight CPU cost.
5556 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5557 as requested by Marc Sherman.
5559 SC/10 Eximstats: added histograms for user specified patterns as requested
5562 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5564 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5565 fopen() in the content-scanning modules that did not already have it.
5571 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5572 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5574 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5576 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5578 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5579 to test Sieve filters that use "vacation".
5581 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5582 that changes the way the GnuTLS parameters are stored in the cache file.
5583 The new format can be generated externally. For backward compatibility,
5584 if the data in the cache doesn't make sense, Exim assumes it has read an
5585 old-format file, and it generates new data and writes a new file. This
5586 means that you can't go back to an older release without removing the
5589 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5590 work if there are any delivery delays because "one_time" forces the
5591 parent to be marked "delivered", so its unseen clone is never tried
5592 again. For this reason, Exim now forbids the simultaneous setting of
5595 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5596 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5597 entry if you want to know the details. The fix, however, neglected to
5598 consider the case when local delivery batching is involved. The test for
5599 "previously delivered" was not happening when checking to see if an
5600 address could be batched with a previous (undelivered) one; under
5601 certain circumstances this could lead to multiple deliveries to the same
5604 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5605 in its include files, and this causes problems building Exim.
5607 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5608 header_syntax) but Exim was just ignoring anything given after a slash.
5609 In particular, this caused confusion with an attempt to use "verify =
5610 reverse_host_lookup/defer_ok". An error is now given when options are
5611 supplied for verify items that do not have them. (Maybe reverse_host_
5612 lookup should have a defer_ok option, but that's a different point.)
5614 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5615 defined by RFC 821) to 2048, because there were problems with some AUTH
5616 commands, and RFC 1869 says the size should be increased for extended
5617 SMTP commands that take arguments.
5619 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5622 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5623 "unknown" error; now it says that the functionality isn't in the binary.
5625 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5626 an address' error message when a string expansion fails (syntax or
5627 whatever). Otherwise the password may appear in the log. Following change
5628 PH/42 below, there is no longer a chance of it appearing in a bounce
5631 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5633 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5634 of its addresses was ever tried. (Bugzilla bug #2.)
5636 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5637 the result incorrectly in the debug output. (It correctly added a newline
5638 to what was transported.)
5640 TF/01 Added $received_time.
5642 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5643 commented out examples of how to interface to a virus scanner and to
5644 SpamAssassin. Also added commented examples of av_scanner and
5645 spamd_address settings.
5647 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5648 and controls are allowed in which ACLs. There were a couple of minor
5649 errors. Some of the entries in the conditions table (which is a table of
5650 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5651 negation of where the condition IS allowed.
5653 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5655 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5656 header file does not have a version number, so I've had to invent a new
5657 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5658 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5659 radiusclient), but it was contributed by a Radius user.
5661 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5662 files or directories, for OpenSSL.
5664 PH/19 When an Exim process that is running as root has to create an Exim log
5665 file, it does so in a subprocess that runs as exim:exim so as to get the
5666 ownership right at creation (otherwise, other Exim processes might see
5667 the file with the wrong ownership). There was no test for failure of this
5668 fork() call, which would lead to the process getting stuck as it waited
5669 for a non-existent subprocess. Forks do occasionally fail when resources
5670 run out. I reviewed all the other calls to fork(); they all seem to check
5673 PH/20 When checking for unexpected SMTP input at connect time (before writing
5674 the banner), Exim was not dealing correctly with a non-positive return
5675 from the read() function. If the client had disconnected by this time,
5676 the result was a log entry for a synchronization error with an empty
5677 string after "input=" when read() returned zero. If read() returned -1
5678 (an event I could not check), uninitialized data bytes were printed.
5679 There were reports of junk text (parts of files, etc) appearing after
5682 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5684 PH/22 Added support for macro redefinition, and (re)definition in between
5685 driver and ACL definitions.
5687 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5688 forgetting to use the resulting value; it was using the unexpanded value.
5690 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5691 hadn't been configured. The fix is from Juergen Kreileder, who
5692 understands it better than I do:
5694 "Here's what I see happening with three configured cyrus_sasl
5695 authenticators configured (plain, login, cram-md5):
5697 On startup auth_cyrus_sasl_init() gets called for each of these.
5698 This means three calls to sasl_listmech() without a specified mech_list.
5699 => SASL tests which mechs of all available mechs actually work
5700 => three warnings about OTP not working
5701 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5703 With the patch, sasl_listmech() also gets called three times. But now
5704 SASL's mech_list option is set to the server_mech specified in the the
5705 authenticator. Or in other words, the answer from sasl_listmech()
5706 gets limited to just the mech you're testing for (which is different
5708 => the return list contains just 'plain' or 'login', 'cram-md5' or
5709 nothing depending on the value of ob->server_mech.
5711 I've just tested the patch: Authentication still works fine,
5712 unavailable mechs specified in the exim configuration are still
5713 caught, and the auth.log warnings about OTP are gone."
5715 PH/25 When debugging is enabled, the contents of the command line are added
5716 to the debugging output, even when log_selector=+arguments is not
5719 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5720 answer is "GNU", and only if the return is "GNU/something" is the answer
5723 PH/27 $acl_verify_message is now set immediately after the failure of a
5724 verification in an ACL, and so is available in subsequent modifiers. In
5725 particular, the message can be preserved by coding like this:
5727 warn !verify = sender
5728 set acl_m0 = $acl_verify_message
5730 Previously, $acl_verify_message was set only while expanding "message"
5731 and "log_message" when a very denied access.
5733 PH/28 Modified OS/os.c-Linux with
5735 -#ifndef OS_LOAD_AVERAGE
5736 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5738 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5739 nomenclature these days.)
5741 PH/29 Installed patch from the Sieve maintainer that adds the options
5742 sieve_useraddress and sieve_subaddress to the redirect router.
5744 PH/30 In these circumstances:
5745 . Two addresses routed to the same list of hosts;
5746 . First host does not offer TLS;
5747 . First host accepts first address;
5748 . First host gives temporary error to second address;
5749 . Second host offers TLS and a TLS session is established;
5750 . Second host accepts second address.
5751 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5752 and peerdn, if requested) that were in fact used only for the second
5755 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5756 attention to any local part prefix or suffix that was matched by the
5757 router that accepted the address. It now behaves in the same way as it
5758 does for delivery: the affixes are removed from the local part unless
5759 rcpt_include_affixes is set on the transport.
5761 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5762 timeout during the DATA phase of an incoming message.
5764 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5765 applied a patch sanctioned by the Sieve maintainer.
5767 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5768 the uid or gid is negative. A case of a negative gid caused this to be
5769 noticed. The fix allows for either to be negative.
5771 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5772 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5775 PH/36 The change PH/12 above was broken. Fixed it.
5777 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5778 the grounds that routing the same address twice would always produce the
5779 same answer. This might have been true once, but it is certainly no
5780 longer true now. Routing a child address may depend on the previous
5781 routing that produced that child. Some complicated redirection strategies
5782 went wrong when messages had multiple recipients, and made Exim's
5783 behaviour dependent on the order in which the addresses were given.
5785 I have moved the duplicate checking until after the routing is complete.
5786 Exim scans the addresses that are assigned to local and remote
5787 transports, and removes any duplicates. This means that more work will be
5788 done, as duplicates will always all be routed, but duplicates are
5789 presumably rare, so I don't expect this is of any significance.
5791 For deliveries to pipes, files, and autoreplies, the duplicate checking
5792 still happens during the routing process, since they are not going to be
5795 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5796 It corrects a timeout issue with spamd. This is Ian's comment: "The
5797 background is that sometimes spamd either never reads data from a
5798 connection it has accepted, or it never writes response data. The exiscan
5799 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5800 blindly assumes that writes won't block so it may never time out."
5802 PH/39 Allow G after quota size as well as K and M.
5804 PH/40 The value set for $authenticated_id in an authenticator may not contain
5805 binary zeroes or newlines because the value is written to log lines and
5806 to spool files. There was no check on this. Now the value is run through
5807 the string_printing() function so that such characters are converted to
5808 printable escape sequences.
5810 PH/41 $message_linecount is a new variable that contains the total number of
5811 lines in the message. Compare $body_linecount, which is the count for the
5814 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5815 bounce and delay warning messages, except in certain special cases, which
5818 (a) An SMTP error message from a remote host;
5819 (b) A message specified in a :fail: redirection;
5820 (c) A message specified in a "fail" command in a system filter;
5821 (d) A message specified in a FAIL return from the queryprogram router;
5822 (e) A message specified by the cannot_route_message router option.
5824 In these cases only, Exim does include the error details in bounce and
5825 warning messages. There are also a few cases where bland messages such
5826 as "unrouteable address" or "local delivery error" are given.
5828 PH/43 $value is now also set for the "else" part of a ${run expansion.
5830 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5831 being worked on, but at least Exim now implements the latest version to
5834 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5835 process to complete was treated as a delivery failure, a timeout while
5836 writing the message to the pipe was logged, but erroneously treated as a
5837 successful delivery. Such timeouts include transport filter timeouts. For
5838 consistency with the overall process timeout, these timeouts are now
5839 treated as errors, giving rise to delivery failures by default. However,
5840 there is now a new Boolean option for the pipe transport called
5841 timeout_defer, which, if set TRUE, converts the failures into defers for
5842 both kinds of timeout. A transport filter timeout is now identified in
5845 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5846 systems where "make" and "gmake" are different, calling "gmake" at top
5847 level broke things. I've arranged for the value of $(MAKE) to be passed
5848 from the Makefile to this script so that it can call the same version of
5852 A note about Exim versions 4.44 and 4.50
5853 ----------------------------------------
5855 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5856 changes of various kinds. As a consequence, a big documentation update was
5857 needed. This delayed the release for rather longer than seemed good, especially
5858 in the light of a couple of (minor) security issues. Therefore, the changes
5859 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5860 release. So 4.44 and 4.50 are in effect two different branches that both start
5863 I have left the 4.50 change log unchanged; it contains all the changes since
5864 4.43. The change log for 4.44 is below; many of its items are identical to
5865 those for 4.50. This seems to be the most sensible way to preserve the
5866 historical information.
5872 1. Minor wording change to the doc/README.SIEVE file.
5874 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5875 computation of the current number of files was incorrect.
5877 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5878 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5881 4. Give more explanation in the error message when the command for a transport
5882 filter fails to execute.
5884 5. There are several places where Exim runs a non-Exim command in a
5885 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5886 was being done only for the command run by the queryprogram router. It is
5887 now done for all such subprocesses. The other cases are: ${run, transport
5888 filters, and the commands run by the lmtp and pipe transports.
5890 6. Added CONFIGURE_GROUP build-time option.
5892 7. Some older OS have a limit of 256 on the maximum number of file
5893 descriptors. Exim was using setrlimit() to set 1000 as a large value
5894 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5895 systems. I've change it so that if it can't get 1000, it tries for 256.
5897 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5898 was an oversight, and furthermore, ever since the addition of extra
5899 controls (e.g. 4.43/32), the checks on when to allow different forms of
5900 "control" were broken. There should now be diagnostics for all cases when a
5901 control that does not make sense is encountered.
5903 9. Added the /retain_sender option to "control=submission".
5905 10. $recipients is now available in the predata ACL (oversight).
5907 11. Tidy the search cache before the fork to do a delivery from a message
5908 received from the command line. Otherwise the child will trigger a lookup
5909 failure and thereby defer the delivery if it tries to use (for example) a
5910 cached ldap connection that the parent has called unbind on.
5912 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5913 of $address_data from the recipient verification was clobbered by the
5914 sender verification.
5916 13. The value of address_data from a sender verification is now available in
5917 $sender_address_data in subsequent conditions in the ACL statement.
5919 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5921 15. Added a new option "connect=<time>" to callout options, to set a different
5924 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5925 was its contents. (It was OK if the option was not defined at all.)
5927 17. A "Completed" log line is now written for messages that are removed from
5928 the spool by the -Mrm option.
5930 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5931 information about exactly what failed.
5933 19. Added -dd to debug only the daemon process.
5935 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5936 handles timeouts, both on the server side and network timeouts. Renamed the
5937 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5939 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5940 It is now set to "smtps".
5942 22. $host_address is now set to the target address during the checking of
5943 ignore_target_hosts.
5945 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5946 being passed; this would have caused $sender_host_name to have been used if
5947 matching the list had actually called for a host name (not very likely,
5948 since this list is usually IP addresses). A host name is now passed as
5951 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5952 code that specifies a non-restarting handler (typically sigaction() in
5953 modern systems) in an attempt to fix a rare and obscure crash bug.
5955 25. Narrowed the window for a race in the daemon that could cause it to ignore
5956 SIGCHLD signals. This is not a major problem, because they are used only to
5957 wake it up if nothing else does.
5959 26. A malformed maildirsize file could cause Exim to calculate negative values
5960 for the mailbox size or file count. Odd effects could occur as a result.
5961 The maildirsize information is now recalculated if the size or filecount
5964 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5965 support for a long time. Removed HAVE_SYS_VFS_H.
5967 28. Installed the latest version of exipick from John Jetmore.
5969 29. In an address list, if the pattern was not a regular expression, an empty
5970 subject address (from a bounce message) matched only if the pattern was an
5971 empty string. Non-empty patterns were not even tested. This was the wrong
5972 because it is perfectly reasonable to use an empty address as part of a
5973 database query. An empty address is now tested by patterns that are
5974 lookups. However, all the other forms of pattern expect the subject to
5975 contain a local part and a domain, and therefore, for them, an empty
5976 address still always fails if the pattern is not itself empty.
5978 30. Exim went into a mad DNS loop when attempting to do a callout where the
5979 host was specified on an smtp transport, and looking it up yielded more
5980 than one IP address.
5982 31. Re-factored the code for checking spool and log partition space into a
5983 function that finds that data and another that does the check. The former
5984 is then used to implement four new variables: $spool_space, $log_space,
5985 $spool_inodes, and $log_inodes.
5987 32. The RFC2047 encoding function was originally intended for short strings
5988 such as real names; it was not keeping to the 75-character limit for
5989 encoded words that the RFC imposes. It now respects the limit, and
5990 generates multiple encoded words if necessary. To be on the safe side, I
5991 have increased the buffer size for the ${rfc2047: expansion operator from
5994 33. It is now permitted to omit both strings after an "if" condition; if the
5995 condition is true, the result is "true". As before, when the second string
5996 is omitted, a false condition yields an empty string. This makes it less
5997 cumbersome to write custom ACL and router conditions.
5999 34. Failure to deliver a bounce message always caused it to be frozen, even if
6000 there was an errors_to setting on the router. The errors_to setting is now
6003 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6004 canonical form (fully expanded) before being placed in
6005 $sender_host_address.
6007 36. The table in the code that translates DNS record types into text (T_A to
6008 "A" for instance) was missing entries for NS and CNAME. It is just possible
6009 that this could have caused confusion if both these types were looked up
6010 for the same domain, because the text type is used as part of Exim's
6011 per-process caching. But the chance of anyone hitting this buglet seems
6014 37. The dnsdb lookup has been extended in a number of ways.
6016 (1) There is a new type, "zns", which walks up the domain tree until it
6017 finds some nameserver records. It should be used with care.
6019 (2) There is a new type, "mxh", which is like "mx" except that it returns
6020 just the host names, not the priorities.
6022 (3) It is now possible to give a list of domains (or IP addresses) to be
6023 looked up. The behaviour when one of the lookups defers can be
6024 controlled by a keyword.
6026 (4) It is now possible to specify the separator character for use when
6027 multiple records are returned.
6029 38. The dnslists ACL condition has been extended: it is now possible to supply
6030 a list of IP addresses and/or domains to be looked up in a particular DNS
6033 39. Added log_selector=+queue_time_overall.
6035 40. When running the queue in the test harness, wait just a tad after forking a
6036 delivery process, to get repeatability of debugging output.
6038 41. Include certificate and key file names in error message when GnuTLS fails
6039 to set them up, because the GnuTLS error message doesn't include the name
6040 of the failing file when there is a problem reading it.
6042 42. Allow both -bf and -bF in the same test run.
6044 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6046 44. The "Exiscan patch" is now merged into the mainline Exim source.
6048 45. Sometimes the final signoff response after QUIT could fail to get
6049 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6050 < 0 before doing a fflush(). This bug looks as though it goes back to the
6051 introduction of TLS in release 3.20, but "sometimes" must have been rare
6052 because the tests only now provoked it.
6054 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6055 (this can affect the format of dates).
6057 47. exim_tidydb, when checking for the continued existence of a message for
6058 which it has found a message-specific retry record, was not finding
6059 messages that were in split spool directories. Consequently, it was
6060 deleting retry records that should have stayed in existence.
6062 48. Steve fixed some bugs in eximstats.
6064 49. The SPA authentication driver was not abandoning authentication and moving
6065 on to the next authenticator when an expansion was forced to fail,
6066 contradicting the general specification for all authenticators. Instead it
6067 was generating a temporary error. It now behaves as specified.
6069 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6070 (the order specifies the preference for clients). The order is now AES256,
6071 AES128, 3DES, ARCFOUR128.
6073 51. Small patch to Sieve code - explicitly set From: when generating an
6076 52. Exim crashed if a remote delivery caused a very long error message to be
6077 recorded - for instance if somebody sent an entire SpamAssassin report back
6078 as a large number of 550 error lines. This bug was coincidentally fixed by
6079 increasing the size of one of Exim's internal buffers (big_buffer) that
6080 happened as part of the Exiscan merge. However, to be on the safe side, I
6081 have made the code more robust (and fixed the comments that describe what
6084 53. Now that there can be additional text after "Completed" in log lines (if
6085 the queue_time_overall log selector is set), a one-byte patch to exigrep
6086 was needed to allow it to recognize "Completed" as not the last thing in
6089 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6090 patch that reportedly fixes this has been added. I am not expert enough to
6091 create a test for it. This is what the patch creator wrote:
6093 "I found a little strange behaviour of ldap code when working with
6094 Windows 2003 AD Domain, where users was placed in more than one
6095 Organization Units. When I tried to give exim partial DN, the exit code
6096 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6097 But simultaneously result of request was absolutely normal ldap result,
6098 so I produce this patch..."
6100 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6101 REFERENCE, so I have modified the code to exclude the patch when that macro
6104 55. Some experimental protocols are using DNS PTR records for new purposes. The
6105 keys for these records are domain names, not reversed IP addresses. The
6106 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6107 leaves it alone. Component reversal etc. now happens only for IP addresses.
6110 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6112 57. Double the size of the debug message buffer (to 2048) so that more of very
6113 long debug lines gets shown.
6115 58. The exicyclog utility now does better if the number of log files to keep
6116 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6118 59. Two changes related to the smtp_active_hostname option:
6120 (1) $smtp_active_hostname is now available as a variable.
6121 (2) The default for smtp_banner uses $smtp_active_hostname instead
6122 of $primary_hostname.
6124 60. The host_aton() function is supposed to be passed a string that is known
6125 to be a valid IP address. However, in the case of IPv6 addresses, it was
6126 not checking this. This is a hostage to fortune. Exim now panics and dies
6127 if the condition is not met. A case was found where this could be provoked
6128 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6129 components; fortuitously, this particular loophole had already been fixed
6130 by change 4.50/55 above.
6132 If there are any other similar loopholes, the new check in host_aton()
6133 itself should stop them being exploited. The report I received stated that
6134 data on the command line could provoke the exploit when Exim was running as
6135 exim, but did not say which command line option was involved. All I could
6136 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6137 running as the user.
6140 61. There was a buffer overflow vulnerability in the SPA authentication code
6141 (which came originally from the Samba project). I have added a test to the
6142 spa_base64_to_bits() function which I hope fixes it.
6145 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6146 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6148 63. The daemon start-up calls getloadavg() while still root for those OS that
6149 need the first call to be done as root, but it missed one case: when
6150 deliver_queue_load_max is set with deliver_drop_privilege. This is
6151 necessary for the benefit of the queue runner, because there is no re-exec
6152 when deliver_drop_privilege is set.
6154 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6155 This has been fixed.
6157 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6158 was in use, was not putting the data itself into the right store pool;
6159 consequently, it could be overwritten for a subsequent message in the same
6160 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6163 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6165 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6166 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6167 the function were treating the return as a boolean value, which happened to
6168 work because 0=false and not-0=true, but is not correct code.
6170 68. The host_aton() function was not handling scoped IPv6 addresses (those
6171 with, for example, "%eth0" on the end) correctly.
6173 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6174 negated items (that is, ~something) in unsigned ints. Some compilers
6175 apparently mutter when there is no cast.
6177 70. If an address verification called from an ACL failed, and did not produce a
6178 user-specific message (i.e. there was only a "system" message), nothing was
6179 put in $acl_verify_message. In this situation, it now puts the system
6182 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6183 session; change 4.31/43 added the unwanted input to the log line - except
6184 that it did not do this in the start of session case. It now does.
6186 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6187 This is wrong and can cause the other end to generate a synchronization
6188 error if it is another Exim or anything else that does the synchronization
6189 check. A QUIT command is no longer sent after a timeout.
6191 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6192 during host lookups.
6194 74. The defer_ok option of callout verification was not working if it was used
6195 when verifying addresses in header lines, that is, for this case:
6197 verify = header_sender/callout=defer_ok
6199 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6200 those file descriptors could be used for SMTP connections. If anything
6201 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6202 could be sent to the SMTP client, causing chaos. The daemon now opens
6203 stdin, stdout, and stderr to /dev/null when it puts itself into the
6206 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6207 log by default. The user can override this with suitable Perl magic.
6209 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6210 the log message when discard triggers, was not working for the DATA ACL or
6211 for the non-SMTP ACL.
6213 78. Error message wording change in sieve.c.
6215 79. If smtp_accept_max_per_host was set, the number of connections could be
6216 restricted to fewer than expected, because the daemon was trying to set up
6217 a new connection before checking whether the processes handling previous
6218 connections had finished. The check for completed processes is now done
6219 earlier. On busy systems, this bug wouldn't be noticed because something
6220 else would have woken the daemon, and it would have reaped the completed
6223 80. If a message was submitted locally by a user whose login name contained one
6224 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6225 It caused a spool format error. I have fixed the spool reading code. A
6226 related problem was that the "from" clause in the Received: line became
6227 illegal because of the space(s). It is now covered by ${quote_local_part.
6229 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6232 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6234 83. Patch from Sieve maintainer to fix unterminated string problem in
6235 "vacation" handling.
6237 84. Some minor changes to the Linux configuration files to help with other
6238 OS variants using glibc.
6240 85. One more patch for Sieve to update vacation handling to latest spec.
6243 ----------------------------------------------------
6244 See the note above about the 4.44 and 4.50 releases.
6245 ----------------------------------------------------
6251 1. Change 4.43/35 introduced a bug that caused file counts to be
6252 incorrectly computed when quota_filecount was set in an appendfile
6255 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6256 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6259 3. Give more explanation in the error message when the command for a transport
6260 filter fails to execute.
6262 4. There are several places where Exim runs a non-Exim command in a
6263 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6264 was being done only for the command run by the queryprogram router. It is
6265 now done for all such subprocesses. The other cases are: ${run, transport
6266 filters, and the commands run by the lmtp and pipe transports.
6268 5. Some older OS have a limit of 256 on the maximum number of file
6269 descriptors. Exim was using setrlimit() to set 1000 as a large value
6270 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6271 systems. I've change it so that if it can't get 1000, it tries for 256.
6273 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6274 was an oversight, and furthermore, ever since the addition of extra
6275 controls (e.g. 4.43/32), the checks on when to allow different forms of
6276 "control" were broken. There should now be diagnostics for all cases when a
6277 control that does not make sense is encountered.
6279 7. $recipients is now available in the predata ACL (oversight).
6281 8. Tidy the search cache before the fork to do a delivery from a message
6282 received from the command line. Otherwise the child will trigger a lookup
6283 failure and thereby defer the delivery if it tries to use (for example) a
6284 cached ldap connection that the parent has called unbind on.
6286 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6287 of $address_data from the recipient verification was clobbered by the
6288 sender verification.
6290 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6291 was its contents. (It was OK if the option was not defined at all.)
6293 11. A "Completed" log line is now written for messages that are removed from
6294 the spool by the -Mrm option.
6296 12. $host_address is now set to the target address during the checking of
6297 ignore_target_hosts.
6299 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6300 being passed; this would have caused $sender_host_name to have been used if
6301 matching the list had actually called for a host name (not very likely,
6302 since this list is usually IP addresses). A host name is now passed as
6305 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6306 code that specifies a non-restarting handler (typically sigaction() in
6307 modern systems) in an attempt to fix a rare and obscure crash bug.
6309 15. Narrowed the window for a race in the daemon that could cause it to ignore
6310 SIGCHLD signals. This is not a major problem, because they are used only to
6311 wake it up if nothing else does.
6313 16. A malformed maildirsize file could cause Exim to calculate negative values
6314 for the mailbox size or file count. Odd effects could occur as a result.
6315 The maildirsize information is now recalculated if the size or filecount
6318 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6319 support for a long time. Removed HAVE_SYS_VFS_H.
6321 18. Updated exipick to current release from John Jetmore.
6323 19. Allow an empty sender to be matched against a lookup in an address list.
6324 Previously the only cases considered were a regular expression, or an
6327 20. Exim went into a mad DNS lookup loop when doing a callout where the
6328 host was specified on the transport, if the DNS lookup yielded more than
6331 21. The RFC2047 encoding function was originally intended for short strings
6332 such as real names; it was not keeping to the 75-character limit for
6333 encoded words that the RFC imposes. It now respects the limit, and
6334 generates multiple encoded words if necessary. To be on the safe side, I
6335 have increased the buffer size for the ${rfc2047: expansion operator from
6338 22. Failure to deliver a bounce message always caused it to be frozen, even if
6339 there was an errors_to setting on the router. The errors_to setting is now
6342 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6343 canonical form (fully expanded) before being placed in
6344 $sender_host_address.
6346 24. Updated eximstats to version 1.33
6348 25. Include certificate and key file names in error message when GnuTLS fails
6349 to set them up, because the GnuTLS error message doesn't include the name
6350 of the failing file when there is a problem reading it.
6352 26. Expand error message when OpenSSL has problems setting up cert/key files.
6355 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6356 (this can affect the format of dates).
6358 28. exim_tidydb, when checking for the continued existence of a message for
6359 which it has found a message-specific retry record, was not finding
6360 messages that were in split spool directories. Consequently, it was
6361 deleting retry records that should have stayed in existence.
6363 29. eximstats updated to version 1.35
6364 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6365 1.35 - bugfix such that pie charts by volume are generated correctly
6367 30. The SPA authentication driver was not abandoning authentication and moving
6368 on to the next authenticator when an expansion was forced to fail,
6369 contradicting the general specification for all authenticators. Instead it
6370 was generating a temporary error. It now behaves as specified.
6372 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6373 (the order specifies the preference for clients). The order is now AES256,
6374 AES128, 3DES, ARCFOUR128.
6376 31. Small patch to Sieve code - explicitly set From: when generating an
6379 32. Exim crashed if a remote delivery caused a very long error message to be
6380 recorded - for instance if somebody sent an entire SpamAssassin report back
6381 as a large number of 550 error lines. This bug was coincidentally fixed by
6382 increasing the size of one of Exim's internal buffers (big_buffer) that
6383 happened as part of the Exiscan merge. However, to be on the safe side, I
6384 have made the code more robust (and fixed the comments that describe what
6387 33. Some experimental protocols are using DNS PTR records for new purposes. The
6388 keys for these records are domain names, not reversed IP addresses. The
6389 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6390 leaves it alone. Component reversal etc. now happens only for IP addresses.
6393 34. The host_aton() function is supposed to be passed a string that is known
6394 to be a valid IP address. However, in the case of IPv6 addresses, it was
6395 not checking this. This is a hostage to fortune. Exim now panics and dies
6396 if the condition is not met. A case was found where this could be provoked
6397 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6398 components; fortuitously, this particular loophole had already been fixed
6399 by change 4.50/55 or 4.44/33 above.
6401 If there are any other similar loopholes, the new check in host_aton()
6402 itself should stop them being exploited. The report I received stated that
6403 data on the command line could provoke the exploit when Exim was running as
6404 exim, but did not say which command line option was involved. All I could
6405 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6406 running as the user.
6409 35. There was a buffer overflow vulnerability in the SPA authentication code
6410 (which came originally from the Samba project). I have added a test to the
6411 spa_base64_to_bits() function which I hope fixes it.
6414 36. The daemon start-up calls getloadavg() while still root for those OS that
6415 need the first call to be done as root, but it missed one case: when
6416 deliver_queue_load_max is set with deliver_drop_privilege. This is
6417 necessary for the benefit of the queue runner, because there is no re-exec
6418 when deliver_drop_privilege is set.
6420 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6421 was in use, was not putting the data itself into the right store pool;
6422 consequently, it could be overwritten for a subsequent message in the same
6423 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6426 38. Sometimes the final signoff response after QUIT could fail to get
6427 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6428 < 0 before doing a fflush(). This bug looks as though it goes back to the
6429 introduction of TLS in release 3.20, but "sometimes" must have been rare
6430 because the tests only now provoked it.
6436 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6437 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6438 Now it returns an integer. A number of calls were still expecting a T/F
6439 return. Fortuitously, in all cases, the tests worked in OK situations,
6440 which is the norm. However, things would have gone wrong on any write
6441 failures on the smtp file descriptor. This function is used when sending
6442 messages over SMTP and also when doing verify callouts.
6444 2. When Exim is called to do synchronous delivery of a locally submitted
6445 message (the -odf or -odi options), it no longer closes stderr before doing
6448 3. Implemented the mua_wrapper option.
6450 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6452 5. Implemented the functions header_remove(), header_testname(),
6453 header_add_at_position(), and receive_remove_recipient(), and exported them
6456 6. If an ACL "warn" statement specified the addition of headers, Exim already
6457 inserted X-ACL-Warn: at the start if there was no header name. However, it
6458 was not making this test for the second and subsequent header lines if
6459 there were newlines in the string. This meant that an invalid header could
6460 be inserted if Exim was badly configured.
6462 7. Allow an ACL "warn" statement to add header lines at the start or after all
6463 the Received: headers, as well as at the end.
6465 8. Added the rcpt_4xx retry error code.
6467 9. Added postmaster_mailfrom=xxx to callout verification option.
6469 10. Added mailfrom=xxxx to the callout verification option, for verify=
6472 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6473 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6475 12. Inserted some casts to stop certain compilers warning when using pointer
6476 differences as field lengths or precisions in printf-type calls (mostly
6477 affecting debugging statements).
6479 13. Added optional readline() support for -be (dynamically loaded).
6481 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6482 same clock tick as a message's arrival, so that its received time was the
6483 same as the "first fail" time on the retry record, and that message
6484 remained on the queue past the ultimate address timeout, every queue runner
6485 would try a delivery (because it was past the ultimate address timeout) but
6486 after another failure, the ultimate address timeout, which should have then
6487 bounced the address, did not kick in. This was a "< instead of <=" error;
6488 in most cases the first failure would have been in the next clock tick
6489 after the received time, and all would be well.
6491 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6492 being recognized when the domain list was tested by the match_domain
6493 condition in an expansion string.
6495 16. Added the ${str2b64: operator.
6497 17. Exim was always calling setrlimit() to set a large limit for the number of
6498 processes, without checking whether the existing limit was already
6499 adequate. (It did check for the limit on file descriptors.) Furthermore,
6500 errors from getrlimit() and setrlimit() were being ignored. Now they are
6501 logged to the main and panic logs, but Exim does carry on, to try to do its
6502 job under whatever limits there are.
6504 18. Imported PCRE 5.0.
6506 19. Trivial typo in log message " temporarily refused connection" (the leading
6509 20. If the log selector return_path_on_delivery was set and an address was
6510 redirected to /dev/null, the delivery process crashed because it assumed
6511 that a return path would always be set for a "successful" delivery. In this
6512 case, the whole delivery is bypassed as an optimization, and therefore no
6515 21. Internal re-arrangement: the function for sending a challenge and reading
6516 a response while authentication was assuming a zero-terminated challenge
6517 string. It's now changed to take a pointer and a length, to allow for
6518 binary data in such strings.
6520 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6522 23. Exim was not respecting finduser_retries when seeking the login of the
6523 uid under which it was called; it was always trying 10 times. (The default
6524 setting of finduser_retries is zero.) Also, it was sleeping after the final
6525 failure, which is pointless.
6527 24. Implemented tls_on_connect_ports.
6529 25. Implemented acl_smtp_predata.
6531 26. If the domain in control=submission is set empty, Exim assumes that the
6532 authenticated id is a complete email address when it generates From: or
6533 Sender: header lines.
6535 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6536 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6537 chown and chgrp in /bin and hostname in /usr/bin.
6539 28. Exim was keeping the "process log" file open after each use, just as it
6540 does for the main log. This opens the possibility of it remaining open for
6541 long periods when the USR1 signal hits a daemon. Occasional processlog
6542 errors were reported, that could have been caused by this. Anyway, it seems
6543 much more sensible not to leave this file open at all, so that is what now
6546 29. The long-running daemon process does not normally write to the log once it
6547 has entered its main loop, and it closes the log before doing so. This is
6548 so that log files can straightforwardly be renamed and moved. However,
6549 there are a couple of unusual error situations where the daemon does write
6550 log entries, and I had neglected to close the log afterwards.
6552 30. The text of an SMTP error response that was received during a remote
6553 delivery was being truncated at 512 bytes. This is too short for some of
6554 the long messages that one sometimes sees. I've increased the limit to
6557 31. It is now possible to make retry rules that apply only when a message has a
6558 specific sender, in particular, an empty sender.
6560 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6561 it possible to be selective about when SMTP synchronization is enforced.
6563 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6565 32. Implemented hosts_connection_nolog.
6567 33. Added an ACL for QUIT.
6569 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6572 35. Added mailbox_size and mailbox_filecount to appendfile.
6574 36. Added control = no_multiline_responses to ACLs.
6576 37. There was a bug in the logic of the code that waits for the clock to tick
6577 in the case where the clock went backwards by a substantial amount such
6578 that the microsecond fraction of "now" was more than the microsecond
6579 fraction of "then" (but the whole seconds number was less).
6581 38. Added support for the libradius Radius client library this is found on
6582 FreeBSD (previously only the radiusclient library was supported).
6588 1. When certain lookups returned multiple values in the form name=value, the
6589 quoting of the values was not always being done properly. Specifically:
6590 (a) If the value started with a double quote, but contained no whitespace,
6592 (b) If the value contained whitespace other than a space character (i.e.
6593 tabs or newlines or carriage returns) it was not quoted.
6594 This fix has been applied to the mysql and pgsql lookups by writing a
6595 separate quoting function and calling it from the lookup code. The fix
6596 should probably also be applied to nisplus, ibase and oracle lookups, but
6597 since I cannot test any of those, I have not disturbed their existing code.
6599 2. A hit in the callout cache for a specific address caused a log line with no
6600 reason for rejecting RCPT. Now it says "Previous (cached) callout
6601 verification failure".
6603 3. There was an off-by-one bug in the queryprogram router. An over-long
6604 return line was truncated at 256 instead of 255 characters, thereby
6605 overflowing its buffer with the terminating zero. As well as fixing this, I
6606 have increased the buffer size to 1024 (and made a note to document this).
6608 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6609 when Exim is waiting for an SMTP response from a remote server, Exim
6610 restarts its select() call on the socket, thereby resetting its timeout.
6611 This is not a problem when such interrupts are rare. Somebody set up a cron
6612 job to run exiwhat every 2 minutes, which is less than the normal select()
6613 timeout (5 or 10 minutes). This meant that the select() timeout never
6614 kicked in because it was always reset. I have fixed this by comparing the
6615 time when an interrupt arrives with the time at the start of the first call
6616 to select(). If more time than the timeout has elapsed, the interrupt is
6617 treated as a timeout.
6619 5. Some internal re-factoring in preparation for the addition of Sieve
6620 extensions (by MH). In particular, the "personal" test is moved to a
6621 separate function, and given an option for scanning Cc: and Bcc: (which is
6622 not set for Exim filters).
6624 6. When Exim created an email address using the login of the caller as the
6625 local part (e.g. when creating a From: or Sender: header line), it was not
6626 quoting the local part when it contained special characters such as @.
6628 7. Installed new OpenBSD configuration files.
6630 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6631 try to make them clearer.
6633 9. Callout options, other than the timeout value, were being ignored when
6634 verifying sender addresses in header lines. For example, when using
6636 verify = header_sender/callout=no_cache
6638 the cache was (incorrectly) being used.
6640 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6641 only the Cygwin environment.
6643 11. When return_path_on_delivery was set as a log selector, if different remote
6644 addresses in the same message used different return paths and parallel
6645 remote delivery occurred, the wrong values would sometimes be logged.
6646 (Whenever a remote delivery process finished, the return path value from
6647 the most recently started remote delivery process was logged.)
6649 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6650 lines when AUTH and/or TLS are in use. This is the "received protocol"
6651 field. Exim used to use "asmtp" for authenticated SMTP, without any
6652 indication (in the protocol name) for TLS use. Now it follows the RFC and
6653 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6654 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6655 names appear in log lines as well as in Received: header lines.
6657 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6658 extensions, and comparison tests, and to fix some bugs.
6660 14. Changes to the "personal" filter test:
6662 (1) The test was buggy in that it was just doing the equivalent of
6663 "contains" tests on header lines. For example, if a user's address was
6664 anne@some.where, the "personal" test would incorrectly be true for
6666 To: susanne@some.where
6668 This test is now done by extracting each address from the header in turn,
6669 and checking the entire address. Other tests that are part of "personal"
6670 are now done using regular expressions (for example, to check local parts
6671 of addresses in From: header lines).
6673 (2) The list of non-personal local parts in From: addresses has been
6674 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6675 taken from the Sieve specification recommendations.
6677 (3) If the message contains any header line starting with "List-" it is
6678 treated as non-personal.
6680 (4) The test for "circular" in the Subject: header line has been removed
6681 because it now seems ill-conceived.
6683 15. Minor typos in src/EDITME comments corrected.
6685 16. Installed latest exipick from John Jetmore.
6687 17. If headers_add on a router specified a text string that was too long for
6688 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6689 of string_sprintf() is now avoided.
6691 18. $message_body_size was not set (it was always zero) when running the DATA
6692 ACL and the local_scan() function.
6694 19. For the "mail" command in an Exim filter, no default was being set for
6695 the once_repeat time, causing a random time value to be used if "once" was
6696 specified. (If the value happened to be <= 0, no repeat happened.) The
6697 default is now 0s, meaning "never repeat". The "vacation" command was OK
6698 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6699 (I found it when inspecting the code).
6701 20. There is now an overall timeout for performing a callout verification. It
6702 defaults to 4 times the callout timeout, which applies to individual SMTP
6703 commands during the callout. The overall timeout applies when there is more
6704 than one host that can be tried. The timeout is checked before trying the
6705 next host. This prevents very long delays if there are a large number of
6706 hosts and all are timing out (e.g. when the network connections are timing
6707 out). The value of the overall timeout can be changed by specifying an
6708 additional sub-option for "callout", called "maxwait". For example:
6710 verify = sender/callout=5s,maxwait=20s
6712 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6713 to the end before writing, but this should make it even safer).
6715 22. Exim was forgetting that it had advertised PIPELINING for the second and
6716 subsequent messages on an SMTP connection. It was also not resetting its
6717 memory on STARTTLS and an internal HELO.
6719 23. When Exim logs an SMTP synchronization error within a session, it now
6720 records whether PIPELINING has been advertised or not.
6722 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6723 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6724 rather than long int.
6726 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6728 26. Added the never_mail option to autoreply.
6734 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6735 crash if the getsockname() call failed; this can happen if a connection is
6736 closed very soon after it is established. The problem was simply in the
6737 order in which certain operations were done, causing Exim to try to write
6738 to the SMTP stream before it had set up the file descriptor. The bug has
6739 been fixed by making things happen in the correct order.
6745 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6746 before the connection was closed, thus losing the rejection response.
6748 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6749 some early Solaris releases, but causes trouble in current releases where
6750 socklen_t is defined.
6752 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6755 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6758 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6759 that releasing the top of it at the end releases what was used for sub-
6760 expansions (unless the block got too big). However, discard this block if
6761 the first thing is a variable or header, so that we can use its block when
6762 it is dynamic (useful for very large $message_headers, for example).
6764 6. Lookups now cache *every* query, not just the most recent. A new, separate
6765 store pool is used for this. It can be recovered when all lookup caches are
6766 flushed. Lookups now release memory at the end of their result strings.
6767 This has involved some general refactoring of the lookup sources.
6769 7. Some code has been added to the store_xxx() functions to reduce the amount
6770 of flapping under certain conditions.
6772 8. log_incoming_interface used to affect only the <= reception log lines. Now
6773 it causes the local interface and port to be added to several more SMTP log
6774 lines, for example "SMTP connection from", and rejection lines.
6776 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6778 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6780 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6781 could be overwritten at the end of the current message (or the start of a
6782 new message if it was set in a HELO ACL). The value is now preserved for
6783 the duration of the SMTP connection.
6785 12. If a transport had a headers_rewrite setting, and a matching header line
6786 contained an unqualified address, that address was qualified, even if it
6787 did not match any rewriting rules. The underlying bug was that the values
6788 of the flags that permit the existence of unqualified sender and recipient
6789 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6790 non-local messages, and by -bnq for local messages) were not being
6791 preserved with the message after it was received.
6793 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6794 "next input=" as part of the text comprising the host identity instead of
6795 the correct text. The code was using the same buffer for two different
6796 strings. However, depending on which order the printing function evaluated
6797 its arguments, the bug did not always show up. Under Linux, for example, my
6798 test suite worked just fine.
6800 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6801 doesn't work with some older versions of Perl. It has been changed to "my",
6802 which in any case is probably the better facility to use.
6804 15. A really picky compiler found some instances of statements for creating
6805 error messages that either had too many or two few arguments for the format
6808 16. The size of the buffer for calls to the DNS resolver has been increased
6809 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6810 for addresses that have a lot of PTR records. This alleviates a problem; it
6811 does not fully solve it.
6813 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6814 buffer now truncates the list and logs the incident, which is the same
6815 action as happens when Exim is looking up a host name and its aliases.
6816 Previously in this situation something unpredictable would happen;
6817 sometimes it was "internal error: store_reset failed".
6819 18. If a server dropped the connection unexpectedly when an Exim client was
6820 using GnuTLS and trying to read a response, the client delivery process
6821 crashed while trying to generate an error log message.
6823 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6824 string, for example:
6826 warn message = H1: something\nH2: something
6828 the text was added as a single header line from Exim's point of view
6829 though it ended up OK in the delivered message. However, searching for the
6830 second and subsequent header lines using $h_h2: did not work. This has been
6831 fixed. Similarly, if a system filter added multiple headers in this way,
6832 the routers could not see them.
6834 20. Expanded the error message when iplsearch is called with an invalid key to
6835 suggest using net-iplsearch in a host list.
6837 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6838 ACLs are no longer actually imposed (and a message to that effect is
6841 22. If a "gecos" field in a passwd entry contained escaped characters, in
6842 particular, if it contained a \" sequence, Exim got it wrong when building
6843 a From: or a Sender: header from that name. A second bug also caused
6844 incorrect handling when an unquoted " was present following a character
6845 that needed quoting.
6847 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6848 was not being matched caselessly.
6850 24. Arranged for all hyphens in the exim.8 source to be escaped with
6853 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6854 properly. Recipient callout cache records were still being keyed to include
6855 the sender, even when use_sender was set false. This led to far more
6856 callouts that were necessary. The sender is no longer included in the key
6857 when use_sender is false.
6859 26. Added "control = submission" modifier to ACLs.
6861 27. Added the ${base62d: operator to decode base 62 numbers.
6863 28. dnsdb lookups can now access SRV records.
6865 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6866 the configuration file.
6868 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6869 (-v) mode. This makes the output for a verbose queue run more intelligible.
6871 31. Added a use_postmaster feature to recipient callouts.
6873 32. Added the $body_zerocount variable, containing the number of binary zero
6874 bytes in the message body.
6876 33. The time of last modification of the "new" subdirectory is now used as the
6877 "mailbox time last read" when there is a quota error for a maildir
6880 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6882 35. Added +ignore_unknown as a special item in host lists.
6884 36. Code for decoding IPv6 addresses in host lists is now included, even if
6885 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6886 address was recognized as an IP address, but was then not correctly decoded
6887 into binary, causing unexpected and incorrect effects when compared with
6894 1. Very minor rewording of debugging text in manualroute to say "list of
6895 hosts" instead of "hostlist".
6897 2. If verify=header_syntax was set, and a header line with an unqualified
6898 address (no domain) and a large number of spaces between the end of the
6899 name and the colon was received, the reception process suffered a buffer
6900 overflow, and (when I tested it) crashed. This was caused by some obsolete
6901 code that should have been removed. The fix is to remove it!
6903 3. When running in the test harness, delay a bit after writing a bounce
6904 message to get a bit more predictability in the log output.
6906 4. Added a call to search_tidyup() just before forking a reception process. In
6907 theory, someone could use a lookup in the expansion of smtp_accept_max_
6908 per_host which, without the tidyup, could leave open a database connection.
6910 5. Added the variables $recipient_data and $sender_data which get set from a
6911 lookup success in an ACL "recipients" or "senders" condition, or a router
6912 "senders" option, similar to $domain_data and $local_part_data.
6914 6. Moved the writing of debug_print from before to after the "senders" test
6917 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6918 problems for message scanning, either using a data ACL, or using
6919 local_scan() because the Received: header was not generated till after they
6920 were called (in order to set the time as the time of reception completion).
6921 I have revised the way this works. The header is now generated after the
6922 body is received, but before the ACL or local_scan() are called. After they
6923 are run, the timestamp in the header is updated.
6929 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6930 before starting a queue runner without re-exec. This happened only when
6931 deliver_drop_privilege was set or when the Exim user was set to root. The
6932 effect of the bug was that timeouts during subsequent deliveries caused
6933 crashes instead of being properly handled. The handler is now left at its
6934 default (and expected) setting.
6936 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6937 message, again when deliver_drop_privilege is set or Exim is run as root.
6938 The bug described in (1) was not present in this case, but the tidying up
6939 of the other signals was missing. I have made the two cases consistent.
6941 3. The ignore_target_hosts setting on a manualroute router was being ignored
6942 for hosts that were looked up using the /MX notation.
6944 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6947 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6948 operated on the sender address. After changing the $sender_address to <>
6949 for the sender address verify, Exim was re-instated it as the original
6950 (before rewriting) address, but remembering that it had rewritten it, so it
6951 wasn't rewriting it again. This bug also had the effect of breaking the
6952 sender address verification caching when the sender address was rewritten.
6954 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6955 This has been changed so that if the ip literal address matches
6956 ignore_target_hosts, the router declines.
6958 7. Added expansion conditions match_domain, match_address, and match_local_
6959 part (NOT match_host).
6961 8. The placeholder for the Received: header didn't have a length field set.
6963 9. Added code to Exim itself and to exim_lock to test for a specific race
6964 condition that could lead to file corruption when using MBX delivery. The
6965 issue is with the lockfile that is created in /tmp. If this file is removed
6966 after a process has opened it but before that process has acquired a lock,
6967 there is the potential for a second process to recreate the file and also
6968 acquire a lock. This could lead to two Exim processes writing to the file
6969 at the same time. The added code performs the same test as UW imapd; it
6970 checks after acquiring the lock that its file descriptor still refers to
6971 the same named file.
6973 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6974 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6977 11. Added the smtp_active_hostname option. If used, this will typically be made
6978 to depend on the incoming interface address. Because $interface_address is
6979 not set up until the daemon has forked a reception process, error responses
6980 that can happen earlier (such as "too many connections") no longer contain
6983 12. If an expansion in a condition on a "warn" statement fails because a lookup
6984 defers, the "warn" statement is abandoned, and the next ACL statement is
6985 processed. Previously this caused the whole ACL to be aborted.
6987 13. Added the iplsearch lookup type.
6989 14. Added ident_timeout as a log selector.
6991 15. Added tls_certificate_verified as a log selector.
6993 16. Added a global option tls_require_ciphers (compare the smtp transport
6994 option of the same name). This controls incoming TLS connections.
6996 17. I finally figured out how to make tls_require_ciphers do a similar thing
6997 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6998 before starting the TLS session.
7000 18. Tabs are now shown as \t in -bP output.
7002 19. If the log selector return_path_on_delivery was set, Exim crashed when
7003 bouncing a message because it had too many Received: header lines.
7005 20. If two routers both had headers_remove settings, and the first one included
7006 a superfluous trailing colon, the final name in the first list and the
7007 first name in the second list were incorrectly joined into one item (with a
7008 colon in the middle).
7014 1. Added -C and -D options to the exinext utility, mainly to make it easier
7015 to include in the automated testing, but these could be helpful when
7016 multiple configurations are in use.
7018 2. The exinext utility was not formatting the output nicely when there was
7019 an alternate port involved in the retry record key, nor when there was a
7020 message id as well (for retries that were specific to a specific message
7021 and a specific host). It was also confused by IPv6 addresses, because of
7022 the additional colons they contain. I have fixed the IPv4 problem, and
7023 patched it up to do a reasonable job for IPv6.
7025 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7026 delivery, the log line now contains "pipelined" if PIPELINING was used.
7028 4. An SMTP transport process used to panic and die if the bind() call to set
7029 an explicit outgoing interface failed. This has been changed; it is now
7030 treated in the same way as a connect() failure.
7032 5. A reference to $sender_host_name in the part of a conditional expansion
7033 that was being skipped was still causing a DNS lookup. This no longer
7036 6. The def: expansion condition was not recognizing references to header lines
7037 that used bh_ and bheader_.
7039 7. Added the _cache feature to named lists.
7041 8. The code for checking quota_filecount in the appendfile transport was
7042 allowing one more file than it should have been.
7044 9. For compatibility with Sendmail, the command line option
7052 and sets the incoming protocol and host name (for trusted callers). The
7053 host name and its colon can be omitted when only the protocol is to be set.
7054 Note the Exim already has two private options, -pd and -ps, that refer to
7055 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7056 "s", but I don't think that's a major issue.
7058 10. A number of refactoring changes to the code, none of which should affect
7061 (a) The number of logging options was getting close to filling up the
7062 32-bit word that was used as a bit map. I have split them into two classes:
7063 those that are passed in the argument to log_write(), and those that are
7064 only ever tested independently outside of that function. These are now in
7065 separate 32-bit words, so there is plenty of room for expansion again.
7066 There is no change in the user interface or the logging behaviour.
7068 (b) When building, for example, log lines, the code previously used a
7069 macro that called string_cat() twice, in order to add two strings. This is
7070 not really sufficiently general. Furthermore, there was one instance where
7071 it was actually wrong because one of the argument was used twice, and in
7072 one call a function was used. (As it happened, calling the function twice
7073 did not affect the overall behaviour.) The macro has been replaced by a
7074 function that can join an arbitrary number of extra strings onto a growing
7077 (c) The code for expansion conditions now uses a table and a binary chop
7078 instead of a serial search (which was left over from when there were very
7079 few conditions). Also, it now recognizes conditions like "pam" even when
7080 the relevant support is not compiled in: a suitably worded error message is
7081 given if an attempt is made to use such a condition.
7083 11. Added ${time_interval:xxxxx}.
7085 12. A bug was causing one of the ddress fields not to be passed back correctly
7086 from remote delivery subprocesses. The field in question was not being
7087 subsequently used, so this caused to problems in practice.
7089 13. Added new log selectors queue_time and deliver_time.
7091 14. Might have fixed a bug in maildirsizefile handling that threw up
7092 "unexpected character" debug warnings, and recalculated the data
7093 unnecessarily. In any case, I expanded the warning message to give more
7096 15. Added the message "Restricted characters in address" to the statements in
7097 the default ACL that block characters like @ and % in local parts.
7099 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7100 Three changes have been made:
7102 (a) There was a serious bug; a negative response to MAIL caused the whole
7103 recipient domain to be cached as invalid, thereby blocking all messages
7104 to all local parts at the same domain, from all senders. This bug has
7105 been fixed. The domain is no longer cached after a negative response to
7106 MAIL if the sender used is not empty.
7108 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7111 (c) A new callout option, "use_sender" has been added for people who want
7112 the modified behaviour.
7118 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7121 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7122 indeed breaks things for older releases.
7124 3. Added additional logging to the case where there is a problem reading data
7125 from a filter that is running in a subprocess using a pipe, in order to
7126 try to track down a specific problem.
7128 4. Testing facility fudge: when running in the test harness and attempting
7129 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7130 getting "No route to host". Convert this to a timeout.
7132 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7135 6. Some OS don't have socklen_t but use size_t instead. This affects the
7136 fifth argument of getsockopt() amongst other things. This is now
7137 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7138 can be set for individual OS. I have set it for SunOS5, OSF1, and
7139 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7140 some earlier ones do not.
7142 7. Change 4.30/15 was not doing the test caselessly.
7144 8. The standard form for an IPv6 address literal was being rejected by address
7145 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7146 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7147 this, as well as the form without the "IPv6" on the front (but only when
7148 address literals are enabled, of course).
7150 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7152 10. Exim crashed if a message with an empty sender address specified by -f
7153 encountered a router with an errors_to setting. This could be provoked only
7154 by a command such as
7158 where an empty string was supplied; "<>" did not hit this bug.
7160 11. Installed PCRE release 4.5.
7162 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7163 remained set. It is now erased.
7165 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7166 times from message ids (which are base 36 rather than the normal 62).
7168 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7169 were being counted as actual protocol errors, and logged if the log
7170 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7171 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7172 and DATA following a set of rejected RCPTs do not count as protocol errors.
7173 In other words, Exim assumes they were pipelined, though this may not
7174 actually be the case. Of course, in all cases the client gets an
7175 appropriate error code.
7177 15. If a lookup fails in an ACL condition, a message about the failure may
7178 be available; it is used if testing the ACL cannot continue, because most
7179 such messages specify what the cause of the deferral is. However, some
7180 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7181 that caused an old message to be retained and used if a later statement
7182 caused a defer, replacing the real cause of the deferral.
7184 16. If an IP address had so many PTR records that the DNS lookup buffer
7185 was not large enough to hold them, Exim could crash while trying to process
7186 the truncated data. It now detects and logs this case.
7188 17. Further to 4.21/58, another change has been made: if (and only if) the
7189 first line of a message (the first header line) ends with CRLF, a bare LF
7190 in a subsequent header line has a space inserted after it, so as not to
7191 terminate the header.
7193 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7194 unnecessarily, used atoi() instead of strtol(), and didn't check the
7195 termination when getting file sizes from file names by regex.
7197 19. Completely re-implemented the support for maildirsize files, in the light
7198 of a number of problems with the previous contributed implementation
7199 (4.30/29). In particular:
7201 . If the quota is zero, the maildirsize file is maintained, but no quota is
7204 . If the maildir directory does not exist, it is created before any attempt
7205 to write a maildirsize file.
7207 . The quota value in the file is just a cache; if the quota is changed in
7208 the transport, the new value overrides.
7210 . A regular expression is available for excluding directories from the
7213 20. The autoreply transport checks the characters in options that define the
7214 message's headers; it allows continued headers, but it was checking with
7215 isspace() after an embedded newline instead of explicitly looking for a
7218 21. If all the "regular" hosts to which an address was routed had passed their
7219 expiry times, and had not reached their retry times, the address was
7220 bounced, even if fallback hosts were defined. Now Exim should go on to try
7223 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7224 equivalent code in the SMTP transport. Some hosts send humungous responses
7225 to HELO/EHLO, more than 1024 it seems.
7227 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7228 gives compiler warnings in some environments. I've now done it "properly",
7231 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7232 (because of problems with the built-in one) was declared to return uschar *
7233 instead of char *, causing compiler failure.
7235 25. Fixed a file descriptor leak when processing alias/forward files.
7237 26. Fixed a minor format string issue in dbfn.c.
7239 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7241 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7242 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7245 29. When a sender address is verified, it is cached, to save repeating the test
7246 when there is more than one recipient in a message. However, when the
7247 verification involves a callout, it is possible for different callout
7248 options to be set for different recipients. It is too complicated to keep
7249 track of this in the cache, so now Exim always runs a verification when a
7250 callout is required, relying on the callout cache for the optimization.
7251 The overhead is duplication of the address routing, but this should not be
7254 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7255 to be verified with callout=postmaster, and the main callout worked but the
7256 postmaster check failed, the verification correctly failed. However, if a
7257 subsequent RCPT command asked for sender verification *without* the
7258 postmaster check, incorrect caching caused this verification also to fail,
7261 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7262 it was not caching the DNS options (qualify_single, search_parents) that
7263 were used when the lookup failed. A subsequent lookup with different
7264 options therefore always gave the same answer, though there were cases
7265 where it should not have. (Example: a "domains = !$mx_any" option on a
7266 dnslookup router: the "domains" option is always processed without any
7267 widening, but the router might have qualify_single set.) Now Exim uses the
7268 cached value only when the same options are set.
7270 32. Added John Jetmore's "exipick" utility to the distribution.
7272 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7273 than a timeout (e.g. a certificate is required, and is not provided), an
7274 Exim server now closes the connection immediately. Previously it waited for
7275 the client to close - but if the client is SSL, it seems that they each
7276 wait for each other, leading to a delay before one of them times out.
7278 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7279 maintained 0.8.x compatibility because I don't think many are using it, and
7280 it is clearly obsolete.
7282 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7285 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7286 client certificate was expired. A simple patch fixes this, though I don't
7287 understand the full logic of why the verify callback is called multiple
7290 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7291 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7292 which causes problems with some clients (such as the Certicom SSL Plus
7293 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7294 disables the coutermeasure allowing Eudora to connect."
7296 38. Exim was not checking that a write() to a log file succeeded. This could
7297 lead to Bad Things if a log got too big, in particular if it hit a file
7298 size limit. Exim now panics and dies if it cannot write to a log file, just
7299 as it does if it cannot open a log file.
7301 39. Modified OS/Makefile-Linux so that it now contains
7303 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7305 The two -D definitions ensure that Exim is compiled with large file
7306 support, which makes it possible to handle log files that are bigger than
7309 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7310 instance) a domain was checked against a named list that involved a lookup,
7311 causing $domain_data to be set, then another domain was checked against the
7312 same list, then the first domain was re-checked, the value of $domain_data
7313 after the final check could be wrong. In particular, if the second check
7314 failed, it could be set empty. This bug probably also applied to
7317 41. The strip_trailing_dot option was not being applied to the address given
7318 with the -f command-line option.
7320 42. The code for reading a message's header from the spool was incrementing
7321 $received_count, but never initializing it. This meant that the value was
7322 incorrect (doubled) while delivering a message in the same process in which
7323 it was received. In the most common configuration of Exim, this never
7324 happens - a fresh exec is done - but it can happen when
7325 deliver_drop_privilege is set.
7327 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7328 it now includes up to 150 characters of the unexpected data in the log
7331 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7332 and building data strings. The size of both of these buffers was 10 000
7333 bytes - far larger than anybody would *ever* want, thought I. Needless to
7334 say, somebody hit the limit. I have increased the maximum line length to
7335 20 000 and the maximum data length of concatenated lines to 100 000. I have
7336 also fixed two bugs, because there was no checking on these buffers. Tsk,
7337 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7338 buffer is too small.
7340 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7341 lsearch lookups. Now it does.
7343 46. When parsing a route_list item in a manualroute router, a fixed-length
7344 buffer was used for the list of hosts. I made this 1024 bytes long,
7345 thinking that nobody would ever have a list of hosts that long. Wrong.
7346 Somebody had a whole pile of complicated expansion conditions, and the
7347 string was silently truncated, leading to an expansion error. It turns out
7348 that it is easier to change to an unlimited length (owing to other changes
7349 that have happened since this code was originally written) than to build
7350 structure for giving a limitation error. The length of the item that
7351 expands into the list of hosts is now unlimited.
7353 47. The lsearch lookup could not handle data where the length of text line was
7354 more than 4095 characters. Such lines were truncated, leading to shortened
7355 data being returned. It should now handle lines of any length.
7357 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7358 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7361 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7362 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7363 no longer happens in comment lines. A list of replacements is now placed
7364 at the head of all of the source files, except those whose only change is
7365 to replace PERL_COMMAND in the very first #! line.
7367 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7368 messages on the queue, with a bottom-up merge sort, using code contributed
7369 by Michael Haardt. This should make operations like -bp somewhat faster on
7370 large queues. It won't affect queue runners, except when queue_run_in_order
7373 51. Installed eximstats 1.31 in the distribution.
7375 52. Added support for SRV lookups to the dnslookup router.
7377 53. If an ACL referred to $message_body or $message_body_end, the value was not
7378 reset for any messages that followed in the same SMTP session.
7380 54. The store-handling optimization for building very long strings was not
7381 differentiating between the different store pools. I don't think this
7382 actually made any difference in practice, but I've tidied it.
7384 55. While running the routers to verify a sender address, $sender_address
7385 was still set to the sender address. This is wrong, because when routing to
7386 send a bounce to the sender, it would be empty. Therefore, I have changed
7387 it so that, while verifying a sender address, $sender_address is set to <>.
7388 (There is no change to what happens when verifying a recipient address.)
7390 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7391 target A or AAAA records (if not already returned) without resetting the
7392 qualify_single or search_parents options of the DNS resolver. These are
7393 inappropriate in this case because the targets of MX and SRV records must
7394 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7395 target that, when qualified, matched something in the local domain. These
7396 two options are now turned off when doing these lookups.
7398 57. It seems that at least some releases of Reiserfs (which does not have the
7399 concept of a fixed number of inodes) returns zero and not -1 for the
7400 number of available inodes. This interacted badly with check_spool_inodes,
7401 which assumed that -1 was the "no such thing" setting. What I have done is
7402 to check that the total number of inodes is greater than zero before doing
7403 the test of how many are available.
7405 58. When a "warn" ACL statement has a log_message modifier, the message is
7406 remembered, and not repeated. This is to avoid a lot of repetition when a
7407 message has many recipients that cause the same warning to be written.
7408 However, Exim was preserving the list of already written lines for an
7409 entire SMTP session, which doesn't seem right. The memory is now reset if a
7410 new message is started.
7412 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7413 debugging output unless log_rewrite was also set.
7415 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7416 of (int)(handle) when we know that handle contains (void *)(-1).
7418 61. The Exim daemon panic-logs an error return when it closes the incoming
7419 connection. However "connection reset by peer" seems to be common, and
7420 isn't really an error worthy of noting specially, so that particular error
7423 62. When Exim is trying to find all the local interfaces, it used to panic and
7424 die if the ioctl to get the interface flags failed. However, it seems that
7425 on at least one OS (Solaris 9) it is possible to have an interface that is
7426 included in the list of interfaces, but for which you get a failure error
7427 for this call. This happens when the interface is not "plumbed" into a
7428 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7429 failure of the "get flags" call assumes that the interface is down.
7431 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7432 makes life easier for people who are doing arithmetic on fields extracted
7433 from dates, where you often get leading zeros that should not be
7434 interpreted as octal.
7436 64. Added qualify_domain to the redirect router, to override the global
7439 65. If a pathologically long header line contained very many addresses (the
7440 report of this problem mentioned 10 000) and each of them was rewritten,
7441 Exim could use up a very large amount of memory. (It kept on making new
7442 copies of the header line as it rewrote, and never released the old ones.)
7443 At the expense of a bit more processing, the header rewriting function has
7444 been changed so that it no longer eats memory in this way.
7446 66. The generation of the Received: header has been moved from the time that a
7447 message starts to be received, to the time that it finishes. The timestamp
7448 in the Received: header should now be very close to that of the <= log
7449 line. There are two side-effects of this change:
7451 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7452 logged header lines no longer include the local Received: line, because
7453 it has not yet been created. The same applies to a copy of the message
7454 that is returned to a non-SMTP sender when a message is rejected.
7456 (b) When a filter file is tested using -bf, no additional Received: header
7457 is added to the test message. After some thought, I decided that this
7460 This change does not affect the value of $received_for. It is still set
7461 after address rewriting, but before local_scan() is called.
7463 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7465 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7466 gave an unhelpful panic error message, and a defer error. I have managed to
7467 change this behaviour so that it now rejects any supplied certificate,
7468 which seems right, as the list of acceptable certificates is empty.
7470 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7471 gave an unhelpful defer error. I have not managed to make this reject any
7472 supplied certificates, but the error message it gives is "no certificate
7473 supplied", which is not helpful.
7475 70. exigrep's output now also includes lines that are not associated with any
7476 message, but which match the given pattern. Implemented by a patch from
7477 Martin Sluka, which also tidied up the Perl a bit.
7479 71. Recipient callout verification, like sender verification, was using <> in
7480 the MAIL FROM command. This isn't really the right thing, since the actual
7481 sender may affect whether the remote host accepts the recipient or not. I
7482 have changed it to use the actual sender in the callout; this means that
7483 the cache record is now keyed on a recipient/sender pair, not just the
7484 recipient address. There doesn't seem to be a real danger of callout loops,
7485 since a callout by the remote host to check the sender would use <>.
7486 [SEE ABOVE: changed after hitting problems.]
7488 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7489 temporary errors. However, in the case of such a code being given after
7490 the end of a data transmission (i.e. after ".") Exim was failing to write
7491 a retry record for the message. (Yes, there was some broken host that was
7492 actually sending 8xx at this point.)
7494 73. An unknown lookup type in a host list could cause Exim to panic-die when
7495 the list was checked. (An example that provoked this was putting <; in the
7496 middle of a list instead of at the start.) If this happened during a DATA
7497 ACL check, a -D file could be left lying around. This kind of configuration
7498 error no longer causes Exim to die; instead it causes a defer error. The
7499 incident is still logged to the main and panic logs.
7501 74. Buglet left over from Exim 3 conversion. The message "too many messages
7502 in one connection" was written to the rejectlog but not the mainlog, except
7503 when address rewriting (yes!) was being logged.
7505 75. Added write_rejectlog option.
7507 76. When a system filter was run not as root (that is, when system_filter_user
7508 was set), the values of the $n variables were not being returned to the
7509 main process; thus, they were not subsequently available in the $sn
7512 77. Added +return_path_on_delivery log selector.
7514 78. A connection timeout was being treated differently from recipients deferred
7515 when testing hosts_max_try with a message that was older than the host's
7516 retry timeout. (The host should not be counted, thus allowing all hosts to
7517 be tried at least once before bouncing.) This may have been the cause of an
7518 occasionally reported bug whereby a message would remain on the queue
7519 longer than the retry timeout, but would be bounced if a delivery was
7520 forced. I say "may" because I never totally pinned down the problem;
7521 setting up timeout/retry tests is difficult. See also the next item.
7523 79. The ultimate address timeout was not being applied to errors that involved
7524 a combination of host plus message (for example, a timeout on a MAIL
7525 command). When an address resolved to a number of possible hosts, and they
7526 were not all tried for each delivery (e.g. because of hosts_max_try), a
7527 message could remain on the queue longer than the retry timeout.
7529 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7532 81. Fixed an obscure SMTP outgoing bug which required at least the following
7533 conditions: (a) there was another message waiting for the same server;
7534 (b) the server returned 5xx to all RCPT commands in the first message so
7535 that the message was not completed; (c) the server dropped the connection
7536 or gave a negative response to the RSET that Exim sends to abort the
7537 transaction. The observed case was a dropped connection after DATA that had
7538 been sent in pipelining mode. That is, the server had advertised PIPELINING
7539 but was not implementing it correctly. The effect of the bug was incorrect
7540 behaviour, such as trying another host, and this could lead to a crash.
7546 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7547 and daemon.c were passed as pointers to ints; they should have been
7548 pointers to socklen_t variables (which are typically unsigned ints).
7550 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7553 3. Fixed a really odd bug that affected only the testing scheme; patching a
7554 certain fixed string in the binary changed the value of another string that
7555 happened to be identical to the end of the original first string.
7557 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7558 name", it returns that address as the IP address. On some operating
7559 systems (e.g. Solaris), it also passes back the IP address string as the
7560 "host name". However, on others (e.g. Linux), it passes back an empty
7561 string. Exim wasn't checking for this, and was changing the host name to an
7562 empty string, assuming it had been canonicalized.
7564 5. Although rare, it is permitted to have more than one PTR record for a given
7565 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7566 all the names associated with an address, because they do in Solaris.
7567 However, it seems that they do not in Linux for data that comes from the
7568 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7569 I found this out when I moved to a new Linux workstation and tried to run
7570 the Exim test suite.
7572 To get round this problem I have changed the code so that it now does its
7573 own call to the DNS to look up PTR records when searching for a host name.
7574 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7575 addresses that are only in /etc/hosts are still found.
7577 This behaviour is, however, controlled by an option called host_lookup_
7578 order, which defaults to "bydns:byaddr". If people want to use the other
7579 order, or indeed, just use one or the other means of lookup, they can
7580 specify it in this variable.
7582 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7583 some operating systems, this comes back from gethostbyaddr() as an empty
7584 string, and this is what Exim used to test for. However, it seems that in
7585 other systems, "." is yielded. Exim now tests for this case too.
7587 7. The values of check_spool_space and check_log_space are now held internally
7588 as a number of kilobytes instead of an absolute number of bytes. If a
7589 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7590 kilobyte. This means that much larger values can be stored.
7592 8. Exim monitor: an attempt to get the action menu when not actually pointing
7593 at a message produces an empty menu entitled "No message selected". This
7594 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7595 no entries in it ("Shell widget menu has zero width and/or height"). So I
7596 have added a single, blank menu entry in this case.
7598 9. Added ${quote_local_part.
7600 10. MIME decoding is now applied to the contents of Subject: header lines when
7603 11. Now that a reference to $sender_host_address automatically causes a reverse
7604 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7605 host lookup before query-style lookups in lists that might use this
7606 variable. This has therefore been abolished, and the "net-" prefix is no
7607 longer necessary for query-style lookups.
7609 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7610 have been a typo for LFLAGS, so it has been changed.
7612 13. The install script calls Exim with "-C /dev/null" in order to find the
7613 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7614 to be output. However, since Exim outputs its version number before the
7615 error, it didn't break the script. It just looked ugly. I fixed this by
7616 always allowing "-C /dev/null" if the caller is root.
7618 14. Ignore overlarge ACL variable number when reading spool file - insurance
7619 against a later release with more variables having written the file.
7621 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7622 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7623 this, as well as the form without the "IPv6" on the front.
7625 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7626 OS/Makefile-Darwin file.
7628 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7629 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7631 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7632 variable. However, some OS use ints for inodes. I've added cast to long int
7633 to get rid of the compiler warning.
7635 19. I had forgotten to lock out "/../" in configuration file names when
7636 ALT_CONFIG_PREFIX was set.
7638 20. Routers used for verification do not need to specify transports. However,
7639 if such a router generated a host list, and callout was configured, Exim
7640 crashed, because it could not find a port number from the (non-existent)
7641 transport. It now assumes port 25 in this circumstance.
7643 21. Added the -t option to exigrep.
7645 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7646 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7647 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7648 them is set, LOOKUP_LSEARCH is forced.
7650 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7652 24. Added sender and host information to the "rejected by local_scan()" log
7653 line; previously there was no indication of these.
7655 25. Added .include_if_exists.
7657 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7658 Linux. It turns out that not all file systems support this. Apparently some
7659 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7660 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7661 not supported on the file descriptor, is now ignored when Exim is trying to
7662 sync a directory. This applies only to Linux.
7664 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7666 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7667 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7670 29. Added support for maildirsize files from supplied patch (modified a bit).
7672 30. The use of :fail: followed by an empty string could lead Exim to respond to
7673 sender verification failures with (e.g.):
7675 550 Verification failed for <xxx>
7676 550 Sender verify failed
7678 where the first response line was missing the '-' that indicates it is not
7679 the final line of the response.
7681 31. The loop for finding the name of the user that called Exim had a hardwired
7682 limit of 10; it now uses the value of finduser_retries, which is used for
7683 all other user lookups.
7685 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7688 33. Exim was neglecting to zero errno before one call of strtol() when
7689 expanding a string and expecting an integer value. On some systems this
7690 resulted in spurious "integer overflow" errors. Also, it was casting the
7691 result into an int without checking.
7693 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7694 did not work. The code looks as if it has *never* worked, though it appears
7695 to have been documented since at least release 1.62. I have made it work.
7697 35. The "timeout_DNS" error in retry rules, also documented since at least
7698 1.62, also never worked. As it isn't clear exactly what this means, and
7699 clearly it isn't a major issue, I have abolished the feature by treating it
7700 as "timeout", and writing a warning to the main and panic logs.
7702 36. The display of retry rules for -brt wasn't always showing the error code
7705 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7706 timeout_connect_A, timeout_connect_MX.
7708 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7709 to the empty sender.
7711 39. The daemon was not analysing the content of -oX till after it had closed
7712 stderr and disconnected from the controlling terminal. This meant that any
7713 syntax errors were only noted on the panic log, and the return code from
7714 the command was 0. By re-arranging the code a little, I've made the
7715 decoding happen first, so such errors now appear on stderr, and the return
7716 code is 1. However, the actual setting up of the sockets still happens in
7717 the disconnected process, so errors there are still only recorded on the
7720 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7721 connections (as happens on some IP stacks) was logged at start up time as
7722 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7723 it from "IPv6 and IPv4", which means that two separate sockets are being
7726 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7727 says whether AF_INET or AF_INET6 was passed as an argument.
7729 42. Exiwhat output was messed up when time zones were included in log
7732 43. Exiwhat now gives more information about the daemon's listening ports,
7733 and whether -tls-on-connect was used.
7735 44. The "port" option of the smtp transport is now expanded.
7737 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7738 silently ignored. Now an error message is written to the main and panic
7741 46. There's a new ACL modifier called "logwrite" which writes to a log file
7742 as soon as it is encountered.
7744 47. Added $local_user_uid and $local_user_gid at routing time.
7746 48. Exim crashed when trying to verify a sender address that was being
7749 49. Exim was recognizing only a space character after ".include". It now also
7750 recognizes a tab character.
7752 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7753 extracting the relevant information from the specification. The man page no
7754 longer contains scrambled data for the -d option, and I've added a section
7755 at the front about calling Exim under different names.
7757 51. Added "extra_headers" argument to the "mail" command in filter files.
7759 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7762 53. Installed eximstats 1.29.
7764 54. Added transport_filter_timeout as a generic transport option.
7766 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7767 Cc: header lines. This was required by RFC 822, but it not required by RFC
7770 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7771 incoming messages that did not have them. Now it does so only if the
7772 message originates locally, that is, if there is no associated remote host
7773 address. When Resent- header lines are present, this applies to the Resent-
7774 lines rather than the non-Resent- lines.
7776 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7777 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7779 58. Messages for configuration errors now include the name of the main
7780 configuration file - useful now that there may be more than one file in a
7781 list (.included file names were always shown).
7783 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7784 for those rare installations that do not start the daemon as root or run it
7785 setuid root. I've cut out the call to initgroups() if the daemon is not
7788 60. The Exim user and group can now be bound into the binary as text strings
7789 that are looked up at the start of Exim's processing.
7791 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7793 62. Added $mailstore_basename variable.
7795 63. Installed patch to sieve.c from Michael Haardt.
7797 64. When Exim failed to open the panic log after failing to open the main log,
7798 the original message it was trying to log was written to stderr and debug
7799 output, but if they were not available (the usual case in production), it
7800 was lost. Now it is written to syslog before the two lines that record the
7801 failures to open the logs.
7803 65. Users' Exim filters run in subprocesses under the user's uid. It is
7804 possible for a "deliver" command or an alias in a "personal" command to
7805 provoke an address rewrite. If logging of address rewriting is configured,
7806 this fails because the process is not running as root or exim. There may be
7807 a better way of dealing with this, but for the moment (because 4.30 needs
7808 to be released), I have disabled address rewrite logging when running a
7809 filter in a non-root, non-exim process.
7815 1. The buildconfig auxiliary program wasn't quoting the value set for
7816 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7817 not defined. This bug was masked in 4.22 by the effect that was fixed in
7820 2. Some messages that were rejected after a message id was allocated were
7821 shown as "incomplete" by exigrep. It no longer does this for messages that
7822 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7824 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7825 have allow_domain_literals set, the ID did not get logged in the <= line.
7826 Domain literals are now always recognized in Message-ID: header lines.
7828 4. The first argument for a ${extract expansion item is the key name or field
7829 number. Leading and trailing spaces in this item were not being ignored,
7830 causing some misleading effects.
7832 5. When deliver_drop_privilege was set, single queue runner processes started
7833 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7834 same command in the process it spins off) were not dropping privilege.
7836 6. When the daemon running as "exim" started a queue runner, it always
7837 re-executed Exim in the spun-off process. This is a waste of effort when
7838 deliver_drop_privilege is set. The new process now just calls the
7839 queue-runner function directly.
7845 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7848 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7849 ignored. Though the use of -oP was forcing the writing of a pid file, it
7850 was always written to the default place.
7852 3. If the message "no IP address found for host xxxx" is generated during
7853 incoming verification, it is now followed by identification of the incoming
7854 connection (so you can more easily find what provoked it).
7856 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7858 5. Added some features to "harden" Exim a bit more against certain attacks:
7860 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7861 be put in Local/Makefile. This is like the never_users runtime option,
7862 but it cannot be overridden. The default setting is "root".
7864 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7865 prefix string with which any file named in a -C command line option
7868 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7869 is retained for -C and -D only if the caller of Exim is root. Without
7870 it, the exim user may also use -C and -D and retain privilege.
7872 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7873 command line option is disabled.
7875 6. Macro names set by the -D option must start with an upper case letter, just
7876 like macro names defined in the configuration file.
7878 7. Added "dereference=" facility to LDAP.
7880 8. Two instances of the typo "uknown" in the source files are fixed.
7882 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7883 the Configure-Makefile script screwed up while processing it.
7885 10. Incorporated PCRE 4.4.
7887 11. The SMTP synchronization check was not operating right at the start of an
7888 SMTP session. For example, it could not catch a HELO sent before the client
7889 waited for the greeting. There is now a check for outstanding input at the
7890 point when the greeting is written. Because of the duplex, asynchronous
7891 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7892 way, but not yet received, when the check is performed.
7894 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7895 on TCP/IP sockets, because this apparently causes some broken clients to
7898 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7899 unchanged) from the Cygwin maintainer.
7901 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7902 was supported instead of testing for mbx. Effectively a typo.
7904 15. The spa authenticator server code was not checking that the input it
7905 received was valid base64.
7907 16. The debug output line for the "set" modifier in ACLs was not showing the
7908 name of the variable that was being set.
7910 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7912 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7913 lookup on its own. Something else was needed to trigger the lookup. For
7914 example, a match in host_lookup or the need for a host name in a host list.
7915 Now, if $sender_host_name is referenced and the host name has not yet been
7916 looked up, a lookup is performed. If the lookup fails, the variable remains
7917 empty, and $host_lookup_failed is set to "1".
7919 19. Added "eqi" as a case-independent comparison operator.
7921 20. The saslauthd authentication condition could segfault if neither service
7922 nor realm was specified.
7924 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7925 error message that was logged was misleading, and incoming SMTP
7926 connections were dropped. The message is now more accurate, and temporary
7927 errors are given to SMTP connections.
7929 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7930 (except RSET) with a 421 error, until QUIT is received. However, it was
7931 failing to send a response to QUIT.
7933 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7934 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7935 ACL is now run after the helo_try_verify_hosts code.
7937 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7938 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7939 case-independent, so other case variants are also recognized). Apparently
7940 some systems use these upper case variants.
7942 25. If more than two messages were waiting for the same host, and a transport
7943 filter was specified for the transport, Exim sent two messages over the
7944 same TCP/IP connection, and then failed with "socket operation on non-
7945 socket" when it tried to send the third.
7947 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7949 27. The extern definition of crypt16() in expand.c was not being excluded when
7950 the OS had its own crypt16() function.
7952 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7953 as a preferred synonym for return_size_limit, both as an option and as an
7956 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7958 30. Changed the default configuration ACL to relax the local part checking rule
7959 for addresses that are not in any local domains. For these addresses,
7960 slashes and pipe symbols are allowed within local parts, but the sequence
7961 /../ is explicitly forbidden.
7963 31. SPA server authentication was not clearing the challenge buffer before
7966 32. log_message in a "warn" ACL statement was writing to the reject log as
7967 well as to the main log, which contradicts the documentation and doesn't
7968 seem right (because no rejection is happening). So I have stopped it.
7970 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7971 However, I am unable to do any testing of this.
7973 34. Fixed an infelicity in the appendfile transport. When checking directories
7974 for a mailbox, to see if any needed to be created, it was accidentally
7975 using path names with one or more superfluous leading slashes; tracing
7976 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7978 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7979 contents are added to the log line that is written for every discarded
7980 recipient. (Previously a log_message setting was ignored.)
7982 36. The ${quote: operator now quotes the string if it is empty.
7984 37. The install script runs exim in order to find its version number. If for
7985 some reason other than non-existence or emptiness, which it checks, it
7986 could not run './exim', it was installing it with an empty version number,
7987 i.e. as "exim-". This error state is now caught, and the installation is
7990 38. An argument was missing from the function that creates an error message
7991 when Exim fails to connect to the socket for saslauthd authentication.
7992 This could cause Exim to crash, or give a corrupted message.
7994 39. Added isip, isip4, and isip6 to ${if conditions.
7996 40. The ACL variables $acl_xx are now saved with the message, and can be
7997 accessed later in routers, transports, and filters.
7999 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8000 strings in the file are not string-expanded.
8002 42. If a MAIL command specified a SIZE value that was too large to fit into an
8003 int variable, the check against message_size_limit failed. Such values are
8004 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8005 day this will have to be increased, but I don't think I want to be around
8006 when emails are that large.
8013 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8014 iconv() is not standard in FreeBSD.
8016 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8017 IPv6 enabled. The observed symptom was a segmentation fault on return from
8018 the function os_common_find_running_interfaces() in src/os.c.
8020 3. In the check_special_case() function in daemon.c I had used "errno" as an
8021 argument name, which causes warnings on some systems. This was basically a
8022 typo, since it was named "eno" in the comments!
8024 4. The code that waits for the clock to tick (at a resolution of some fraction
8025 of a second) so as to ensure message-id uniqueness was always waiting for
8026 at least one whole tick, when it could have waited for less. [This is
8027 almost certainly not relevant at current processor speeds, where it is
8028 unlikely to ever wait at all. But we try to future-proof.]
8030 5. The function that sleeps for a time interval that includes fractions of a
8031 second contained a race. It did not block SIGALRM between setting the
8032 timer, and suspending (a couple of lines later). If the interval was short
8033 and the sigsuspend() was delayed until after it had expired, the suspension
8034 never ended. On busy systems this could lead to processes getting stuck for
8037 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8038 process, before it forks any delivery processes. The open lookup caching
8039 mechanism meant that the open file or database connection was passed into
8040 the delivery process. The problem was that delivery processes always tidy
8041 up cached lookup data. This could cause a problem for the next delivery
8042 process started by the queue runner, because the external queue runner
8043 process does not know about the closure. So the next delivery process
8044 still has data in the lookup cache. In the case of a file lookup, there was
8045 no problem because closing a file descriptor in a subprocess doesn't affect
8046 the parent. However, if the lookup was caching a connection to a database,
8047 the connection was closed, and the second delivery process was likely to
8048 see errors such as "PGSQL: query failed: server closed the connection
8049 unexpectedly". The problem has been fixed by closing all cached lookups
8050 in a queue runner before running a delivery process.
8052 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8053 seem to have the "const" qualifier which it has on other OS. I've
8056 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8057 *of the same type* (client or server) with the same public name that an
8058 error should be diagnosed.
8060 9. When Exim looked up a host name for an IP address, but failed to find the
8061 original IP address when looking up the host name (a safety check), it
8062 output the message "<ip address> does not match any IP for NULL", which was
8063 confusing, to say the least. The bug was that the host name should have
8064 appeared instead of "NULL".
8066 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8067 user that is built into the binary, and the -C or -D options is used, root
8068 privilege is dropped before the configuration file is read. In addition,
8069 logging is switched to stderr instead of the normal log files. If the
8070 configuration then re-defines the Exim user, the unprivileged environment
8071 is probably not what is expected, so Exim logs a panic warning message (but
8074 However, if deliver_drop_privilege is set, the unprivileged state may well
8075 be exactly what is intended, so the warning has been cut out in that case,
8076 and Exim is allowed to try to write to its normal log files.
8082 1. smtp_return_error_details was not giving details for temporary sender
8083 or receiver verification errors.
8085 2. Diagnose a configuration error if two authenticators have the same public
8088 3. Exim used not to create the message log file for a message until the first
8089 delivery attempt. This could be confusing when incoming messages were held
8090 for policy or load reasons. The message log file is now created at the time
8091 the message is received, and an initial "Received" line is written to it.
8093 4. The automatically generated man page for command line options had a minor
8094 bug that caused no ill effects; however, a more serious problem was that
8095 the procedure for building the man page automatically didn't always
8096 operate. Consequently, release 4.20 contains an out-of-date version. This
8097 shouldn't happen again.
8099 5. When building Exim with embedded Perl support, the script that builds the
8100 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8101 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8103 6. The freeze_tell option was not being used for messages that were frozen on
8104 arrival, either by an ACL or by local_scan().
8106 7. Added the smtp_incomplete_transaction log selector.
8108 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8109 was accepting AUTH without a new EHLO.
8111 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8112 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8115 10. Logging of TCP/IP connections (when configured) now happens in the main
8116 daemon process instead of the child process, so that the TCP/IP connection
8117 count is more accurate (but it can never be perfect).
8119 11. The use of "drop" in a nested ACL was not being handled correctly in the
8120 outer ACL. Now, if condition failure induced by the nested "drop" causes
8121 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8122 or "require"), the connection is dropped.
8124 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8125 that yield "discard" can now be used with an "accept" or a "discard" verb,
8126 but an error is generated for any others (because I can't see a useful way
8127 to define what should happen).
8129 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8130 are now processed in the same way as lines in the Exim configuration file.
8131 In particular, continuation lines are supported.
8133 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8135 15. Added -ti meaning -t -i.
8137 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8138 domains, and warn by logging if others are found.
8140 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8141 returned from GIFCONF when Exim is trying to find the list of interfaces on
8142 a host. The code in os.c has been modified to copy each ifreq to an aligned
8143 structure in all cases.
8145 Also, in some cases, the returned ifreq's were being copied to a 'struct
8146 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8147 means the last couple of bytes of an IPv6 address could be chopped if the
8148 ifreq contained only a normal sockaddr (14 bytes storage).
8150 18. Named domain lists were not supported in the hosts_treat_as_local option.
8151 An entry such as +xxxx was not recognized, and was treated as a literal
8154 19. Ensure that header lines added by a DATA ACL are included in the reject log
8155 if the ACL subsequently rejects the message.
8157 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8158 MD5 (which is deprecated).
8160 21. When testing a filter file using -bf, Exim was writing a message when it
8161 took the sender from a "From " line in the message, but it was not doing so
8162 when it took $return_path from a Return-Path: header line. It now does.
8164 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8165 with a valid header line field name (a series of printing characters
8166 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8168 23. Changed "disc" in the source to "disk" to conform to the documentation and
8169 the book and for uniformity.
8171 24. Ignore Sendmail's -Ooption=value command line item.
8173 25. When execve() failed while trying to run a command in a pipe transport,
8174 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8175 could be confused with a return value of 69 from the command itself. This
8176 has been changed to 127, the value the shell returns if it is asked to run
8177 a non-existent command. The wording for the related log line suggests a
8178 non-existent command as the problem.
8180 26. If received_header_text expands to an empty string, do not add a Received:
8181 header line to the message. (Well, it adds a token one on the spool, but
8182 marks it "old" so that it doesn't get used or transmitted.)
8184 27. Installed eximstats 1.28 (addition of -nt option).
8186 28. There was no check for failure on the call to getsockname() in the daemon
8187 code. This can fail if there is a shortage of resources on the system, with
8188 ENOMEM, for example. A temporary error is now given on failure.
8190 29. Contrary to the C standard, it seems that in some environments, the
8191 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8192 program. Exim now does this explicitly; it affects the formatting of
8193 timestamps using strftime().
8195 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8196 complaints. I've now initialized all the variables, to avoid this.
8198 32. Header lines added by a system filter were not being "seen" during
8199 transport-time rewrites.
8201 33. The info_callback() function passed to OpenSSL is set up with type void
8202 (*)(SSL *, int, int), as described somewhere. However, when calling the
8203 function (actually a macro) that sets it up, the type void(*)() is
8204 expected. I've put in a cast to prevent warnings from picky compilers.
8206 34. If a DNS black list lookup found a CNAME record, but there were no A
8207 records associated with the domain it pointed at, Exim crashed.
8209 35. If a DNS black list lookup returned more than one A record, Exim ignored
8210 all but the first. It now scans all returned addresses if a particular IP
8211 value is being sought. In this situation, the contents of the
8212 $dnslist_value variable are a list of all the addresses, separated by a
8215 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8216 to accept a host name and all its aliases if the forward lookup for any of
8217 them yielded the IP address of the incoming connection. Now it accepts only
8218 those names whose forward lookup yields the correct IP address. Any other
8219 names are discarded. This closes a loophole whereby a rogue DNS
8220 administrator could create reverse DNS records to break through a
8221 wildcarded host restriction in an ACL.
8223 37. If a user filter or a system filter that ran in a subprocess used any of
8224 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8225 the wrong values were passed to the pipe command ($thisaddress had the
8226 value of $0, $0 had the value of $1, etc). This bug was introduced by
8227 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8229 38. Improved the line breaking for long SMTP error messages from ACLs.
8230 Previously, if there was no break point between 40 and 75 characters, Exim
8231 left the rest of the message alone. Two changes have been made: (a) I've
8232 reduced the minimum length to 35 characters; (b) if it can't find a break
8233 point between 35 and 75 characters, it looks ahead and uses the first one
8234 that it finds. This may give the occasional overlong line, but at least the
8235 remaining text gets split now.
8237 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8238 file descriptors might be low, and that setting 1000 would always raise it.
8239 It turns out that in some environments, the limit is already over 1000 and
8240 that lowering it causes trouble. So now Exim takes care not to decrease it.
8242 40. When delivering a message, the value of $return_path is set to $sender_
8243 address at the start of routing (routers may change the value). By an
8244 oversight, this default was not being set up when an address was tested by
8245 -bt or -bv, which affected the outcome if any router or filter referred to
8248 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8249 log only when "message" or "log_message" are set. However, if one of the
8250 conditions was an address verification, or a call to a nested ACL, the
8251 messages generated by the underlying test were being passed through. This
8252 no longer happens. The underlying message is available in $acl_verify_
8253 message for both "message" and "log_message" expansions, so it can be
8254 passed through if needed.
8256 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8257 new expansion $bh_ to give the encoded byte string without charset
8258 translation. Translation happens only if iconv() is available; HAVE_ICONV
8259 indicates this at build time. HEADERS_CHARSET gives the charset to
8260 translate to; headers_charset can change it in the configuration, and
8261 "headers charset" can change it in an individual filter file.
8263 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8264 that creates RFC 2047 encoded "words" labels them as that charset instead
8265 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8266 expansion operator; (ii) when Exim creates a From: line for a local
8267 message; (iii) when a header line is rewritten to include a "phrase" part.
8269 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8270 buggy, causing it to skip the first lines of messages whose message ID
8271 ended in 'D'. This would not have bitten before Exim release 4.14, because
8272 message IDs were unlikely to end in 'D' before then. The effect was to have
8273 incorrect size information for certain domains.
8275 45. #include "config.h" was missing at the start of the crypt16.c module. This
8276 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8279 46. If there was a timeout during a "random" callout check, Exim treated it as
8280 a failure of the random address, and carried on sending RSET and the real
8281 address. If the delay was just some slowness somewhere, the response to the
8282 original RCPT would be taken as a response to RSET and so on, causing
8283 mayhem of various kinds.
8285 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8286 when I implemented it. It didn't allow for the fact that some option values
8287 may legitimately be negative (e.g. size_addition), and it didn't even do
8288 the right test for positive values.
8290 48. Domain names in DNS records are case-independent. Exim always looks them up
8291 in lower case. Some resolvers return domain names in exactly the case they
8292 appear in the zone file, that is, they may contain uppercase letters. Not
8293 all resolvers do this - some return always lower case. Exim was treating a
8294 change of case by a resolver as a change of domain, similar to a widening
8295 of a domain abbreviation. This triggered its re-routing code and so it was
8296 trying to route what was effectively the same domain again. This normally
8297 caused routing to fail (because the router wouldn't handle the domain
8298 twice). Now Exim checks for this case specially, and just changes the
8299 casing of the domain that it ultimately uses when it transmits the message
8302 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8305 50. If a filter generated a file delivery with a non-absolute name (possible if
8306 no home directory exists for the router), the forbid_file option was not
8309 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8310 the existing equality matching.
8312 52. Exim was using ints instead of ino_t variables in some places where it was
8313 dealing with inode numbers.
8315 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8316 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8317 called TMPDIR, and if it finds it is different, it changes its value.
8319 54. The smtp_printf() function is now made available to local_scan() so
8320 additional output lines can be written before returning. There is also an
8321 smtp_fflush() function to enable the detection of a dropped connection.
8322 The variables smtp_input and smtp_batched_input are exported to
8325 55. Changed the default runtime configuration: the message "Unknown user"
8326 has been removed from the ACL, and instead placed on the localuser router,
8327 using the cannot_route_message feature. This means that any verification
8328 failures that generate their own messages won't get overridden. Similarly,
8329 the "Unrouteable address" message that was in the ACL for unverifiable
8330 relay addresses has also been removed.
8332 56. Added hosts_avoid_esmtp to the smtp transport.
8334 57. The exicyclog script was not checking for the esoteric option
8335 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8336 will work only if exicyclog is run under the appropriate euid.
8338 58. Following a discussion on the list, the rules by which Exim recognises line
8339 endings on incoming messages have been changed. The -dropcr and drop_cr
8340 options are now no-ops, retained only for backwards compatibility. The
8341 following line terminators are recognized: LF CRLF CR. However, special
8342 processing applies to CR:
8344 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8345 nor a local message in the state where . is a terminator.
8347 (ii) If a bare CR is encountered in a header line, an extra space is added
8348 after the line terminator so as not to end the header. The reasoning
8349 behind this is that bare CRs in header lines are most likely either
8350 to be mistakes, or people trying to play silly games.
8352 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8353 was being incorrectly given as 18 bytes larger than it should have been.
8354 This is a VOB (very old bug).
8356 60. This may never have affected anything current, but just in case it has:
8357 When the local host is found other than at the start of a list of hosts,
8358 the local host, those with the same MX, and any that follow, are discarded.
8359 When the list in question was part of a longer list of hosts, the following
8360 hosts (not currently being processed) were also being discarded. This no
8361 longer happens. I'm not sure if this situation could ever has previously
8364 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8367 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8368 header. This is something that is recommended in a new Internet Draft, and
8369 is something that is documented as being done by Sendmail. There are two
8370 possible values. For messages generated by the autoreply transport, Exim
8373 Auto-Submitted: auto-replied
8375 whereas for all other generated messages (e.g. bounces) it adds
8377 Auto-Submitted: auto-generated
8379 63. The "personal" condition in filters now includes a test for the
8380 Auto-Submitted: header. If it contains the string "auto-" the message it
8381 not considered personal.
8383 64. Added rcpt_include_affixes as a generic transport option.
8385 65. Added queue_only_override (default true).
8387 66. Added the syslog_duplication option.
8389 67. If what should have been the first header line of a message consisted of
8390 a space followed by a colon, Exim was mis-interpreting it as a header line.
8391 It isn't of course - it is syntactically invalid and should therefore be
8392 treated as the start of the message body. The misbehaviour could have
8393 caused a number of strange effects, including loss of data in subsequent
8394 header lines, and spool format errors.
8396 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8397 client host had authenticated. This control can now be exercised by an ACL
8398 for more flexibility.
8400 69. By default, callouts do not happen when testing with -bh. There is now a
8401 variant, -bhc, which does actually run the callout code, including
8402 consulting and updating the callout cache.
8404 70. Added support for saslauthd authentication, courtesy of Alexander
8407 71. If statvfs() failed on the spool or log directories while checking their
8408 size for availability, Exim confusingly gave the error "space shortage".
8409 Furthermore, in debugging mode it crashed with a floating point exception.
8410 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8411 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8412 serious problem, Exim now writes to the main and panic logs when this
8413 happens, with details of the failure. It then refuses to accept the
8414 incoming message, giving the message "spool directory problem" or "log
8415 directory problem" with a 421 code for SMTP messages.
8417 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8418 0, 1, and 2 exist, because some OS complain for execs without them (see
8419 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8420 descriptors. However, the code omitted to check that the open succeeded,
8421 causing mysterious errors if for some reason the permissions on /dev/null
8422 got screwed. Now Exim writes a message to the main and panic logs, and
8423 bombs out if it can't open /dev/null.
8425 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8426 interact so that it is all more flexible. It is supposed to remain
8427 backwards compatible. Also added extra_local_interfaces.
8429 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8430 to bomb out with an assertion failure - to the client this appears as a
8431 connection drop. This problem occurs in the part of the code that was taken
8432 from the Samba project. Fortunately, the assertion is in a very simple
8433 function, so I have fixed this by reproducing the function inline in the
8434 one place where it is called, and arranging for authentication to fail
8435 instead of killing the process with assert().
8437 75. The SPA client code was not working when the server requested OEM rather
8438 than Unicode encoding.
8440 76. Added code to make require_files with a specific uid setting more usable in
8441 the case where statting the file as root fails - usually a non-root-mounted
8442 NFS file system. When this happens and the failure is EACCES, Exim now
8443 forks a subprocess and does the per-uid checking as the relevant uid.
8445 77. Added process_log_path.
8447 78. If log_file_path was not explicitly set, a setting of check_log_space or
8448 check_log_inodes was ignored.
8450 79. If a space check for the spool or log partitions fails, the incident is now
8451 logged. Of course, in the latter case the data may get lost...
8453 80. Added the %p formatting code to string_format() so that it can be used to
8454 print addresses in debug_print(). Adjusted all the address printing in the
8455 debugging in store.c to use %p rather than %d.
8457 81. There was a concern that a line of code in smtp_in.c could overflow a
8458 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8459 initially expressed, the concern was not well-founded, because trailing
8460 spaces are removed early. However, if the trailing spaces were followed by
8461 a NULL, they did not get removed, so the overflow was possible. Two fixes
8464 (a) I re-wrote the offending code in a cleaner fashion.
8465 (b) If an incoming SMTP command contains a NULL character, it is rejected
8468 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8469 runs initgroups(), so that if the Exim user is in any additional groups,
8470 they will be used during message reception.
8476 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob