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 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
47 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
48 SMTP connection" log lines.
50 JH/02 Option default value updates:
51 - queue_fast_ramp (main) true (was false)
52 - remote_max_parallel (main) 4 (was 2)
54 JH/03 Cache static regex pattern compilations, for use by ACLs.
56 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
57 Make the rewrite never match and keep the logging. Trust the
58 admin to be using verify=header-syntax (to actually reject the message).
60 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
61 (under Linux) we watch the dir containing the final file; previously
62 it would be the dir with the first symlink. We still do not monitor
65 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
66 dn_expand() is happy to pass them through.
68 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
69 the file with the proof had an unchanged name, the new proof(s) were
70 loaded on top of the old ones (and nover used; the old ones were stapled).
72 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
73 more than one message arrived in a single connection a reference from
74 the earlier message could be re-used. Often a sigsegv resulted.
75 These variables were introduced in Exim 4.87.
76 Debug help from Graeme Fowler.
78 JH/09 Fix ${filter } for conditions that modify $value. Previously the
79 modified version would be used in construction the result, and a memory
82 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
83 Find and fix by Jasen Betts.
85 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
86 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
87 the systemwide configuration to override the Exim config.
89 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
90 API changes in libopendmarc.
92 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
93 pid 1, in the normal "background daemon" mode, having to drop process-
94 group leadership also lost track of needing to create listener sockets.
96 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
97 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
99 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
100 a capture group which obtained no text (eg. "(abc)*" matching zero
101 occurrences) could cause a segfault if the corresponding $<n> was
104 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
105 included a close-brace character (eg. it itself used an expansion) an
108 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
109 starting TLS. Previously it was after, meaning that attackers on such
110 ports had to be screened using the host_reject_connection main config
111 option. The new sequence aligns better with the STARTTLS behaviour, and
112 permits defences against crypto-processing load attacks, even though it
113 is strictly an incompatible change.
114 Also, avoid sending any SMTP fail response for either the connect ACL
115 or host_reject_connection, for TLS-on-connect ports.
117 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
118 Previously this was not permitted, but it makes reasonable sense.
119 While there, restore a restriction on using it from a connect ACL; given
120 the change JH/16 it could only return false (and before 4.91 was not
123 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
124 was exactly sized compared to the log buffer, a crash occurred with the
125 misleading message "bad memory reference; pool not found".
126 Found and traced by Jasen Betts.
128 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
129 dns_again_means_nonexist included an element causing a DNS lookup which
130 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
131 included (though probably not limited to) a process crash from stack
132 memory limit, or from excessive open files. Replace this with a paniclog
133 whine (as this is likely a configuration error), and returning
136 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
137 this always failed, probably leading to the usual downgrade to in-clear
140 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
141 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
142 to not checking that list for these lookups.
144 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
147 JH/23 Fix crash in string expansions. Previously, if an empty variable was
148 immediately followed by an expansion operator, a null-indirection read
149 was done, killing the process.
151 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
152 include an SMTP response string which is longer than that supported
153 by the delivering transport. Alleviate by wrapping such lines before
156 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
157 chars (RFC limit). Previously a limit of 12 items was made, which with
158 a not-impossible References: in the message being bounced could still
159 be over-large and get stopped in the transport.
161 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
162 close. Previously a bare socket close was done.
164 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
167 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
168 message being considered for delivery was already being handled by
169 another process, and having an SMTP connection already open, the function
170 to close it tried to use an uninitialized variable. This would afftect
171 high-volume sites more, especially when running mailing-list-style loads.
172 Pollution of logs was the major effect, as the other process delivered
173 the message. Found and partly investigated by Graeme Fowler.
175 JH/29 Change format of the internal ID used for message identification. The old
176 version only supported 31 bits for a PID element; the new 64 (on systems
177 which can use Base-62 encoding, which is all currently supported ones
178 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
179 and must use Base-36). The new ID is 23 characters rather than 16, and is
180 visible in various places - notably logs, message headers, and spool file
181 names. Various of the ancillary utilities also have to know the format.
182 As well as the expanded PID portion, the sub-second part of the time
183 recorded in the ID is expanded to support finer precision. Theoretically
184 this permits a receive rate from a single comms channel of better than the
186 The major timestamp part of the ID is not changed; at 6 characters it is
187 usable until about year 3700.
188 Updating from previously releases is fully supported: old-format spool
189 files are still usable, and the utilities support both formats. New
190 message will use the new format. The one hints-DB file type which uses
191 message-IDs (the transport wait- DB) will be discarded if an old-format ID
192 is seen; new ones will be built with only new-format IDs.
193 Optionally, a utility can be used to convert spool files from old to new,
194 but this is only an efficiency measure not a requirement for operation
195 Downgrading from new to old requires running a provided utility, having
196 first stopped all operations. This will convert any spool files from new
197 back to old (losing time-precision and PID information) and remove any
198 wait- hints databases.
200 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
201 we treated them as item separators when parsing for a list item, but they
202 need to be protected by the doublequotes. While there, add handling for
205 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
206 Found and fixed by Jasen Betts. No testcase for this as my usual text
207 editor insists on emitting only valid UTF-8.
209 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
210 a null-indirection SIGSEGV for the receive process.
212 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
213 -bh use. Internal checking would spot this and take a panic.
215 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
216 In 4.96 this would expand to empty.
218 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
219 certificate. Find and fix by Andreas Metzler.
221 JH/36 Add ARC info to DMARC hostory records.
223 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
224 or fakedefer. Previously the sender could discover that the message
225 had in fact been accepted.
227 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
228 sequences. Previously the input was not noted as being tainted; notably
229 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
230 bad coding of authenticators.
232 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
233 and ${tr...}. Found and diagnosed by Heiko Schlichting.
235 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
236 could be triggered by externally-supplied input. Found by Trend Micro.
239 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
240 be triggered by externally-controlled input. Found by Trend Micro.
243 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
244 be triggered by externally-controlled input. Found by Trend Micro.
247 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
248 Make the rewrite never match and keep the logging. Trust the
249 admin to be using verify=header-syntax (to actually reject the message).
251 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
254 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
260 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
261 after reception to before a subsequent reception. This should
262 mean slightly faster delivery, and also confirmation of reception
265 JH/02 Move from using the pcre library to pcre2. The former is no longer
266 being developed or supported (by the original developer).
268 JH/03 Constification work in the filters module required a major version
269 bump for the local-scan API. Specifically, the "headers_charset"
270 global which is visible via the API is now const and may therefore
271 not be modified by local-scan code.
273 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
274 sendfile() didi not account for the way the ClamAV driver code called it.
276 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
277 time check was being done for every character; replace that with one
280 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
281 sent was prefixed with a length byte.
283 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
284 RFC 5321. Previously Dovecot (at least) would log errors during
287 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
288 in 4.61 without a reason logged. Binaries will be bigger, which might
289 matter on diskspace-constrained systems, but debug is easier.
291 JH/09 Fix macro-definition during "-be" expansion testing. The move to
292 write-protected store for macros had not accounted for these runtime
293 additions; fix by removing this protection for "-be" mode.
295 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
296 handing out large-numbered file descriptors, violating the usual Unix
297 assumption (and required by Posix) that the lowest possible number will be
298 allocated by the kernel when a new one is needed. In the daemon, and any
299 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
300 are not useable for FD_SET() [and hence select()] and overwrite the stack.
301 Assorted crashes happen.
303 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
304 main-section options or in a connect ACL, the value from the first ever
305 connection was never replaced for subsequent connections. Found by
308 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
309 though only once PCRE2 was introduced: the memory accounting used under
310 debug offset allocations by an int, giving a hard trap in early startup.
311 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
313 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
314 with underbars is given. The write-protection of configuration introduced
315 in 4.95 trapped when normalisation was applied to an option not needing
318 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
320 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
321 failing to close the certificates directory, every hour or any time it
324 JH/16 Debugging initiated by an ACL control now continues through into routing
325 and transport processes. Previously debugging stopped any time Exim
326 re-execs, or for processing a queued message.
328 JH/17 The "expand" debug selector now gives more detail, specifically on the
329 result of expansion operators and items.
331 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
332 bad comparison between the option value and the name of the file to
333 be included was done, and a mismatch was wrongly identified.
334 4.88 to 4.95 are affected.
336 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
338 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
339 supplied as the db file-pair basename. Previously, if a directory
340 path was given, for example via the autoreply "once" option, the DB
341 file.pag and file.dir files would be created in that directory's
344 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
345 "taint" log_selector. These were previously deprecated.
347 JH/22 Fix static address-list lookups to properly return the matched item.
348 Previously only the domain part was returned.
350 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
351 the call into OpenSSL to send a TLS Close was being repeated; this
352 resulted in the library waiting for the peer's Close. If that was never
353 sent we waited forever. Fix by tracking send calls.
355 JH/24 The ${run} expansion item now expands its command string elements after
356 splitting. Previously it was before; the new ordering makes handling
357 zero-length arguments simpler. The old ordering can be obtained by
358 appending a new option "preexpand", after a comma, to the "run".
360 JH/25 Taint-check exec arguments for transport-initiated external processes.
361 Previously, tainted values could be used. This affects "pipe", "lmtp" and
362 "queryprogram" transport, transport-filter, and ETRN commands.
363 The ${run} expansion is also affected: in "preexpand" mode no part of
364 the command line may be tainted, in default mode the executable name
367 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
368 the the facility was not passed across execs, and only the first message
369 passed over a connection could use BDAT; any further ones using DATA.
371 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
372 uses $sending_ip_address and an interface is specified.
373 Previously any use of the local address in the EHLO name disabled
374 PIPECONNECT, the common case being to use the rDNS of it.
376 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
377 resumption. Previously verify failed because no certificate status is
378 passed on the wire for the restarted session. Fix by using the recorded
379 ocsp status of the stored session for the new connection.
381 JH/29 TLS resumption: the key for session lookup in the client now includes
382 more info that a server could potentially use in configuring a TLS
383 session, avoiding oferring mismatching sessions to such a server.
384 Previously only the server IP was used.
386 JH/30 Fix string_copyn() for limit greater than actual string length.
387 Previously the copied amount was the limit, which could result in a
388 overlapping memcpy for newly allocated destination soon after a
389 source string shorter than the limit. Found/investigated by KM.
391 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
392 close; it may be needed for a subsequent connection. This caused a
393 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
395 JH/32 Fix CHUNKING for a second message on a connection when the first was
396 rejected. Previously we did not reset the chunking-offered state, and
397 erroneously rejected the BDAT command. Investigation help from
400 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
401 an empty address. Previously the expansion returned an error.
403 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
404 proxy. Previously these were misparsed, leading to paniclog entries.
410 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
411 related applications. Previously an "H" was used where available info
412 says that "M" should be, so change to match.
414 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
415 as arguments, so an implementation trying to copy these into a local
416 buffer was taking a taint-enforcement trap. Fix by using dynamically
417 created buffers. Similar fix for radius expansion condition.
419 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
420 reasonable, eg. to count headers. Fix by using dynamically created
421 buffers rather than a local. Do similar fixes for ACL actions "dcc",
422 "log_reject_target", "malware" and "spam"; the arguments are expanded
423 so could be handling tainted values.
425 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
426 broken the (no-op) support for this sendmail command. Restore it
427 to doing nothing, silently, and returning good status.
429 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
430 record path was given (or the default used) without a leading directory
431 path, an error occurred on trying to open it. Use the transport's working
434 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
435 transport. Previously it was the name on the DNS A-record; use instead
436 the head of the CNAME chain leading there (if there is one). This seems
437 to align better with RFC 6125.
439 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
440 smtp_accept_max_per_host allocated resources which were not released
441 when the limit was exceeded. This eventually crashed the daemon. Fix
442 by adding a release action in that path.
444 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
445 expanded; previously using tainted values was rejected. Fix by using
446 dynamically-created buffers.
448 JH/09 Relax restrictions on ACL verify condition needing access to message
449 headers. Previously they were only permitted in data and non-smtp ACLs;
450 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
451 not_blind, header_sender and header_names_ascii verification.
453 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
454 Previously a macro used one argument twice; when called with the
455 argument as an expression having side-effects, incorrect operation
456 resulted. Use an inlineable function.
458 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
459 held open for a verify callout. Previously this wan not accounted for
460 and a corrupt onward SMTP conversation resulted.
462 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
463 Previously they were embedded naively in the constructed address; when
464 needed, strip the quoting and quote the entire local_part.
465 Also make the inbound_srs expansion condition handle quoting.
467 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
468 excluded, not matching the documentation.
470 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
471 was given for the sqlite_dbfile a trap resulted.
473 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
474 "name" argument resulted in a trap. There is no reason to disallow such;
475 this was a coding error.
477 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
478 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
479 spent suspended, ignoring the POSIX definition. Previously we assumed
480 it did and a constant offset from real time could be used as a correction.
481 Change to using the same clock source for the start-of-message and the
482 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
483 exists, just to get a clock slightly more aligned to reality.
485 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
486 RFC says it is optional some validators care. The missing char was not
487 intended but triggered by a line-wrap alignment. Discovery and fix by
488 Guillaume Outters, hacked on by JH.
490 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
491 name being quoted was tainted a trap would be taken. Fix by using
492 dynamically created buffers. The routine could have been called by a
493 rewrite with the "h" flag, by using the "-F" command-line option, or
494 by using a "name=" option on a control=submission ACL modifier.
496 JH/19 SPF: change the Authentication-Results expansion component to give
497 smtp.helo when the sender domain is empty. Previously it gave
500 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
501 not in 127.0.0.0/8 to help in spotting list domains taken over by a
502 domain-parking registrar.
504 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
505 Previously when a whitespace character was specified it was not inserted
506 after removing the newline.
508 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
509 the domain part of the recipient address. This overrides any tls_sni
510 option set, which was previously used.
512 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
515 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
516 is_tainted() had an off-by-one error in the overenthusiastic direction.
517 Find and fix by Gavan. Although NetBSD is not a supported platform for
518 4.94 this bug could affect other platforms.
520 PP/01 Fix default prime selection to be consistent.
521 One path used ike23 still, instead of exim.dev.20160529.3; now both
522 execution flows will use the same DH primes (currently
523 exim.dev.20160529.3).
525 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
526 option in smtp transport, to match the documentation. Previously
527 verification was not being done in some cases where it should have been.
529 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
530 than one server was defined and depending on the platform memory layout
531 details, an internal consistency trap could be hit while walking the list
534 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
535 files. The value is used by the authresults expansion item. Previously
536 if this was used in a router or transport, a crash could result.
538 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
539 have a sufficient incidence of EINTR returns from select that an
540 interaction with other operations done by the main daemon loop exposed
541 a bug in the error-handling. This was benign apart from the log
544 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
545 for consistency with delivered "=>" and failed "**" lines. While we're
546 there, handle PRX and TFO.
548 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
549 applied. This resulted, if any header-line rewrite rules were configured,
550 in a panic-log triggerable by sending a message with a long address in
551 a header. Fix by increasing the arbitrary limit to larger than a single
552 (dewrapped) 5322 header line maximum size.
554 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
555 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
556 RFC 6648 which deprecates X- options in protocols as a general practice.
557 Changeover between the implementations is handled by the mechanisms
560 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
561 a remote which uses callout/hold. Previously the local was queued.
563 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
566 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
567 log line. The asterisk is now consistently appended to the remote IP
568 (and port, if given), and will also be provided on defer and fail log
569 lines. Previously it could be placed on the local IP if that was being
570 logged, and was only provided on delivery lines.
572 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
574 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
575 authenticator client_send option. Previously the next char, after a pair
576 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
577 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
578 documentation. There is still no way to get a leading ^ immediately
579 after a NUL (ie. for the password of a PLAIN method authenticator.
581 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
582 files. For bad sizes read, delete the record and whine to paniclog.
584 JH/38 When logging an AUTH failure, as server, do not include sensitive
585 information. Previously, the credentials would be included if given
586 as part of the AUTH command line and an ACL denied authentication.
588 JH/39 Bug 2691: fix $local_part_data. When the matching list element
589 referred to a file, bad data was returned. This likely also affected
592 JH/40 The gsasl authenticator now supports caching of the salted password
593 generated by the client-side implementation. This required the addition
594 of a new variable: $auth4.
596 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
597 left undeleted; the attempt to re-create it then failed - resulting in
598 the usual "SIGHUP tp have daemon reload configuration" to not work.
599 This affected any platform not supporting "abstract" Unix-domain
600 sockets (i.e. not Linux).
602 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
603 recipients" response to RCPT in a later response, with a 250. The
604 previous coding assumed this would not happen, and under PIPELINING
605 would result in both lost and duplicate recipients for a message.
607 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
608 Previously the weighting was incorrectly applied. Similar fix for socks
609 proxies. Found and fixed by Heiko Schlichting.
611 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
612 not handle sub-lists included using the +namedlist syntax. While
613 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
614 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
617 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
618 Previously the permanent pool was used, so the sore could not be freed.
619 This meant a connection with many messages would use continually-growing
622 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
623 per-pool so as not to waste too much space. Previously a constant size
624 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
625 DOS attacks harder. The cost is wasted memory use in the larger blocks.
627 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
628 same space repeatedly, and a smaller process footprint.
630 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
631 for GnuTLS. Previously the name started "%s" which, while not a bug,
632 looked as if if might be one.
634 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
635 "continued connection") make the $proxy_* variables available. Previously
636 the information was not passed across the exec() call for subsequent
637 transport executions. This also mean that the log lines for the
638 messages can show the proxy information.
640 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
641 receive time. With modern systems the difference is significant.
642 The historical behaviour can be restored by disabling (a new) log_selector
643 "queue_time_exclusive".
645 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
646 content) were specifically tested for. Now, also cover expansions
647 resulting in ACL names and inline ACL content.
649 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
650 making it unusable in complex expressions.
652 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
653 fail with a taint-check on the spoolfile name, and leave the message
656 HS/01 Enforce absolute PID file path name.
658 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
660 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
662 PP/02 Bug 2643: Correct TLS DH constants.
663 A missing NUL termination in our code-generation tool had led to some
664 incorrect Diffie-Hellman constants in the Exim source.
665 Reported by kylon94, code-gen tool fix by Simon Arlott.
667 PP/03 Impose security length checks on various command-line options.
668 Fixes CVE-2020-SPRSS reported by Qualys.
670 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
671 better. Reported by Qualys.
673 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
674 providing a particularly obnoxious sender full name.
677 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
679 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
682 PP/08 Change default for recipients_max from unlimited to 50,000.
684 PP/09 Fix security issue with too many recipients on a message (to remove a
685 known security problem if someone does set recipients_max to unlimited,
686 or if local additions add to the recipient list).
687 Fixes CVE-2020-RCPTL reported by Qualys.
689 PP/10 Fix security issue in SMTP verb option parsing
690 Fixes CVE-2020-EXOPT reported by Qualys.
692 PP/11 Fix security issue in BDAT state confusion.
693 Ensure we reset known-good where we know we need to not be reading BDAT
694 data, as a general case fix, and move the places where we switch to BDAT
695 mode until after various protocol state checks.
696 Fixes CVE-2020-BDATA reported by Qualys.
698 HS/03 Die on "/../" in msglog file names
700 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
701 the Exim runtime user are allowed to create files.
703 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
706 QS/03 When reading the output from interpreted forward files we do not
707 pass the pipe between the parent and the interpreting process to
708 executed child processes (if any).
710 QS/04 Always die if requested from internal logging, even is logging is
713 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
714 the API; compilation noo longer completes with DMARC support included.
715 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
716 been reported on other platforms.
718 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
720 JH/56 Make the majority of info read from config files readonly, for defence-in-
721 depth against exploits. Suggestion by Qualys.
722 Not supported on Solaris 10.
724 JH/57 Fix control=fakreject for a custom message containing tainted data.
725 Previously this resulted in a log complaint, due to a re-expansion present
726 since fakereject was originally introduced.
728 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
729 modifier was given, a loop resulted.
731 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
732 pipelined SMTP command followed the BDAT LAST then it would be
733 incorrectly treated as part of the message body, causing a verification
736 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
737 require looser validation rules than those for 821-level addresses,
738 which only permit IP addresses.
744 JH/01 Avoid costly startup code when not strictly needed. This reduces time
745 for some exim process initialisations. It does mean that the logging
746 of TLS configuration problems is only done for the daemon startup.
748 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
750 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
751 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
754 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
755 DKIM signing is being done. Previously a transport_filter would always
756 disable CHUNKING, falling back to traditional DATA.
758 JH/05 Regard command-line recipients as tainted.
760 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
762 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
763 PAM library frees one of the arguments given to it, despite the
764 documentation. Therefore a plain malloc must be used.
766 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
767 on-stack buffers were used, resulting in a taint trap when DSN information
768 copied from a received message was written into the buffer.
770 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
771 the ordering of its ARC headers. This caused a crash.
773 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
774 a new record was being constructed with information from the peer, a trap
777 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
778 installation would get error messages from DMARC verify, when it hit the
779 nonexistent file indicated by the default. Distros wanting DMARC enabled
780 should both provide the file and set the option.
781 Also enforce no DMARC verification for command-line sourced messages.
783 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
784 could, depending on the platform, hang at the STARTTLS response.
786 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
787 message on a connection. Previously if one message had ARC headers and
788 the following one did not, a crash could result when adding an
789 Authentication-Results: header.
791 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
792 Exim main code and Exim-related utities. The introduction of taint
793 tracking also did many adjustments to string handling. Since then, eximon
794 frequently terminated with an assert failure.
796 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
797 check for 452 responses. This slightly helps the inefficieny of doing
798 a large alias-expansion into a recipient-limited target. The max_rcpt
799 transport option still applies (and at the current default, will override
800 the new feature). The check is done for either cause of synch, and forces
801 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
802 connection. The new facility is not tunable at this time.
804 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
805 library live data was being used, so the results became garbage. Make
806 copies while it is still usable.
808 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
809 on delivery deferred (==) and failed (**) lines (if a delivery was
810 attemtped). Previously it was only on completion (=>) lines.
812 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
813 for the expansion of the server_scram_iter and server_scram_salt options.
815 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
816 are now specifically given a NO_DATA response without hitting the system
817 resolver. The library goes on to do the now-standard TXT lookup.
818 Use of dnsdb lookups is not affected.
820 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
821 only retrieve the errormessage once. Previously two calls to dlerror()
822 were used, and the second one (for mainlog/paniclog) retrieved null
825 JH/20 Taint checking: disallow use of tainted data for
826 - the appendfile transport file and directory options
827 - the pipe transport command
828 - the autoreply transport file, log and once options
829 - file names used by the redirect router (including filter files)
831 - paths used by single-key lookups
832 Previously this was permitted.
834 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
835 adjusted the size of a major service buffer; this failed because the
836 buffer was in use at the time. Change to a compile-time increase in the
837 buffer size, when this authenticator is compiled into exim.
839 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
840 previous fast-mode was untenable in the face of glibs using mmap to
841 support larger malloc requests.
843 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
844 New values supported, if defined on system where compiled:
845 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
846 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
848 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
849 running a limited number of proceses in parallel, a benefit is gained. The
850 amount varies with the platform hardware and load. The use of the option
851 queue_run_in_order means we cannot do this, as ordering becomes
854 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
855 had introduced a string-copy (for ensuring NUL-termination) which was not
856 appropriate for that case, which can include embedded NUL bytes in the
857 block of data. Investigation showed the copy to actually be needless, the
858 data being length-specified.
860 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
861 done during a receiving connection, and both used TLS, global info was
862 used rather than per-connection info for tracking the state of data
863 queued for transmission. This could result in a connection hang.
865 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
866 Previously, when delivering serveral messages down a single connection
867 only the first would provide a SIZE. This was due to the size information
868 not being properly tracked.
870 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
871 TAI (at 37 seconds currently), pretend to be in UTC for time-related
872 expansion and logging. Previously, spurious values such as a future
873 minute could be seen.
875 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
876 it could crash from a null-deref. This could also affect the
877 ${addresses: } operator and ${readsock } item.
879 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
880 message following a mime one, the variable was not reset.
882 JH/30 When an pipelined-connect fails at the first response, assume incorrect
883 cached capability (perhaps the peer reneged?) and immediately retry in
886 JH/31 Fix spurious detection of timeout while writing to transport filter.
888 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
889 an attempt to copy the string was made before checking it.
891 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
892 taint of the lookup key was maintained; we now regard the presence in the
893 filesystem as sufficient validation.
895 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
896 argument is supplied.
898 JH/35 The dsearch lookup now requires that the directory is an absolute path.
899 Previously this was not checked, and nonempty relative paths made an
900 access under Exim's current working directory.
902 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
903 Previously no event was raised.
905 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
906 parameter supplied by the sender MAIL FROM command. Previously it was
907 ignored, and only the check_spool_space option value for the required
910 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
911 the size of the signing public-key. Previously it was instead giving
912 the size of the signature hash.
914 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
915 the default. See the (new) dkim_verify_min_keysizes option.
917 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
918 and an ACL use a lookup for checking either the local_part or domain,
919 stale data could be accessed. Ensure that variable references are
920 dropped between messages.
922 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
923 by the client was not checked as pointing within response data before
924 being used. A malicious client could thus cause an out-of-bounds read and
925 possibly gain authentication. Fix by adding the check.
927 JH/42 Internationalisation: change the default for downconversion in the smtp
928 transport to be "if needed". Previously it was "as previously set" for
929 the message, which usually meant "if needed" for message-submission but
930 "no" for everything else. However, MTAs have been seen using SMTPUTF8
931 even when the envelope addresses did not need it, resulting in forwarding
932 failures to non-supporting MTAs. A downconvert in such cases will be
933 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
934 transport. The change does mean that addresses needing conversion will
935 be converted when previously a delivery failure would occur.
937 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
938 response was received it would be used unchecked in a fail-DSN, violating
939 standards on line-length limits. Truncate if needed.
941 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
942 doesn't work. (Additionally add a new main config option to configure the
949 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
950 side, to decode a TLS 1.3 packet capture.
952 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
953 Previously the default library behaviour applied, sending two, each in
956 JH/03 Debug output for ACL now gives the config file name and line number for
959 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
961 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
963 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
964 buffer overrun for (non-chunking) other transports.
966 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
967 TLS1.3, means that a server rejecting a client certificate is not visible
968 to the client until the first read of encrypted data (typically the
969 response to EHLO). Add detection for that case and treat it as a failed
970 TLS connection attempt, so that the normal retry-in-clear can work (if
971 suitably configured).
973 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
974 and/or domain. Found and fixed by Jason Betts.
976 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
977 configuration). If a CNAME target was not a wellformed name pattern, a
980 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
981 the OS reports them interleaved with other addresses.
983 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
984 used both for input and for a verify callout, both encrypted, SMTP
985 responses being sent by the server could be lost. This resulted in
986 dropped connections and sometimes bounces generated by a peer sending
989 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
990 string. Previously it was possible to cause undefined behaviour in a
991 library routine (usually a crash). Found by "zerons".
993 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
996 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
997 API was removed, so update to use the newer ones.
999 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1000 any timeout set, is taking a long time. Previously we would hang on to a
1001 rotated logfile "forever" if the input was arriving with long gaps
1002 (a previous attempt to fix addressed lack, for a long time, of initial
1005 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1006 shared (NFS) environment. The length of the tempfile name is now
1007 4 + 16 ("hdr.$message_exim_id") which might break on file
1008 systems which restrict the file name length to lower values.
1009 (It was "hdr.$pid".)
1011 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1012 shared (NFS) environment.
1014 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1015 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1018 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1019 on some platforms for bit 31.
1021 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1022 to changes apparently associated with TLS1.3 handling some of the APIs
1023 previously used were either nonfunctional or inappropriate. Strings
1024 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1025 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1026 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1027 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1028 and the use of specific cipher names in the encrypted= ACL condition.
1030 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1032 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1033 verification result was not updated unless hosts_require_ocsp applied.
1035 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1036 queue_list_requires_admin set to false, non-admin users were denied the
1039 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1040 directory-of-certs mode. Previously they were advertised despite the
1043 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1044 A single TCP connection by a client will now hold a TLS connection open
1045 for multiple message deliveries, by default. Previously the default was to
1048 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1049 default. If built with the facility, DANE will be used. The facility
1050 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1052 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1053 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1054 must be defined and you must still, unless you define DISABLE_TLS, manage
1055 the the include-dir and library-file requirements that go with that
1056 choice. Non-TLS builds are still supported.
1058 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1061 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1062 default. If the platform supports and has the facility enabled, it will
1063 be requested on all coneections.
1065 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1066 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1068 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1070 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1071 success-DSN messages. Previously the From: header was always the default
1072 one for these; the option was ignored.
1074 JH/28 Fix the timeout on smtp response to apply to the whole response.
1075 Previously it was reset for every read, so a teergrubing peer sending
1076 single bytes within the time limit could extend the connection for a
1077 long time. Credit to Qualsys Security Advisory Team for the discovery.
1079 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1080 delivery address, which leaked information of the results of local
1081 forwarding. Change to the original envelope recipient address, per
1084 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1085 requested. Previously not bounce was generated and a log entry of
1086 error ignored was made.
1088 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1090 JH/32 Introduce a general tainting mechanism for values read from the input
1091 channel, and values derived from them. Refuse to expand any tainted
1092 values, to catch one form of exploit.
1094 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1095 was unused and the unexpanded text used for the test. Found and
1096 fixed by Ruben Jenster.
1098 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1099 an attempt to use a TLS library read routine dereffed a nul pointer,
1102 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1103 them in our smtp response.
1105 JH/36 Have the general router option retry_use_local_part default to true when
1106 any of the restrictive preconditions are set (to anything). Previously it
1107 was only for check_local user. The change removes one item of manual
1108 configuration which is required for proper retries when a remote router
1109 handles a subset of addresses for a domain.
1111 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1112 link count into consideration.
1114 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1115 caused the extension of big_buffer, the following lines were ignored.
1117 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1118 accordance with RFC 2308. Previously there was no expiry, so a longlived
1119 receive process (eg. due to ACL delays) versus a short SOA value could
1122 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1124 JH/39 Promote DMARC support to mainline.
1126 JH/40 Bug 2452: Add a References: header to DSNs.
1128 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1129 parameters. The relevant library call is documented as "Deprecated: This
1130 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1131 3.6.0, DH parameters are negotiated following RFC7919."
1133 HS/06 Change the default of dnssec_request_domains to "*"
1135 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1136 carried on and emitted a BDAT command, even when PIPELINING was not
1139 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1140 buffer was used for the filename, resulting in a trap when tainted
1141 arguments (eg. $domain) were used.
1143 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1144 recommended to avoid a possible server-load attack. The feature can be
1145 re-enabled via the openssl_options main cofiguration option.
1147 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1148 for version 4.90 - adding a "more data" boolean to the arguments.
1149 Bumped the ABI version number also, this having been missed previously;
1150 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1151 respect of usage of smtp_printf() by either local_scan code or libraries
1152 accessed via the ${dlfunc } expansion item. Both will need coding
1153 adjustment for any calls to smtp_printf() to match the new function
1154 signature; a FALSE value for the new argument is always safe.
1156 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1157 the file-offset (which the Linux syscall does, and exim expects); this
1158 resulted in an indefinite loop.
1160 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1161 to do ARC verification. The Authentication-Results: header line added
1162 by the configuration then had no ARC item.
1164 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1165 an incorrect mode was used for reading the result, resulting in it being
1172 JH/01 Remove code calling the customisable local_scan function, unless a new
1173 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1175 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1176 non-signal-safe functions being used.
1178 JH/03 Bug 2269: When presented with a received message having a stupidly large
1179 number of DKIM-Signature headers, disable DKIM verification to avoid
1180 a resource-consumption attack. The limit is set at twenty.
1182 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1183 report of oldest_pass in ${authres } in consequence, and separate out
1184 some descriptions of reasons for verification fail.
1186 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1187 files in the spool were present and unlocked. A queue-runner could spot
1188 them, resulting in a duplicate delivery. Fix that by doing the unlock
1189 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1190 add more error-checking on spoolfile handling while that code is being
1193 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1194 No known attacks, no CVE, this is defensive hardening.
1196 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1197 a queue-runner could start a delivery while other operations were ongoing.
1198 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1199 Found and investigated by Tim Stewart. Fix by using the open message data
1200 file handle rather than opening another, and not locally closing it (which
1201 releases a lock) for that case, while creating the temporary .eml format
1202 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1204 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1205 $sender_verify_failure/$recipient_verify_failure to "random".
1207 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1210 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1211 Previously this would segfault.
1213 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1216 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1217 like zero, since the resolver should be doing this for us, But we need one
1218 as a CNAME but no MX presence gets the CNAME returned; we need to check
1219 that doesn't point to an MX to declare it "no MX returned" rather than
1220 "error, loop". A new main option is added so the older capability of
1221 following some limited number of chain links is maintained.
1223 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1225 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1226 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1227 now-deprecated earlier definitions used only specified the range up to TLS
1228 1.2 (in the older-version library docs).
1230 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1232 JH/15 Rework TLS client-side context management. Stop using a global, and
1233 explicitly pass a context around. This enables future use of TLS for
1234 connections to service-daemons (eg. malware scanning) while a client smtp
1235 connection is using TLS; with cutthrough connections this is quite likely.
1237 JH/16 Fix ARC verification to do AS checks in reverse order.
1239 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1241 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1242 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1243 Previously the "utf8" would be re-prepended for every additional message.
1245 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1246 Previously thery were accepted, resulting in issues when attempting to
1247 forward messages to a non-supporting MTA.
1249 PP/02 Let -n work with printing macros too, not just options.
1251 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1252 one parent address was copied, and bogus data was used at delivery-logging
1253 time. Either a crash (after delivery) or bogus log data could result.
1254 Discovery and analysis by Tim Stewart.
1256 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1257 Previously if the string ended mid-character, we did not insert the
1258 promised '?' replacement.
1260 PP/04 Documentation: current string operators work on bytes, not codepoints.
1262 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1263 should pack well giving a smaller memory footprint so better caching and
1264 therefore performance. Group the declarations where this can't be done so
1265 that the byte-sized flag variables are not interspersed among pointer
1266 variables, giving a better chance of good packing by the compiler.
1268 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1269 non-null, to avoid issues with sites running BATV. Previously reports were
1270 sent with an empty envelope sender so looked like bounces.
1272 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1273 The ignore_error flag wasn't being returned from the filter subprocess so
1274 was not set for later routers. Investigation and fix by Matthias Kurz.
1276 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1277 and a msg:complete for the whole, when a message is manually removed using
1278 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1280 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1281 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1282 hope that is portable enough.
1284 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1285 requiring a known-CA anchor certificate; make it now rely entirely on the
1286 TLSA as an anchor. Checking the name on the leaf cert against the name
1287 on the A-record for the host is still done for TA (but not for EE mode).
1289 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1290 included in delivery lines for non-proxied connections, when compiled with
1291 SUPPORT_SOCKS and running with proxy logging enabled.
1293 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1294 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1295 move the existing event to fire before the normal logging of message
1296 failure so that custom logging is bracketed by normal logging.
1298 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1299 msg:fail:internal event. Developement by Matthias Kurz.
1301 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1302 far too small for todays use of crypto signatures stored there. Go all
1303 the way to the max DNS message size of 64kB, even though this might be
1304 overmuch for IOT constrained device use.
1306 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1307 copy a string over itself. The library routine is documented as not
1308 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1310 JH/32 For main options check_spool_space and check_inode_space, where the
1311 platform supports 64b integers, support more than the previous 2^31 kB
1312 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1313 the previous G, M, k.
1315 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1316 $authenticated_fail_id variable on authentication failure. Previously
1319 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1320 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1321 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1322 GNUTLS_SEC_PARAM_MEDIUM.
1324 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1325 side. Previously we would continue as if no SNI had been received.
1327 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1328 "<" character, which should be a list-separator specification, we walked
1329 off past the nul-terimation.
1331 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1332 causes) even when the retry time is not yet met. Previously they were
1333 not, meaning that when (say) an account was over-quota and temp-rejecting,
1334 and multiple senders' messages were queued, only one sender would get
1335 notified on each configured delay_warning cycle.
1337 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1339 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1340 cipher-suites, an error can be left on the stack even for a succeeding
1341 accept; this results in impossible error messages when a later operation
1344 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1345 return error codes indicating retry. Under TLS1.3 this becomes required.
1347 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1348 it only wrote the new authenticators, resulting in a lack of tracking of
1349 peer changes of ESMTP extensions until the next cache flush.
1351 JH/41 Fix the loop reading a message header line to check for integer overflow,
1352 and more-often against header_maxsize. Previously a crafted message could
1353 induce a crash of the recive process; now the message is cleanly rejected.
1355 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1356 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1362 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1363 When redis_servers is set to a list of > 1 element, and the Redis servers
1364 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1365 case of MOVED into a DEFER case instead, thus moving the query onto the
1366 next server in the list. For a cluster of N elements, all N servers must
1367 be defined in redis_servers.
1369 GF/02 Catch and remove uninitialized value warning in exiqsumm
1370 Check for existence of @ARGV before looking at $ARGV[0]
1372 JH/01 Replace the store_release() internal interface with store_newblock(),
1373 which internalises the check required to safely use the old one, plus
1374 the allocate and data copy operations duplicated in both (!) of the
1375 extant use locations.
1377 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1378 modifier. This matches the restriction on the commandline.
1380 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1381 Previously only the last row was returned.
1383 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1384 we assumed that tags in the header were well-formed, and parsed the
1385 element content after inspecting only the first char of the tag.
1386 Assumptions at that stage could crash the receive process on malformed
1389 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1390 While running the DKIM ACL we operate on the Permanent memory pool so that
1391 variables created with "set" persist to the DATA ACL. Also (at any time)
1392 DNS lookups that fail create cache records using the Permanent pool. But
1393 expansions release any allocations made on the current pool - so a dnsdb
1394 lookup expansion done in the DKIM ACL releases the memory used for the
1395 DNS negative-cache, and bad things result. Solution is to switch to the
1396 Main pool for expansions.
1397 While we're in that code, add checks on the DNS cache during store_reset,
1398 active in the testsuite.
1399 Problem spotted, and debugging aided, by Wolfgang Breyha.
1401 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1402 When none of the hosts presented to a transport match an already-open
1403 connection, close it and proceed with the list. Previously we would
1404 queue the message. Spotted by Lena with Yahoo, probably involving
1407 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1408 Previously a spurious "250 OK id=" response was appended to the proper
1411 JH/08 The "support for" informational output now, which built with Content
1412 Scanning support, has a line for the malware scanner interfaces compiled
1413 in. Interface can be individually included or not at build time.
1415 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1416 by the template makefile "src/EDITME". The "STREAM" support for an older
1417 ClamAV interface method is removed.
1419 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1420 rows affected is given instead).
1422 JH/11 The runtime Berkeley DB library version is now additionally output by
1423 "exim -d -bV". Previously only the compile-time version was shown.
1425 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1426 SMTP connection. Previously, when one had more recipients than the
1427 first, an abortive onward connection was made. Move to full support for
1428 multiple onward connections in sequence, handling cutthrough connection
1429 for all multi-message initiating connections.
1431 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1432 routers. Previously, a multi-recipient message would fail to match the
1433 onward-connection opened for the first recipient, and cause its closure.
1435 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1436 a timeout on read on a GnuTLS initiating connection, resulting in the
1437 initiating connection being dropped. This mattered most when the callout
1438 was marked defer_ok. Fix to keep the two timeout-detection methods
1441 JH/15 Relax results from ACL control request to enable cutthrough, in
1442 unsupported situations, from error to silently (except under debug)
1443 ignoring. This covers use with PRDR, frozen messages, queue-only and
1446 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1448 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1449 metadata, resulting in a crash in free().
1451 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1452 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1453 Broken also in d185889f4, with init system revamp.
1455 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1456 Previously we abruptly closed the connection after reading a malware-
1457 found indication; now we go on to read the "scan ok" response line,
1460 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1461 ACL. Previously, a crash would result.
1463 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1464 macro text after a replacement (previously it was only once per line) and
1465 by skipping builtin macros when searching for an uppercase lead character.
1467 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1468 for the build is renamed.
1470 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1471 was allocated for every new TLS startup, meaning one per message. Fix
1472 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1474 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1475 reported the original. Fix to report (as far as possible) the ACL
1476 result replacing the original.
1478 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1479 OpenSSL. Certificate information is loaded for every new TLS startup,
1480 and the resources needed to be freed.
1482 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1484 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1487 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1488 DATA response info to the (existing) per-recipient response info for
1489 the "C=" log element. It can have useful tracking info from the
1490 destination system. Patch from Simon Arlott.
1492 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1493 length value. Previously this would segfault.
1495 HS/02 Support Avast multiline protoocol, this allows passing flags to
1496 newer versions of the scanner.
1498 JH/28 Ensure that variables possibly set during message acceptance are marked
1499 dead before release of memory in the daemon loop. This stops complaints
1500 about them when the debug_store option is enabled. Discovered specifically
1501 for sender_rate_period, but applies to a whole set of variables.
1502 Do the same for the queue-runner and queue-list loops, for variables set
1503 from spool message files. Do the same for the SMTP per-message loop, for
1504 certain variables indirectly set in ACL operations.
1506 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1507 as a multi-recipient message from a mailinglist manager). The coding had
1508 an arbitrary cutoff number of characters while checking for more input;
1509 enforced by writing a NUL into the buffer. This corrupted long / fast
1510 input. The problem was exposed more widely when more pipelineing of SMTP
1511 responses was introduced, and one Exim system was feeding another.
1512 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1513 receiving system, and refused recipients seen by the sending system
1514 (propating to people being dropped from mailing lists).
1515 Discovered and pinpointed by David Carter.
1517 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1518 replaced by the ${authresults } expansion.
1520 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1522 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1523 allows proper process termination in container environments.
1525 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1526 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1528 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1529 and "err_perm", deprecated since 4.83 when the RFC-defined words
1530 "temperror" and "permerror" were introduced.
1532 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1533 transport-filters or DKIM-signing. The restriction was lost in the
1534 consolidation of verify-callout and delivery SMTP handling.
1535 Extend the restriction to also cover ARC-signing.
1537 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1538 in defer=pass mode supply a 450 to the initiator. Previously the message
1541 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1542 tls_require_ciphers is used as before.
1544 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1545 "pass_unscanned". Only tmpfails from the scanner are written to
1546 the paniclog, as they may require admin intervention (permission
1547 denied, license issues). Other scanner errors (like decompression
1548 bombs) do not cause a paniclog entry.
1550 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1551 Previously it was possible to log spurious information in receive log
1554 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1555 triggered odd behaviour from Outlook Express clients.
1557 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1560 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1561 since the IETF WG has not yet settled on that versus the original
1562 "bare" representation.
1564 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1565 Previously the millisecond value corrupted the output.
1566 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1567 corrupted the output.
1573 JH/01 Rework error string handling in TLS interface so that the caller in
1574 more cases is responsible for logging. This permits library-sourced
1575 string to be attached to addresses during delivery, and collapses
1576 pairs of long lines into single ones.
1578 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1579 during configuration. Wildcards are allowed and expanded.
1581 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1582 This permits better logging.
1584 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1585 do not close it down and have the child transport start it up again on
1586 the passed-on TCP connection. Instead, proxy the child (and any
1587 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1588 affected: the continued delivery log lines do not have any DNSSEC, TLS
1589 Certificate or OCSP information. TLS cipher information is still logged.
1591 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1592 identical IP addresses on different listening ports. Will also affect
1595 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1596 add noisy ifdef guards to special-case this sillyness.
1597 Patch from Bernd Kuhls.
1599 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1600 than 255 are no longer allowed.
1602 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1603 Disable the session-cache too, which might reduce our load. Since we
1604 currrectly use a new context for every connection, both as server and
1605 client, there is no benefit for these.
1606 GnuTLS appears to not support tickets server-side by default (we don't
1607 call gnutls_session_ticket_enable_server()) but client side is enabled
1608 by default on recent versions (3.1.3 +) unless the PFS priority string
1611 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1612 <https://reproducible-builds.org/specs/source-date-epoch/>.
1614 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1615 the check for any unsuccessful recipients did not notice the limit, and
1616 erroneously found still-pending ones.
1618 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1619 MSG_MORE. Only in-clear (not on TLS connections).
1621 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1622 transport-filter is involved we can buffer the headers in memory for
1623 creating the signature, and read the spool data file once for the
1624 signature and again for transmission.
1626 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1627 4.77 as the kernel support then wasn't solid, having issues in 64bit
1628 mode. Now, it's been long enough. Add support for FreeBSD also.
1630 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1631 case where the routing stage had gathered several addresses to send to
1632 a host before calling the transport for the first, we previously failed
1633 to close down TLS in the old transport process before passing the TCP
1634 connection to the new process. The new one sent a STARTTLS command
1635 which naturally failed, giving a failed delivery and bloating the retry
1636 database. Investigation and fix prototype from Wolfgang Breyha.
1638 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1639 false-negatives in the check that the sender had not preempted a response
1640 or prompt from Exim (running as a server), due to that code's lack of
1641 awareness of the SMTP input buffering.
1643 PP/04 Add commandline_checks_require_admin option.
1644 Exim drops privileges sanely, various checks such as -be aren't a
1645 security problem, as long as you trust local users with access to their
1646 own account. When invoked by services which pass untrusted data to
1647 Exim, this might be an issue. Set this option in main configuration
1648 AND make fixes to the calling application, such as using `--` to stop
1651 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1652 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1653 responses to those, into a single TLS record each way (this usually means
1654 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1657 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1658 affects you only if you're dancing at the edge of the param size limits.
1659 If you are, and this message makes sense to you, then: raise the
1660 configured limit or use OpenSSL 1.1. Nothing we can do for older
1663 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1664 cmdline element to get the documented default one. Previously it was
1667 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1668 are unsuable for non-message-specific reasons - eg. connection timeout,
1669 banner-time rejection.
1671 JH/16 Fix logging of delivery remote port, when specified by router, under
1674 PP/06 Repair manualroute's ability to take options in any order, even if one
1675 is the name of a transport.
1678 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1680 JH/17 Change the list-building routines interface to use the expanding-string
1681 triplet model, for better allocation and copying behaviour.
1683 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1684 Previously it was constructed the first time a possibly-matching string
1685 was met in the configuration file input during startup; now it is done
1688 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1689 compatible one, to avoid the (poorly documented) possibility of a config
1690 file in the working directory redirecting the DB files, possibly correpting
1691 some existing file. CVE-2017-10140 assigned for BDB.
1693 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1694 cache-hot. Previously, although the result was properly cached, the
1695 initial verify call returned a defer.
1697 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1698 the main verify for receipient in uncached-mode.
1700 JH/22 Retire historical build files to an "unsupported" subdir. These are
1701 defined as "ones for which we have no current evidence of testing".
1703 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1704 if present. Previously it was ignored.
1706 JH/24 Start using specified-initialisers in C structure init coding. This is
1707 a C99 feature (it's 2017, so now considered safe).
1709 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1710 if was a fixed-sized field and bitmask ops via macros; it is now more
1713 PP/07 GitHub PR 56: Apply MariaDB build fix.
1714 Patch provided by Jaroslav Škarvada.
1716 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1717 during Coverity cleanups [4.87 JH/47]
1718 Diagnosis and fix provided by Michael Fischer v. Mollard.
1720 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1721 the right size to place the terminating semicolon on its own folded
1722 line, the header hash was calculated to an incorrect value thanks to
1723 the (relaxed) space the fold became.
1725 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1726 and confused the parent.
1728 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1729 which could crash as a result. This could lead to undeliverable messages.
1731 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1734 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1735 matters on fast-turnover and PID-randomising systems, which were getting
1736 out-of-order delivery.
1738 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1739 a possibly-overlapping copy. The symptom was that "Remote host closed
1740 connection in response to HELO" was logged instead of the actual 4xx
1743 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1744 Previously only that bufferd was discarded, resulting in SYMTP command
1747 JH/32 DKIM: when a message has multiple signatures matching an identity given
1748 in dkim_verify_signers, run the dkim acl once for each. Previously only
1749 one run was done. Bug 2189.
1751 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1752 "panic the current process" to "deliberately defer". The panic log is
1753 still written with the problem list name; the mail and reject logs now
1754 get a temp-reject line for the message that was being handled, saying
1755 something like "domains check lookup or other defer". The SMTP 451
1756 message is still "Temporary local problem".
1758 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1759 A crafted sequence of BDAT commands could result in in-use memory beeing
1760 freed. CVE-2017-16943.
1762 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1763 from SMTP input. Previously it was always done; now only done for DATA
1764 and not BDAT commands. CVE-2017-16944.
1766 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1767 to the message (such as an overlong header line). Previously this was
1768 not done and we did not exit BDAT mode. Followon from the previous item
1769 though a different problem.
1775 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1776 than -2003 did; needs libidn2 in addition to libidn.
1778 JH/02 The path option on a pipe transport is now expanded before use.
1780 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1781 Patch provided by "Björn", documentation fix added too.
1783 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1784 missing a wire-to-host endian conversion.
1786 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1787 close after a BDAT command line could be taken as a following command,
1788 giving a synch failure. Fix by only checking for synch immediately
1789 before acknowledging the chunk.
1791 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1792 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1793 macro. Patches provided by Josh Soref.
1795 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1796 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1797 by IANA as a service extension. However, John Klensin suggests that we
1800 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1801 the dkim code may be unix-mode line endings rather than smtp wire-format
1802 CRLF, so prepend a CR to any bare LF.
1804 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1805 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1807 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1808 the first were themselves being wrongly included in the feed into dkim
1809 processing; with most chunk sizes in use this resulted in an incorrect
1810 body hash calculated value.
1812 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1813 DKIM signature block, for verification. Although advised against by
1814 standards it is specifically not ruled illegal.
1816 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1818 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1819 missing a body hash (the bh= tag).
1821 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1822 It seems that HAProxy sends the Proxy Protocol information in clear and
1823 only then does a TLS startup, so do the same.
1825 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1826 TCP connections (such as for Spamd) unless the daemon successfully set
1827 Fast Open mode on its listening sockets. This fixes breakage seen on
1828 too-old kernels or those not configured for Fast Open, at the cost of
1829 requiring both directions being enabled for TFO, and TFO never being used
1830 by non-daemon-related Exim processes.
1832 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1833 endings, at least on the first header line. Try to canonify any that get
1834 past that check, despite the cost.
1836 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1837 now limited to an arbitrary five deep, while parsing addresses with the
1838 strip_excess_angle_brackets option enabled.
1840 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1841 instead leave the unprompted TLS handshake in socket buffer for the
1842 TLS library to consume.
1844 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1846 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1848 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1849 region in one operation, for speed, and this leaves assigned pointers
1850 dangling. Add checks run only under the testsuite which checks all
1851 variables at a store-reset and panics on a dangling pointer; add code
1852 explicitly nulling out all the variables discovered. Fixes one known
1853 bug: a transport crash, where a dangling pointer for $sending_ip_address
1854 originally assigned in a verify callout, is re-used.
1856 PP/06 Drop '.' from @INC in various Perl scripts.
1858 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1860 PP/08 Reduce a number of compilation warnings under clang; building with
1861 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1862 should be warning-free.
1864 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1866 HS/01 Fix portability problems introduced by PP/08 for platforms where
1867 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1869 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1870 chunk. This allows us to accept broken chunked messages. We need a more
1871 general solution here.
1873 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1874 already-broken messages in the queue.
1876 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1878 JH/19 Fix reference counting bug in routing-generated-address tracking.
1884 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1885 supports it and a size is available (ie. the sending peer gave us one).
1887 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1888 years of being deprecated). The replacements are the ACLs
1889 acl_smtp_mime and acl_not_smtp_mime.
1891 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1892 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1893 or even an in-clear connection were permitted. Now, if the host lookup
1894 was dnssec and dane was requested then the host is only used if the
1895 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1896 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1897 if one fails this test.
1898 This means that a poorly-configured remote DNS will make it incommunicado;
1899 but it protects against a DNS-interception attack on it.
1901 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1902 non-noisy when a race steals the message being considered.
1904 JH/05 If main configuration option tls_certificate is unset, generate a
1905 self-signed certificate for inbound TLS connections.
1907 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1908 in rewrites and routers.
1910 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1911 and logged a warning sing 4.83; now they are a configuration file error.
1913 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1914 (lacking @domain). Apply the same qualification processing as RCPT.
1916 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1918 JH/10 Support ${sha256:} applied to a string (as well as the previous
1921 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1922 a cutthrough deliver is pending, as we always want to make a connection.
1923 This also avoids re-routing the message when later placing the cutthrough
1924 connection after a verify cache hit.
1925 Do not update it with the verify result either.
1927 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1928 when routing results in more than one destination address.
1930 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1931 signing (which inhibits the cutthrough capability). Previously only
1932 the presence of an option was tested; now an expansion evaluating as
1933 empty is permissible (obviously it should depend only on data available
1934 when the cutthrough connection is made).
1936 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1937 the relevant preceding SMTP command did not note the pipelining mode.
1939 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1940 Previously they were not counted.
1942 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1943 as one having no matching records. Previously we deferred the message
1944 that needed the lookup.
1946 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1947 distinguished as "(=".
1949 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1950 for missing MX records. Previously it only worked for missing A records.
1952 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1954 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1955 after the data-go-ahead and data-ack. Patch from Jason Betts.
1957 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1958 even for a "none" policy. Patch from Tony Meyer.
1960 JH/22 Fix continued use of a connection for further deliveries. If a port was
1961 specified by a router, it must also match for the delivery to be
1964 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1965 When one of the recipients of a message was unsuitable for the connection
1966 (has no matching addresses), we lost track of needing to mark it
1967 deferred. As a result mail would be lost.
1969 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1971 JH/25 Decoding ACL controls is now done using a binary search; the source code
1972 takes up less space and should be simpler to maintain. Merge the ACL
1973 condition decode tables also, with similar effect.
1975 JH/26 Fix problem with one_time used on a redirect router which returned the
1976 parent address unchanged. A retry would see the parent address marked as
1977 delivered, so not attempt the (identical) child. As a result mail would
1980 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1981 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1982 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1985 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1986 Main config options check_{log,spool}_{inodes,space} are now
1987 100 inodes, 10MB unless set otherwise in the configuration.
1989 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1990 Previously it only applied to the main-section connection policy
1993 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1995 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1996 by me. Added RFC7919 DH primes as an alternative.
1998 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1999 are not in the system include path.
2001 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2002 GnuTLS, when a session startup failed (eg because the client disconnected)
2003 Exim did stdio operations after fclose. This was exposed by a recent
2004 change which nulled out the file handle after the fclose.
2006 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2007 signed directly by the cert-signing cert, rather than an intermediate
2008 OCSP-signing cert. This is the model used by LetsEncrypt.
2010 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2012 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2013 an incoming connection.
2015 HS/02 Bug 1802: Do not half-close the connection after sending a request
2018 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2019 fallback to "prime256v1".
2021 JH/34 SECURITY: Use proper copy of DATA command in error message.
2022 Could leak key material. Remotely exploitable. CVE-2016-9963.
2028 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2029 and 3.4.4 - once the server is enabled to respond to an OCSP request
2030 it does even when not requested, resulting in a stapling non-aware
2031 client dropping the TLS connection.
2033 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2034 support variable-length bit vectors. No functional change.
2036 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2037 The I= interface field on outgoing lines is now after the H= remote
2038 host field, same as incoming lines. There is a separate
2039 outgoing_interface log selector which allows you to disable the
2042 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2043 If not running log_selector +smtp_connection the mainlog would be held
2044 open indefinitely after a "too many connections" event, including to a
2045 deleted file after a log rotate. Leave the per net connection logging
2046 leaving it open for efficiency as that will be quickly detected by the
2047 check on the next write.
2049 HS/01 Bug 1671: Fix post transport crash.
2050 Processing the wait-<transport> messages could crash the delivery
2051 process if the message IDs didn't exist for some reason. When
2052 using 'split_spool_directory=yes' the construction of the spool
2053 file name failed already, exposing the same netto behaviour.
2055 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2056 mime_regex ACL conditions.
2058 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2059 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2060 to HELO, local diagnostic string.
2062 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2063 log line to debug. Even when configured with a tls authenticator many
2064 client connections are expected to not authenticate in this way, so
2065 an authenticate fail is not an error.
2067 HS/02 Add the Exim version string to the process info. This way exiwhat
2068 gives some more detail about the running daemon.
2070 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2071 matter for fast-change records such as DNSBLs.
2073 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2074 retry record, even if constant. There may be multiple transports with
2075 different interface settings and the retry behaviour needs to be kept
2078 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2080 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2082 JH/10 Bug 840: fix log_defer_output option of pipe transport
2084 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2085 hosts_randomize. This matters a lot when combined with mua_wrapper.
2087 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2088 ${quote_pgsql:<string>} operator.
2090 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2092 JH/14 Change header file name for freeradius-client. Relevant if compiling
2093 with Radius support; from the Gentoo tree and checked under Fedora.
2095 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2097 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2098 option transport_filter, meaning no filtering.
2100 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2102 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2103 defaults to "*" (all hosts). The variable is now available when not built
2104 with TLS, default unset, mainly to enable keeping the testsuite sane.
2105 If a server certificate is not supplied (via tls_certificate) an error is
2106 logged, and clients will find TLS connections fail on startup. Presumably
2107 they will retry in-clear.
2108 Packagers of Exim are strongly encouraged to create a server certificate
2109 at installation time.
2111 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2112 with the $config_file variable.
2114 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2115 in transport context, after the attempt, and per-recipient. The latter type
2116 is per host attempted. The event data is the error message, and the errno
2117 information encodes the lookup type (A vs. MX) used for the (first) host,
2118 and the trailing two digits of the smtp 4xx response.
2120 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2121 to write to mainlog (or rejectlog, paniclog) in the window between file
2122 creation and permissions/ownership being changed. Particularly affects
2123 installations where exicyclog is run as root, rather than exim user;
2124 result is that the running daemon panics and dies.
2126 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2128 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2129 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2130 "pri" and "weight". Note that the previous implicit priority given by the
2131 list order is no longer honoured.
2133 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2134 for DKIM processing.
2136 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2137 by defining SUPPORT_SOCKS.
2139 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2140 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2141 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2142 variables are renamed to proxy_{local,external}_{address,port}.
2144 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2145 for a build by defining SUPPORT_I18N
2147 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2148 of the query string, and make ${quote_redis:} do that quoting.
2150 JH/29 Move Events support from Experimental to mainline, enabled by default
2151 and removable for a build by defining DISABLE_EVENT.
2153 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2155 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2156 cached by the daemon.
2158 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2159 by defining LOOKUP_REDIS. The libhiredis library is required.
2161 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2162 keys are given for lookup.
2164 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2165 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2166 only supported when built with TLS support. The PolarSSL SHA routines
2167 are still used when the TLS library is too old for convenient support.
2169 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2170 openssl_options), for security. OpenSSL forces this from version 1.1.0
2171 server-side so match that on older versions.
2173 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2174 allocation for $value could be released as the expansion processing
2175 concluded, but leaving the global pointer active for it.
2177 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2178 and to use the domains and local_parts ACL conditions.
2180 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2181 incorrectly not doubled on cutthrough transmission, hence seen as a
2182 body-termination at the receiving system - resulting in truncated mails.
2183 Commonly the sender saw a TCP-level error, and retransmitted the message
2184 via the normal store-and-forward channel. This could result in duplicates
2185 received - but deduplicating mailstores were liable to retain only the
2186 initial truncated version.
2188 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2190 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2192 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2193 we're in there, support oversigning also; bug 1309.
2195 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2197 HS/04 Add support for keep_environment and add_environment options.
2199 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2200 either intentional arithmetic overflow during PRNG, or testing config-
2203 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2204 delivery resulted in actual delivery. Cancel cutthrough before DATA
2207 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2208 rejects a recipient: pass the reject to the originator.
2210 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2211 Many were false-positives and ignorable, but it's worth fixing the
2214 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2215 for the new environment-manipulation done at startup. Move the routines
2216 from being local to tls.c to being global via the os.c file.
2218 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2219 an extract embedded as result-arg for a map, the first arg for extract
2220 is unavailable so we cannot tell if this is a numbered or keyed
2221 extraction. Accept either.
2227 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2230 JH/02 The smtp transport option "multi_domain" is now expanded.
2232 JH/03 The smtp transport now requests PRDR by default, if the server offers
2235 JH/04 Certificate name checking on server certificates, when exim is a client,
2236 is now done by default. The transport option tls_verify_cert_hostnames
2237 can be used to disable this per-host. The build option
2238 EXPERIMENTAL_CERTNAMES is withdrawn.
2240 JH/05 The value of the tls_verify_certificates smtp transport and main options
2241 default to the word "system" to access the system default CA bundle.
2242 For GnuTLS, only version 3.0.20 or later.
2244 JH/06 Verification of the server certificate for a TLS connection is now tried
2245 (but not required) by default. The verification status is now logged by
2246 default, for both outbound TLS and client-certificate supplying inbound
2249 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2252 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2253 Status Notification (bounce) messages are now MIME format per RFC 3464.
2254 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2255 under the control of the dsn_advertise_hosts option, and routers may
2256 have a dsn_lasthop option.
2258 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2259 default, modifiable by a malware= option. The list separator for
2260 the options can now be changed in the usual way. Bug 68.
2262 JH/10 The smtp_receive_timeout main option is now expanded before use.
2264 JH/11 The incoming_interface log option now also enables logging of the
2265 local interface on delivery outgoing connections.
2267 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2268 if the interface and destination host and port all match.
2270 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2273 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2274 Patch from Andrew Lewis.
2276 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2277 now supports optional time-restrictions, weighting, and priority
2278 modifiers per server. Patch originally by <rommer@active.by>.
2280 JH/16 The spamd_address main option now supports a mixed list of local
2281 and remote servers. Remote servers can be IPv6 addresses, and
2282 specify a port-range.
2284 JH/17 Bug 68: The spamd_address main option now supports an optional
2285 timeout value per server.
2287 JH/18 Bug 1581: Router and transport options headers_add/remove can
2288 now have the list separator specified.
2290 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2293 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2296 JH/21 Support for the A6 type of dns record is withdrawn.
2298 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2299 rather than the verbs used.
2301 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2302 from 255 to 1024 chars.
2304 JH/24 Verification callouts now attempt to use TLS by default.
2306 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2307 are generic router options now. The defaults didn't change.
2309 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2310 Original patch from Alexander Shikoff, worked over by JH.
2312 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2313 files if ZCAT_COMMAND is not executable.
2315 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2317 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2319 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2320 Normally benign, it bites when the pair was led to by a CNAME;
2321 modern usage is to not canonicalize the domain to a CNAME target
2322 (and we were inconsistent anyway for A-only vs AAAA+A).
2324 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2326 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2327 when evaluating $sender_host_dnssec.
2329 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2330 $sender_helo_dnssec variable.
2332 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2334 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2336 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2338 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2339 documented as working, but never had. Support all but $spam_report.
2341 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2342 added for tls authenticator.
2344 HS/03 Add perl_taintmode main config option
2350 TL/01 When running the test suite, the README says that variables such as
2351 no_msglog_check are global and can be placed anywhere in a specific
2352 test's script, however it was observed that placement needed to be near
2353 the beginning for it to behave that way. Changed the runtest perl
2354 script to read through the entire script once to detect and set these
2355 variables, reset to the beginning of the script, and then run through
2356 the script parsing/test process like normal.
2358 TL/02 The BSD's have an arc4random API. One of the functions to induce
2359 adding randomness was arc4random_stir(), but it has been removed in
2360 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2361 function when detected.
2363 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2364 cause callback expansion.
2366 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2367 syntax errors in an expansion can be treated as a string instead of
2368 logging or causing an error, due to the internal use of bool_lax
2369 instead of bool when processing it.
2371 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2372 server certificates when making smtp deliveries.
2374 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2376 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2378 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2380 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2381 Merged patch from Sebastian Wiedenroth.
2383 JH/05 Fix results-pipe from transport process. Several recipients, combined
2384 with certificate use, exposed issues where response data items split
2385 over buffer boundaries were not parsed properly. This eventually
2386 resulted in duplicates being sent. This issue only became common enough
2387 to notice due to the introduction of connection certificate information,
2388 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2390 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2391 size buffer was used, resulting in syntax errors when an expansion
2394 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2395 version 3.3.6 or later.
2397 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2398 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2399 both become "event_action", the variables become $event_name, $event_data
2400 and $event_defer_errno. There is a new variable $verify_mode, usable in
2401 routers, transports and related events. The tls:cert event is now also
2402 raised for inbound connections, if the main configuration event_action
2405 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2406 early OCSP support, but no stapling (appears to be less than 1.0.0).
2408 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2409 server certificate names available under the smtp transport option
2410 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2413 JH/10 Time-related extraction expansions from certificates now use the main
2414 option "timezone" setting for output formatting, and are consistent
2415 between OpenSSL and GnuTLS compilations. Bug 1541.
2417 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2418 encoded parameter in the incoming message. Bug 1558.
2420 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2421 include certificate info, eximon was claiming there were spoolfile
2424 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2426 JH/14 Log delivery-related information more consistently, using the sequence
2427 "H=<name> [<ip>]" wherever possible.
2429 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2430 are problematic for Debian distribution, omit them from the release
2433 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2435 JH/16 Fix string representation of time values on 64bit time_t architectures.
2438 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2439 output list separator was used.
2444 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2445 checkers that were complaining about end of non-void function with no
2448 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2449 This was a regression introduced in 4.83 by another bugfix.
2451 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2453 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2454 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2460 TF/01 Correctly close the server side of TLS when forking for delivery.
2462 When a message was received over SMTP with TLS, Exim failed to clear up
2463 the incoming connection properly after forking off the child process to
2464 deliver the message. In some situations the subsequent outgoing
2465 delivery connection happened to have the same fd number as the incoming
2466 connection previously had. Exim would try to use TLS and fail, logging
2467 a "Bad file descriptor" error.
2469 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2470 utilities have not been installed.
2472 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2473 temporary space as the ACL may create new global variables.
2475 TL/01 LDAP support uses per connection or global context settings, depending
2476 upon the detected version of the libraries at build time.
2478 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2479 to extract and use the src ip:port in logging and expansions as if it
2480 were a direct connection from the outside internet. PPv2 support was
2481 updated based on HAProxy spec change in May 2014.
2483 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2485 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2486 Properly escape header and check for NULL return.
2488 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2491 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2493 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2494 characters in header names, implemented as a verify condition.
2495 Contributed by Michael Fischer v. Mollard.
2497 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2498 results permerror and temperror. Previous values are deprecated but
2499 still accepted. In a future release, err_perm and err_temp will be
2500 completely removed, which will be a backward incompatibility if the
2501 ACL tests for either of these two old results. Patch contributed by
2502 user bes-internal on the mailing list.
2504 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2506 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2507 selectors, in both main and reject logs.
2509 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2512 JH/07 Add malware type "sock" for talking to simple daemon.
2514 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2516 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2517 routers/transports under cutthrough routing.
2519 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2520 numbers. Touch up "bool" conditional to keep the same definition.
2522 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2524 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2526 JH/12 Expand items in router/transport headers_add or headers_remove lists
2527 individually rather than the list as a whole. Bug 1452.
2529 Required for reasonable handling of multiple headers_ options when
2530 they may be empty; requires that headers_remove items with embedded
2531 colons must have them doubled (or the list-separator changed).
2533 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2534 view the policy declared in the DMARC record. Currently, $dmarc_status
2535 is a combined value of both the record presence and the result of the
2538 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2540 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2541 dnslookup router and the smtp transport (applying to the forward
2544 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2545 of ldap servers used for a specific lookup. Patch provided by Heiko
2548 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2549 New variable $lookup_dnssec_authenticated for observability.
2551 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2552 Patch submitted by Lars Timman.
2554 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2556 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2557 Requires trusted mode and valid format message id, aborts otherwise.
2558 Patch contributed by Heiko Schlichting.
2560 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2561 certextract with support for various fields. Bug 1358.
2563 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2564 is requested by default, modifiable by smtp transport option
2567 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2568 operate on certificate variables to give certificate fingerprints
2569 Also new ${sha256:cert_variable}.
2571 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2573 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2574 Christian Aistleitner.
2576 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2578 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2579 file. Patch from Wolfgang Breyha.
2581 JH/25 Expand the coverage of the delivery $host and $host_address to
2582 client authenticators run in verify callout. Bug 1476.
2584 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2585 align with daemon_smtp_ports. Bug 72.
2587 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2588 support and error reporting did not work properly.
2590 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2591 and is readable. Patch from Andrew Colin Kissa.
2593 TL/14 Enhance documentation of ${run expansion and how it parses the
2594 commandline after expansion, particularly in the case when an
2595 unquoted variable expansion results in an empty value.
2597 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2599 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2600 by renaming away from it. Addresses GH issue 15, reported by
2603 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2605 TL/15 SECURITY: prevent double expansion in math comparison functions
2606 (can expand unsanitized data). Not remotely exploitable.
2613 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2615 PP/02 Make -n do something, by making it not do something.
2616 When combined with -bP, the name of an option is not output.
2618 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2621 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2622 $sender_host_name and config options to manage this, and basic check
2625 PP/05 DSCP support for outbound connections and control modifier for inbound.
2627 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2628 (Only plugin which currently uses this is kerberos4, which nobody should
2629 be using, but we should make it available and other future plugins might
2630 conceivably use it, even though it would break NAT; stuff *should* be
2631 using channel bindings instead).
2633 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2634 name; added for Sendmail compatibility; requires admin caller.
2635 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2636 just ignore it); requires trusted caller.
2637 Also parse but ignore: -Ac -Am -X<logfile>
2640 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2642 TL/02 Add +smtp_confirmation as a default logging option.
2644 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2645 Patch by Magnus Holmgren from 2007-02-20.
2647 TL/04 Bugzilla 1281 - Spec typo.
2648 Bugzilla 1283 - Spec typo.
2649 Bugzilla 1290 - Spec grammar fixes.
2651 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2653 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2655 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2656 mailing list by by Dmitry Isaikin.
2658 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2660 JH/02 Support "G" suffix to numbers in ${if comparisons.
2662 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2664 NM/01 Bugzilla 1197 - Spec typo
2665 Bugzilla 1196 - Spec examples corrections
2667 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2669 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2670 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2673 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2674 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2676 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2677 "acl {{name}{arg}...}", and optional args on acl condition
2680 JH/05 Permit multiple router/transport headers_add/remove lines.
2682 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2684 JH/07 Avoid using a waiting database for a single-message-only transport.
2685 Performance patch from Paul Fisher. Bugzilla 1262.
2687 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2690 JH/09 Add $headers_added variable, with content from use of ACL modifier
2691 add_header (but not yet added to the message). Bugzilla 199.
2693 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2694 Pulled from Bugzilla 817 by Wolfgang Breyha.
2696 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2698 (nb: this is the same fix as in Exim 4.80.1)
2700 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2703 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2706 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2708 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2709 advertise SMTP AUTH mechanism to us, instead of a generic
2710 protocol violation error. Also, make Exim more robust to bad
2711 data from the Dovecot auth socket.
2713 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2715 When a queue runner is handling a message, Exim first routes the
2716 recipient addresses, during which it prunes them based on the retry
2717 hints database. After that it attempts to deliver the message to
2718 any remaining recipients. It then updates the hints database using
2721 So if a recipient address works intermittently, it can get repeatedly
2722 deferred at routing time. The retry hints record remains fresh so the
2723 address never reaches the final cutoff time.
2725 This is a fairly common occurrence when a user is bumping up against
2726 their storage quota. Exim had some logic in its local delivery code
2727 to deal with this. However it did not apply to per-recipient defers
2728 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2730 This change adds a proper retry rule check during routing so that the
2731 final cutoff time is checked against the message's age. We only do
2732 this check if there is an address retry record and there is not a
2733 domain retry record; this implies that previous attempts to handle
2734 the address had the retry_use_local_parts option turned on. We use
2735 this as an approximation for the destination being like a local
2736 delivery, as in LMTP.
2738 I suspect this new check makes the old local delivery cutoff check
2739 redundant, but I have not verified this so I left the code in place.
2741 TF/02 Correct gecos expansion when From: is a prefix of the username.
2743 Test 0254 submits a message to Exim with the header
2747 When I ran the test suite under the user fanf2, Exim expanded
2748 the header to contain my full name, whereas it should have added
2749 a Resent-Sender: header. It erroneously treats any prefix of the
2750 username as equal to the username.
2752 This change corrects that bug.
2754 GF/01 DCC debug and logging tidyup
2755 Error conditions log to paniclog rather than rejectlog.
2756 Debug lines prefixed by "DCC: " to remove any ambiguity.
2758 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2760 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2761 Bug spotted by Jeremy Harris; was flawed since initial commit.
2762 Would have resulted in OCSP responses post-SNI triggering an Exim
2763 NULL dereference and crash.
2765 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2767 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2768 Bug detection, analysis and fix by Samuel Thibault.
2769 Bugzilla 1331, Debian bug #698092.
2771 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2773 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2774 Server implementation by Todd Lyons, client by JH.
2775 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2776 config variable "prdr_enable" controls whether the server
2777 advertises the facility. If the client requests PRDR a new
2778 acl_data_smtp_prdr ACL is called once for each recipient, after
2779 the body content is received and before the acl_smtp_data ACL.
2780 The client is controlled by both of: a hosts_try_prdr option
2781 on the smtp transport, and the server advertisement.
2782 Default client logging of deliveries and rejections involving
2783 PRDR are flagged with the string "PRDR".
2785 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2786 fclose(). Diagnosis by Todd Lyons.
2788 PP/17 Update configure.default to handle IPv6 localhost better.
2789 Patch by Alain Williams (plus minor tweaks).
2792 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2793 This is now consistent with GnuTLS, and is now documented: the
2794 previous undocumented portable approach to treating the option as
2795 unset was to force an expansion failure. That still works, and
2796 an empty string is now equivalent.
2798 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2799 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2800 not performing validation itself.
2802 PP/20 Added force_command boolean option to pipe transport.
2803 Patch from Nick Koston, of cPanel Inc.
2805 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2808 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2810 PP/21 Fix eximon continuous updating with timestamped log-files.
2811 Broken in a format-string cleanup in 4.80, missed when I repaired the
2812 other false fix of the same issue.
2813 Report and fix from Heiko Schlichting.
2816 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2817 Report from Prashanth Katuri.
2819 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2820 It's SecureTransport, so affects any MacOS clients which use the
2821 system-integrated TLS libraries, including email clients.
2823 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2824 using a MIME ACL for non-SMTP local injection.
2825 Report and assistance in diagnosis by Warren Baker.
2827 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2829 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2831 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2832 last id that failed so it may be referenced in subsequent ACL's.
2834 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2837 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2838 ldap library initialization, allowing self-signed CA's to be
2839 used. Also properly sets require_cert option later in code by
2840 using NULL (global ldap config) instead of ldap handle (per
2841 session). Bug diagnosis and testing by alxgomz.
2843 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2844 the src/util/ subdirectory.
2846 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2847 renamed to Transport Post Delivery Action by Jeremy Harris, as
2850 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2851 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2852 redis_servers = needs to be configured which will be used by the redis
2853 lookup. Patch from Warren Baker, of The Packet Hub.
2855 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2857 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2858 hostname or reverse DNS when processing a host list. Used suggestions
2859 from multiple comments on this bug.
2861 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2863 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2864 Missed a few lines, added it to make the runtest require no keyboard
2867 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2868 contains upper case chars. Make router use caseful_local_part.
2870 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2871 support when GnuTLS has been built with p11-kit.
2877 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2879 This, or similar/improved, will also be change PP/11 of 4.82.
2885 PP/01 Handle short writes when writing local log-files.
2886 In practice, only affects FreeBSD (8 onwards).
2887 Bugzilla 1053, with thanks to Dmitry Isaikin.
2889 NM/01 Bugzilla 949 - Documentation tweak
2891 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2894 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2896 PP/02 Implemented gsasl authenticator.
2898 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2900 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2901 `pkg-config foo` for cflags/libs.
2903 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2904 with rest of GSASL and with heimdal_gssapi.
2906 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2907 `pkg-config foo` for cflags/libs for the TLS implementation.
2909 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2910 properties get this fed in as external SSF. A number of robustness
2911 and debugging improvements to the cyrus_sasl authenticator.
2913 PP/08 cyrus_sasl server now expands the server_realm option.
2915 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2916 Patch by Jeremy Harris.
2918 PP/10 Added dbmjz lookup type.
2920 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2922 PP/12 MAIL args handles TAB as well as SP, for better interop with
2923 non-compliant senders.
2924 Analysis and variant patch by Todd Lyons.
2926 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2927 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2928 Patch from Dirk Mueller <dmueller@suse.com>
2930 PP/13 tls_peerdn now print-escaped for spool files.
2931 Observed some $tls_peerdn in wild which contained \n, which resulted
2932 in spool file corruption.
2934 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2935 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2936 or write after TLS renegotiation, which otherwise led to messages
2939 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2940 as a tracking header (ie: a signed header comes before the signature).
2941 Patch from Wolfgang Breyha.
2943 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2944 comma-sep list; embedded commas doubled.
2946 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2948 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2950 Report and patch from Dmitry Banschikov.
2952 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2953 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2954 failures. We appear to now support TLS1.1+ with Exim.
2956 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2957 lets Exim select keys and certificates based upon TLS SNI from client.
2958 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2959 before an outbound SMTP session. New log_selector, +tls_sni.
2961 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2962 NULL dereference. Report and patch from Alun Jones.
2964 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2965 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2966 Not seeing resolver debug output on NetBSD, but suspect this is a
2967 resolver implementation change.
2969 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2970 Left warnings. Added "eximon gdb" invocation mode.
2972 PP/21 Defaulting "accept_8bitmime" to true, not false.
2974 PP/22 Added -bw for inetd wait mode support.
2976 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2977 locate the relevant includes and libraries. Made this the default.
2979 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2980 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2982 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2983 This may cause build issues on older platforms.
2985 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2986 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2987 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2988 Added SNI support via GnuTLS too.
2989 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2991 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2993 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2994 Applied second patch from Janne, implementing suggestion to default
2995 multiple-strings-in-record handling to match SPF spec.
2997 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2999 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3000 read-only, out of scope).
3001 Patch from Wolfgang Breyha, report from Stuart Northfield.
3003 PP/29 Fix three issues highlighted by clang analyser static analysis.
3004 Only crash-plausible issue would require the Cambridge-specific
3005 iplookup router and a misconfiguration.
3006 Report from Marcin Mirosław.
3008 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3010 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3011 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3012 As part of this, removing so much warning spew let me fix some minor
3013 real issues in debug logging.
3015 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3016 assignment on my part. Fixed.
3018 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3019 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3020 Janne Snabb (who went above and beyond: thank you).
3022 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3023 string otherwise requires a connection and a bunch more work and it's
3024 relatively easy to get wrong. Should also expose TLS library linkage
3027 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3028 64-bit ${eval} (JH/03).
3030 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3031 GNU libc to support some of the 64-bit stuff, should not lead to
3032 conflicts. Defined before os.h is pulled in, so if a given platform
3033 needs to override this, it can.
3035 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3036 protection layer was required, which is not implemented.
3037 Bugzilla 1254, patch from Wolfgang Breyha.
3039 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3040 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3041 tls_dhparam take prime identifiers. Also unbreak combination of
3042 OpenSSL+DH_params+TLSSNI.
3044 PP/39 Disable SSLv2 by default in OpenSSL support.
3050 PP/01 Solaris build fix for Oracle's LDAP libraries.
3051 Bugzilla 1109, patch from Stephen Usher.
3053 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3055 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3058 TF/02 Fix a couple more cases where we did not log the error message
3059 when unlink() failed. See also change 4.74-TF/03.
3061 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3062 lock up or crash if it happened to be inside a call to libc when it
3063 got a SIGUSR1 from exiwhat.
3065 The SIGUSR1 handler appends the current process status to the process
3066 log which is later printed by exiwhat. It used to use the general
3067 purpose logging code to do this, but several functions it calls are
3068 not safe for signals.
3070 The new output code in the SIGUSR1 handler is specific to the process
3071 log, and simple enough that it's easy to inspect for signal safety.
3072 Removing some special cases also simplifies the general logging code.
3073 Removing the spurious timestamps from the process log simplifies
3076 TF/04 Improved ratelimit ACL condition.
3078 The /noupdate option has been deprecated in favour of /readonly which
3079 has clearer semantics. The /leaky, /strict, and /readonly update modes
3080 are mutually exclusive. The update mode is no longer included in the
3081 database key; it just determines when the database is updated. (This
3082 means that when you upgrade Exim will forget old rate measurements.)
3084 Exim now checks that the per_* options are used with an update mode that
3085 makes sense for the current ACL. For example, when Exim is processing a
3086 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3087 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3088 must specify per_mail/readonly. If you omit the update mode it defaults to
3089 /leaky where that makes sense (as before) or /readonly where required.
3091 The /noupdate option is now undocumented but still supported for
3092 backwards compatibility. It is equivalent to /readonly except that in
3093 ACLs where /readonly is required you may specify /leaky/noupdate or
3094 /strict/noupdate which are treated the same as /readonly.
3096 A useful new feature is the /count= option. This is a generalization
3097 of the per_byte option, so that you can measure the throughput of other
3098 aggregate values. For example, the per_byte option is now equivalent
3099 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3101 The per_rcpt option has been generalized using the /count= mechanism
3102 (though it's more complicated than the per_byte equivalence). When it is
3103 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3104 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3105 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3106 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3107 Note that using per_rcpt with a non-readonly update mode in more than
3108 one ACL will cause the recipients to be double-counted. (The per_mail
3109 and per_byte options don't have this problem.)
3111 The handling of very low rates has changed slightly. If the computed rate
3112 is less than the event's count (usually one) then this event is the first
3113 after a long gap. In this case the rate is set to the same as this event's
3114 count, so that the first message of a spam run is counted properly.
3116 The major new feature is a mechanism for counting the rate of unique
3117 events. The new per_addr option counts the number of different
3118 recipients that someone has sent messages to in the last time period. It
3119 behaves like per_rcpt if all the recipient addresses are different, but
3120 duplicate recipient addresses do not increase the measured rate. Like
3121 the /count= option this is a general mechanism, so the per_addr option
3122 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3123 example, measure the rate that a client uses different sender addresses
3124 with the options per_mail/unique=$sender_address. There are further
3125 details in the main documentation.
3127 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3129 TF/06 Removed a few PCRE remnants.
3131 TF/07 Automatically extract Exim's version number from tags in the git
3132 repository when doing development or release builds.
3134 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3135 Bugzilla 879. Patch from Paul Fisher.
3137 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3138 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3141 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3143 PP/05 Variable $av_failed, true if the AV scanner deferred.
3144 Bugzilla 1078. Patch from John Horne.
3146 PP/06 Stop make process more reliably on build failure.
3147 Bugzilla 1087. Patch from Heiko Schlittermann.
3149 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3150 Bugzilla 1089. Patch from Heiko Schlittermann.
3152 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3153 Bugzilla 1131. Patch from Holger Weiß.
3155 PP/09 Handle IPv6 addresses with SPF.
3156 Bugzilla 860. Patch from Wolfgang Breyha.
3158 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3160 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3163 PP/11 match_* no longer expand right-hand-side by default.
3164 New compile-time build option, EXPAND_LISTMATCH_RHS.
3165 New expansion conditions, "inlist", "inlisti".
3167 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3169 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3171 PP/14 fix log_write() format string regression from TF/03.
3172 Bugzilla 1152. Patch from Dmitry Isaikin.
3178 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3180 PP/02 Harmonised TLS library version reporting; only show if debugging.
3181 Layout now matches that introduced for other libraries in 4.74 PP/03.
3183 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3185 PP/04 New "dns_use_edns0" global option.
3187 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3190 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3191 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3193 TK/01 Updated PolarSSL code to 0.14.2.
3194 Bugzilla 1097. Patch from Andreas Metzler.
3196 PP/07 Catch divide-by-zero in ${eval:...}.
3197 Fixes bugzilla 1102.
3199 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3202 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3203 format-string attack -- SECURITY: remote arbitrary code execution.
3205 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3206 time unintentionally subject to list matching rules, letting the header
3207 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3208 arbitrary string expansion). This allowed for information disclosure.
3210 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3211 INT_MIN/-1 -- value coerced to INT_MAX.
3217 NM/01 Workaround for PCRE version dependency in version reporting
3220 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3221 This fixes portability to compilers other than gcc, notably
3222 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3224 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3225 makefiles for portability to HP-UX and POSIX correctness.
3227 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3228 Also via indented variable definition in the Makefile.
3229 (Debugging by Oliver Heesakkers).
3231 PP/02 Restore caching of spamd results with expanded spamd_address.
3232 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3234 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3235 Improves build reliability. Fix from: Frank Elsner
3237 NM/02 Fix wide character breakage in the rfc2047 coding
3238 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3240 NM/03 Allow underscore in dnslist lookups
3241 Fixes bug 1026. Patch from Graeme Fowler
3243 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3244 Code patches from Adam Ciarcinski of NetBSD.
3246 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3249 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3250 is logged, to avoid truncation. Patch from John Horne.
3252 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3253 Patch from Jakob Hirsch.
3255 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3256 SQL string expansion failure details.
3257 Patch from Andrey Oktyabrski.
3259 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3260 Patch from Simon Arlott.
3262 PP/09 New lookups functionality failed to compile on old gcc which rejects
3263 extern declarations in function scope.
3264 Patch from Oliver Fleischmann
3266 PP/10 Use sig_atomic_t for flags set from signal handlers.
3267 Check getgroups() return and improve debugging.
3268 Fixed developed for diagnosis in bug 927 (which turned out to be
3271 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3272 Patch from Mark Zealey.
3274 PP/12 Bugzilla 1056: Improved spamd server selection.
3275 Patch from Mark Zealey.
3277 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3278 Based on patch from Heiko Schlittermann.
3280 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3281 Patch from Uwe Doering, sign-off by Michael Haardt.
3283 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3284 variable declaration deep within a block. Bug and patch from
3287 PP/15 lookups-Makefile IRIX compatibility coercion.
3289 PP/16 Make DISABLE_DKIM build knob functional.
3291 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3292 Patch by Simon Arlott
3294 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3295 variable argument macros. Our copy now differs from upstream.
3301 TF/01 Failure to get a lock on a hints database can have serious
3302 consequences so log it to the panic log.
3304 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3305 controlled using the smtp_confirmation log selector.
3307 TF/03 Include the error message when we fail to unlink a spool file.
3309 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3310 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3311 for maintaining out-of-tree patches for some time.
3313 PP/01 Bugzilla 139: Documentation and portability issues.
3314 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3315 Handle per-OS dynamic-module compilation flags.
3317 PP/02 Let /dev/null have normal permissions.
3318 The 4.73 fixes were a little too stringent and complained about the
3319 permissions on /dev/null. Exempt it from some checks.
3320 Reported by Andreas M. Kirchwitz.
3322 PP/03 Report version information for many libraries, including
3323 Exim version information for dynamically loaded libraries. Created
3324 version.h, now support a version extension string for distributors
3325 who patch heavily. Dynamic module ABI change.
3327 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3328 privilege escalation vulnerability whereby the Exim run-time user
3329 can cause root to append content of the attacker's choosing to
3332 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3335 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3336 If dropping privileges for untrusted macros, we disabled normal logging
3337 on the basis that it would fail; for the Exim run-time user, this is not
3338 the case, and it resulted in successful deliveries going unlogged.
3339 Fixed. Reported by Andreas Metzler.
3345 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3346 only prepend for the Resent-* case. Fixes regression introduced in
3347 Exim 4.70 by NM/22 for Bugzilla 607.
3349 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3350 increasing numbers of administrators be bitten by this.
3352 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3354 PP/03 Bugzilla 994: added openssl_options main configuration option.
3356 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3358 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3360 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3362 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3363 without a peer certificate, leading to a segfault because of an
3364 assumption that peers always have certificates. Be a little more
3365 paranoid. Problem reported by Martin Tscholak.
3367 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3368 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3369 NB: ClamAV planning to remove STREAM in "middle of 2010".
3370 CL also introduces -bmalware, various -d+acl logging additions and
3371 more caution in buffer sizes.
3373 PP/09 Implemented reverse_ip expansion operator.
3375 PP/10 Bugzilla 937: provide a "debug" ACL control.
3377 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3379 PP/12 Bugzilla 973: Implement --version.
3381 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3383 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3385 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3387 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3388 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3389 ignore trailing whitespace.
3391 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3393 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3396 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3397 Notification from Dr Andrew Aitchison.
3399 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3400 ExtendedDetectionInfo response format.
3401 Notification from John Horne.
3403 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3406 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3407 XSL and documented dependency on system catalogs, with examples of how
3410 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3413 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3414 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3415 configuration file which is writeable by the Exim user or group.
3417 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3418 of configuration files to cover files specified with the -C option if
3419 they are going to be used with root privileges, not just the default
3422 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3423 option (effectively making it always true).
3425 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3426 files to be used while preserving root privileges.
3428 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3429 that rogue child processes cannot use them.
3431 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3432 run-time user, instead of root.
3434 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3435 Exim run-time user without dropping privileges.
3437 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3438 result string, instead of calling string_vformat() twice with the same
3441 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3442 for other users. Others should always drop root privileges if they use
3443 -C on the command line, even for a whitelisted configure file.
3445 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3447 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3453 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3454 $data_path, and $header_path variables; fixed documentation bugs and
3457 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3458 exipick to access non-standard spools, including the "frozen" queue
3461 NM/01 Bugzilla 965: Support mysql stored procedures.
3462 Patch from Alain Williams
3464 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3466 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3467 Patch from Andreas Metzler
3469 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3470 Patch from Kirill Miazine
3472 NM/05 Bugzilla 671: Added umask to procmail example.
3474 JJ/03 installed exipick 20100323.0, fixing doc bug
3476 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3477 directory. Notification and patch from Dan Rosenberg.
3479 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3481 TK/02 Improve log output when DKIM signing operation fails.
3483 MH/01 Treat the transport option dkim_domain as a colon separated
3484 list, not as a single string, and sign the message with each element,
3485 omitting multiple occurences of the same signer.
3487 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3488 Bugzilla 985, 986. Patch by Simon Arlott
3490 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3491 Patch by Simon Arlott
3493 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3494 MBX locking. Notification from Dan Rosenberg.
3500 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3502 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3504 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3506 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3508 NM/04 Bugzilla 915: Flag broken perl installation during build.
3514 TK/01 Added patch by Johannes Berg that expands the main option
3515 "spamd_address" if it starts with a dollar sign.
3517 TK/02 Write list of recipients to X-Envelope-Sender header when building
3518 the mbox-format spool file for content scanning (suggested by Jakob
3521 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3522 (http://www.dcc-servers.net/) support via dccifd. Activated by
3523 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3525 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3526 by Mark Daniel Reidel <mr@df.eu>.
3528 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3529 When building exim an external PCRE library is now needed -
3530 PCRE is a system library on the majority of modern systems.
3531 See entry on PCRE_LIBS in EDITME file.
3533 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3534 conversation. Added nologin parameter to request.
3535 Patch contributed by Kirill Miazine.
3537 TF/01 Do not log submission mode rewrites if they do not change the address.
3539 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3541 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3542 log files in place. Contributed by Roberto Lima.
3544 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3546 TF/03 Bugzilla 615: When checking the local_parts router precondition
3547 after a local_part_suffix or local_part_prefix option, Exim now
3548 does not use the address's named list lookup cache, since this
3549 contains cached lookups for the whole local part.
3551 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3552 Robert Millan. Documentation is in experimental-spec.txt.
3554 TF/04 Bugzilla 668: Fix parallel build (make -j).
3556 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3558 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3559 Patch provided by Jan Srzednicki.
3561 TF/05 Leading white space used to be stripped from $spam_report which
3562 wrecked the formatting. Now it is preserved.
3564 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3565 that they are available at delivery time.
3567 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3569 TF/08 TLS error reporting now respects the incoming_interface and
3570 incoming_port log selectors.
3572 TF/09 Produce a more useful error message if an SMTP transport's hosts
3573 setting expands to an empty string.
3575 NM/06 Bugzilla 744: EXPN did not work under TLS.
3576 Patch provided by Phil Pennock.
3578 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3579 Patch provided by Richard Godbee.
3581 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3582 acl_smtp_notquit, added index entry.
3584 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3585 Patch provided by Eugene Bujak.
3587 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3588 accept(). Patch provided by Maxim Dounin.
3590 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3591 Patch provided by Phil Pennock.
3593 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3595 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3596 Patch provided by Brad "anomie" Jorsch.
3598 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3599 Patch provided by Dean Brooks.
3601 TK/05 Add native DKIM support (does not depend on external libraries).
3603 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3604 Patch provided by Graeme Fowler.
3606 NM/16 Bugzilla 851: Documentation example syntax fix.
3608 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3610 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3613 NM/19 Bugzilla 745: TLS version reporting.
3614 Patch provided by Phil Pennock.
3616 NM/20 Bugzilla 167: bool: condition support.
3617 Patch provided by Phil Pennock.
3619 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3620 clients. Patch provided by Phil Pennock.
3622 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3623 Patch provided by Brad "anomie" Jorsch.
3625 NM/23 Bugzilla 687: Fix misparses in eximstats.
3626 Patch provided by Heiko Schlittermann.
3628 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3629 Patch provided by Heiko Schlittermann.
3631 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3632 plus update to original patch.
3634 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3636 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3637 Patch provided by David Brownlee.
3639 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3641 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3643 NM/30 Bugzilla 888: TLS documentation bugfixes.
3645 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3647 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3648 Unlike the original bugzilla I have changed all shell scripts in src tree.
3650 NM/33 Bugzilla 898: Transport filter timeout fix.
3651 Patch by Todd Rinaldo.
3653 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3654 Patch by Serge Demonchaux.
3656 NM/35 Bugzilla 39: Base64 decode bug fixes.
3657 Patch by Jakob Hirsch.
3659 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3661 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3663 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3665 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3671 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3672 ALT-N's libdkim that I have put here:
3673 http://duncanthrax.net/exim-experimental/
3675 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3676 were called 'true' and it seems that C99 defines that as a reserved
3677 keyword to be used with 'bool' variable types. That means you could
3678 not include C99-style headers which use bools without triggering
3679 build errors in sieve.c.
3681 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3682 as mailq or other aliases. Changed the --help handling significantly
3683 to do whats expected. exim_usage() emits usage/help information.
3685 SC/01 Added the -bylocaldomain option to eximstats.
3687 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3689 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3691 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3697 PH/01 Another patch from the Sieve maintainer.
3699 PH/02 When an IPv6 address is converted to a string for single-key lookup
3700 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3701 dots are used instead of colons so that keys in lsearch files need not
3702 contain colons. This was done some time before quoting was made available
3703 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3704 (notated using the quote facility) so as to distinguish them from IPv4
3705 keys. This meant that lookups for IP addresses in host lists did not work
3706 for iplsearch lookups.
3708 This has been fixed by arranging for IPv6 addresses to be expressed with
3709 colons if the lookup type is iplsearch. This is not incompatible, because
3710 previously such lookups could never work.
3712 The situation is now rather anomalous, since one *can* have colons in
3713 ordinary lsearch keys. However, making the change in all cases is
3714 incompatible and would probably break a number of configurations.
3716 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3719 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3720 conversion specification without a maximum field width, thereby enabling
3721 a rogue spamd server to cause a buffer overflow. While nobody in their
3722 right mind would setup Exim to query an untrusted spamd server, an
3723 attacker that gains access to a server running spamd could potentially
3724 exploit this vulnerability to run arbitrary code as the Exim user.
3726 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3727 $primary_hostname instead of what libspf2 thinks the hosts name is.
3729 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3730 a directory entry by the name of the lookup key. Previously, if a
3731 symlink pointed to a non-existing file or a file in a directory that
3732 Exim lacked permissions to read, a lookup for a key matching that
3733 symlink would fail. Now it is enough that a matching directory entry
3734 exists, symlink or not. (Bugzilla 503.)
3736 PH/03 The body_linecount and body_zerocount variables are now exported in the
3739 PH/04 Added the $dnslist_matched variable.
3741 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3742 This means they are set thereafter only if the connection becomes
3745 PH/06 Added the client_condition to authenticators so that some can be skipped
3746 by clients under certain conditions.
3748 PH/07 The error message for a badly-placed control=no_multiline_responses left
3749 "_responses" off the end of the name.
3751 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3753 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3754 (without spaces) instead of just copying the configuration text.
3756 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3758 PH/11 Added $max_received_linelength.
3760 PH/12 Added +ignore_defer and +include_defer to host lists.
3762 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3763 way in which PCRE > 7.0 is built.
3765 PH/14 Implemented queue_only_load_latch.
3767 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3768 MAIL command. The effect was to mangle the value on 64-bit systems.
3770 PH/16 Another patch from the Sieve maintainer.
3772 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3774 PH/18 If a system quota error occurred while trying to create the file for
3775 a maildir delivery, the message "Mailbox is full" was not appended to the
3776 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3777 only to a quota excession during the actual writing of the file.
3779 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3780 characters?) which causes problems in log lines. The DN values are now
3781 passed through string_printing() before being added to log lines.
3783 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3784 and InterBase are left for another time.)
3786 PH/21 Added message_body_newlines option.
3788 PH/22 Guard against possible overflow in moan_check_errorcopy().
3790 PH/23 POSIX allows open() to be a macro; guard against that.
3792 PH/24 If the recipient of an error message contained an @ in the local part
3793 (suitably quoted, of course), incorrect values were put in $domain and
3794 $local_part during the evaluation of errors_copy.
3800 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3801 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3804 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3805 issue a MAIL command.
3807 PH/02 In an ACL statement such as
3809 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3811 if a client was not listed at all, or was listed with a value other than
3812 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3813 the condition was not true (as it should be), so access was not denied.
3814 The bug was that the ! inversion was incorrectly passed on to the second
3815 item. This has been fixed.
3817 PH/03 Added additional dnslists conditions == and =& which are different from
3818 = and & when the dns lookup returns more than one IP address.
3820 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3821 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3823 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3824 FSYNC, which compiles an option called disable_fsync that allows for
3825 bypassing fsync(). The documentation is heavily laced with warnings.
3827 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3829 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3830 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3831 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3832 including adding "make clean"; (3) Added -fPIC when compiling the test
3833 dynamically loaded module, to get rid of a warning.
3835 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3836 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3837 = 0s. The bug is otherwise harmless.
3839 PH/07 There was a bug in the dovecot authenticator such that the value of
3840 $auth1 could be overwritten, and so not correctly preserved, after a
3841 successful authentication. This usually meant that the value preserved by
3842 the server_setid option was incorrect.
3844 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3846 PH/09 Installed PCRE release 7.0.
3848 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3849 run for batched SMTP input. It is now run at the start of every message
3850 in the batch. While fixing this I discovered that the process information
3851 (output by running exiwhat) was not always getting set for -bs and -bS
3852 input. This is fixed, and it now also says "batched" for BSMTP.
3854 PH/11 Added control=no_pipelining.
3856 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3857 patch, slightly modified), and move the expansion of helo_data till after
3858 the connection is made in the smtp transport (so it can use these
3861 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3863 PH/14 Added log_selector = +pid.
3865 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3867 PH/16 Add ${if forany and ${if forall.
3869 PH/17 Added dsn_from option to vary the From: line in DSNs.
3871 PH/18 Flush SMTP output before performing a callout, unless control =
3872 no_callout_flush is set.
3874 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3875 was true (the default) a successful delivery failed to delete the retry
3876 item, thus causing premature timeout of the address. The bug is now
3879 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3881 PH/21 Long custom messages for fakedefer and fakereject are now split up
3882 into multiline responses in the same way that messages for "deny" and
3883 other ACL rejections are.
3885 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3886 with slight modification.
3888 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3889 draft, changing the syntax and factoring some duplicate code".
3891 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3892 for deliveries of the second and subsequent messages over the same SMTP
3895 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3896 ${reduce, with only minor "tidies".
3898 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3900 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3901 expansion side effects.
3903 PH/27 When a message times out after an over-quota error from an Exim-imposed
3904 quota, the bounce message says "mailbox is full". This message was not
3905 being given when it was a system quota that was exceeded. It now should
3908 MH/03 Made $recipients available in local_scan(). local_scan() already has
3909 better access to the recipient list through recipients_list[], but
3910 $recipients can be useful in postmaster-provided expansion strings.
3912 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3913 in the case of a MAIL command with additional options following the
3914 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3915 were accidentally chopped off.
3917 PH/29 SMTP synchronization checks are implemented when a command is read -
3918 there is a check that no more input is waiting when there shouldn't be
3919 any. However, for some commands, a delay in an ACL can mean that it is
3920 some time before the response is written. In this time, more input might
3921 arrive, invalidly. So now there are extra checks after an ACL has run for
3922 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3923 pipelining has not been advertised.
3925 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3927 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3928 returned in the SMTP response when smtp_return_error_details was false.
3929 This has been fixed.
3931 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3932 instead of the C I/O that was originally supplied, because problems were
3933 reported on Solaris.
3935 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3936 Exim which did not show up earlier: it was assuming that a call to
3937 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3938 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3939 was a macro that became an assignment, so it seemed to work. This has
3940 changed to a proper function call with a void return, hence the compile
3941 error. Exim's code has been fixed.
3943 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3946 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3948 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3950 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3951 because it turns out to be too big or there is a timeout) there is no
3952 "Completed" line in the log. When some messages of this type were
3953 selected by exigrep, they were listed as "not completed". Others were
3954 picked up by some special patterns. I have improved the selection
3955 criteria to be more general.
3957 PH/38 The host_find_failed option in the manualroute router can now be set
3958 to "ignore", to completely ignore a host whose IP address cannot be
3959 found. If all hosts are ignored, the behaviour is controlled by the new
3960 host_all_ignored option.
3962 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3963 homing or because of multiple MX records with /mx) generated more than
3964 one IP address, and the following item turned out to be the local host,
3965 all the secondary addresses of the first item were incorrectly removed
3966 from the list, along with the local host and any following hosts (which
3967 is what is supposed to happen).
3969 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3970 whoever called Exim into the -H file. In the case of the daemon it was
3971 behaving confusingly. When first started, it used values for whoever
3972 started the daemon, but after a SIGHUP it used the Exim user (because it
3973 calls itself on a restart). I have changed the code so that it now always
3976 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3977 message are rejected with the same error (e.g. no authentication or bad
3978 sender address), and a DATA command is nevertheless sent (as can happen
3979 with PIPELINING or a stupid MUA), the error message that was given to the
3980 RCPT commands is included in the rejection of the DATA command. This is
3981 intended to be helpful for MUAs that show only the final error to their
3984 PH/42 Another patch from the Sieve maintainer.
3986 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3987 Eximstats - Fixed some broken HTML links and added missing column headers
3989 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3990 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3992 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3994 PH/43 Yet another patch from the Sieve maintainer.
3996 PH/44 I found a way to check for a TCP/IP connection going away before sending
3997 the response to the final '.' that terminates a message, but only in the
3998 case where the client has not sent further data following the '.'
3999 (unfortunately, this is allowed). However, in many cases there won't be
4000 any further data because there won't be any more messages to send. A call
4001 to select() can be used: if it shows that the input is "ready", there is
4002 either input waiting, or the socket has been closed. An attempt to read
4003 the next input character can distinguish the two cases. Previously, Exim
4004 would have sent an OK response which the client would never have see.
4005 This could lead to message repetition. This fix should cure that, at
4006 least in a lot of common cases.
4008 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4009 advertised in response to EHLO.
4015 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4016 fixed by 4.65/MH/01 (is this a record?) are fixed:
4018 (i) An empty string was always treated as zero by the numeric comparison
4019 operators. This behaviour has been restored.
4021 (ii) It is documented that the numeric comparison operators always treat
4022 their arguments as decimal numbers. This was broken in that numbers
4023 starting with 0 were being interpreted as octal.
4025 While fixing these problems I realized that there was another issue that
4026 hadn't been noticed. Values of message_size_limit (both the global option
4027 and the transport option) were treated as octal if they started with 0.
4028 The documentation was vague. These values are now always treated as
4029 decimal, and I will make that clear in the documentation.
4035 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4036 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4039 MH/01 Don't check that the operands of numeric comparison operators are
4040 integers when their expansion is in "skipping" mode (fixes bug
4041 introduced by 4.64-PH/07).
4043 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4044 child addresses, Exim now panics and dies. Previously, because the count
4045 is held in a short int, deliveries were likely to be lost. As such a
4046 large number of recipients for a single message is ridiculous
4047 (performance will be very, very poor), I have chosen to impose a limit
4048 rather than extend the field.
4054 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4055 leftover -K file (the existence of which was triggered by #402).
4056 While we were at it, introduced process PID as part of the -K
4057 filename. This should rule out race conditions when creating
4060 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4061 processing considerably. Previous code took too long for large mails,
4062 triggering a timeout which in turn triggers #401.
4064 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4065 in the DK code in transports.c. sendfile() is not really portable,
4066 hence the _LINUX specificness.
4068 TF/01 In the add_headers option to the mail command in an Exim filter,
4069 there was a bug that Exim would claim a syntax error in any
4070 header after the first one which had an odd number of characters
4073 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4074 callout verification, Exim cached a "reject" for the entire domain. This
4075 is correct for most verifications, but it is not correct for a recipient
4076 verification with use_sender or use_postmaster set, because in that case
4077 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4078 case of MAIL FROM:<> rejection from other early rejections (e.g.
4079 rejection of HELO). When verifying a recipient using a non-null MAIL
4080 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4081 Whatever the result of the callout, the value of the domain cache is
4082 left unchanged (for any other kind of callout, getting as far as trying
4083 RCPT means that the domain itself is ok).
4085 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4088 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4089 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4090 session, the daemon ignores ECONNECT errors and logs others; it now
4091 ignores EPIPE as well.
4093 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4094 (quoted-printable decoding).
4096 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4097 later the small subsequent patch to fix an introduced bug.
4099 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4101 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4103 PH/08 An error is now given if message_size_limit is specified negative.
4105 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4106 to be given (somewhat) arbitrary names.
4108 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4111 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4112 miscellaneous code fixes
4114 PH/10 Added the log_reject_target ACL modifier to specify where to log
4117 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4118 hostname. This is wrong, because it relates to the incoming message (and
4119 probably the interface on which it is arriving) and not to the outgoing
4120 callout (which could be using a different interface). This has been
4121 changed to use the value of the helo_data option from the smtp transport
4122 instead - this is what is used when a message is actually being sent. If
4123 there is no remote transport (possible with a router that sets up host
4124 addresses), $smtp_active_hostname is used.
4126 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4127 tweaks were necessary in order to get it to work (see also 21 below):
4128 (a) The code assumed that strncpy() returns a negative number on buffer
4129 overflow, which isn't the case. Replaced with Exim's string_format()
4131 (b) There were several signed/unsigned issues. I just did the minimum
4132 hacking in of casts. There is scope for a larger refactoring.
4133 (c) The code used strcasecmp() which is not a standard C function.
4134 Replaced with Exim's strcmpic() function.
4135 (d) The code set only $1; it now sets $auth1 as well.
4136 (e) A simple test gave the error "authentication client didn't specify
4137 service in request". It would seem that Dovecot has changed its
4138 interface. Fortunately there's a specification; I followed it and
4139 changed what the client sends and it appears to be working now.
4141 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4144 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4145 address A is aliased to B and C, where B exists and C does not. Without
4146 -v the output is "A verified" because verification stops after a
4147 successful redirection if more than one address is generated. However,
4148 with -v the child addresses are also verified. Exim was outputting "A
4149 failed to verify" and then showing the successful verification for C,
4150 with its parentage. It now outputs "B failed to verify", showing B's
4151 parentage before showing the successful verification of C.
4153 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4154 look up a TXT record in a specific list after matching in a combined
4157 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4158 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4159 they consult the DNS. I had assumed they would set it the way they
4160 wanted; and indeed my experiments on Linux seem to show that in some
4161 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4162 To be on the safe side, however, I have now made the interface to
4163 host_find_byname() similar to host_find_bydns(), with an argument
4164 containing the DNS resolver options. The host_find_byname() function now
4165 sets these options at its start, just as host_find_bydns() does. The smtp
4166 transport options dns_qualify_single and dns_search_parents are passed to
4167 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4168 of host_find_byname() use the default settings of RES_DEFNAMES
4169 (qualify_single) but not RES_DNSRCH (search_parents).
4171 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4172 spool_read_header() do less string testing, by means of a preliminary
4173 switch on the second character of optional "-foo" lines. (This is
4174 overdue, caused by the large number of possibilities that now exist.
4175 Originally there were few.) While I was there, I also converted the
4176 str(n)cmp tests so they don't re-test the leading "-" and the first
4177 character, in the hope this might squeeze out yet more improvement.
4179 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4180 flag allowing group syntax was set by the header_syntax check but not
4181 turned off, possible causing trouble later; (2) The flag was not being
4182 set at all for the header_verify test, causing "group"-style headers to
4183 be rejected. I have now set it in this case, and also caused header_
4184 verify to ignore an empty address taken from a group. While doing this, I
4185 came across some other cases where the code for allowing group syntax
4186 while scanning a header line wasn't quite right (mostly, not resetting
4187 the flag correctly in the right place). These bugs could have caused
4188 trouble for malformed header lines. I hope it is now all correct.
4190 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4191 with the "reply" argument non-NULL. The code, however (which originally
4192 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4193 but it didn't always do it. This confused somebody who was copying the
4194 code for some other use. I have removed all the tests.
4196 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4197 feature that was used to support insecure browsers during the U.S. crypto
4198 embargo. It requires special client support, and Exim is probably the
4199 only MTA that supported it -- and would never use it because real RSA is
4200 always available. This code has been removed, because it had the bad
4201 effect of slowing Exim down by computing (never used) parameters for the
4202 RSA_EXPORT functionality.
4204 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4205 authenticator to fail if there's a tab character in the incoming data
4206 (there should never be unless someone is messing about, as it's supposed
4207 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4208 if the connection is using TLS or if the remote IP is the same as the
4209 local IP, and the "valid-client-cert option" if a client certificate has
4212 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4213 authenticators. This can be used for authorization after authentication
4214 succeeds. (In the case of plaintext, it servers for both authentication
4217 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4218 if any retry times were supplied.
4220 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4221 connection, where there is no client IP address to check. In this
4222 situation, the verify now always succeeds.
4224 PH/25 Applied John Jetmore's -Mset patch.
4226 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4228 PH/27 In a string expansion for a processed (not raw) header when multiple
4229 headers of the same name were present, leading whitespace was being
4230 removed from all of them, but trailing whitespace was being removed only
4231 from the last one. Now trailing whitespace is removed from each header
4232 before concatenation. Completely empty headers in a concatenation (as
4233 before) are ignored.
4235 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4236 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4238 PH/29 [Removed. This was a change that I later backed out, and forgot to
4239 correct the ChangeLog entry (that I had efficiently created) before
4240 committing the later change.]
4242 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4243 address errors (4xx response to RCPT) over the same connection as other
4244 messages routed to the same hosts. Such deliveries are always "forced",
4245 so retry times are not inspected. This resulted in far too many retries
4246 for the affected addresses. The effect occurred only when there were more
4247 hosts than the hosts_max_try setting in the smtp transport when it had
4248 the 4xx errors. Those hosts that it had tried were not added to the list
4249 of hosts for which the message was waiting, so if all were tried, there
4250 was no problem. Two fixes have been applied:
4252 (i) If there are any address or message errors in an SMTP delivery, none
4253 of the hosts (tried or untried) are now added to the list of hosts
4254 for which the message is waiting, so the message should not be a
4255 candidate for sending over the same connection that was used for a
4256 successful delivery of some other message. This seems entirely
4257 reasonable: after all the message is NOT "waiting for some host".
4258 This is so "obvious" that I'm not sure why it wasn't done
4259 previously. Hope I haven't missed anything, but it can't do any
4260 harm, as the worst effect is to miss an optimization.
4262 (ii) If, despite (i), such a delivery is accidentally attempted, the
4263 routing retry time is respected, so at least it doesn't keep
4264 hammering the server.
4266 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4267 in ${readsocket because some servers need this prod.
4269 PH/32 Added some extra debug output when updating a wait-xxx database.
4271 PH/33 The hint "could be header name not terminated by colon", which has been
4272 given for certain expansion errors for a long time, was not being given
4273 for the ${if def:h_colon_omitted{... case.
4275 PH/34 The spec says: "With one important exception, whenever a domain list is
4276 being scanned, $domain contains the subject domain." There was at least
4277 one case where this was not true.
4279 PH/35 The error "getsockname() failed: connection reset by peer" was being
4280 written to the panic log as well as the main log, but it isn't really
4281 panic-worthy as it just means the connection died rather early on. I have
4282 removed the panic log writing for the ECONNRESET error when getsockname()
4285 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4286 runs only) independently of the message's sender address. This meant
4287 that, if the 4xx error was in fact related to the sender, a different
4288 message to the same recipient with a different sender could confuse
4289 things. In particular, this can happen when sending to a greylisting
4290 server, but other circumstances could also provoke similar problems.
4291 I have changed the default so that the retry time for these errors is now
4292 based a combination of the sender and recipient addresses. This change
4293 can be overridden by setting address_retry_include_sender=false in the
4296 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4297 remote server are returned as part of bounce messages. This was not
4298 happening for LMTP over a pipe (the lmtp transport), but now it is the
4299 same for both kinds of LMTP.
4301 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4302 in header lines that were in fact CNAMEs. This is no longer the case.
4304 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4305 and queue runs started by the daemon processed all messages. This has
4306 been fixed so that -R and -S can now usefully be given with -q<time>.
4308 PH/40 Import PCRE release 6.7 (fixes some bugs).
4310 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4312 PH/42 Give an error if -q is specified more than once.
4314 PH/43 Renamed the variables $interface_address and $interface_port as
4315 $received_ip_address and $received_port, to make it clear that these
4316 values apply to message reception, and not to the outgoing interface when
4317 a message is delivered. (The old names remain recognized, of course.)
4319 PH/44 There was no timeout on the connect() call when using a Unix domain
4320 socket in the ${readsocket expansion. There now is.
4322 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4323 be meaningful with "accept".
4325 SC/01 Eximstats V1.43
4326 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4328 SC/02 Eximstats V1.44
4329 Use a glob alias rather than an array ref in the generated
4330 parser. This improves both readability and performance.
4332 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4333 Collect SpamAssassin and rejection statistics.
4334 Don't display local sender or destination tables unless
4335 there is data to show.
4336 Added average volumes into the top table text output.
4338 SC/04 Eximstats V1.46
4339 Collect data on the number of addresses (recipients)
4340 as well as the number of messages.
4342 SC/05 Eximstats V1.47
4343 Added 'Message too big' to the list of mail rejection
4344 reasons (thanks to Marco Gaiarin).
4346 SC/06 Eximstats V1.48
4347 Mainlog lines which have GMT offsets and are too short to
4348 have a flag are now skipped.
4350 SC/07 Eximstats V1.49 (Alain Williams)
4351 Added the -emptyok flag.
4353 SC/08 Eximstats V1.50
4354 Fixes for obtaining the IP address from reject messages.
4356 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4357 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4358 whitespace changes from 4.64-PH/27
4360 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4363 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4364 are found, allow negative numbers in numeric criteria)
4366 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4368 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4369 to match changes made in 4.64-PH/43
4371 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4373 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4374 "domain" is actually an IP address, and force a failure. This locks out
4375 those revolvers/nameservers that support "A-for-A" lookups, in
4376 contravention of the specifications.
4378 PH/48 When a host name was looked up from an IP address, and the subsequent
4379 forward lookup of the name timed out, the host name was left in
4380 $sender_host_name, contrary to the specification.
4382 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4383 restricted to single-key lookups, Exim was not diagnosing an error if
4384 * or *@ was used with a query-style lookup.
4386 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4388 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4389 long ago, but noone interested enough thought of it. Let's just say that
4390 the "1.1" means that there are some new functions that weren't there at
4391 some point in the past.
4393 PH/51 Error processing for expansion failure of helo_data from an smtp
4394 transport during callout processing was broken.
4396 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4397 tested/used via the -bh/-bhc/-bs options.
4399 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4400 bug, fixed in subsequent PCRE releases).
4402 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4403 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4405 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4411 SC/01 Use a glob alias rather than an array ref in eximstats generated
4412 parser. This improves both readability and performance.
4414 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4415 Don't display local sender or destination tables in eximstats unless
4416 there is data to show.
4417 Added average volumes into the eximstats top table text output.
4419 SC/03 Collect data on the number of addresses (recipients) as well
4420 as the number of messages in eximstats.
4422 TF/01 Correct an error in the documentation for the redirect router. Exim
4423 does (usually) call initgroups() when daemonizing.
4425 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4426 with consistent privilege compared to when running as a daemon.
4428 TF/03 Note in the spec that $authenticated_id is not set for local
4429 submissions from trusted users.
4431 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4432 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4434 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4435 by adding some example configuration directives to the default
4436 configuration file. A little bit of work is required to uncomment the
4437 directives and define how usernames and passwords are checked, but
4438 there is now a framework to start from.
4440 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4441 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4442 without this. I don't know how relevant this is to other LDAP libraries.
4444 PH/02 Add the verb name to the "unknown ACL verb" error.
4446 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4448 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4450 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4451 directory not expanded when it should be if an expanded home directory
4452 was set for the address (which is overridden by the transport).
4454 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4457 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4458 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4459 because it is too late at that time, and has no effect.
4461 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4462 security issue with \' (bugzilla #107). I could not use the
4463 PQescapeStringConn() function, because it needs a PGconn value as one of
4466 PH/08 When testing addresses using -bt, indicate those final addresses that
4467 are duplicates that would not cause an additional delivery. At least one
4468 person was confused, thinking that -bt output corresponded to deliveries.
4469 (Suppressing duplicates isn't a good idea as you lose the information
4470 about possibly different redirections that led to the duplicates.)
4472 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4473 systems where poll() doesn't work, in particular OS X.
4475 PH/10 Added more information to debugging output for retry time not reached.
4477 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4478 operations in malware.c.
4480 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4483 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4484 syslog_duplication set false, log lines that would normally be written
4485 both the the main log and to the reject log were not written to syslog at
4488 PH/14 In the default configuration, change the use of "message" in ACL warn
4489 statements to "add_header".
4491 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4492 not followed by a command (e.g. "seen endif").
4494 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4495 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4498 PH/17 Added extra conditions to the default value of delay_warning_condition
4502 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4503 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4504 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4507 The Auto-Submitted: and various List- headers are standardised, whereas I
4508 don't think Precedence: ever was.
4510 PH/18 Refactored debugging code in route_finduser() to show more information,
4511 in particular, the error code if getpwnam() issues one.
4513 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4514 This is apparently needed in addition to the PH/07 change above to avoid
4515 any possible encoding problems.
4517 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4518 but not after initializing Perl.
4520 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4521 output them only if debugging. By default they are written stderr,
4522 apparently, which is not desirable.
4524 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4527 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4530 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4532 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4533 authenticated or an ident call has been made. Suppress the default
4534 values for $authenticated_id and $authenticated_sender (but permit -oMai
4535 and -oMas) when testing with -bh.
4537 PH/24 Re-jigged the order of the tests in the default configuration so that the
4538 tests for valid domains and recipients precede the DNS black list and CSA
4539 tests, on the grounds that those ones are more expensive.
4541 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4542 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4543 This bug exists in every version of Exim that I still have, right back to
4546 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4547 However, an attempt to turn on case-sensitivity in a regex key by
4548 including (?-i) didn't work because the subject string was already
4549 lowercased, and the effects were non-intuitive. It turns out that a
4550 one-line patch can be used to allow (?-i) to work as expected.
4556 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4557 other effects) broke the use of negated acl sub-conditions.
4559 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4562 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4563 "Deny" causes Exim to reject the incoming connection with a 554 error.
4564 Unfortunately, if there is a major crisis, such as a disk failure,
4565 tcp-wrappers gives "deny", whereas what one would like would be some
4566 kind of temporary error. A kludge has been added to help with this.
4567 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4568 554 error is used if errno is still zero or contains ENOENT (which occurs
4569 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4572 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4574 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4575 errors. Otherwise a message that provokes a temporary error (when other
4576 messages do not) can cause a whole host to time out.
4578 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4579 addresses were routed directly to files or pipes from a redirect router.
4580 File deliveries just didn't batch; pipe deliveries might have suffered
4583 PH/06 A failure to get a lock for a hints database would erroneously always say
4584 "Failed to get write lock", even when it was really a read lock.
4586 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4587 of 0600. This has been changed to use the value of the lockfile_mode
4588 option (which defaults to 0600).
4590 PH/08 Applied small patch from the Sieve maintainer.
4592 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4593 folder from quota calculations, a direct delivery into this folder messed
4594 up the contents of the maildirsize file. This was because the regex was
4595 used only to exclude .Trash (or whatever) when the size of the mailbox
4596 was calculated. There was no check that a delivery was happening into an
4597 excluded directory. This bug has been fixed by ignoring all quota
4598 processing for deliveries into excluded directories.
4600 PH/10 Added the maildirfolder_create_regex option to appendfile.
4606 PH/01 The code for finding all the local interface addresses on a FreeBSD
4607 system running IPv6 was broken. This may well have applied to all BSD
4608 systems, as well as to others that have similar system calls. The broken
4609 code found IPv4 interfaces correctly, but gave incorrect values for the
4610 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4611 that it would not match correctly against @[] and not recognize the IPv6
4614 PH/02 The ipliteral router was not recognizing addresses of the form user@
4615 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4617 PH/03 Added disable_ipv6.
4619 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4620 decoded form, because it is most often used to construct To: headers
4621 lines in autoreplies, and the decoded form may well be syntactically
4622 invalid. However, $reply_address has leading white space removed, and all
4623 newlines turned into spaces so that the autoreply transport does not
4626 PH/05 If group was specified without a user on a router, and no group or user
4627 was specified on a transport, the group from the router was ignored.
4629 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4630 for visible compile-time settings that can be used to change these
4631 numbers, for those that want even more. Backwards compatibility with old
4632 spool files has been maintained. However, going back to a previous Exim
4633 release will lost any variables that are in spool files.
4635 PH/07 Two small changes when running in the test harness: increase delay when
4636 passing a TCP/IP connection to a new process, in case the original
4637 process has to generate a bounce, and remove special handling of
4638 127.0.0.2 (sic), which is no longer necessary.
4640 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4641 be the same on different OS.
4643 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4646 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4647 whether --show-vars was specified or not
4649 JJ/02 exipick: Added support for new ACL variable spool format introduced
4652 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4653 syntactically invalid From: or Reply-to: line, and a filter used this to
4654 generate an autoreply, and therefore failed to obtain an address for the
4655 autoreply, Exim could try to deliver to a non-existent relative file
4656 name, causing unrelated and misleading errors. What now happens is that
4657 it logs this as a hard delivery error, but does not attempt to create a
4660 PH/11 The exinext utility has a -C option for testing purposes, but although
4661 the given file was scanned by exinext itself; it wasn't being passed on
4662 when Exim was called.
4664 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4665 an end-of-file indication when reading a command response.
4667 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4668 compiled. In many other places in Exim, IPv6 addresses are always
4669 recognized, so I have changed this. It also means that IPv4 domain
4670 literals of the form [IPV4:n.n.n.n] are now always recognized.
4672 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4673 used if the router is not running as root, for example, when verifying at
4674 ACL time, or when using -bh. The debugging output from this situation was
4675 non-existent - all you got was a failure to exec. I have made two
4678 (a) Failures to set uid/gid, the current directory, or a process leader
4679 in a subprocess such as that created by queryprogram now generate
4680 suitable debugging output when -d is set.
4682 (b) The queryprogram router detects when it is not running as root,
4683 outputs suitable debugging information if -d is set, and then runs
4684 the subprocess without attempting to change uid/gid.
4686 PH/15 Minor change to Makefile for building test_host (undocumented testing
4689 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4690 additional section of a DNS packet that returns MX or SRV records.
4691 Instead, it always explicitly searches for A/AAAA records. This avoids
4692 major problems that occur when a DNS server includes only records of one
4693 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4694 fixed another bug: if SRV records were looked up and the corresponding
4695 address records were *not* found in the additional section, the port
4696 values from the SRV records were lost.
4698 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4699 using the correct key (the original address) when searching the retry
4700 rules in order to find which one to use for generating the retry hint.
4702 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4703 adding the default one. Similarly, if it contains a Reply-To: header, the
4704 errors_reply_to option, if set, is not used.
4706 PH/19 When calculating a retry time, Exim used to measure the "time since
4707 failure" by looking at the "first failed" field in the retry record. Now
4708 it does not use this if it is later than than the arrival time of the
4709 message. Instead it uses the arrival time. This makes for better
4710 behaviour in cases where some deliveries succeed, thus re-setting the
4711 "first failed" field. An example is a quota failure for a huge message
4712 when small messages continue to be delivered. Without this change, the
4713 "time since failure" will always be short, possible causing more frequent
4714 delivery attempts for the huge message than are intended.
4715 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4717 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4718 $1, $2, $3) because the numerical variables can be reset during some
4719 expansion items (e.g. "match"), thereby losing the authentication data.
4721 PH/21 Make -bV show the size of off_t variables so that the test suite can
4722 decide whether to run tests for quotas > 2G.
4724 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4725 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4726 filecount value is greater than 2G or if a quota value is greater than 2G
4727 on a system where the size of off_t is not greater than 4, a panic error
4730 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4731 never match. The debug and -bh output now contains an explicit error
4732 message indicating a malformed IPv4 address or mask.
4734 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4735 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4736 PH/23 above applies.
4738 PH/25 Do not write to syslog when running in the test harness. The only
4739 occasion when this arises is a failure to open the main or panic logs
4740 (for which there is an explicit test).
4742 PH/26 Added the /no_tell option to "control=freeze".
4744 PH/27 If a host name lookup failed very early in a connection, for example, if
4745 the IP address matched host_lookup and the reverse lookup yielded a name
4746 that did not have a forward lookup, an error message of the form "no IP
4747 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4748 could be logged. Now it outputs the IP address instead of "NULL".
4750 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4751 allows the sender and the authenticated sender to be set when
4752 submitting a message from within Exim. Since child_open_exim() is
4753 documented for local_scan(), the new function should be too.
4755 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4756 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4757 results in an empty string is now treated as unset.
4759 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4761 PH/31 Added sender_verify_fail logging option.
4763 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4764 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4765 tidied the source and removed it altogether.
4767 PH/33 When a queue run was abandoned because the load average was too high, a
4768 log line was always written; now it is written only if the queue_run log
4769 selector is set. In addition, the log line for abandonment now contains
4770 information about the queue run such as the pid. This is always present
4771 in "start" and "stop" lines but was omitted from the "abandon" line.
4773 PH/34 Omit spaces between a header name and the colon in the error message that
4774 is given when verify = headers_syntax fails (if there are lots of them,
4775 the message gets confusing).
4777 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4778 names, as there are now some PTR records that contain slashes. This check
4779 is only to protect against broken name servers that fall over on strange
4780 characters, so the fact that it applies to all lookups doesn't matter.
4782 PH/36 Now that the new test suite is complete, we can remove some of the
4783 special code in Exim that was needed for the old test suite. For example,
4784 sorting DNS records because real resolvers return them in an arbitrary
4785 order. The new test suite's fake resolver always returns records in the
4788 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4789 bounces) except when queue_only is set, to avoid logging races between
4790 the different processes.
4792 PH/38 Panic-die if .include specifies a non-absolute path.
4794 PH/39 A tweak to the "H" retry rule from its user.
4796 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4797 a label. They prevented compilation on older perls.
4799 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4800 a warning to be raised on newish perls.
4802 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4803 on queue. Changes to match documented behaviour of showing count of
4804 messages matching specified criteria.
4806 PH/40 Changed the default ident timeout from 30s to 5s.
4808 PH/41 Added support for the use of login_cap features, on those BSD systems
4809 that have them, for controlling the resources used by pipe deliveries.
4811 PH/42 The content-scanning code uses fopen() to create files in which to put
4812 message data. Previously it was not paying any attention to the mode of
4813 the files. Exim runs with umask(0) because the rest of the code creates
4814 files with open(), and sets the required mode explicitly. Thus, these
4815 files were ending up world-writeable. This was not a big issue, because,
4816 being within the spool directory, they were not world-accessible. I have
4817 created a function called modefopen, which takes an additional mode
4818 argument. It sets umask(777), creates the file, chmods it to the required
4819 mode, then resets the umask. All the relevant calls to fopen() in the
4820 content scanning code have been changed to use this function.
4822 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4823 to 24 hours. This avoids potential overflow problems when processing G
4824 and H retry rules. I suspect nobody ever tinkers with this value.
4826 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4828 PH/45 When the plaintext authenticator is running as a client, the server's
4829 challenges are checked to ensure they are valid base64 strings. By
4830 default, the authentication attempt is cancelled if an invalid string is
4831 received. Setting client_ignore_invalid_base64 true ignores these errors.
4832 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4833 they are received. Thus, the responses can be made to depend on the
4834 challenges. If an invalid string is ignored, an empty string is placed in
4837 PH/46 Messages that are created by the autoreply transport now contains a
4838 References: header, in accordance with RFCs 2822 and 3834.
4840 PH/47 Added authenticated_sender_force to the smtp transport.
4842 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4844 PH/49 Installed latest patch from the Sieve maintainer.
4846 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4847 was also set, the appendfile transport was unnecessarily scanning a
4848 directory of message files (e.g. for maildir delivery) to find the count
4849 of files (along with the size), even though it did not need this
4850 information. It now does the scan only if it needs to find either the
4851 size of the count of files.
4853 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4855 PH/52 Two bugs concerned with error handling when the smtp transport is
4858 (i) Exim was not creating retry information for temporary errors given
4859 for individual recipients after the DATA command when the smtp transport
4860 was used in LMTP mode. This meant that they could be retried too
4861 frequently, and not timed out correctly.
4863 (ii) Exim was setting the flag that allows error details to be returned
4864 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4865 recipients that were returned after the DATA command.
4867 PH/53 This is related to PH/52, but is more general: for any failing address,
4868 when detailed error information was permitted to be returned to the
4869 sender, but the error was temporary, then after the final timeout, only
4870 "retry timeout exceeded" was returned. Now it returns the full error as
4871 well as "retry timeout exceeded".
4873 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4874 do this, and (what is worse) MTAs that accept it.
4876 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4877 will now be deprecated.
4879 PH/56 New os.c-cygwin from the Cygwin maintainer.
4881 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4882 formats (previously only available in exim formats via -bpr, -bpru,
4883 and -bpra. Now also available in native and exiqgrep formats)
4885 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4886 with very large, slow to parse queues
4888 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4890 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4892 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4893 responses to authentication challenges, though it was showing the
4894 challenges; (ii) I've removed the CR characters from the debug output for
4897 PH/58 Allow for the insertion of a newline as well as a space when a string
4898 is turned into more than one encoded-word during RFC 2047 encoding. The
4899 Sieve code now uses this.
4901 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4902 data_4xx, lost_connection, tls_required.
4904 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4905 message was being sent as an SMTP response.
4907 PH/61 Add -l and -k options to exicyclog.
4909 PH/62 When verifying, if an address was redirected to one new address, so that
4910 verification continued, and the new address failed or deferred after
4911 having set something in $address_data, the value of $address_data was not
4912 passed back to the ACL. This was different to the case when no
4913 redirection occurred. The value is now passed back in both cases.
4915 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4916 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4917 use login_cap.h, so on its own it isn't the distinguishing feature. The
4918 new name refers directly to the setclassresources() function.
4920 PH/65 Added configuration files for NetBSD3.
4922 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4924 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4925 is preferred over IPv4.
4927 PH/68 The bounce_return_message and bounce_return_body options were not being
4928 honoured for bounces generated during the reception of non-SMTP messages.
4929 In particular, this applied to messages rejected by the ACL. This bug has
4930 been fixed. However, if bounce_return_message is true and bounce_return_
4931 body is false, the headers that are returned for a non-SMTP message
4932 include only those that have been read before the error was detected.
4933 (In the case of an ACL rejection, they have all been read.)
4935 PH/69 The HTML version of the specification is now built in a directory called
4936 spec_html instead of spec.html, because the latter looks like a path with
4937 a MIME-type, and this confuses some software.
4939 PH/70 Catch two compiler warnings in sieve.c.
4941 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4942 function verify_get_ident() calls ip_connect() to connect a socket, but
4943 if the "connect()" function timed out, ip_connect() used to close the
4944 socket. However, verify_get_ident() also closes the socket later, and in
4945 between Exim writes to the log, which may get opened at this point. When
4946 the socket was closed in ip_connect(), the log could get the same file
4947 descriptor number as the socket. This naturally causes chaos. The fix is
4948 not to close the socket in ip_connect(); the socket should be closed by
4949 the function that creates it. There was only one place in the code where
4950 this was missing, in the iplookup router, which I don't think anybody now
4951 uses, but I've fixed it anyway.
4953 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4954 well as to direct DNS lookups. Otherwise the handling of names in host
4955 lists is inconsistent and therefore confusing.
4961 PH/01 Two changes to the default runtime configuration:
4963 (1) Move the checks for relay_from_hosts and authenticated clients from
4964 after to before the (commented out) DNS black list checks.
4966 (2) Add control=submission to the relay_from_hosts and authenticated
4967 clients checks, on the grounds that messages accepted by these
4968 statements are most likely to be submissions.
4970 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4972 (1) Generate an error if the third argument for the ${prvs expansion is
4975 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4978 (3) Reset the variables that are obtained from the first argument of
4979 ${prvscheck and used in the second argument before leaving the code,
4980 because their memory is reclaimed, so using them afterwards may do
4983 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4984 one (it's much easier than Tom thought :-).
4986 (5) Because of (4), we can now allow for the use of $prvscheck_result
4987 inside the third argument.
4989 PH/03 For some reason, the default setting of PATH when running a command from
4990 a pipe transport was just "/usr/bin". I have changed it to
4993 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4994 anything to be listed in the output from -bV.
4996 PH/05 When a filter generated an autoreply, the entire To: header line was
4997 quoted in the delivery log line, like this:
4999 => >A.N.Other <ano@some.domain> <original@ddress> ...
5001 This has been changed so that it extracts the operative address. There
5002 may be more than one such address. If so, they are comma-separated, like
5005 => >ano@some.domain,ona@other.domain <original@ddress> ...
5007 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5008 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5009 not being looked up in the rDNS to get a host name, Exim was showing the
5010 IP address twice in Received: lines, even though the IP addresses were
5011 identical. For example:
5013 Received: from [1.2.3.4] (helo=[1.2.3.4])
5015 However, if the real host name was known, it was omitting the HELO data
5016 if it matched the actual IP address. This has been tidied up so that it
5017 doesn't show the same IP address twice.
5019 PH/07 When both +timestamp and +memory debugging was on, the value given by
5020 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5021 called by the debug printing, thereby overwriting the timestamp buffer.
5022 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5024 PH/08 When the original message was included in an autoreply transport, it
5025 always said "this is a copy of the message, including all the headers",
5026 even if body_only or headers_only was set. It now gives an appropriate
5029 PH/09 Applied a patch from the Sieve maintainer which:
5031 o fixes some comments
5032 o adds the (disabled) notify extension core
5033 o adds some debug output for the result of if/elsif tests
5034 o points to the current vacation draft in the documentation
5035 and documents the missing references header update
5039 o fixes a bug in processing the envelope test (when testing
5040 multiple envelope elements, the last element determined the
5043 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5044 Electronic Mail") by including:
5046 Auto-submitted: auto-generated
5048 in the messages that it generates (bounce messages and others, such as
5049 warnings). In the case of bounce messages for non-SMTP messages, there was
5050 also a typo: it was using "Auto_submitted" (underscore instead of
5051 hyphen). Since every message generated by Exim is necessarily in response
5052 to another message, thes have all been changed to:
5054 Auto-Submitted: auto-replied
5056 in accordance with these statements in the RFC:
5058 The auto-replied keyword:
5060 - SHOULD be used on messages sent in direct response to another
5061 message by an automatic process,
5063 - MUST NOT be used on manually-generated messages,
5065 - MAY be used on Delivery Status Notifications (DSNs) and Message
5066 Disposition Notifications (MDNs),
5068 - MUST NOT be used on messages generated by automatic or periodic
5069 processes, except for messages which are automatic responses to
5072 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5073 to the default Received: header definition.
5075 PH/12 Added log selector acl_warn_skipped (default on).
5077 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5078 variables because (a) they are in the wrong storage pool and (b) even if
5079 they were copied, it wouldn't work properly because of the caching.
5081 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5082 checking when decoding. Apparently there are clients that generate
5083 overlong encoded strings. Why am I not surprised?
5085 PH/15 If the first argument of "${if match_address" was not empty, but did not
5086 contain an "@" character, Exim crashed. Now it writes a panic log message
5087 and treats the condition as false.
5089 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5091 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5092 extension "envelope-auth". The code is finished and in agreement with
5093 other implementations, but there is no documentation so far and in fact,
5094 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5095 not changing the active code.
5097 Print executed "if" and "elsif" statements when debugging is used. This
5098 helps a great deal to understand what a filter does.
5100 Document more things not specified clearly in RFC3028. I had all this
5101 sorted out, when out of a sudden new issues came to my mind. Oops."
5103 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5106 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5107 canonical form (as documented). However, after a host name lookup from
5108 the IP address, check_host() was doing a simple string comparison with
5109 addresses acquired from the DNS when checking that the found name did
5110 have the original IP as one of its addresses. Since any found IPv6
5111 addresses are likely to be in abbreviated form, the comparison could
5112 fail. Luckily, there already exists a function for doing the comparison
5113 by converting both addresses to binary, so now that is used instead of
5114 the text comparison.
5116 PH/20 There was another similar case to PH/19, when a complete host name was
5117 given in a host list; looking up its IP address could give an abbreviated
5118 form, whereas the current host's name might or might not be abbreviated.
5119 The same fix has been applied.
5125 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5126 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5129 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5130 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5132 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5134 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5135 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5136 Also, for "Auto-Submitted", treat anything other than "no" as
5137 non-personal, in accordance with RFC 3834. (Previously it treated
5138 anything starting "auto-" as non-personal.)
5140 TF/01 The control=submission/name=... option had a problem with syntax
5141 errors if the name included a slash character. The /name= option
5142 now slurps the rest of the string, so it can include any characters
5143 but it must come last in the list of options (after /sender_retain
5146 PH/05 Some modifications to the interface to the fake nameserver for the new
5154 TK/01 Added the "success_on_redirect" address verification option. See
5155 NewStuff for rationale and an example.
5157 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5159 PH/02 Patch to exigrep to allow it to work on syslog lines.
5161 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5162 fread() to skip over the body file's header line, because in Cygwin the
5163 header line is locked and is inaccessible.
5165 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5166 co-exist for some time) to make it clear that it is the Exim ID that is
5167 referenced, not the Message-ID: header line.
5169 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5170 string_format() function, because snprintf() does not exist on all
5173 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5174 ${stat: expansion item.
5176 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5177 protocol synchronization error", to keep the pedants happy.
5179 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5180 well as for IRIX systems, when gcc is being used. See the host.c source
5183 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5185 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5188 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5189 transport and to the smtp transport in LMTP mode.
5191 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5193 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5194 run a filter in a subprocess. This could lead to confusion in subsequent
5195 lookups in the parent process. There should also be a search_tidyup() at
5196 the end of the subprocess.
5198 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5199 only if the host matched helo_try_verify_hosts, which caused the
5200 verification to occur when the EHLO/HELO command was issued. The ACL just
5201 tested the remembered result. Now, if a previous verification attempt has
5202 not happened, "verify = helo" does it there and then.
5204 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5206 TK/03 Fix log output including CR from clamd.
5208 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5209 exist provoked a memory error which could cause a segfault.
5211 PH/15 Installed PCRE 6.2
5213 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5215 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5216 of the problem. Specifically, suggested +O2 rather than +O1 for the
5219 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5221 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5222 record, the port was not correctly logged when the outgoing_port log
5223 selector was set (it logged the transort's default port).
5225 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5226 fallback_hosts, and "hosts" in the smtp transport.
5228 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5229 host errors such as "Connection refused".
5231 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5232 authentication with radiusclient 0.4.9:
5234 - Error returned from rc_read_config was caught wrongly
5235 - Username/password not passed on to radius server due to wrong length.
5237 The presumption is that some radiusclient API changes for 4.51/PH/17
5238 were not taken care of correctly. The code is still untested by me (my
5239 Linux distribution still has 0.3.2 of radiusclient), but it was
5240 contributed by a Radius user.
5242 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5243 expanding the "port" option of the smtp transport.
5245 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5246 while reading a MIME header. Thanks to Tom Hughes for a patch.
5248 PH/24 Include config.h inside local_scan.h so that configuration settings are
5251 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5252 that in an ACL for RCPT (for example), you can examine exactly what was
5255 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5256 commands, but it was not correctly comparing the address with the actual
5257 client host address. Thus, it would show the EHLO address in Received:
5258 header lines when this was not necessary.
5260 PH/27 Added the % operator to ${eval:}.
5262 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5263 it should be ignoring failures (because with -C, for example, it has lost
5264 privilege). It wasn't ignoring creation failures other than "already
5267 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5270 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5271 because an input line was too long, either on its own, or by virtue of
5272 too many continuations, the temporary file was not being removed, and the
5273 return code was incorrect.
5275 PH/31 Missing "BOOL" in function definition in filtertest.c.
5277 PH/32 Applied Sieve patches from the maintainer.
5279 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5281 PH/33 Added "verify = not_blind".
5283 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5284 Local/Makefile (with some defaults set). These are used in built scripts
5285 such as exicyclog, but they have never been used in the exim_install
5286 script (though there are many overriding facilities there). I have
5287 arranged that the exim_install script now takes note of these two
5290 PH/35 Installed configuration files for Dragonfly.
5292 PH/36 When a locally submitted message by a trusted user did not contain a
5293 From: header, and the sender address was obtained from -f or from an SMTP
5294 MAIL command, and the trusted user did not use -F to supply a sender
5295 name, $originator_name was incorrectly used when constructing a From:
5296 header. Furthermore, $originator_name was used for submission mode
5297 messages from external hosts without From: headers in a similar way,
5298 which is clearly wrong.
5300 PH/37 Added control=suppress_local_fixups.
5302 PH/38 When log_selector = +received_sender was set, and the addition of the
5303 sender made the log line's construction buffer exactly full, or one byte
5304 less than full, an overflow happened when the terminating "\n" was
5307 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5308 when the result of a list match is failure because a DNS lookup failed.
5310 PH/40 RM_COMMAND is now used in the building process.
5312 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5313 the "build-* directories that it finds.
5315 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5316 address in a domain literal was a prefix of an interface address.
5318 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5319 when verifying a sender address, unless rewrite_headers is false.
5321 PH/44 Wrote a long comment about why errors_to addresses are verified as
5322 recipients, not senders.
5324 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5325 the ratelimit ACL was added.
5327 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5329 PH/46 Added extra information about PostgreSQL errors to the error string.
5331 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5332 avoiding the need to install special zones in a real server. This is
5333 backwards compatible; if it can't find the fake resolver, it drops back.
5334 Thus, both old and new test suites can be run.
5336 TF/02 Added util/ratelimit.pl
5338 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5339 clock is set back in time.
5341 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5342 Candler <B.Candler@pobox.com>.
5344 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5345 OK for addresses that are the result of redirections.
5347 PH/48 A number of further additions for the benefit of the new test suite,
5348 including a fake gethostbyname() that interfaces to the fake DNS resolver
5351 TF/06 The fix for widen_domains has also been applied to qualify_single and
5352 search_parents which are the other dnslookup options that can cause
5355 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5358 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5360 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5361 service" if it encountered an MX record with an empty target hostname.
5362 The message is now "an MX or SRV record indicated no SMTP service".
5364 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5365 if the DNS of the sending site is misconfigured. This is quite a
5366 common situation. This change restores the behaviour of treating a
5367 helo verification defer as a failure.
5369 PH/51 If self=fail was set on a router, the bounce message did not include the
5370 actual error message.
5376 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5378 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5379 command itself ended in error, the underlying message about the transport
5380 filter timeout was being overwritten with the pipe command error. Now the
5381 underlying error message should be appended to the second error message.
5383 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5385 PH/02 Reduce the amount of output that "make" produces by default. Full output
5386 can still be requested.
5388 PH/03 The warning log line about a condition test deferring for a "warn" verb
5389 was being output only once per connection, rather than after each
5390 occurrence (because it was using the same function as for successful
5391 "warn" verbs). This seems wrong, so I have changed it.
5393 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5394 it should not have, which might have caused a crash in the right
5395 circumstances, but probably never did.
5397 PH/04 Installed a modified version of Tony Finch's patch to make submission
5398 mode fix the return path as well as the Sender: header line, and to
5399 add a /name= option so that you can make the user's friendly name appear
5402 TF/03 Added the control = fakedefer ACL modifier.
5404 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5405 Mark Lowes for thorough testing.
5407 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5409 TK/03 Merged latest SRS patch from Miles Wilton.
5411 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5412 with the definition in sysexits.h (which is #included earlier).
5413 Fortunately, Exim does not actually use EX_OK. The code used to try to
5414 preserve the sysexits.h value, by assuming that macro definitions were
5415 scanned for macro replacements. I have been disabused of this notion,
5416 so now the code just undefines EX_OK before #including unistd.h.
5418 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5419 in the smtp transport. When a block could not be written in a single
5420 write() function, the timeout was being re-applied to each part-write.
5421 This seems wrong - if the receiver was accepting one byte at a time it
5422 would take for ever. The timeout is now adjusted when this happens. It
5423 doesn't have to be particularly precise.
5425 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5426 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5428 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5429 without a domain if the check to <postmaster@domain> fails.
5431 SC/01 Eximstats: added -xls and the ability to specify output files
5432 (patch written by Frank Heydlauf).
5434 SC/02 Eximstats: use FileHandles for outputting results.
5436 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5438 SC/04 Eximstats: fixed display of large numbers with -nvr option
5440 SC/05 Eximstats: fixed merging of reports with empty tables.
5442 SC/06 Eximstats: added the -include_original_destination flag
5444 SC/07 Eximstats: removed tabs and trailing whitespace.
5446 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5448 TK/06 MBOX spool code: Add real "From " MBOX separator line
5449 so the .eml file is really in mbox format (even though
5450 most programs do not really care). Patch from Alex Miller.
5452 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5453 The latter is generated from $received_to and is only set if the
5454 message has one envelope recipient. SA can use these headers,
5455 obviously out-of-the-box. Patch from Alex Miller.
5457 PH/08 The ${def test on a variable was returning false if the variable's
5458 value was "0", contrary to what the specification has always said!
5459 The result should be true unless the variable is empty.
5461 PH/09 The syntax error of a character other than { following "${if
5462 def:variable_name" (after optional whitespace) was not being diagnosed.
5463 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5464 accidental colon was present, for example, could give incorrect results.
5466 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5467 result is used (not including appendfile, where other changes are about
5470 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5471 This involved changing a lot of size variables from int to off_t. It
5472 should work with maildirs and everything.
5474 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5475 spamd dying while we are connected to it.
5477 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5480 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5481 function for BDB 4.3.
5483 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5485 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5486 the sources for PCRE have been re-arranged and more files are now
5489 PH/15 The code I had for printing potentially long long variables in PH/11
5490 above was not the best (it lost precision). The length of off_t variables
5491 is now inspected at build time, and an appropriate printing format (%ld
5492 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5493 to be "long long int" or "long int". This is needed for the internal
5494 formatting function string_vformat().
5496 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5497 the configuration file to be ":syslog", then the script "guesses" where
5498 the logs files are, rather than using the compiled in default. In our
5499 case the guess is not the same as the compiled default, so the script
5500 suddenly stopped working when I started to use syslog. The patch checks
5501 to see if log_file_path is "". If so, it attempts to read it from exim
5502 with no configuration file to get the compiled in version, before it
5503 falls back to the previous guessing code."
5505 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5506 implementing BATV in an Exim configuration. See NewStuff for the gory
5509 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5510 Makefile that are specific to HP-UX.
5512 PH/18 If the "use_postmaster" option was set for a recipient callout together
5513 with the "random" option, the postmaster address was used as the MAIL
5514 FROM address for the random test, but not for the subsequent recipient
5515 test. It is now used for both.
5517 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5518 patch removes a few documentation additions to RFC 3028, because the
5519 latest draft now contains them. It adds the new en;ascii-case comparator
5520 and a new error check for 8bit text in MIME parts. Comparator and
5521 require names are now matched exactly. I enabled the subaddress
5522 extension, but it is not well tested yet (read: it works for me)."
5524 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5525 rework some of the code of TK/09 above to avoid the hardwired use of
5526 "%lld" and "long long". Replaced the call to snprintf() with a call to
5529 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5530 records point to non-existent hosts", "retry timeout exceeded", and
5531 "retry time not reached for any host after a long failure period".
5533 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5534 experimental DomainKeys support:
5536 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5537 (2) On an error such as an illegally used "control", the wrong name for
5538 the control was given.
5540 These problems did NOT occur unless DomainKeys support was compiled.
5542 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5544 PH/24 Added ${if match_ip condition.
5546 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5547 databases so that it will be absolutely obvious if a crash occurs in the
5548 DB library. This is a regular occurrence (often caused by mis-matched
5551 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5552 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5553 picked up on a user's system that detects such things. There doesn't seem
5554 to be a gcc warning option for this - only an attribute that has to be
5555 put on the function's prototype. It seems that in Fedora Core 4 they have
5556 set this on a number of new functions. No doubt there will be more in due
5559 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5560 specify a transport. However, if an address that was verified by such a
5561 router was the subject of a callout, Exim crashed because it tried to
5562 read the rcpt_include_affixes from the non-existent transport. Now it
5563 just assumes that the setting of that option is false. This bug was
5564 introduced by 4.51/PH/31.
5566 PH/28 Changed -d+all to exclude +memory, because that information is very
5567 rarely of interest, but it makes the output a lot bigger. People tend to
5568 do -d+all out of habit.
5570 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5571 code in os-type was giving problems when libc.so lives in lib64, like on
5574 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5575 aren't the modern standard, and it seems that some systems' include files
5576 don't always have them. Exim was already checking for some of the newer
5577 ones like T_AAAA, and defining it itself. I've added checks for all the
5578 record types that Exim uses.
5580 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5581 not automatically generating a new one, as it is supposed to. This
5582 prevented TLS from working. If the file did exist, but contained invalid
5583 data, a new version was generated, as expected. It was only the case of a
5584 non-existent file that was broken.
5586 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5587 with a change in libdomainkeys > 0.64.
5589 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5590 from DNS. If the selector record carries the flag, it now has
5591 precedence over the domain-wide flag.
5593 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5595 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5596 the use of a port name that isn't defined in /etc/services) occurred, the
5597 message was deferred as in a normal delivery, and thus remained on the
5598 spool, instead of being failed because of the mua_wrapper setting. This
5599 is now fixed, and I tidied up some of the mua_wrapper messages at the
5602 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5603 the messages in a hash of arrays rather than using individual hashes.
5604 This is a bit cleaner and results in dramatic memory savings, albeit
5605 at a slight CPU cost.
5607 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5608 as requested by Marc Sherman.
5610 SC/10 Eximstats: added histograms for user specified patterns as requested
5613 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5615 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5616 fopen() in the content-scanning modules that did not already have it.
5622 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5623 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5625 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5627 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5629 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5630 to test Sieve filters that use "vacation".
5632 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5633 that changes the way the GnuTLS parameters are stored in the cache file.
5634 The new format can be generated externally. For backward compatibility,
5635 if the data in the cache doesn't make sense, Exim assumes it has read an
5636 old-format file, and it generates new data and writes a new file. This
5637 means that you can't go back to an older release without removing the
5640 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5641 work if there are any delivery delays because "one_time" forces the
5642 parent to be marked "delivered", so its unseen clone is never tried
5643 again. For this reason, Exim now forbids the simultaneous setting of
5646 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5647 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5648 entry if you want to know the details. The fix, however, neglected to
5649 consider the case when local delivery batching is involved. The test for
5650 "previously delivered" was not happening when checking to see if an
5651 address could be batched with a previous (undelivered) one; under
5652 certain circumstances this could lead to multiple deliveries to the same
5655 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5656 in its include files, and this causes problems building Exim.
5658 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5659 header_syntax) but Exim was just ignoring anything given after a slash.
5660 In particular, this caused confusion with an attempt to use "verify =
5661 reverse_host_lookup/defer_ok". An error is now given when options are
5662 supplied for verify items that do not have them. (Maybe reverse_host_
5663 lookup should have a defer_ok option, but that's a different point.)
5665 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5666 defined by RFC 821) to 2048, because there were problems with some AUTH
5667 commands, and RFC 1869 says the size should be increased for extended
5668 SMTP commands that take arguments.
5670 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5673 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5674 "unknown" error; now it says that the functionality isn't in the binary.
5676 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5677 an address' error message when a string expansion fails (syntax or
5678 whatever). Otherwise the password may appear in the log. Following change
5679 PH/42 below, there is no longer a chance of it appearing in a bounce
5682 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5684 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5685 of its addresses was ever tried. (Bugzilla bug #2.)
5687 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5688 the result incorrectly in the debug output. (It correctly added a newline
5689 to what was transported.)
5691 TF/01 Added $received_time.
5693 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5694 commented out examples of how to interface to a virus scanner and to
5695 SpamAssassin. Also added commented examples of av_scanner and
5696 spamd_address settings.
5698 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5699 and controls are allowed in which ACLs. There were a couple of minor
5700 errors. Some of the entries in the conditions table (which is a table of
5701 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5702 negation of where the condition IS allowed.
5704 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5706 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5707 header file does not have a version number, so I've had to invent a new
5708 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5709 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5710 radiusclient), but it was contributed by a Radius user.
5712 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5713 files or directories, for OpenSSL.
5715 PH/19 When an Exim process that is running as root has to create an Exim log
5716 file, it does so in a subprocess that runs as exim:exim so as to get the
5717 ownership right at creation (otherwise, other Exim processes might see
5718 the file with the wrong ownership). There was no test for failure of this
5719 fork() call, which would lead to the process getting stuck as it waited
5720 for a non-existent subprocess. Forks do occasionally fail when resources
5721 run out. I reviewed all the other calls to fork(); they all seem to check
5724 PH/20 When checking for unexpected SMTP input at connect time (before writing
5725 the banner), Exim was not dealing correctly with a non-positive return
5726 from the read() function. If the client had disconnected by this time,
5727 the result was a log entry for a synchronization error with an empty
5728 string after "input=" when read() returned zero. If read() returned -1
5729 (an event I could not check), uninitialized data bytes were printed.
5730 There were reports of junk text (parts of files, etc) appearing after
5733 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5735 PH/22 Added support for macro redefinition, and (re)definition in between
5736 driver and ACL definitions.
5738 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5739 forgetting to use the resulting value; it was using the unexpanded value.
5741 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5742 hadn't been configured. The fix is from Juergen Kreileder, who
5743 understands it better than I do:
5745 "Here's what I see happening with three configured cyrus_sasl
5746 authenticators configured (plain, login, cram-md5):
5748 On startup auth_cyrus_sasl_init() gets called for each of these.
5749 This means three calls to sasl_listmech() without a specified mech_list.
5750 => SASL tests which mechs of all available mechs actually work
5751 => three warnings about OTP not working
5752 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5754 With the patch, sasl_listmech() also gets called three times. But now
5755 SASL's mech_list option is set to the server_mech specified in the the
5756 authenticator. Or in other words, the answer from sasl_listmech()
5757 gets limited to just the mech you're testing for (which is different
5759 => the return list contains just 'plain' or 'login', 'cram-md5' or
5760 nothing depending on the value of ob->server_mech.
5762 I've just tested the patch: Authentication still works fine,
5763 unavailable mechs specified in the exim configuration are still
5764 caught, and the auth.log warnings about OTP are gone."
5766 PH/25 When debugging is enabled, the contents of the command line are added
5767 to the debugging output, even when log_selector=+arguments is not
5770 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5771 answer is "GNU", and only if the return is "GNU/something" is the answer
5774 PH/27 $acl_verify_message is now set immediately after the failure of a
5775 verification in an ACL, and so is available in subsequent modifiers. In
5776 particular, the message can be preserved by coding like this:
5778 warn !verify = sender
5779 set acl_m0 = $acl_verify_message
5781 Previously, $acl_verify_message was set only while expanding "message"
5782 and "log_message" when a very denied access.
5784 PH/28 Modified OS/os.c-Linux with
5786 -#ifndef OS_LOAD_AVERAGE
5787 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5789 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5790 nomenclature these days.)
5792 PH/29 Installed patch from the Sieve maintainer that adds the options
5793 sieve_useraddress and sieve_subaddress to the redirect router.
5795 PH/30 In these circumstances:
5796 . Two addresses routed to the same list of hosts;
5797 . First host does not offer TLS;
5798 . First host accepts first address;
5799 . First host gives temporary error to second address;
5800 . Second host offers TLS and a TLS session is established;
5801 . Second host accepts second address.
5802 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5803 and peerdn, if requested) that were in fact used only for the second
5806 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5807 attention to any local part prefix or suffix that was matched by the
5808 router that accepted the address. It now behaves in the same way as it
5809 does for delivery: the affixes are removed from the local part unless
5810 rcpt_include_affixes is set on the transport.
5812 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5813 timeout during the DATA phase of an incoming message.
5815 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5816 applied a patch sanctioned by the Sieve maintainer.
5818 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5819 the uid or gid is negative. A case of a negative gid caused this to be
5820 noticed. The fix allows for either to be negative.
5822 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5823 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5826 PH/36 The change PH/12 above was broken. Fixed it.
5828 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5829 the grounds that routing the same address twice would always produce the
5830 same answer. This might have been true once, but it is certainly no
5831 longer true now. Routing a child address may depend on the previous
5832 routing that produced that child. Some complicated redirection strategies
5833 went wrong when messages had multiple recipients, and made Exim's
5834 behaviour dependent on the order in which the addresses were given.
5836 I have moved the duplicate checking until after the routing is complete.
5837 Exim scans the addresses that are assigned to local and remote
5838 transports, and removes any duplicates. This means that more work will be
5839 done, as duplicates will always all be routed, but duplicates are
5840 presumably rare, so I don't expect this is of any significance.
5842 For deliveries to pipes, files, and autoreplies, the duplicate checking
5843 still happens during the routing process, since they are not going to be
5846 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5847 It corrects a timeout issue with spamd. This is Ian's comment: "The
5848 background is that sometimes spamd either never reads data from a
5849 connection it has accepted, or it never writes response data. The exiscan
5850 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5851 blindly assumes that writes won't block so it may never time out."
5853 PH/39 Allow G after quota size as well as K and M.
5855 PH/40 The value set for $authenticated_id in an authenticator may not contain
5856 binary zeroes or newlines because the value is written to log lines and
5857 to spool files. There was no check on this. Now the value is run through
5858 the string_printing() function so that such characters are converted to
5859 printable escape sequences.
5861 PH/41 $message_linecount is a new variable that contains the total number of
5862 lines in the message. Compare $body_linecount, which is the count for the
5865 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5866 bounce and delay warning messages, except in certain special cases, which
5869 (a) An SMTP error message from a remote host;
5870 (b) A message specified in a :fail: redirection;
5871 (c) A message specified in a "fail" command in a system filter;
5872 (d) A message specified in a FAIL return from the queryprogram router;
5873 (e) A message specified by the cannot_route_message router option.
5875 In these cases only, Exim does include the error details in bounce and
5876 warning messages. There are also a few cases where bland messages such
5877 as "unrouteable address" or "local delivery error" are given.
5879 PH/43 $value is now also set for the "else" part of a ${run expansion.
5881 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5882 being worked on, but at least Exim now implements the latest version to
5885 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5886 process to complete was treated as a delivery failure, a timeout while
5887 writing the message to the pipe was logged, but erroneously treated as a
5888 successful delivery. Such timeouts include transport filter timeouts. For
5889 consistency with the overall process timeout, these timeouts are now
5890 treated as errors, giving rise to delivery failures by default. However,
5891 there is now a new Boolean option for the pipe transport called
5892 timeout_defer, which, if set TRUE, converts the failures into defers for
5893 both kinds of timeout. A transport filter timeout is now identified in
5896 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5897 systems where "make" and "gmake" are different, calling "gmake" at top
5898 level broke things. I've arranged for the value of $(MAKE) to be passed
5899 from the Makefile to this script so that it can call the same version of
5903 A note about Exim versions 4.44 and 4.50
5904 ----------------------------------------
5906 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5907 changes of various kinds. As a consequence, a big documentation update was
5908 needed. This delayed the release for rather longer than seemed good, especially
5909 in the light of a couple of (minor) security issues. Therefore, the changes
5910 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5911 release. So 4.44 and 4.50 are in effect two different branches that both start
5914 I have left the 4.50 change log unchanged; it contains all the changes since
5915 4.43. The change log for 4.44 is below; many of its items are identical to
5916 those for 4.50. This seems to be the most sensible way to preserve the
5917 historical information.
5923 1. Minor wording change to the doc/README.SIEVE file.
5925 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5926 computation of the current number of files was incorrect.
5928 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5929 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5932 4. Give more explanation in the error message when the command for a transport
5933 filter fails to execute.
5935 5. There are several places where Exim runs a non-Exim command in a
5936 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5937 was being done only for the command run by the queryprogram router. It is
5938 now done for all such subprocesses. The other cases are: ${run, transport
5939 filters, and the commands run by the lmtp and pipe transports.
5941 6. Added CONFIGURE_GROUP build-time option.
5943 7. Some older OS have a limit of 256 on the maximum number of file
5944 descriptors. Exim was using setrlimit() to set 1000 as a large value
5945 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5946 systems. I've change it so that if it can't get 1000, it tries for 256.
5948 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5949 was an oversight, and furthermore, ever since the addition of extra
5950 controls (e.g. 4.43/32), the checks on when to allow different forms of
5951 "control" were broken. There should now be diagnostics for all cases when a
5952 control that does not make sense is encountered.
5954 9. Added the /retain_sender option to "control=submission".
5956 10. $recipients is now available in the predata ACL (oversight).
5958 11. Tidy the search cache before the fork to do a delivery from a message
5959 received from the command line. Otherwise the child will trigger a lookup
5960 failure and thereby defer the delivery if it tries to use (for example) a
5961 cached ldap connection that the parent has called unbind on.
5963 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5964 of $address_data from the recipient verification was clobbered by the
5965 sender verification.
5967 13. The value of address_data from a sender verification is now available in
5968 $sender_address_data in subsequent conditions in the ACL statement.
5970 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5972 15. Added a new option "connect=<time>" to callout options, to set a different
5975 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5976 was its contents. (It was OK if the option was not defined at all.)
5978 17. A "Completed" log line is now written for messages that are removed from
5979 the spool by the -Mrm option.
5981 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5982 information about exactly what failed.
5984 19. Added -dd to debug only the daemon process.
5986 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5987 handles timeouts, both on the server side and network timeouts. Renamed the
5988 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5990 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5991 It is now set to "smtps".
5993 22. $host_address is now set to the target address during the checking of
5994 ignore_target_hosts.
5996 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5997 being passed; this would have caused $sender_host_name to have been used if
5998 matching the list had actually called for a host name (not very likely,
5999 since this list is usually IP addresses). A host name is now passed as
6002 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6003 code that specifies a non-restarting handler (typically sigaction() in
6004 modern systems) in an attempt to fix a rare and obscure crash bug.
6006 25. Narrowed the window for a race in the daemon that could cause it to ignore
6007 SIGCHLD signals. This is not a major problem, because they are used only to
6008 wake it up if nothing else does.
6010 26. A malformed maildirsize file could cause Exim to calculate negative values
6011 for the mailbox size or file count. Odd effects could occur as a result.
6012 The maildirsize information is now recalculated if the size or filecount
6015 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6016 support for a long time. Removed HAVE_SYS_VFS_H.
6018 28. Installed the latest version of exipick from John Jetmore.
6020 29. In an address list, if the pattern was not a regular expression, an empty
6021 subject address (from a bounce message) matched only if the pattern was an
6022 empty string. Non-empty patterns were not even tested. This was the wrong
6023 because it is perfectly reasonable to use an empty address as part of a
6024 database query. An empty address is now tested by patterns that are
6025 lookups. However, all the other forms of pattern expect the subject to
6026 contain a local part and a domain, and therefore, for them, an empty
6027 address still always fails if the pattern is not itself empty.
6029 30. Exim went into a mad DNS loop when attempting to do a callout where the
6030 host was specified on an smtp transport, and looking it up yielded more
6031 than one IP address.
6033 31. Re-factored the code for checking spool and log partition space into a
6034 function that finds that data and another that does the check. The former
6035 is then used to implement four new variables: $spool_space, $log_space,
6036 $spool_inodes, and $log_inodes.
6038 32. The RFC2047 encoding function was originally intended for short strings
6039 such as real names; it was not keeping to the 75-character limit for
6040 encoded words that the RFC imposes. It now respects the limit, and
6041 generates multiple encoded words if necessary. To be on the safe side, I
6042 have increased the buffer size for the ${rfc2047: expansion operator from
6045 33. It is now permitted to omit both strings after an "if" condition; if the
6046 condition is true, the result is "true". As before, when the second string
6047 is omitted, a false condition yields an empty string. This makes it less
6048 cumbersome to write custom ACL and router conditions.
6050 34. Failure to deliver a bounce message always caused it to be frozen, even if
6051 there was an errors_to setting on the router. The errors_to setting is now
6054 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6055 canonical form (fully expanded) before being placed in
6056 $sender_host_address.
6058 36. The table in the code that translates DNS record types into text (T_A to
6059 "A" for instance) was missing entries for NS and CNAME. It is just possible
6060 that this could have caused confusion if both these types were looked up
6061 for the same domain, because the text type is used as part of Exim's
6062 per-process caching. But the chance of anyone hitting this buglet seems
6065 37. The dnsdb lookup has been extended in a number of ways.
6067 (1) There is a new type, "zns", which walks up the domain tree until it
6068 finds some nameserver records. It should be used with care.
6070 (2) There is a new type, "mxh", which is like "mx" except that it returns
6071 just the host names, not the priorities.
6073 (3) It is now possible to give a list of domains (or IP addresses) to be
6074 looked up. The behaviour when one of the lookups defers can be
6075 controlled by a keyword.
6077 (4) It is now possible to specify the separator character for use when
6078 multiple records are returned.
6080 38. The dnslists ACL condition has been extended: it is now possible to supply
6081 a list of IP addresses and/or domains to be looked up in a particular DNS
6084 39. Added log_selector=+queue_time_overall.
6086 40. When running the queue in the test harness, wait just a tad after forking a
6087 delivery process, to get repeatability of debugging output.
6089 41. Include certificate and key file names in error message when GnuTLS fails
6090 to set them up, because the GnuTLS error message doesn't include the name
6091 of the failing file when there is a problem reading it.
6093 42. Allow both -bf and -bF in the same test run.
6095 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6097 44. The "Exiscan patch" is now merged into the mainline Exim source.
6099 45. Sometimes the final signoff response after QUIT could fail to get
6100 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6101 < 0 before doing a fflush(). This bug looks as though it goes back to the
6102 introduction of TLS in release 3.20, but "sometimes" must have been rare
6103 because the tests only now provoked it.
6105 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6106 (this can affect the format of dates).
6108 47. exim_tidydb, when checking for the continued existence of a message for
6109 which it has found a message-specific retry record, was not finding
6110 messages that were in split spool directories. Consequently, it was
6111 deleting retry records that should have stayed in existence.
6113 48. Steve fixed some bugs in eximstats.
6115 49. The SPA authentication driver was not abandoning authentication and moving
6116 on to the next authenticator when an expansion was forced to fail,
6117 contradicting the general specification for all authenticators. Instead it
6118 was generating a temporary error. It now behaves as specified.
6120 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6121 (the order specifies the preference for clients). The order is now AES256,
6122 AES128, 3DES, ARCFOUR128.
6124 51. Small patch to Sieve code - explicitly set From: when generating an
6127 52. Exim crashed if a remote delivery caused a very long error message to be
6128 recorded - for instance if somebody sent an entire SpamAssassin report back
6129 as a large number of 550 error lines. This bug was coincidentally fixed by
6130 increasing the size of one of Exim's internal buffers (big_buffer) that
6131 happened as part of the Exiscan merge. However, to be on the safe side, I
6132 have made the code more robust (and fixed the comments that describe what
6135 53. Now that there can be additional text after "Completed" in log lines (if
6136 the queue_time_overall log selector is set), a one-byte patch to exigrep
6137 was needed to allow it to recognize "Completed" as not the last thing in
6140 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6141 patch that reportedly fixes this has been added. I am not expert enough to
6142 create a test for it. This is what the patch creator wrote:
6144 "I found a little strange behaviour of ldap code when working with
6145 Windows 2003 AD Domain, where users was placed in more than one
6146 Organization Units. When I tried to give exim partial DN, the exit code
6147 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6148 But simultaneously result of request was absolutely normal ldap result,
6149 so I produce this patch..."
6151 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6152 REFERENCE, so I have modified the code to exclude the patch when that macro
6155 55. Some experimental protocols are using DNS PTR records for new purposes. The
6156 keys for these records are domain names, not reversed IP addresses. The
6157 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6158 leaves it alone. Component reversal etc. now happens only for IP addresses.
6161 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6163 57. Double the size of the debug message buffer (to 2048) so that more of very
6164 long debug lines gets shown.
6166 58. The exicyclog utility now does better if the number of log files to keep
6167 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6169 59. Two changes related to the smtp_active_hostname option:
6171 (1) $smtp_active_hostname is now available as a variable.
6172 (2) The default for smtp_banner uses $smtp_active_hostname instead
6173 of $primary_hostname.
6175 60. The host_aton() function is supposed to be passed a string that is known
6176 to be a valid IP address. However, in the case of IPv6 addresses, it was
6177 not checking this. This is a hostage to fortune. Exim now panics and dies
6178 if the condition is not met. A case was found where this could be provoked
6179 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6180 components; fortuitously, this particular loophole had already been fixed
6181 by change 4.50/55 above.
6183 If there are any other similar loopholes, the new check in host_aton()
6184 itself should stop them being exploited. The report I received stated that
6185 data on the command line could provoke the exploit when Exim was running as
6186 exim, but did not say which command line option was involved. All I could
6187 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6188 running as the user.
6191 61. There was a buffer overflow vulnerability in the SPA authentication code
6192 (which came originally from the Samba project). I have added a test to the
6193 spa_base64_to_bits() function which I hope fixes it.
6196 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6197 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6199 63. The daemon start-up calls getloadavg() while still root for those OS that
6200 need the first call to be done as root, but it missed one case: when
6201 deliver_queue_load_max is set with deliver_drop_privilege. This is
6202 necessary for the benefit of the queue runner, because there is no re-exec
6203 when deliver_drop_privilege is set.
6205 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6206 This has been fixed.
6208 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6209 was in use, was not putting the data itself into the right store pool;
6210 consequently, it could be overwritten for a subsequent message in the same
6211 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6214 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6216 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6217 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6218 the function were treating the return as a boolean value, which happened to
6219 work because 0=false and not-0=true, but is not correct code.
6221 68. The host_aton() function was not handling scoped IPv6 addresses (those
6222 with, for example, "%eth0" on the end) correctly.
6224 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6225 negated items (that is, ~something) in unsigned ints. Some compilers
6226 apparently mutter when there is no cast.
6228 70. If an address verification called from an ACL failed, and did not produce a
6229 user-specific message (i.e. there was only a "system" message), nothing was
6230 put in $acl_verify_message. In this situation, it now puts the system
6233 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6234 session; change 4.31/43 added the unwanted input to the log line - except
6235 that it did not do this in the start of session case. It now does.
6237 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6238 This is wrong and can cause the other end to generate a synchronization
6239 error if it is another Exim or anything else that does the synchronization
6240 check. A QUIT command is no longer sent after a timeout.
6242 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6243 during host lookups.
6245 74. The defer_ok option of callout verification was not working if it was used
6246 when verifying addresses in header lines, that is, for this case:
6248 verify = header_sender/callout=defer_ok
6250 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6251 those file descriptors could be used for SMTP connections. If anything
6252 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6253 could be sent to the SMTP client, causing chaos. The daemon now opens
6254 stdin, stdout, and stderr to /dev/null when it puts itself into the
6257 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6258 log by default. The user can override this with suitable Perl magic.
6260 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6261 the log message when discard triggers, was not working for the DATA ACL or
6262 for the non-SMTP ACL.
6264 78. Error message wording change in sieve.c.
6266 79. If smtp_accept_max_per_host was set, the number of connections could be
6267 restricted to fewer than expected, because the daemon was trying to set up
6268 a new connection before checking whether the processes handling previous
6269 connections had finished. The check for completed processes is now done
6270 earlier. On busy systems, this bug wouldn't be noticed because something
6271 else would have woken the daemon, and it would have reaped the completed
6274 80. If a message was submitted locally by a user whose login name contained one
6275 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6276 It caused a spool format error. I have fixed the spool reading code. A
6277 related problem was that the "from" clause in the Received: line became
6278 illegal because of the space(s). It is now covered by ${quote_local_part.
6280 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6283 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6285 83. Patch from Sieve maintainer to fix unterminated string problem in
6286 "vacation" handling.
6288 84. Some minor changes to the Linux configuration files to help with other
6289 OS variants using glibc.
6291 85. One more patch for Sieve to update vacation handling to latest spec.
6294 ----------------------------------------------------
6295 See the note above about the 4.44 and 4.50 releases.
6296 ----------------------------------------------------
6302 1. Change 4.43/35 introduced a bug that caused file counts to be
6303 incorrectly computed when quota_filecount was set in an appendfile
6306 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6307 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6310 3. Give more explanation in the error message when the command for a transport
6311 filter fails to execute.
6313 4. There are several places where Exim runs a non-Exim command in a
6314 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6315 was being done only for the command run by the queryprogram router. It is
6316 now done for all such subprocesses. The other cases are: ${run, transport
6317 filters, and the commands run by the lmtp and pipe transports.
6319 5. Some older OS have a limit of 256 on the maximum number of file
6320 descriptors. Exim was using setrlimit() to set 1000 as a large value
6321 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6322 systems. I've change it so that if it can't get 1000, it tries for 256.
6324 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6325 was an oversight, and furthermore, ever since the addition of extra
6326 controls (e.g. 4.43/32), the checks on when to allow different forms of
6327 "control" were broken. There should now be diagnostics for all cases when a
6328 control that does not make sense is encountered.
6330 7. $recipients is now available in the predata ACL (oversight).
6332 8. Tidy the search cache before the fork to do a delivery from a message
6333 received from the command line. Otherwise the child will trigger a lookup
6334 failure and thereby defer the delivery if it tries to use (for example) a
6335 cached ldap connection that the parent has called unbind on.
6337 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6338 of $address_data from the recipient verification was clobbered by the
6339 sender verification.
6341 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6342 was its contents. (It was OK if the option was not defined at all.)
6344 11. A "Completed" log line is now written for messages that are removed from
6345 the spool by the -Mrm option.
6347 12. $host_address is now set to the target address during the checking of
6348 ignore_target_hosts.
6350 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6351 being passed; this would have caused $sender_host_name to have been used if
6352 matching the list had actually called for a host name (not very likely,
6353 since this list is usually IP addresses). A host name is now passed as
6356 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6357 code that specifies a non-restarting handler (typically sigaction() in
6358 modern systems) in an attempt to fix a rare and obscure crash bug.
6360 15. Narrowed the window for a race in the daemon that could cause it to ignore
6361 SIGCHLD signals. This is not a major problem, because they are used only to
6362 wake it up if nothing else does.
6364 16. A malformed maildirsize file could cause Exim to calculate negative values
6365 for the mailbox size or file count. Odd effects could occur as a result.
6366 The maildirsize information is now recalculated if the size or filecount
6369 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6370 support for a long time. Removed HAVE_SYS_VFS_H.
6372 18. Updated exipick to current release from John Jetmore.
6374 19. Allow an empty sender to be matched against a lookup in an address list.
6375 Previously the only cases considered were a regular expression, or an
6378 20. Exim went into a mad DNS lookup loop when doing a callout where the
6379 host was specified on the transport, if the DNS lookup yielded more than
6382 21. The RFC2047 encoding function was originally intended for short strings
6383 such as real names; it was not keeping to the 75-character limit for
6384 encoded words that the RFC imposes. It now respects the limit, and
6385 generates multiple encoded words if necessary. To be on the safe side, I
6386 have increased the buffer size for the ${rfc2047: expansion operator from
6389 22. Failure to deliver a bounce message always caused it to be frozen, even if
6390 there was an errors_to setting on the router. The errors_to setting is now
6393 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6394 canonical form (fully expanded) before being placed in
6395 $sender_host_address.
6397 24. Updated eximstats to version 1.33
6399 25. Include certificate and key file names in error message when GnuTLS fails
6400 to set them up, because the GnuTLS error message doesn't include the name
6401 of the failing file when there is a problem reading it.
6403 26. Expand error message when OpenSSL has problems setting up cert/key files.
6406 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6407 (this can affect the format of dates).
6409 28. exim_tidydb, when checking for the continued existence of a message for
6410 which it has found a message-specific retry record, was not finding
6411 messages that were in split spool directories. Consequently, it was
6412 deleting retry records that should have stayed in existence.
6414 29. eximstats updated to version 1.35
6415 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6416 1.35 - bugfix such that pie charts by volume are generated correctly
6418 30. The SPA authentication driver was not abandoning authentication and moving
6419 on to the next authenticator when an expansion was forced to fail,
6420 contradicting the general specification for all authenticators. Instead it
6421 was generating a temporary error. It now behaves as specified.
6423 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6424 (the order specifies the preference for clients). The order is now AES256,
6425 AES128, 3DES, ARCFOUR128.
6427 31. Small patch to Sieve code - explicitly set From: when generating an
6430 32. Exim crashed if a remote delivery caused a very long error message to be
6431 recorded - for instance if somebody sent an entire SpamAssassin report back
6432 as a large number of 550 error lines. This bug was coincidentally fixed by
6433 increasing the size of one of Exim's internal buffers (big_buffer) that
6434 happened as part of the Exiscan merge. However, to be on the safe side, I
6435 have made the code more robust (and fixed the comments that describe what
6438 33. Some experimental protocols are using DNS PTR records for new purposes. The
6439 keys for these records are domain names, not reversed IP addresses. The
6440 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6441 leaves it alone. Component reversal etc. now happens only for IP addresses.
6444 34. The host_aton() function is supposed to be passed a string that is known
6445 to be a valid IP address. However, in the case of IPv6 addresses, it was
6446 not checking this. This is a hostage to fortune. Exim now panics and dies
6447 if the condition is not met. A case was found where this could be provoked
6448 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6449 components; fortuitously, this particular loophole had already been fixed
6450 by change 4.50/55 or 4.44/33 above.
6452 If there are any other similar loopholes, the new check in host_aton()
6453 itself should stop them being exploited. The report I received stated that
6454 data on the command line could provoke the exploit when Exim was running as
6455 exim, but did not say which command line option was involved. All I could
6456 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6457 running as the user.
6460 35. There was a buffer overflow vulnerability in the SPA authentication code
6461 (which came originally from the Samba project). I have added a test to the
6462 spa_base64_to_bits() function which I hope fixes it.
6465 36. The daemon start-up calls getloadavg() while still root for those OS that
6466 need the first call to be done as root, but it missed one case: when
6467 deliver_queue_load_max is set with deliver_drop_privilege. This is
6468 necessary for the benefit of the queue runner, because there is no re-exec
6469 when deliver_drop_privilege is set.
6471 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6472 was in use, was not putting the data itself into the right store pool;
6473 consequently, it could be overwritten for a subsequent message in the same
6474 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6477 38. Sometimes the final signoff response after QUIT could fail to get
6478 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6479 < 0 before doing a fflush(). This bug looks as though it goes back to the
6480 introduction of TLS in release 3.20, but "sometimes" must have been rare
6481 because the tests only now provoked it.
6487 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6488 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6489 Now it returns an integer. A number of calls were still expecting a T/F
6490 return. Fortuitously, in all cases, the tests worked in OK situations,
6491 which is the norm. However, things would have gone wrong on any write
6492 failures on the smtp file descriptor. This function is used when sending
6493 messages over SMTP and also when doing verify callouts.
6495 2. When Exim is called to do synchronous delivery of a locally submitted
6496 message (the -odf or -odi options), it no longer closes stderr before doing
6499 3. Implemented the mua_wrapper option.
6501 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6503 5. Implemented the functions header_remove(), header_testname(),
6504 header_add_at_position(), and receive_remove_recipient(), and exported them
6507 6. If an ACL "warn" statement specified the addition of headers, Exim already
6508 inserted X-ACL-Warn: at the start if there was no header name. However, it
6509 was not making this test for the second and subsequent header lines if
6510 there were newlines in the string. This meant that an invalid header could
6511 be inserted if Exim was badly configured.
6513 7. Allow an ACL "warn" statement to add header lines at the start or after all
6514 the Received: headers, as well as at the end.
6516 8. Added the rcpt_4xx retry error code.
6518 9. Added postmaster_mailfrom=xxx to callout verification option.
6520 10. Added mailfrom=xxxx to the callout verification option, for verify=
6523 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6524 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6526 12. Inserted some casts to stop certain compilers warning when using pointer
6527 differences as field lengths or precisions in printf-type calls (mostly
6528 affecting debugging statements).
6530 13. Added optional readline() support for -be (dynamically loaded).
6532 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6533 same clock tick as a message's arrival, so that its received time was the
6534 same as the "first fail" time on the retry record, and that message
6535 remained on the queue past the ultimate address timeout, every queue runner
6536 would try a delivery (because it was past the ultimate address timeout) but
6537 after another failure, the ultimate address timeout, which should have then
6538 bounced the address, did not kick in. This was a "< instead of <=" error;
6539 in most cases the first failure would have been in the next clock tick
6540 after the received time, and all would be well.
6542 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6543 being recognized when the domain list was tested by the match_domain
6544 condition in an expansion string.
6546 16. Added the ${str2b64: operator.
6548 17. Exim was always calling setrlimit() to set a large limit for the number of
6549 processes, without checking whether the existing limit was already
6550 adequate. (It did check for the limit on file descriptors.) Furthermore,
6551 errors from getrlimit() and setrlimit() were being ignored. Now they are
6552 logged to the main and panic logs, but Exim does carry on, to try to do its
6553 job under whatever limits there are.
6555 18. Imported PCRE 5.0.
6557 19. Trivial typo in log message " temporarily refused connection" (the leading
6560 20. If the log selector return_path_on_delivery was set and an address was
6561 redirected to /dev/null, the delivery process crashed because it assumed
6562 that a return path would always be set for a "successful" delivery. In this
6563 case, the whole delivery is bypassed as an optimization, and therefore no
6566 21. Internal re-arrangement: the function for sending a challenge and reading
6567 a response while authentication was assuming a zero-terminated challenge
6568 string. It's now changed to take a pointer and a length, to allow for
6569 binary data in such strings.
6571 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6573 23. Exim was not respecting finduser_retries when seeking the login of the
6574 uid under which it was called; it was always trying 10 times. (The default
6575 setting of finduser_retries is zero.) Also, it was sleeping after the final
6576 failure, which is pointless.
6578 24. Implemented tls_on_connect_ports.
6580 25. Implemented acl_smtp_predata.
6582 26. If the domain in control=submission is set empty, Exim assumes that the
6583 authenticated id is a complete email address when it generates From: or
6584 Sender: header lines.
6586 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6587 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6588 chown and chgrp in /bin and hostname in /usr/bin.
6590 28. Exim was keeping the "process log" file open after each use, just as it
6591 does for the main log. This opens the possibility of it remaining open for
6592 long periods when the USR1 signal hits a daemon. Occasional processlog
6593 errors were reported, that could have been caused by this. Anyway, it seems
6594 much more sensible not to leave this file open at all, so that is what now
6597 29. The long-running daemon process does not normally write to the log once it
6598 has entered its main loop, and it closes the log before doing so. This is
6599 so that log files can straightforwardly be renamed and moved. However,
6600 there are a couple of unusual error situations where the daemon does write
6601 log entries, and I had neglected to close the log afterwards.
6603 30. The text of an SMTP error response that was received during a remote
6604 delivery was being truncated at 512 bytes. This is too short for some of
6605 the long messages that one sometimes sees. I've increased the limit to
6608 31. It is now possible to make retry rules that apply only when a message has a
6609 specific sender, in particular, an empty sender.
6611 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6612 it possible to be selective about when SMTP synchronization is enforced.
6614 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6616 32. Implemented hosts_connection_nolog.
6618 33. Added an ACL for QUIT.
6620 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6623 35. Added mailbox_size and mailbox_filecount to appendfile.
6625 36. Added control = no_multiline_responses to ACLs.
6627 37. There was a bug in the logic of the code that waits for the clock to tick
6628 in the case where the clock went backwards by a substantial amount such
6629 that the microsecond fraction of "now" was more than the microsecond
6630 fraction of "then" (but the whole seconds number was less).
6632 38. Added support for the libradius Radius client library this is found on
6633 FreeBSD (previously only the radiusclient library was supported).
6639 1. When certain lookups returned multiple values in the form name=value, the
6640 quoting of the values was not always being done properly. Specifically:
6641 (a) If the value started with a double quote, but contained no whitespace,
6643 (b) If the value contained whitespace other than a space character (i.e.
6644 tabs or newlines or carriage returns) it was not quoted.
6645 This fix has been applied to the mysql and pgsql lookups by writing a
6646 separate quoting function and calling it from the lookup code. The fix
6647 should probably also be applied to nisplus, ibase and oracle lookups, but
6648 since I cannot test any of those, I have not disturbed their existing code.
6650 2. A hit in the callout cache for a specific address caused a log line with no
6651 reason for rejecting RCPT. Now it says "Previous (cached) callout
6652 verification failure".
6654 3. There was an off-by-one bug in the queryprogram router. An over-long
6655 return line was truncated at 256 instead of 255 characters, thereby
6656 overflowing its buffer with the terminating zero. As well as fixing this, I
6657 have increased the buffer size to 1024 (and made a note to document this).
6659 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6660 when Exim is waiting for an SMTP response from a remote server, Exim
6661 restarts its select() call on the socket, thereby resetting its timeout.
6662 This is not a problem when such interrupts are rare. Somebody set up a cron
6663 job to run exiwhat every 2 minutes, which is less than the normal select()
6664 timeout (5 or 10 minutes). This meant that the select() timeout never
6665 kicked in because it was always reset. I have fixed this by comparing the
6666 time when an interrupt arrives with the time at the start of the first call
6667 to select(). If more time than the timeout has elapsed, the interrupt is
6668 treated as a timeout.
6670 5. Some internal re-factoring in preparation for the addition of Sieve
6671 extensions (by MH). In particular, the "personal" test is moved to a
6672 separate function, and given an option for scanning Cc: and Bcc: (which is
6673 not set for Exim filters).
6675 6. When Exim created an email address using the login of the caller as the
6676 local part (e.g. when creating a From: or Sender: header line), it was not
6677 quoting the local part when it contained special characters such as @.
6679 7. Installed new OpenBSD configuration files.
6681 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6682 try to make them clearer.
6684 9. Callout options, other than the timeout value, were being ignored when
6685 verifying sender addresses in header lines. For example, when using
6687 verify = header_sender/callout=no_cache
6689 the cache was (incorrectly) being used.
6691 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6692 only the Cygwin environment.
6694 11. When return_path_on_delivery was set as a log selector, if different remote
6695 addresses in the same message used different return paths and parallel
6696 remote delivery occurred, the wrong values would sometimes be logged.
6697 (Whenever a remote delivery process finished, the return path value from
6698 the most recently started remote delivery process was logged.)
6700 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6701 lines when AUTH and/or TLS are in use. This is the "received protocol"
6702 field. Exim used to use "asmtp" for authenticated SMTP, without any
6703 indication (in the protocol name) for TLS use. Now it follows the RFC and
6704 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6705 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6706 names appear in log lines as well as in Received: header lines.
6708 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6709 extensions, and comparison tests, and to fix some bugs.
6711 14. Changes to the "personal" filter test:
6713 (1) The test was buggy in that it was just doing the equivalent of
6714 "contains" tests on header lines. For example, if a user's address was
6715 anne@some.where, the "personal" test would incorrectly be true for
6717 To: susanne@some.where
6719 This test is now done by extracting each address from the header in turn,
6720 and checking the entire address. Other tests that are part of "personal"
6721 are now done using regular expressions (for example, to check local parts
6722 of addresses in From: header lines).
6724 (2) The list of non-personal local parts in From: addresses has been
6725 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6726 taken from the Sieve specification recommendations.
6728 (3) If the message contains any header line starting with "List-" it is
6729 treated as non-personal.
6731 (4) The test for "circular" in the Subject: header line has been removed
6732 because it now seems ill-conceived.
6734 15. Minor typos in src/EDITME comments corrected.
6736 16. Installed latest exipick from John Jetmore.
6738 17. If headers_add on a router specified a text string that was too long for
6739 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6740 of string_sprintf() is now avoided.
6742 18. $message_body_size was not set (it was always zero) when running the DATA
6743 ACL and the local_scan() function.
6745 19. For the "mail" command in an Exim filter, no default was being set for
6746 the once_repeat time, causing a random time value to be used if "once" was
6747 specified. (If the value happened to be <= 0, no repeat happened.) The
6748 default is now 0s, meaning "never repeat". The "vacation" command was OK
6749 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6750 (I found it when inspecting the code).
6752 20. There is now an overall timeout for performing a callout verification. It
6753 defaults to 4 times the callout timeout, which applies to individual SMTP
6754 commands during the callout. The overall timeout applies when there is more
6755 than one host that can be tried. The timeout is checked before trying the
6756 next host. This prevents very long delays if there are a large number of
6757 hosts and all are timing out (e.g. when the network connections are timing
6758 out). The value of the overall timeout can be changed by specifying an
6759 additional sub-option for "callout", called "maxwait". For example:
6761 verify = sender/callout=5s,maxwait=20s
6763 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6764 to the end before writing, but this should make it even safer).
6766 22. Exim was forgetting that it had advertised PIPELINING for the second and
6767 subsequent messages on an SMTP connection. It was also not resetting its
6768 memory on STARTTLS and an internal HELO.
6770 23. When Exim logs an SMTP synchronization error within a session, it now
6771 records whether PIPELINING has been advertised or not.
6773 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6774 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6775 rather than long int.
6777 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6779 26. Added the never_mail option to autoreply.
6785 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6786 crash if the getsockname() call failed; this can happen if a connection is
6787 closed very soon after it is established. The problem was simply in the
6788 order in which certain operations were done, causing Exim to try to write
6789 to the SMTP stream before it had set up the file descriptor. The bug has
6790 been fixed by making things happen in the correct order.
6796 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6797 before the connection was closed, thus losing the rejection response.
6799 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6800 some early Solaris releases, but causes trouble in current releases where
6801 socklen_t is defined.
6803 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6806 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6809 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6810 that releasing the top of it at the end releases what was used for sub-
6811 expansions (unless the block got too big). However, discard this block if
6812 the first thing is a variable or header, so that we can use its block when
6813 it is dynamic (useful for very large $message_headers, for example).
6815 6. Lookups now cache *every* query, not just the most recent. A new, separate
6816 store pool is used for this. It can be recovered when all lookup caches are
6817 flushed. Lookups now release memory at the end of their result strings.
6818 This has involved some general refactoring of the lookup sources.
6820 7. Some code has been added to the store_xxx() functions to reduce the amount
6821 of flapping under certain conditions.
6823 8. log_incoming_interface used to affect only the <= reception log lines. Now
6824 it causes the local interface and port to be added to several more SMTP log
6825 lines, for example "SMTP connection from", and rejection lines.
6827 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6829 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6831 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6832 could be overwritten at the end of the current message (or the start of a
6833 new message if it was set in a HELO ACL). The value is now preserved for
6834 the duration of the SMTP connection.
6836 12. If a transport had a headers_rewrite setting, and a matching header line
6837 contained an unqualified address, that address was qualified, even if it
6838 did not match any rewriting rules. The underlying bug was that the values
6839 of the flags that permit the existence of unqualified sender and recipient
6840 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6841 non-local messages, and by -bnq for local messages) were not being
6842 preserved with the message after it was received.
6844 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6845 "next input=" as part of the text comprising the host identity instead of
6846 the correct text. The code was using the same buffer for two different
6847 strings. However, depending on which order the printing function evaluated
6848 its arguments, the bug did not always show up. Under Linux, for example, my
6849 test suite worked just fine.
6851 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6852 doesn't work with some older versions of Perl. It has been changed to "my",
6853 which in any case is probably the better facility to use.
6855 15. A really picky compiler found some instances of statements for creating
6856 error messages that either had too many or two few arguments for the format
6859 16. The size of the buffer for calls to the DNS resolver has been increased
6860 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6861 for addresses that have a lot of PTR records. This alleviates a problem; it
6862 does not fully solve it.
6864 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6865 buffer now truncates the list and logs the incident, which is the same
6866 action as happens when Exim is looking up a host name and its aliases.
6867 Previously in this situation something unpredictable would happen;
6868 sometimes it was "internal error: store_reset failed".
6870 18. If a server dropped the connection unexpectedly when an Exim client was
6871 using GnuTLS and trying to read a response, the client delivery process
6872 crashed while trying to generate an error log message.
6874 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6875 string, for example:
6877 warn message = H1: something\nH2: something
6879 the text was added as a single header line from Exim's point of view
6880 though it ended up OK in the delivered message. However, searching for the
6881 second and subsequent header lines using $h_h2: did not work. This has been
6882 fixed. Similarly, if a system filter added multiple headers in this way,
6883 the routers could not see them.
6885 20. Expanded the error message when iplsearch is called with an invalid key to
6886 suggest using net-iplsearch in a host list.
6888 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6889 ACLs are no longer actually imposed (and a message to that effect is
6892 22. If a "gecos" field in a passwd entry contained escaped characters, in
6893 particular, if it contained a \" sequence, Exim got it wrong when building
6894 a From: or a Sender: header from that name. A second bug also caused
6895 incorrect handling when an unquoted " was present following a character
6896 that needed quoting.
6898 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6899 was not being matched caselessly.
6901 24. Arranged for all hyphens in the exim.8 source to be escaped with
6904 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6905 properly. Recipient callout cache records were still being keyed to include
6906 the sender, even when use_sender was set false. This led to far more
6907 callouts that were necessary. The sender is no longer included in the key
6908 when use_sender is false.
6910 26. Added "control = submission" modifier to ACLs.
6912 27. Added the ${base62d: operator to decode base 62 numbers.
6914 28. dnsdb lookups can now access SRV records.
6916 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6917 the configuration file.
6919 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6920 (-v) mode. This makes the output for a verbose queue run more intelligible.
6922 31. Added a use_postmaster feature to recipient callouts.
6924 32. Added the $body_zerocount variable, containing the number of binary zero
6925 bytes in the message body.
6927 33. The time of last modification of the "new" subdirectory is now used as the
6928 "mailbox time last read" when there is a quota error for a maildir
6931 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6933 35. Added +ignore_unknown as a special item in host lists.
6935 36. Code for decoding IPv6 addresses in host lists is now included, even if
6936 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6937 address was recognized as an IP address, but was then not correctly decoded
6938 into binary, causing unexpected and incorrect effects when compared with
6945 1. Very minor rewording of debugging text in manualroute to say "list of
6946 hosts" instead of "hostlist".
6948 2. If verify=header_syntax was set, and a header line with an unqualified
6949 address (no domain) and a large number of spaces between the end of the
6950 name and the colon was received, the reception process suffered a buffer
6951 overflow, and (when I tested it) crashed. This was caused by some obsolete
6952 code that should have been removed. The fix is to remove it!
6954 3. When running in the test harness, delay a bit after writing a bounce
6955 message to get a bit more predictability in the log output.
6957 4. Added a call to search_tidyup() just before forking a reception process. In
6958 theory, someone could use a lookup in the expansion of smtp_accept_max_
6959 per_host which, without the tidyup, could leave open a database connection.
6961 5. Added the variables $recipient_data and $sender_data which get set from a
6962 lookup success in an ACL "recipients" or "senders" condition, or a router
6963 "senders" option, similar to $domain_data and $local_part_data.
6965 6. Moved the writing of debug_print from before to after the "senders" test
6968 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6969 problems for message scanning, either using a data ACL, or using
6970 local_scan() because the Received: header was not generated till after they
6971 were called (in order to set the time as the time of reception completion).
6972 I have revised the way this works. The header is now generated after the
6973 body is received, but before the ACL or local_scan() are called. After they
6974 are run, the timestamp in the header is updated.
6980 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6981 before starting a queue runner without re-exec. This happened only when
6982 deliver_drop_privilege was set or when the Exim user was set to root. The
6983 effect of the bug was that timeouts during subsequent deliveries caused
6984 crashes instead of being properly handled. The handler is now left at its
6985 default (and expected) setting.
6987 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6988 message, again when deliver_drop_privilege is set or Exim is run as root.
6989 The bug described in (1) was not present in this case, but the tidying up
6990 of the other signals was missing. I have made the two cases consistent.
6992 3. The ignore_target_hosts setting on a manualroute router was being ignored
6993 for hosts that were looked up using the /MX notation.
6995 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6998 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6999 operated on the sender address. After changing the $sender_address to <>
7000 for the sender address verify, Exim was re-instated it as the original
7001 (before rewriting) address, but remembering that it had rewritten it, so it
7002 wasn't rewriting it again. This bug also had the effect of breaking the
7003 sender address verification caching when the sender address was rewritten.
7005 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7006 This has been changed so that if the ip literal address matches
7007 ignore_target_hosts, the router declines.
7009 7. Added expansion conditions match_domain, match_address, and match_local_
7010 part (NOT match_host).
7012 8. The placeholder for the Received: header didn't have a length field set.
7014 9. Added code to Exim itself and to exim_lock to test for a specific race
7015 condition that could lead to file corruption when using MBX delivery. The
7016 issue is with the lockfile that is created in /tmp. If this file is removed
7017 after a process has opened it but before that process has acquired a lock,
7018 there is the potential for a second process to recreate the file and also
7019 acquire a lock. This could lead to two Exim processes writing to the file
7020 at the same time. The added code performs the same test as UW imapd; it
7021 checks after acquiring the lock that its file descriptor still refers to
7022 the same named file.
7024 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7025 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7028 11. Added the smtp_active_hostname option. If used, this will typically be made
7029 to depend on the incoming interface address. Because $interface_address is
7030 not set up until the daemon has forked a reception process, error responses
7031 that can happen earlier (such as "too many connections") no longer contain
7034 12. If an expansion in a condition on a "warn" statement fails because a lookup
7035 defers, the "warn" statement is abandoned, and the next ACL statement is
7036 processed. Previously this caused the whole ACL to be aborted.
7038 13. Added the iplsearch lookup type.
7040 14. Added ident_timeout as a log selector.
7042 15. Added tls_certificate_verified as a log selector.
7044 16. Added a global option tls_require_ciphers (compare the smtp transport
7045 option of the same name). This controls incoming TLS connections.
7047 17. I finally figured out how to make tls_require_ciphers do a similar thing
7048 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7049 before starting the TLS session.
7051 18. Tabs are now shown as \t in -bP output.
7053 19. If the log selector return_path_on_delivery was set, Exim crashed when
7054 bouncing a message because it had too many Received: header lines.
7056 20. If two routers both had headers_remove settings, and the first one included
7057 a superfluous trailing colon, the final name in the first list and the
7058 first name in the second list were incorrectly joined into one item (with a
7059 colon in the middle).
7065 1. Added -C and -D options to the exinext utility, mainly to make it easier
7066 to include in the automated testing, but these could be helpful when
7067 multiple configurations are in use.
7069 2. The exinext utility was not formatting the output nicely when there was
7070 an alternate port involved in the retry record key, nor when there was a
7071 message id as well (for retries that were specific to a specific message
7072 and a specific host). It was also confused by IPv6 addresses, because of
7073 the additional colons they contain. I have fixed the IPv4 problem, and
7074 patched it up to do a reasonable job for IPv6.
7076 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7077 delivery, the log line now contains "pipelined" if PIPELINING was used.
7079 4. An SMTP transport process used to panic and die if the bind() call to set
7080 an explicit outgoing interface failed. This has been changed; it is now
7081 treated in the same way as a connect() failure.
7083 5. A reference to $sender_host_name in the part of a conditional expansion
7084 that was being skipped was still causing a DNS lookup. This no longer
7087 6. The def: expansion condition was not recognizing references to header lines
7088 that used bh_ and bheader_.
7090 7. Added the _cache feature to named lists.
7092 8. The code for checking quota_filecount in the appendfile transport was
7093 allowing one more file than it should have been.
7095 9. For compatibility with Sendmail, the command line option
7103 and sets the incoming protocol and host name (for trusted callers). The
7104 host name and its colon can be omitted when only the protocol is to be set.
7105 Note the Exim already has two private options, -pd and -ps, that refer to
7106 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7107 "s", but I don't think that's a major issue.
7109 10. A number of refactoring changes to the code, none of which should affect
7112 (a) The number of logging options was getting close to filling up the
7113 32-bit word that was used as a bit map. I have split them into two classes:
7114 those that are passed in the argument to log_write(), and those that are
7115 only ever tested independently outside of that function. These are now in
7116 separate 32-bit words, so there is plenty of room for expansion again.
7117 There is no change in the user interface or the logging behaviour.
7119 (b) When building, for example, log lines, the code previously used a
7120 macro that called string_cat() twice, in order to add two strings. This is
7121 not really sufficiently general. Furthermore, there was one instance where
7122 it was actually wrong because one of the argument was used twice, and in
7123 one call a function was used. (As it happened, calling the function twice
7124 did not affect the overall behaviour.) The macro has been replaced by a
7125 function that can join an arbitrary number of extra strings onto a growing
7128 (c) The code for expansion conditions now uses a table and a binary chop
7129 instead of a serial search (which was left over from when there were very
7130 few conditions). Also, it now recognizes conditions like "pam" even when
7131 the relevant support is not compiled in: a suitably worded error message is
7132 given if an attempt is made to use such a condition.
7134 11. Added ${time_interval:xxxxx}.
7136 12. A bug was causing one of the ddress fields not to be passed back correctly
7137 from remote delivery subprocesses. The field in question was not being
7138 subsequently used, so this caused to problems in practice.
7140 13. Added new log selectors queue_time and deliver_time.
7142 14. Might have fixed a bug in maildirsizefile handling that threw up
7143 "unexpected character" debug warnings, and recalculated the data
7144 unnecessarily. In any case, I expanded the warning message to give more
7147 15. Added the message "Restricted characters in address" to the statements in
7148 the default ACL that block characters like @ and % in local parts.
7150 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7151 Three changes have been made:
7153 (a) There was a serious bug; a negative response to MAIL caused the whole
7154 recipient domain to be cached as invalid, thereby blocking all messages
7155 to all local parts at the same domain, from all senders. This bug has
7156 been fixed. The domain is no longer cached after a negative response to
7157 MAIL if the sender used is not empty.
7159 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7162 (c) A new callout option, "use_sender" has been added for people who want
7163 the modified behaviour.
7169 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7172 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7173 indeed breaks things for older releases.
7175 3. Added additional logging to the case where there is a problem reading data
7176 from a filter that is running in a subprocess using a pipe, in order to
7177 try to track down a specific problem.
7179 4. Testing facility fudge: when running in the test harness and attempting
7180 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7181 getting "No route to host". Convert this to a timeout.
7183 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7186 6. Some OS don't have socklen_t but use size_t instead. This affects the
7187 fifth argument of getsockopt() amongst other things. This is now
7188 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7189 can be set for individual OS. I have set it for SunOS5, OSF1, and
7190 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7191 some earlier ones do not.
7193 7. Change 4.30/15 was not doing the test caselessly.
7195 8. The standard form for an IPv6 address literal was being rejected by address
7196 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7197 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7198 this, as well as the form without the "IPv6" on the front (but only when
7199 address literals are enabled, of course).
7201 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7203 10. Exim crashed if a message with an empty sender address specified by -f
7204 encountered a router with an errors_to setting. This could be provoked only
7205 by a command such as
7209 where an empty string was supplied; "<>" did not hit this bug.
7211 11. Installed PCRE release 4.5.
7213 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7214 remained set. It is now erased.
7216 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7217 times from message ids (which are base 36 rather than the normal 62).
7219 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7220 were being counted as actual protocol errors, and logged if the log
7221 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7222 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7223 and DATA following a set of rejected RCPTs do not count as protocol errors.
7224 In other words, Exim assumes they were pipelined, though this may not
7225 actually be the case. Of course, in all cases the client gets an
7226 appropriate error code.
7228 15. If a lookup fails in an ACL condition, a message about the failure may
7229 be available; it is used if testing the ACL cannot continue, because most
7230 such messages specify what the cause of the deferral is. However, some
7231 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7232 that caused an old message to be retained and used if a later statement
7233 caused a defer, replacing the real cause of the deferral.
7235 16. If an IP address had so many PTR records that the DNS lookup buffer
7236 was not large enough to hold them, Exim could crash while trying to process
7237 the truncated data. It now detects and logs this case.
7239 17. Further to 4.21/58, another change has been made: if (and only if) the
7240 first line of a message (the first header line) ends with CRLF, a bare LF
7241 in a subsequent header line has a space inserted after it, so as not to
7242 terminate the header.
7244 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7245 unnecessarily, used atoi() instead of strtol(), and didn't check the
7246 termination when getting file sizes from file names by regex.
7248 19. Completely re-implemented the support for maildirsize files, in the light
7249 of a number of problems with the previous contributed implementation
7250 (4.30/29). In particular:
7252 . If the quota is zero, the maildirsize file is maintained, but no quota is
7255 . If the maildir directory does not exist, it is created before any attempt
7256 to write a maildirsize file.
7258 . The quota value in the file is just a cache; if the quota is changed in
7259 the transport, the new value overrides.
7261 . A regular expression is available for excluding directories from the
7264 20. The autoreply transport checks the characters in options that define the
7265 message's headers; it allows continued headers, but it was checking with
7266 isspace() after an embedded newline instead of explicitly looking for a
7269 21. If all the "regular" hosts to which an address was routed had passed their
7270 expiry times, and had not reached their retry times, the address was
7271 bounced, even if fallback hosts were defined. Now Exim should go on to try
7274 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7275 equivalent code in the SMTP transport. Some hosts send humungous responses
7276 to HELO/EHLO, more than 1024 it seems.
7278 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7279 gives compiler warnings in some environments. I've now done it "properly",
7282 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7283 (because of problems with the built-in one) was declared to return uschar *
7284 instead of char *, causing compiler failure.
7286 25. Fixed a file descriptor leak when processing alias/forward files.
7288 26. Fixed a minor format string issue in dbfn.c.
7290 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7292 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7293 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7296 29. When a sender address is verified, it is cached, to save repeating the test
7297 when there is more than one recipient in a message. However, when the
7298 verification involves a callout, it is possible for different callout
7299 options to be set for different recipients. It is too complicated to keep
7300 track of this in the cache, so now Exim always runs a verification when a
7301 callout is required, relying on the callout cache for the optimization.
7302 The overhead is duplication of the address routing, but this should not be
7305 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7306 to be verified with callout=postmaster, and the main callout worked but the
7307 postmaster check failed, the verification correctly failed. However, if a
7308 subsequent RCPT command asked for sender verification *without* the
7309 postmaster check, incorrect caching caused this verification also to fail,
7312 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7313 it was not caching the DNS options (qualify_single, search_parents) that
7314 were used when the lookup failed. A subsequent lookup with different
7315 options therefore always gave the same answer, though there were cases
7316 where it should not have. (Example: a "domains = !$mx_any" option on a
7317 dnslookup router: the "domains" option is always processed without any
7318 widening, but the router might have qualify_single set.) Now Exim uses the
7319 cached value only when the same options are set.
7321 32. Added John Jetmore's "exipick" utility to the distribution.
7323 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7324 than a timeout (e.g. a certificate is required, and is not provided), an
7325 Exim server now closes the connection immediately. Previously it waited for
7326 the client to close - but if the client is SSL, it seems that they each
7327 wait for each other, leading to a delay before one of them times out.
7329 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7330 maintained 0.8.x compatibility because I don't think many are using it, and
7331 it is clearly obsolete.
7333 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7336 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7337 client certificate was expired. A simple patch fixes this, though I don't
7338 understand the full logic of why the verify callback is called multiple
7341 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7342 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7343 which causes problems with some clients (such as the Certicom SSL Plus
7344 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7345 disables the coutermeasure allowing Eudora to connect."
7347 38. Exim was not checking that a write() to a log file succeeded. This could
7348 lead to Bad Things if a log got too big, in particular if it hit a file
7349 size limit. Exim now panics and dies if it cannot write to a log file, just
7350 as it does if it cannot open a log file.
7352 39. Modified OS/Makefile-Linux so that it now contains
7354 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7356 The two -D definitions ensure that Exim is compiled with large file
7357 support, which makes it possible to handle log files that are bigger than
7360 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7361 instance) a domain was checked against a named list that involved a lookup,
7362 causing $domain_data to be set, then another domain was checked against the
7363 same list, then the first domain was re-checked, the value of $domain_data
7364 after the final check could be wrong. In particular, if the second check
7365 failed, it could be set empty. This bug probably also applied to
7368 41. The strip_trailing_dot option was not being applied to the address given
7369 with the -f command-line option.
7371 42. The code for reading a message's header from the spool was incrementing
7372 $received_count, but never initializing it. This meant that the value was
7373 incorrect (doubled) while delivering a message in the same process in which
7374 it was received. In the most common configuration of Exim, this never
7375 happens - a fresh exec is done - but it can happen when
7376 deliver_drop_privilege is set.
7378 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7379 it now includes up to 150 characters of the unexpected data in the log
7382 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7383 and building data strings. The size of both of these buffers was 10 000
7384 bytes - far larger than anybody would *ever* want, thought I. Needless to
7385 say, somebody hit the limit. I have increased the maximum line length to
7386 20 000 and the maximum data length of concatenated lines to 100 000. I have
7387 also fixed two bugs, because there was no checking on these buffers. Tsk,
7388 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7389 buffer is too small.
7391 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7392 lsearch lookups. Now it does.
7394 46. When parsing a route_list item in a manualroute router, a fixed-length
7395 buffer was used for the list of hosts. I made this 1024 bytes long,
7396 thinking that nobody would ever have a list of hosts that long. Wrong.
7397 Somebody had a whole pile of complicated expansion conditions, and the
7398 string was silently truncated, leading to an expansion error. It turns out
7399 that it is easier to change to an unlimited length (owing to other changes
7400 that have happened since this code was originally written) than to build
7401 structure for giving a limitation error. The length of the item that
7402 expands into the list of hosts is now unlimited.
7404 47. The lsearch lookup could not handle data where the length of text line was
7405 more than 4095 characters. Such lines were truncated, leading to shortened
7406 data being returned. It should now handle lines of any length.
7408 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7409 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7412 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7413 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7414 no longer happens in comment lines. A list of replacements is now placed
7415 at the head of all of the source files, except those whose only change is
7416 to replace PERL_COMMAND in the very first #! line.
7418 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7419 messages on the queue, with a bottom-up merge sort, using code contributed
7420 by Michael Haardt. This should make operations like -bp somewhat faster on
7421 large queues. It won't affect queue runners, except when queue_run_in_order
7424 51. Installed eximstats 1.31 in the distribution.
7426 52. Added support for SRV lookups to the dnslookup router.
7428 53. If an ACL referred to $message_body or $message_body_end, the value was not
7429 reset for any messages that followed in the same SMTP session.
7431 54. The store-handling optimization for building very long strings was not
7432 differentiating between the different store pools. I don't think this
7433 actually made any difference in practice, but I've tidied it.
7435 55. While running the routers to verify a sender address, $sender_address
7436 was still set to the sender address. This is wrong, because when routing to
7437 send a bounce to the sender, it would be empty. Therefore, I have changed
7438 it so that, while verifying a sender address, $sender_address is set to <>.
7439 (There is no change to what happens when verifying a recipient address.)
7441 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7442 target A or AAAA records (if not already returned) without resetting the
7443 qualify_single or search_parents options of the DNS resolver. These are
7444 inappropriate in this case because the targets of MX and SRV records must
7445 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7446 target that, when qualified, matched something in the local domain. These
7447 two options are now turned off when doing these lookups.
7449 57. It seems that at least some releases of Reiserfs (which does not have the
7450 concept of a fixed number of inodes) returns zero and not -1 for the
7451 number of available inodes. This interacted badly with check_spool_inodes,
7452 which assumed that -1 was the "no such thing" setting. What I have done is
7453 to check that the total number of inodes is greater than zero before doing
7454 the test of how many are available.
7456 58. When a "warn" ACL statement has a log_message modifier, the message is
7457 remembered, and not repeated. This is to avoid a lot of repetition when a
7458 message has many recipients that cause the same warning to be written.
7459 However, Exim was preserving the list of already written lines for an
7460 entire SMTP session, which doesn't seem right. The memory is now reset if a
7461 new message is started.
7463 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7464 debugging output unless log_rewrite was also set.
7466 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7467 of (int)(handle) when we know that handle contains (void *)(-1).
7469 61. The Exim daemon panic-logs an error return when it closes the incoming
7470 connection. However "connection reset by peer" seems to be common, and
7471 isn't really an error worthy of noting specially, so that particular error
7474 62. When Exim is trying to find all the local interfaces, it used to panic and
7475 die if the ioctl to get the interface flags failed. However, it seems that
7476 on at least one OS (Solaris 9) it is possible to have an interface that is
7477 included in the list of interfaces, but for which you get a failure error
7478 for this call. This happens when the interface is not "plumbed" into a
7479 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7480 failure of the "get flags" call assumes that the interface is down.
7482 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7483 makes life easier for people who are doing arithmetic on fields extracted
7484 from dates, where you often get leading zeros that should not be
7485 interpreted as octal.
7487 64. Added qualify_domain to the redirect router, to override the global
7490 65. If a pathologically long header line contained very many addresses (the
7491 report of this problem mentioned 10 000) and each of them was rewritten,
7492 Exim could use up a very large amount of memory. (It kept on making new
7493 copies of the header line as it rewrote, and never released the old ones.)
7494 At the expense of a bit more processing, the header rewriting function has
7495 been changed so that it no longer eats memory in this way.
7497 66. The generation of the Received: header has been moved from the time that a
7498 message starts to be received, to the time that it finishes. The timestamp
7499 in the Received: header should now be very close to that of the <= log
7500 line. There are two side-effects of this change:
7502 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7503 logged header lines no longer include the local Received: line, because
7504 it has not yet been created. The same applies to a copy of the message
7505 that is returned to a non-SMTP sender when a message is rejected.
7507 (b) When a filter file is tested using -bf, no additional Received: header
7508 is added to the test message. After some thought, I decided that this
7511 This change does not affect the value of $received_for. It is still set
7512 after address rewriting, but before local_scan() is called.
7514 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7516 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7517 gave an unhelpful panic error message, and a defer error. I have managed to
7518 change this behaviour so that it now rejects any supplied certificate,
7519 which seems right, as the list of acceptable certificates is empty.
7521 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7522 gave an unhelpful defer error. I have not managed to make this reject any
7523 supplied certificates, but the error message it gives is "no certificate
7524 supplied", which is not helpful.
7526 70. exigrep's output now also includes lines that are not associated with any
7527 message, but which match the given pattern. Implemented by a patch from
7528 Martin Sluka, which also tidied up the Perl a bit.
7530 71. Recipient callout verification, like sender verification, was using <> in
7531 the MAIL FROM command. This isn't really the right thing, since the actual
7532 sender may affect whether the remote host accepts the recipient or not. I
7533 have changed it to use the actual sender in the callout; this means that
7534 the cache record is now keyed on a recipient/sender pair, not just the
7535 recipient address. There doesn't seem to be a real danger of callout loops,
7536 since a callout by the remote host to check the sender would use <>.
7537 [SEE ABOVE: changed after hitting problems.]
7539 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7540 temporary errors. However, in the case of such a code being given after
7541 the end of a data transmission (i.e. after ".") Exim was failing to write
7542 a retry record for the message. (Yes, there was some broken host that was
7543 actually sending 8xx at this point.)
7545 73. An unknown lookup type in a host list could cause Exim to panic-die when
7546 the list was checked. (An example that provoked this was putting <; in the
7547 middle of a list instead of at the start.) If this happened during a DATA
7548 ACL check, a -D file could be left lying around. This kind of configuration
7549 error no longer causes Exim to die; instead it causes a defer error. The
7550 incident is still logged to the main and panic logs.
7552 74. Buglet left over from Exim 3 conversion. The message "too many messages
7553 in one connection" was written to the rejectlog but not the mainlog, except
7554 when address rewriting (yes!) was being logged.
7556 75. Added write_rejectlog option.
7558 76. When a system filter was run not as root (that is, when system_filter_user
7559 was set), the values of the $n variables were not being returned to the
7560 main process; thus, they were not subsequently available in the $sn
7563 77. Added +return_path_on_delivery log selector.
7565 78. A connection timeout was being treated differently from recipients deferred
7566 when testing hosts_max_try with a message that was older than the host's
7567 retry timeout. (The host should not be counted, thus allowing all hosts to
7568 be tried at least once before bouncing.) This may have been the cause of an
7569 occasionally reported bug whereby a message would remain on the queue
7570 longer than the retry timeout, but would be bounced if a delivery was
7571 forced. I say "may" because I never totally pinned down the problem;
7572 setting up timeout/retry tests is difficult. See also the next item.
7574 79. The ultimate address timeout was not being applied to errors that involved
7575 a combination of host plus message (for example, a timeout on a MAIL
7576 command). When an address resolved to a number of possible hosts, and they
7577 were not all tried for each delivery (e.g. because of hosts_max_try), a
7578 message could remain on the queue longer than the retry timeout.
7580 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7583 81. Fixed an obscure SMTP outgoing bug which required at least the following
7584 conditions: (a) there was another message waiting for the same server;
7585 (b) the server returned 5xx to all RCPT commands in the first message so
7586 that the message was not completed; (c) the server dropped the connection
7587 or gave a negative response to the RSET that Exim sends to abort the
7588 transaction. The observed case was a dropped connection after DATA that had
7589 been sent in pipelining mode. That is, the server had advertised PIPELINING
7590 but was not implementing it correctly. The effect of the bug was incorrect
7591 behaviour, such as trying another host, and this could lead to a crash.
7597 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7598 and daemon.c were passed as pointers to ints; they should have been
7599 pointers to socklen_t variables (which are typically unsigned ints).
7601 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7604 3. Fixed a really odd bug that affected only the testing scheme; patching a
7605 certain fixed string in the binary changed the value of another string that
7606 happened to be identical to the end of the original first string.
7608 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7609 name", it returns that address as the IP address. On some operating
7610 systems (e.g. Solaris), it also passes back the IP address string as the
7611 "host name". However, on others (e.g. Linux), it passes back an empty
7612 string. Exim wasn't checking for this, and was changing the host name to an
7613 empty string, assuming it had been canonicalized.
7615 5. Although rare, it is permitted to have more than one PTR record for a given
7616 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7617 all the names associated with an address, because they do in Solaris.
7618 However, it seems that they do not in Linux for data that comes from the
7619 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7620 I found this out when I moved to a new Linux workstation and tried to run
7621 the Exim test suite.
7623 To get round this problem I have changed the code so that it now does its
7624 own call to the DNS to look up PTR records when searching for a host name.
7625 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7626 addresses that are only in /etc/hosts are still found.
7628 This behaviour is, however, controlled by an option called host_lookup_
7629 order, which defaults to "bydns:byaddr". If people want to use the other
7630 order, or indeed, just use one or the other means of lookup, they can
7631 specify it in this variable.
7633 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7634 some operating systems, this comes back from gethostbyaddr() as an empty
7635 string, and this is what Exim used to test for. However, it seems that in
7636 other systems, "." is yielded. Exim now tests for this case too.
7638 7. The values of check_spool_space and check_log_space are now held internally
7639 as a number of kilobytes instead of an absolute number of bytes. If a
7640 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7641 kilobyte. This means that much larger values can be stored.
7643 8. Exim monitor: an attempt to get the action menu when not actually pointing
7644 at a message produces an empty menu entitled "No message selected". This
7645 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7646 no entries in it ("Shell widget menu has zero width and/or height"). So I
7647 have added a single, blank menu entry in this case.
7649 9. Added ${quote_local_part.
7651 10. MIME decoding is now applied to the contents of Subject: header lines when
7654 11. Now that a reference to $sender_host_address automatically causes a reverse
7655 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7656 host lookup before query-style lookups in lists that might use this
7657 variable. This has therefore been abolished, and the "net-" prefix is no
7658 longer necessary for query-style lookups.
7660 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7661 have been a typo for LFLAGS, so it has been changed.
7663 13. The install script calls Exim with "-C /dev/null" in order to find the
7664 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7665 to be output. However, since Exim outputs its version number before the
7666 error, it didn't break the script. It just looked ugly. I fixed this by
7667 always allowing "-C /dev/null" if the caller is root.
7669 14. Ignore overlarge ACL variable number when reading spool file - insurance
7670 against a later release with more variables having written the file.
7672 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7673 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7674 this, as well as the form without the "IPv6" on the front.
7676 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7677 OS/Makefile-Darwin file.
7679 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7680 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7682 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7683 variable. However, some OS use ints for inodes. I've added cast to long int
7684 to get rid of the compiler warning.
7686 19. I had forgotten to lock out "/../" in configuration file names when
7687 ALT_CONFIG_PREFIX was set.
7689 20. Routers used for verification do not need to specify transports. However,
7690 if such a router generated a host list, and callout was configured, Exim
7691 crashed, because it could not find a port number from the (non-existent)
7692 transport. It now assumes port 25 in this circumstance.
7694 21. Added the -t option to exigrep.
7696 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7697 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7698 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7699 them is set, LOOKUP_LSEARCH is forced.
7701 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7703 24. Added sender and host information to the "rejected by local_scan()" log
7704 line; previously there was no indication of these.
7706 25. Added .include_if_exists.
7708 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7709 Linux. It turns out that not all file systems support this. Apparently some
7710 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7711 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7712 not supported on the file descriptor, is now ignored when Exim is trying to
7713 sync a directory. This applies only to Linux.
7715 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7717 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7718 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7721 29. Added support for maildirsize files from supplied patch (modified a bit).
7723 30. The use of :fail: followed by an empty string could lead Exim to respond to
7724 sender verification failures with (e.g.):
7726 550 Verification failed for <xxx>
7727 550 Sender verify failed
7729 where the first response line was missing the '-' that indicates it is not
7730 the final line of the response.
7732 31. The loop for finding the name of the user that called Exim had a hardwired
7733 limit of 10; it now uses the value of finduser_retries, which is used for
7734 all other user lookups.
7736 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7739 33. Exim was neglecting to zero errno before one call of strtol() when
7740 expanding a string and expecting an integer value. On some systems this
7741 resulted in spurious "integer overflow" errors. Also, it was casting the
7742 result into an int without checking.
7744 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7745 did not work. The code looks as if it has *never* worked, though it appears
7746 to have been documented since at least release 1.62. I have made it work.
7748 35. The "timeout_DNS" error in retry rules, also documented since at least
7749 1.62, also never worked. As it isn't clear exactly what this means, and
7750 clearly it isn't a major issue, I have abolished the feature by treating it
7751 as "timeout", and writing a warning to the main and panic logs.
7753 36. The display of retry rules for -brt wasn't always showing the error code
7756 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7757 timeout_connect_A, timeout_connect_MX.
7759 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7760 to the empty sender.
7762 39. The daemon was not analysing the content of -oX till after it had closed
7763 stderr and disconnected from the controlling terminal. This meant that any
7764 syntax errors were only noted on the panic log, and the return code from
7765 the command was 0. By re-arranging the code a little, I've made the
7766 decoding happen first, so such errors now appear on stderr, and the return
7767 code is 1. However, the actual setting up of the sockets still happens in
7768 the disconnected process, so errors there are still only recorded on the
7771 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7772 connections (as happens on some IP stacks) was logged at start up time as
7773 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7774 it from "IPv6 and IPv4", which means that two separate sockets are being
7777 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7778 says whether AF_INET or AF_INET6 was passed as an argument.
7780 42. Exiwhat output was messed up when time zones were included in log
7783 43. Exiwhat now gives more information about the daemon's listening ports,
7784 and whether -tls-on-connect was used.
7786 44. The "port" option of the smtp transport is now expanded.
7788 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7789 silently ignored. Now an error message is written to the main and panic
7792 46. There's a new ACL modifier called "logwrite" which writes to a log file
7793 as soon as it is encountered.
7795 47. Added $local_user_uid and $local_user_gid at routing time.
7797 48. Exim crashed when trying to verify a sender address that was being
7800 49. Exim was recognizing only a space character after ".include". It now also
7801 recognizes a tab character.
7803 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7804 extracting the relevant information from the specification. The man page no
7805 longer contains scrambled data for the -d option, and I've added a section
7806 at the front about calling Exim under different names.
7808 51. Added "extra_headers" argument to the "mail" command in filter files.
7810 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7813 53. Installed eximstats 1.29.
7815 54. Added transport_filter_timeout as a generic transport option.
7817 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7818 Cc: header lines. This was required by RFC 822, but it not required by RFC
7821 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7822 incoming messages that did not have them. Now it does so only if the
7823 message originates locally, that is, if there is no associated remote host
7824 address. When Resent- header lines are present, this applies to the Resent-
7825 lines rather than the non-Resent- lines.
7827 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7828 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7830 58. Messages for configuration errors now include the name of the main
7831 configuration file - useful now that there may be more than one file in a
7832 list (.included file names were always shown).
7834 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7835 for those rare installations that do not start the daemon as root or run it
7836 setuid root. I've cut out the call to initgroups() if the daemon is not
7839 60. The Exim user and group can now be bound into the binary as text strings
7840 that are looked up at the start of Exim's processing.
7842 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7844 62. Added $mailstore_basename variable.
7846 63. Installed patch to sieve.c from Michael Haardt.
7848 64. When Exim failed to open the panic log after failing to open the main log,
7849 the original message it was trying to log was written to stderr and debug
7850 output, but if they were not available (the usual case in production), it
7851 was lost. Now it is written to syslog before the two lines that record the
7852 failures to open the logs.
7854 65. Users' Exim filters run in subprocesses under the user's uid. It is
7855 possible for a "deliver" command or an alias in a "personal" command to
7856 provoke an address rewrite. If logging of address rewriting is configured,
7857 this fails because the process is not running as root or exim. There may be
7858 a better way of dealing with this, but for the moment (because 4.30 needs
7859 to be released), I have disabled address rewrite logging when running a
7860 filter in a non-root, non-exim process.
7866 1. The buildconfig auxiliary program wasn't quoting the value set for
7867 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7868 not defined. This bug was masked in 4.22 by the effect that was fixed in
7871 2. Some messages that were rejected after a message id was allocated were
7872 shown as "incomplete" by exigrep. It no longer does this for messages that
7873 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7875 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7876 have allow_domain_literals set, the ID did not get logged in the <= line.
7877 Domain literals are now always recognized in Message-ID: header lines.
7879 4. The first argument for a ${extract expansion item is the key name or field
7880 number. Leading and trailing spaces in this item were not being ignored,
7881 causing some misleading effects.
7883 5. When deliver_drop_privilege was set, single queue runner processes started
7884 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7885 same command in the process it spins off) were not dropping privilege.
7887 6. When the daemon running as "exim" started a queue runner, it always
7888 re-executed Exim in the spun-off process. This is a waste of effort when
7889 deliver_drop_privilege is set. The new process now just calls the
7890 queue-runner function directly.
7896 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7899 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7900 ignored. Though the use of -oP was forcing the writing of a pid file, it
7901 was always written to the default place.
7903 3. If the message "no IP address found for host xxxx" is generated during
7904 incoming verification, it is now followed by identification of the incoming
7905 connection (so you can more easily find what provoked it).
7907 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7909 5. Added some features to "harden" Exim a bit more against certain attacks:
7911 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7912 be put in Local/Makefile. This is like the never_users runtime option,
7913 but it cannot be overridden. The default setting is "root".
7915 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7916 prefix string with which any file named in a -C command line option
7919 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7920 is retained for -C and -D only if the caller of Exim is root. Without
7921 it, the exim user may also use -C and -D and retain privilege.
7923 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7924 command line option is disabled.
7926 6. Macro names set by the -D option must start with an upper case letter, just
7927 like macro names defined in the configuration file.
7929 7. Added "dereference=" facility to LDAP.
7931 8. Two instances of the typo "uknown" in the source files are fixed.
7933 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7934 the Configure-Makefile script screwed up while processing it.
7936 10. Incorporated PCRE 4.4.
7938 11. The SMTP synchronization check was not operating right at the start of an
7939 SMTP session. For example, it could not catch a HELO sent before the client
7940 waited for the greeting. There is now a check for outstanding input at the
7941 point when the greeting is written. Because of the duplex, asynchronous
7942 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7943 way, but not yet received, when the check is performed.
7945 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7946 on TCP/IP sockets, because this apparently causes some broken clients to
7949 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7950 unchanged) from the Cygwin maintainer.
7952 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7953 was supported instead of testing for mbx. Effectively a typo.
7955 15. The spa authenticator server code was not checking that the input it
7956 received was valid base64.
7958 16. The debug output line for the "set" modifier in ACLs was not showing the
7959 name of the variable that was being set.
7961 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7963 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7964 lookup on its own. Something else was needed to trigger the lookup. For
7965 example, a match in host_lookup or the need for a host name in a host list.
7966 Now, if $sender_host_name is referenced and the host name has not yet been
7967 looked up, a lookup is performed. If the lookup fails, the variable remains
7968 empty, and $host_lookup_failed is set to "1".
7970 19. Added "eqi" as a case-independent comparison operator.
7972 20. The saslauthd authentication condition could segfault if neither service
7973 nor realm was specified.
7975 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7976 error message that was logged was misleading, and incoming SMTP
7977 connections were dropped. The message is now more accurate, and temporary
7978 errors are given to SMTP connections.
7980 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7981 (except RSET) with a 421 error, until QUIT is received. However, it was
7982 failing to send a response to QUIT.
7984 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7985 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7986 ACL is now run after the helo_try_verify_hosts code.
7988 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7989 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7990 case-independent, so other case variants are also recognized). Apparently
7991 some systems use these upper case variants.
7993 25. If more than two messages were waiting for the same host, and a transport
7994 filter was specified for the transport, Exim sent two messages over the
7995 same TCP/IP connection, and then failed with "socket operation on non-
7996 socket" when it tried to send the third.
7998 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8000 27. The extern definition of crypt16() in expand.c was not being excluded when
8001 the OS had its own crypt16() function.
8003 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8004 as a preferred synonym for return_size_limit, both as an option and as an
8007 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8009 30. Changed the default configuration ACL to relax the local part checking rule
8010 for addresses that are not in any local domains. For these addresses,
8011 slashes and pipe symbols are allowed within local parts, but the sequence
8012 /../ is explicitly forbidden.
8014 31. SPA server authentication was not clearing the challenge buffer before
8017 32. log_message in a "warn" ACL statement was writing to the reject log as
8018 well as to the main log, which contradicts the documentation and doesn't
8019 seem right (because no rejection is happening). So I have stopped it.
8021 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8022 However, I am unable to do any testing of this.
8024 34. Fixed an infelicity in the appendfile transport. When checking directories
8025 for a mailbox, to see if any needed to be created, it was accidentally
8026 using path names with one or more superfluous leading slashes; tracing
8027 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8029 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8030 contents are added to the log line that is written for every discarded
8031 recipient. (Previously a log_message setting was ignored.)
8033 36. The ${quote: operator now quotes the string if it is empty.
8035 37. The install script runs exim in order to find its version number. If for
8036 some reason other than non-existence or emptiness, which it checks, it
8037 could not run './exim', it was installing it with an empty version number,
8038 i.e. as "exim-". This error state is now caught, and the installation is
8041 38. An argument was missing from the function that creates an error message
8042 when Exim fails to connect to the socket for saslauthd authentication.
8043 This could cause Exim to crash, or give a corrupted message.
8045 39. Added isip, isip4, and isip6 to ${if conditions.
8047 40. The ACL variables $acl_xx are now saved with the message, and can be
8048 accessed later in routers, transports, and filters.
8050 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8051 strings in the file are not string-expanded.
8053 42. If a MAIL command specified a SIZE value that was too large to fit into an
8054 int variable, the check against message_size_limit failed. Such values are
8055 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8056 day this will have to be increased, but I don't think I want to be around
8057 when emails are that large.
8064 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8065 iconv() is not standard in FreeBSD.
8067 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8068 IPv6 enabled. The observed symptom was a segmentation fault on return from
8069 the function os_common_find_running_interfaces() in src/os.c.
8071 3. In the check_special_case() function in daemon.c I had used "errno" as an
8072 argument name, which causes warnings on some systems. This was basically a
8073 typo, since it was named "eno" in the comments!
8075 4. The code that waits for the clock to tick (at a resolution of some fraction
8076 of a second) so as to ensure message-id uniqueness was always waiting for
8077 at least one whole tick, when it could have waited for less. [This is
8078 almost certainly not relevant at current processor speeds, where it is
8079 unlikely to ever wait at all. But we try to future-proof.]
8081 5. The function that sleeps for a time interval that includes fractions of a
8082 second contained a race. It did not block SIGALRM between setting the
8083 timer, and suspending (a couple of lines later). If the interval was short
8084 and the sigsuspend() was delayed until after it had expired, the suspension
8085 never ended. On busy systems this could lead to processes getting stuck for
8088 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8089 process, before it forks any delivery processes. The open lookup caching
8090 mechanism meant that the open file or database connection was passed into
8091 the delivery process. The problem was that delivery processes always tidy
8092 up cached lookup data. This could cause a problem for the next delivery
8093 process started by the queue runner, because the external queue runner
8094 process does not know about the closure. So the next delivery process
8095 still has data in the lookup cache. In the case of a file lookup, there was
8096 no problem because closing a file descriptor in a subprocess doesn't affect
8097 the parent. However, if the lookup was caching a connection to a database,
8098 the connection was closed, and the second delivery process was likely to
8099 see errors such as "PGSQL: query failed: server closed the connection
8100 unexpectedly". The problem has been fixed by closing all cached lookups
8101 in a queue runner before running a delivery process.
8103 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8104 seem to have the "const" qualifier which it has on other OS. I've
8107 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8108 *of the same type* (client or server) with the same public name that an
8109 error should be diagnosed.
8111 9. When Exim looked up a host name for an IP address, but failed to find the
8112 original IP address when looking up the host name (a safety check), it
8113 output the message "<ip address> does not match any IP for NULL", which was
8114 confusing, to say the least. The bug was that the host name should have
8115 appeared instead of "NULL".
8117 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8118 user that is built into the binary, and the -C or -D options is used, root
8119 privilege is dropped before the configuration file is read. In addition,
8120 logging is switched to stderr instead of the normal log files. If the
8121 configuration then re-defines the Exim user, the unprivileged environment
8122 is probably not what is expected, so Exim logs a panic warning message (but
8125 However, if deliver_drop_privilege is set, the unprivileged state may well
8126 be exactly what is intended, so the warning has been cut out in that case,
8127 and Exim is allowed to try to write to its normal log files.
8133 1. smtp_return_error_details was not giving details for temporary sender
8134 or receiver verification errors.
8136 2. Diagnose a configuration error if two authenticators have the same public
8139 3. Exim used not to create the message log file for a message until the first
8140 delivery attempt. This could be confusing when incoming messages were held
8141 for policy or load reasons. The message log file is now created at the time
8142 the message is received, and an initial "Received" line is written to it.
8144 4. The automatically generated man page for command line options had a minor
8145 bug that caused no ill effects; however, a more serious problem was that
8146 the procedure for building the man page automatically didn't always
8147 operate. Consequently, release 4.20 contains an out-of-date version. This
8148 shouldn't happen again.
8150 5. When building Exim with embedded Perl support, the script that builds the
8151 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8152 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8154 6. The freeze_tell option was not being used for messages that were frozen on
8155 arrival, either by an ACL or by local_scan().
8157 7. Added the smtp_incomplete_transaction log selector.
8159 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8160 was accepting AUTH without a new EHLO.
8162 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8163 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8166 10. Logging of TCP/IP connections (when configured) now happens in the main
8167 daemon process instead of the child process, so that the TCP/IP connection
8168 count is more accurate (but it can never be perfect).
8170 11. The use of "drop" in a nested ACL was not being handled correctly in the
8171 outer ACL. Now, if condition failure induced by the nested "drop" causes
8172 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8173 or "require"), the connection is dropped.
8175 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8176 that yield "discard" can now be used with an "accept" or a "discard" verb,
8177 but an error is generated for any others (because I can't see a useful way
8178 to define what should happen).
8180 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8181 are now processed in the same way as lines in the Exim configuration file.
8182 In particular, continuation lines are supported.
8184 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8186 15. Added -ti meaning -t -i.
8188 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8189 domains, and warn by logging if others are found.
8191 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8192 returned from GIFCONF when Exim is trying to find the list of interfaces on
8193 a host. The code in os.c has been modified to copy each ifreq to an aligned
8194 structure in all cases.
8196 Also, in some cases, the returned ifreq's were being copied to a 'struct
8197 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8198 means the last couple of bytes of an IPv6 address could be chopped if the
8199 ifreq contained only a normal sockaddr (14 bytes storage).
8201 18. Named domain lists were not supported in the hosts_treat_as_local option.
8202 An entry such as +xxxx was not recognized, and was treated as a literal
8205 19. Ensure that header lines added by a DATA ACL are included in the reject log
8206 if the ACL subsequently rejects the message.
8208 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8209 MD5 (which is deprecated).
8211 21. When testing a filter file using -bf, Exim was writing a message when it
8212 took the sender from a "From " line in the message, but it was not doing so
8213 when it took $return_path from a Return-Path: header line. It now does.
8215 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8216 with a valid header line field name (a series of printing characters
8217 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8219 23. Changed "disc" in the source to "disk" to conform to the documentation and
8220 the book and for uniformity.
8222 24. Ignore Sendmail's -Ooption=value command line item.
8224 25. When execve() failed while trying to run a command in a pipe transport,
8225 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8226 could be confused with a return value of 69 from the command itself. This
8227 has been changed to 127, the value the shell returns if it is asked to run
8228 a non-existent command. The wording for the related log line suggests a
8229 non-existent command as the problem.
8231 26. If received_header_text expands to an empty string, do not add a Received:
8232 header line to the message. (Well, it adds a token one on the spool, but
8233 marks it "old" so that it doesn't get used or transmitted.)
8235 27. Installed eximstats 1.28 (addition of -nt option).
8237 28. There was no check for failure on the call to getsockname() in the daemon
8238 code. This can fail if there is a shortage of resources on the system, with
8239 ENOMEM, for example. A temporary error is now given on failure.
8241 29. Contrary to the C standard, it seems that in some environments, the
8242 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8243 program. Exim now does this explicitly; it affects the formatting of
8244 timestamps using strftime().
8246 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8247 complaints. I've now initialized all the variables, to avoid this.
8249 32. Header lines added by a system filter were not being "seen" during
8250 transport-time rewrites.
8252 33. The info_callback() function passed to OpenSSL is set up with type void
8253 (*)(SSL *, int, int), as described somewhere. However, when calling the
8254 function (actually a macro) that sets it up, the type void(*)() is
8255 expected. I've put in a cast to prevent warnings from picky compilers.
8257 34. If a DNS black list lookup found a CNAME record, but there were no A
8258 records associated with the domain it pointed at, Exim crashed.
8260 35. If a DNS black list lookup returned more than one A record, Exim ignored
8261 all but the first. It now scans all returned addresses if a particular IP
8262 value is being sought. In this situation, the contents of the
8263 $dnslist_value variable are a list of all the addresses, separated by a
8266 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8267 to accept a host name and all its aliases if the forward lookup for any of
8268 them yielded the IP address of the incoming connection. Now it accepts only
8269 those names whose forward lookup yields the correct IP address. Any other
8270 names are discarded. This closes a loophole whereby a rogue DNS
8271 administrator could create reverse DNS records to break through a
8272 wildcarded host restriction in an ACL.
8274 37. If a user filter or a system filter that ran in a subprocess used any of
8275 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8276 the wrong values were passed to the pipe command ($thisaddress had the
8277 value of $0, $0 had the value of $1, etc). This bug was introduced by
8278 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8280 38. Improved the line breaking for long SMTP error messages from ACLs.
8281 Previously, if there was no break point between 40 and 75 characters, Exim
8282 left the rest of the message alone. Two changes have been made: (a) I've
8283 reduced the minimum length to 35 characters; (b) if it can't find a break
8284 point between 35 and 75 characters, it looks ahead and uses the first one
8285 that it finds. This may give the occasional overlong line, but at least the
8286 remaining text gets split now.
8288 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8289 file descriptors might be low, and that setting 1000 would always raise it.
8290 It turns out that in some environments, the limit is already over 1000 and
8291 that lowering it causes trouble. So now Exim takes care not to decrease it.
8293 40. When delivering a message, the value of $return_path is set to $sender_
8294 address at the start of routing (routers may change the value). By an
8295 oversight, this default was not being set up when an address was tested by
8296 -bt or -bv, which affected the outcome if any router or filter referred to
8299 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8300 log only when "message" or "log_message" are set. However, if one of the
8301 conditions was an address verification, or a call to a nested ACL, the
8302 messages generated by the underlying test were being passed through. This
8303 no longer happens. The underlying message is available in $acl_verify_
8304 message for both "message" and "log_message" expansions, so it can be
8305 passed through if needed.
8307 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8308 new expansion $bh_ to give the encoded byte string without charset
8309 translation. Translation happens only if iconv() is available; HAVE_ICONV
8310 indicates this at build time. HEADERS_CHARSET gives the charset to
8311 translate to; headers_charset can change it in the configuration, and
8312 "headers charset" can change it in an individual filter file.
8314 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8315 that creates RFC 2047 encoded "words" labels them as that charset instead
8316 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8317 expansion operator; (ii) when Exim creates a From: line for a local
8318 message; (iii) when a header line is rewritten to include a "phrase" part.
8320 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8321 buggy, causing it to skip the first lines of messages whose message ID
8322 ended in 'D'. This would not have bitten before Exim release 4.14, because
8323 message IDs were unlikely to end in 'D' before then. The effect was to have
8324 incorrect size information for certain domains.
8326 45. #include "config.h" was missing at the start of the crypt16.c module. This
8327 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8330 46. If there was a timeout during a "random" callout check, Exim treated it as
8331 a failure of the random address, and carried on sending RSET and the real
8332 address. If the delay was just some slowness somewhere, the response to the
8333 original RCPT would be taken as a response to RSET and so on, causing
8334 mayhem of various kinds.
8336 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8337 when I implemented it. It didn't allow for the fact that some option values
8338 may legitimately be negative (e.g. size_addition), and it didn't even do
8339 the right test for positive values.
8341 48. Domain names in DNS records are case-independent. Exim always looks them up
8342 in lower case. Some resolvers return domain names in exactly the case they
8343 appear in the zone file, that is, they may contain uppercase letters. Not
8344 all resolvers do this - some return always lower case. Exim was treating a
8345 change of case by a resolver as a change of domain, similar to a widening
8346 of a domain abbreviation. This triggered its re-routing code and so it was
8347 trying to route what was effectively the same domain again. This normally
8348 caused routing to fail (because the router wouldn't handle the domain
8349 twice). Now Exim checks for this case specially, and just changes the
8350 casing of the domain that it ultimately uses when it transmits the message
8353 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8356 50. If a filter generated a file delivery with a non-absolute name (possible if
8357 no home directory exists for the router), the forbid_file option was not
8360 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8361 the existing equality matching.
8363 52. Exim was using ints instead of ino_t variables in some places where it was
8364 dealing with inode numbers.
8366 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8367 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8368 called TMPDIR, and if it finds it is different, it changes its value.
8370 54. The smtp_printf() function is now made available to local_scan() so
8371 additional output lines can be written before returning. There is also an
8372 smtp_fflush() function to enable the detection of a dropped connection.
8373 The variables smtp_input and smtp_batched_input are exported to
8376 55. Changed the default runtime configuration: the message "Unknown user"
8377 has been removed from the ACL, and instead placed on the localuser router,
8378 using the cannot_route_message feature. This means that any verification
8379 failures that generate their own messages won't get overridden. Similarly,
8380 the "Unrouteable address" message that was in the ACL for unverifiable
8381 relay addresses has also been removed.
8383 56. Added hosts_avoid_esmtp to the smtp transport.
8385 57. The exicyclog script was not checking for the esoteric option
8386 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8387 will work only if exicyclog is run under the appropriate euid.
8389 58. Following a discussion on the list, the rules by which Exim recognises line
8390 endings on incoming messages have been changed. The -dropcr and drop_cr
8391 options are now no-ops, retained only for backwards compatibility. The
8392 following line terminators are recognized: LF CRLF CR. However, special
8393 processing applies to CR:
8395 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8396 nor a local message in the state where . is a terminator.
8398 (ii) If a bare CR is encountered in a header line, an extra space is added
8399 after the line terminator so as not to end the header. The reasoning
8400 behind this is that bare CRs in header lines are most likely either
8401 to be mistakes, or people trying to play silly games.
8403 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8404 was being incorrectly given as 18 bytes larger than it should have been.
8405 This is a VOB (very old bug).
8407 60. This may never have affected anything current, but just in case it has:
8408 When the local host is found other than at the start of a list of hosts,
8409 the local host, those with the same MX, and any that follow, are discarded.
8410 When the list in question was part of a longer list of hosts, the following
8411 hosts (not currently being processed) were also being discarded. This no
8412 longer happens. I'm not sure if this situation could ever has previously
8415 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8418 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8419 header. This is something that is recommended in a new Internet Draft, and
8420 is something that is documented as being done by Sendmail. There are two
8421 possible values. For messages generated by the autoreply transport, Exim
8424 Auto-Submitted: auto-replied
8426 whereas for all other generated messages (e.g. bounces) it adds
8428 Auto-Submitted: auto-generated
8430 63. The "personal" condition in filters now includes a test for the
8431 Auto-Submitted: header. If it contains the string "auto-" the message it
8432 not considered personal.
8434 64. Added rcpt_include_affixes as a generic transport option.
8436 65. Added queue_only_override (default true).
8438 66. Added the syslog_duplication option.
8440 67. If what should have been the first header line of a message consisted of
8441 a space followed by a colon, Exim was mis-interpreting it as a header line.
8442 It isn't of course - it is syntactically invalid and should therefore be
8443 treated as the start of the message body. The misbehaviour could have
8444 caused a number of strange effects, including loss of data in subsequent
8445 header lines, and spool format errors.
8447 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8448 client host had authenticated. This control can now be exercised by an ACL
8449 for more flexibility.
8451 69. By default, callouts do not happen when testing with -bh. There is now a
8452 variant, -bhc, which does actually run the callout code, including
8453 consulting and updating the callout cache.
8455 70. Added support for saslauthd authentication, courtesy of Alexander
8458 71. If statvfs() failed on the spool or log directories while checking their
8459 size for availability, Exim confusingly gave the error "space shortage".
8460 Furthermore, in debugging mode it crashed with a floating point exception.
8461 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8462 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8463 serious problem, Exim now writes to the main and panic logs when this
8464 happens, with details of the failure. It then refuses to accept the
8465 incoming message, giving the message "spool directory problem" or "log
8466 directory problem" with a 421 code for SMTP messages.
8468 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8469 0, 1, and 2 exist, because some OS complain for execs without them (see
8470 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8471 descriptors. However, the code omitted to check that the open succeeded,
8472 causing mysterious errors if for some reason the permissions on /dev/null
8473 got screwed. Now Exim writes a message to the main and panic logs, and
8474 bombs out if it can't open /dev/null.
8476 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8477 interact so that it is all more flexible. It is supposed to remain
8478 backwards compatible. Also added extra_local_interfaces.
8480 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8481 to bomb out with an assertion failure - to the client this appears as a
8482 connection drop. This problem occurs in the part of the code that was taken
8483 from the Samba project. Fortunately, the assertion is in a very simple
8484 function, so I have fixed this by reproducing the function inline in the
8485 one place where it is called, and arranging for authentication to fail
8486 instead of killing the process with assert().
8488 75. The SPA client code was not working when the server requested OEM rather
8489 than Unicode encoding.
8491 76. Added code to make require_files with a specific uid setting more usable in
8492 the case where statting the file as root fails - usually a non-root-mounted
8493 NFS file system. When this happens and the failure is EACCES, Exim now
8494 forks a subprocess and does the per-uid checking as the relevant uid.
8496 77. Added process_log_path.
8498 78. If log_file_path was not explicitly set, a setting of check_log_space or
8499 check_log_inodes was ignored.
8501 79. If a space check for the spool or log partitions fails, the incident is now
8502 logged. Of course, in the latter case the data may get lost...
8504 80. Added the %p formatting code to string_format() so that it can be used to
8505 print addresses in debug_print(). Adjusted all the address printing in the
8506 debugging in store.c to use %p rather than %d.
8508 81. There was a concern that a line of code in smtp_in.c could overflow a
8509 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8510 initially expressed, the concern was not well-founded, because trailing
8511 spaces are removed early. However, if the trailing spaces were followed by
8512 a NULL, they did not get removed, so the overflow was possible. Two fixes
8515 (a) I re-wrote the offending code in a cleaner fashion.
8516 (b) If an incoming SMTP command contains a NULL character, it is rejected
8519 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8520 runs initgroups(), so that if the Exim user is in any additional groups,
8521 they will be used during message reception.
8527 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob