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.
27 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
28 SMTP connection" log lines.
30 JH/02 Option default value updates:
31 - queue_fast_ramp (main) true (was false)
32 - remote_max_parallel (main) 4 (was 2)
34 JH/03 Cache static regex pattern compilations, for use by ACLs.
36 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
37 Make the rewrite never match and keep the logging. Trust the
38 admin to be using verify=header-syntax (to actually reject the message).
40 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
41 (under Linux) we watch the dir containing the final file; previously
42 it would be the dir with the first symlink. We still do not monitor
45 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
46 dn_expand() is happy to pass them through.
48 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
49 the file with the proof had an unchanged name, the new proof(s) were
50 loaded on top of the old ones (and nover used; the old ones were stapled).
52 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
53 more than one message arrived in a single connection a reference from
54 the earlier message could be re-used. Often a sigsegv resulted.
55 These variables were introduced in Exim 4.87.
56 Debug help from Graeme Fowler.
58 JH/09 Fix ${filter } for conditions that modify $value. Previously the
59 modified version would be used in construction the result, and a memory
62 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
63 Find and fix by Jasen Betts.
65 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
66 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
67 the systemwide configuration to override the Exim config.
69 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
70 API changes in libopendmarc.
72 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
73 pid 1, in the normal "background daemon" mode, having to drop process-
74 group leadership also lost track of needing to create listener sockets.
76 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
77 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
79 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
80 a capture group which obtained no text (eg. "(abc)*" matching zero
81 occurrences) could cause a segfault if the corresponding $<n> was
84 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
85 included a close-brace character (eg. it itself used an expansion) an
88 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
89 starting TLS. Previously it was after, meaning that attackers on such
90 ports had to be screened using the host_reject_connection main config
91 option. The new sequence aligns better with the STARTTLS behaviour, and
92 permits defences against crypto-processing load attacks, even though it
93 is strictly an incompatible change.
94 Also, avoid sending any SMTP fail response for either the connect ACL
95 or host_reject_connection, for TLS-on-connect ports.
97 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
98 Previously this was not permitted, but it makes reasonable sense.
99 While there, restore a restriction on using it from a connect ACL; given
100 the change JH/16 it could only return false (and before 4.91 was not
103 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
104 was exactly sized compared to the log buffer, a crash occurred with the
105 misleading message "bad memory reference; pool not found".
106 Found and traced by Jasen Betts.
108 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
109 dns_again_means_nonexist included an element causing a DNS lookup which
110 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
111 included (though probably not limited to) a process crash from stack
112 memory limit, or from excessive open files. Replace this with a paniclog
113 whine (as this is likely a configuration error), and returning
116 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
117 this always failed, probably leading to the usual downgrade to in-clear
120 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
121 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
122 to not checking that list for these lookups.
124 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
127 JH/23 Fix crash in string expansions. Previously, if an empty variable was
128 immediately followed by an expansion operator, a null-indirection read
129 was done, killing the process.
131 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
132 include an SMTP response string which is longer than that supported
133 by the delivering transport. Alleviate by wrapping such lines before
136 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
137 chars (RFC limit). Previously a limit of 12 items was made, which with
138 a not-impossible References: in the message being bounced could still
139 be over-large and get stopped in the transport.
141 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
142 close. Previously a bare socket close was done.
144 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
147 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
148 message being considered for delivery was already being handled by
149 another process, and having an SMTP connection already open, the function
150 to close it tried to use an uninitialized variable. This would afftect
151 high-volume sites more, especially when running mailing-list-style loads.
152 Pollution of logs was the major effect, as the other process delivered
153 the message. Found and partly investigated by Graeme Fowler.
155 JH/29 Change format of the internal ID used for message identification. The old
156 version only supported 31 bits for a PID element; the new 64 (on systems
157 which can use Base-62 encoding, which is all currently supported ones
158 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
159 and must use Base-36). The new ID is 23 characters rather than 16, and is
160 visible in various places - notably logs, message headers, and spool file
161 names. Various of the ancillary utilities also have to know the format.
162 As well as the expanded PID portion, the sub-second part of the time
163 recorded in the ID is expanded to support finer precision. Theoretically
164 this permits a receive rate from a single comms channel of better than the
166 The major timestamp part of the ID is not changed; at 6 characters it is
167 usable until about year 3700.
168 Updating from previously releases is fully supported: old-format spool
169 files are still usable, and the utilities support both formats. New
170 message will use the new format. The one hints-DB file type which uses
171 message-IDs (the transport wait- DB) will be discarded if an old-format ID
172 is seen; new ones will be built with only new-format IDs.
173 Optionally, a utility can be used to convert spool files from old to new,
174 but this is only an efficiency measure not a requirement for operation
175 Downgrading from new to old requires running a provided utility, having
176 first stopped all operations. This will convert any spool files from new
177 back to old (losing time-precision and PID information) and remove any
178 wait- hints databases.
180 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
181 we treated them as item separators when parsing for a list item, but they
182 need to be protected by the doublequotes. While there, add handling for
185 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
186 Found and fixed by Jasen Betts. No testcase for this as my usual text
187 editor insists on emitting only valid UTF-8.
189 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
190 a null-indirection SIGSEGV for the receive process.
192 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
193 -bh use. Internal checking would spot this and take a panic.
195 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
196 In 4.96 this would expand to empty.
198 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
199 certificate. Find and fix by Andreas Metzler.
201 JH/36 Add ARC info to DMARC hostory records.
203 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
204 or fakedefer. Previously the sender could discover that the message
205 had in fact been accepted.
207 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
208 sequences. Previously the input was not noted as being tainted; notably
209 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
210 bad coding of authenticators.
212 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
213 and ${tr...}. Found and diagnosed by Heiko Schlichting.
215 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
216 could be triggered by externally-supplied input. Found by Trend Micro.
219 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
220 be triggered by externally-controlled input. Found by Trend Micro.
223 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
224 be triggered by externally-controlled input. Found by Trend Micro.
227 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
228 Make the rewrite never match and keep the logging. Trust the
229 admin to be using verify=header-syntax (to actually reject the message).
231 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
234 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
240 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
241 after reception to before a subsequent reception. This should
242 mean slightly faster delivery, and also confirmation of reception
245 JH/02 Move from using the pcre library to pcre2. The former is no longer
246 being developed or supported (by the original developer).
248 JH/03 Constification work in the filters module required a major version
249 bump for the local-scan API. Specifically, the "headers_charset"
250 global which is visible via the API is now const and may therefore
251 not be modified by local-scan code.
253 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
254 sendfile() didi not account for the way the ClamAV driver code called it.
256 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
257 time check was being done for every character; replace that with one
260 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
261 sent was prefixed with a length byte.
263 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
264 RFC 5321. Previously Dovecot (at least) would log errors during
267 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
268 in 4.61 without a reason logged. Binaries will be bigger, which might
269 matter on diskspace-constrained systems, but debug is easier.
271 JH/09 Fix macro-definition during "-be" expansion testing. The move to
272 write-protected store for macros had not accounted for these runtime
273 additions; fix by removing this protection for "-be" mode.
275 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
276 handing out large-numbered file descriptors, violating the usual Unix
277 assumption (and required by Posix) that the lowest possible number will be
278 allocated by the kernel when a new one is needed. In the daemon, and any
279 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
280 are not useable for FD_SET() [and hence select()] and overwrite the stack.
281 Assorted crashes happen.
283 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
284 main-section options or in a connect ACL, the value from the first ever
285 connection was never replaced for subsequent connections. Found by
288 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
289 though only once PCRE2 was introduced: the memory accounting used under
290 debug offset allocations by an int, giving a hard trap in early startup.
291 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
293 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
294 with underbars is given. The write-protection of configuration introduced
295 in 4.95 trapped when normalisation was applied to an option not needing
298 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
300 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
301 failing to close the certificates directory, every hour or any time it
304 JH/16 Debugging initiated by an ACL control now continues through into routing
305 and transport processes. Previously debugging stopped any time Exim
306 re-execs, or for processing a queued message.
308 JH/17 The "expand" debug selector now gives more detail, specifically on the
309 result of expansion operators and items.
311 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
312 bad comparison between the option value and the name of the file to
313 be included was done, and a mismatch was wrongly identified.
314 4.88 to 4.95 are affected.
316 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
318 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
319 supplied as the db file-pair basename. Previously, if a directory
320 path was given, for example via the autoreply "once" option, the DB
321 file.pag and file.dir files would be created in that directory's
324 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
325 "taint" log_selector. These were previously deprecated.
327 JH/22 Fix static address-list lookups to properly return the matched item.
328 Previously only the domain part was returned.
330 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
331 the call into OpenSSL to send a TLS Close was being repeated; this
332 resulted in the library waiting for the peer's Close. If that was never
333 sent we waited forever. Fix by tracking send calls.
335 JH/24 The ${run} expansion item now expands its command string elements after
336 splitting. Previously it was before; the new ordering makes handling
337 zero-length arguments simpler. The old ordering can be obtained by
338 appending a new option "preexpand", after a comma, to the "run".
340 JH/25 Taint-check exec arguments for transport-initiated external processes.
341 Previously, tainted values could be used. This affects "pipe", "lmtp" and
342 "queryprogram" transport, transport-filter, and ETRN commands.
343 The ${run} expansion is also affected: in "preexpand" mode no part of
344 the command line may be tainted, in default mode the executable name
347 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
348 the the facility was not passed across execs, and only the first message
349 passed over a connection could use BDAT; any further ones using DATA.
351 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
352 uses $sending_ip_address and an interface is specified.
353 Previously any use of the local address in the EHLO name disabled
354 PIPECONNECT, the common case being to use the rDNS of it.
356 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
357 resumption. Previously verify failed because no certificate status is
358 passed on the wire for the restarted session. Fix by using the recorded
359 ocsp status of the stored session for the new connection.
361 JH/29 TLS resumption: the key for session lookup in the client now includes
362 more info that a server could potentially use in configuring a TLS
363 session, avoiding oferring mismatching sessions to such a server.
364 Previously only the server IP was used.
366 JH/30 Fix string_copyn() for limit greater than actual string length.
367 Previously the copied amount was the limit, which could result in a
368 overlapping memcpy for newly allocated destination soon after a
369 source string shorter than the limit. Found/investigated by KM.
371 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
372 close; it may be needed for a subsequent connection. This caused a
373 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
375 JH/32 Fix CHUNKING for a second message on a connection when the first was
376 rejected. Previously we did not reset the chunking-offered state, and
377 erroneously rejected the BDAT command. Investigation help from
380 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
381 an empty address. Previously the expansion returned an error.
383 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
384 proxy. Previously these were misparsed, leading to paniclog entries.
390 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
391 related applications. Previously an "H" was used where available info
392 says that "M" should be, so change to match.
394 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
395 as arguments, so an implementation trying to copy these into a local
396 buffer was taking a taint-enforcement trap. Fix by using dynamically
397 created buffers. Similar fix for radius expansion condition.
399 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
400 reasonable, eg. to count headers. Fix by using dynamically created
401 buffers rather than a local. Do similar fixes for ACL actions "dcc",
402 "log_reject_target", "malware" and "spam"; the arguments are expanded
403 so could be handling tainted values.
405 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
406 broken the (no-op) support for this sendmail command. Restore it
407 to doing nothing, silently, and returning good status.
409 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
410 record path was given (or the default used) without a leading directory
411 path, an error occurred on trying to open it. Use the transport's working
414 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
415 transport. Previously it was the name on the DNS A-record; use instead
416 the head of the CNAME chain leading there (if there is one). This seems
417 to align better with RFC 6125.
419 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
420 smtp_accept_max_per_host allocated resources which were not released
421 when the limit was exceeded. This eventually crashed the daemon. Fix
422 by adding a release action in that path.
424 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
425 expanded; previously using tainted values was rejected. Fix by using
426 dynamically-created buffers.
428 JH/09 Relax restrictions on ACL verify condition needing access to message
429 headers. Previously they were only permitted in data and non-smtp ACLs;
430 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
431 not_blind, header_sender and header_names_ascii verification.
433 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
434 Previously a macro used one argument twice; when called with the
435 argument as an expression having side-effects, incorrect operation
436 resulted. Use an inlineable function.
438 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
439 held open for a verify callout. Previously this wan not accounted for
440 and a corrupt onward SMTP conversation resulted.
442 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
443 Previously they were embedded naively in the constructed address; when
444 needed, strip the quoting and quote the entire local_part.
445 Also make the inbound_srs expansion condition handle quoting.
447 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
448 excluded, not matching the documentation.
450 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
451 was given for the sqlite_dbfile a trap resulted.
453 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
454 "name" argument resulted in a trap. There is no reason to disallow such;
455 this was a coding error.
457 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
458 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
459 spent suspended, ignoring the POSIX definition. Previously we assumed
460 it did and a constant offset from real time could be used as a correction.
461 Change to using the same clock source for the start-of-message and the
462 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
463 exists, just to get a clock slightly more aligned to reality.
465 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
466 RFC says it is optional some validators care. The missing char was not
467 intended but triggered by a line-wrap alignment. Discovery and fix by
468 Guillaume Outters, hacked on by JH.
470 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
471 name being quoted was tainted a trap would be taken. Fix by using
472 dynamically created buffers. The routine could have been called by a
473 rewrite with the "h" flag, by using the "-F" command-line option, or
474 by using a "name=" option on a control=submission ACL modifier.
476 JH/19 SPF: change the Authentication-Results expansion component to give
477 smtp.helo when the sender domain is empty. Previously it gave
480 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
481 not in 127.0.0.0/8 to help in spotting list domains taken over by a
482 domain-parking registrar.
484 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
485 Previously when a whitespace character was specified it was not inserted
486 after removing the newline.
488 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
489 the domain part of the recipient address. This overrides any tls_sni
490 option set, which was previously used.
492 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
495 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
496 is_tainted() had an off-by-one error in the overenthusiastic direction.
497 Find and fix by Gavan. Although NetBSD is not a supported platform for
498 4.94 this bug could affect other platforms.
500 PP/01 Fix default prime selection to be consistent.
501 One path used ike23 still, instead of exim.dev.20160529.3; now both
502 execution flows will use the same DH primes (currently
503 exim.dev.20160529.3).
505 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
506 option in smtp transport, to match the documentation. Previously
507 verification was not being done in some cases where it should have been.
509 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
510 than one server was defined and depending on the platform memory layout
511 details, an internal consistency trap could be hit while walking the list
514 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
515 files. The value is used by the authresults expansion item. Previously
516 if this was used in a router or transport, a crash could result.
518 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
519 have a sufficient incidence of EINTR returns from select that an
520 interaction with other operations done by the main daemon loop exposed
521 a bug in the error-handling. This was benign apart from the log
524 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
525 for consistency with delivered "=>" and failed "**" lines. While we're
526 there, handle PRX and TFO.
528 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
529 applied. This resulted, if any header-line rewrite rules were configured,
530 in a panic-log triggerable by sending a message with a long address in
531 a header. Fix by increasing the arbitrary limit to larger than a single
532 (dewrapped) 5322 header line maximum size.
534 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
535 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
536 RFC 6648 which deprecates X- options in protocols as a general practice.
537 Changeover between the implementations is handled by the mechanisms
540 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
541 a remote which uses callout/hold. Previously the local was queued.
543 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
546 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
547 log line. The asterisk is now consistently appended to the remote IP
548 (and port, if given), and will also be provided on defer and fail log
549 lines. Previously it could be placed on the local IP if that was being
550 logged, and was only provided on delivery lines.
552 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
554 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
555 authenticator client_send option. Previously the next char, after a pair
556 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
557 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
558 documentation. There is still no way to get a leading ^ immediately
559 after a NUL (ie. for the password of a PLAIN method authenticator.
561 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
562 files. For bad sizes read, delete the record and whine to paniclog.
564 JH/38 When logging an AUTH failure, as server, do not include sensitive
565 information. Previously, the credentials would be included if given
566 as part of the AUTH command line and an ACL denied authentication.
568 JH/39 Bug 2691: fix $local_part_data. When the matching list element
569 referred to a file, bad data was returned. This likely also affected
572 JH/40 The gsasl authenticator now supports caching of the salted password
573 generated by the client-side implementation. This required the addition
574 of a new variable: $auth4.
576 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
577 left undeleted; the attempt to re-create it then failed - resulting in
578 the usual "SIGHUP tp have daemon reload configuration" to not work.
579 This affected any platform not supporting "abstract" Unix-domain
580 sockets (i.e. not Linux).
582 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
583 recipients" response to RCPT in a later response, with a 250. The
584 previous coding assumed this would not happen, and under PIPELINING
585 would result in both lost and duplicate recipients for a message.
587 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
588 Previously the weighting was incorrectly applied. Similar fix for socks
589 proxies. Found and fixed by Heiko Schlichting.
591 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
592 not handle sub-lists included using the +namedlist syntax. While
593 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
594 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
597 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
598 Previously the permanent pool was used, so the sore could not be freed.
599 This meant a connection with many messages would use continually-growing
602 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
603 per-pool so as not to waste too much space. Previously a constant size
604 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
605 DOS attacks harder. The cost is wasted memory use in the larger blocks.
607 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
608 same space repeatedly, and a smaller process footprint.
610 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
611 for GnuTLS. Previously the name started "%s" which, while not a bug,
612 looked as if if might be one.
614 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
615 "continued connection") make the $proxy_* variables available. Previously
616 the information was not passed across the exec() call for subsequent
617 transport executions. This also mean that the log lines for the
618 messages can show the proxy information.
620 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
621 receive time. With modern systems the difference is significant.
622 The historical behaviour can be restored by disabling (a new) log_selector
623 "queue_time_exclusive".
625 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
626 content) were specifically tested for. Now, also cover expansions
627 resulting in ACL names and inline ACL content.
629 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
630 making it unusable in complex expressions.
632 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
633 fail with a taint-check on the spoolfile name, and leave the message
636 HS/01 Enforce absolute PID file path name.
638 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
640 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
642 PP/02 Bug 2643: Correct TLS DH constants.
643 A missing NUL termination in our code-generation tool had led to some
644 incorrect Diffie-Hellman constants in the Exim source.
645 Reported by kylon94, code-gen tool fix by Simon Arlott.
647 PP/03 Impose security length checks on various command-line options.
648 Fixes CVE-2020-SPRSS reported by Qualys.
650 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
651 better. Reported by Qualys.
653 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
654 providing a particularly obnoxious sender full name.
657 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
659 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
662 PP/08 Change default for recipients_max from unlimited to 50,000.
664 PP/09 Fix security issue with too many recipients on a message (to remove a
665 known security problem if someone does set recipients_max to unlimited,
666 or if local additions add to the recipient list).
667 Fixes CVE-2020-RCPTL reported by Qualys.
669 PP/10 Fix security issue in SMTP verb option parsing
670 Fixes CVE-2020-EXOPT reported by Qualys.
672 PP/11 Fix security issue in BDAT state confusion.
673 Ensure we reset known-good where we know we need to not be reading BDAT
674 data, as a general case fix, and move the places where we switch to BDAT
675 mode until after various protocol state checks.
676 Fixes CVE-2020-BDATA reported by Qualys.
678 HS/03 Die on "/../" in msglog file names
680 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
681 the Exim runtime user are allowed to create files.
683 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
686 QS/03 When reading the output from interpreted forward files we do not
687 pass the pipe between the parent and the interpreting process to
688 executed child processes (if any).
690 QS/04 Always die if requested from internal logging, even is logging is
693 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
694 the API; compilation noo longer completes with DMARC support included.
695 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
696 been reported on other platforms.
698 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
700 JH/56 Make the majority of info read from config files readonly, for defence-in-
701 depth against exploits. Suggestion by Qualys.
702 Not supported on Solaris 10.
704 JH/57 Fix control=fakreject for a custom message containing tainted data.
705 Previously this resulted in a log complaint, due to a re-expansion present
706 since fakereject was originally introduced.
708 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
709 modifier was given, a loop resulted.
711 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
712 pipelined SMTP command followed the BDAT LAST then it would be
713 incorrectly treated as part of the message body, causing a verification
716 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
717 require looser validation rules than those for 821-level addresses,
718 which only permit IP addresses.
724 JH/01 Avoid costly startup code when not strictly needed. This reduces time
725 for some exim process initialisations. It does mean that the logging
726 of TLS configuration problems is only done for the daemon startup.
728 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
730 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
731 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
734 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
735 DKIM signing is being done. Previously a transport_filter would always
736 disable CHUNKING, falling back to traditional DATA.
738 JH/05 Regard command-line recipients as tainted.
740 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
742 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
743 PAM library frees one of the arguments given to it, despite the
744 documentation. Therefore a plain malloc must be used.
746 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
747 on-stack buffers were used, resulting in a taint trap when DSN information
748 copied from a received message was written into the buffer.
750 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
751 the ordering of its ARC headers. This caused a crash.
753 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
754 a new record was being constructed with information from the peer, a trap
757 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
758 installation would get error messages from DMARC verify, when it hit the
759 nonexistent file indicated by the default. Distros wanting DMARC enabled
760 should both provide the file and set the option.
761 Also enforce no DMARC verification for command-line sourced messages.
763 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
764 could, depending on the platform, hang at the STARTTLS response.
766 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
767 message on a connection. Previously if one message had ARC headers and
768 the following one did not, a crash could result when adding an
769 Authentication-Results: header.
771 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
772 Exim main code and Exim-related utities. The introduction of taint
773 tracking also did many adjustments to string handling. Since then, eximon
774 frequently terminated with an assert failure.
776 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
777 check for 452 responses. This slightly helps the inefficieny of doing
778 a large alias-expansion into a recipient-limited target. The max_rcpt
779 transport option still applies (and at the current default, will override
780 the new feature). The check is done for either cause of synch, and forces
781 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
782 connection. The new facility is not tunable at this time.
784 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
785 library live data was being used, so the results became garbage. Make
786 copies while it is still usable.
788 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
789 on delivery deferred (==) and failed (**) lines (if a delivery was
790 attemtped). Previously it was only on completion (=>) lines.
792 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
793 for the expansion of the server_scram_iter and server_scram_salt options.
795 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
796 are now specifically given a NO_DATA response without hitting the system
797 resolver. The library goes on to do the now-standard TXT lookup.
798 Use of dnsdb lookups is not affected.
800 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
801 only retrieve the errormessage once. Previously two calls to dlerror()
802 were used, and the second one (for mainlog/paniclog) retrieved null
805 JH/20 Taint checking: disallow use of tainted data for
806 - the appendfile transport file and directory options
807 - the pipe transport command
808 - the autoreply transport file, log and once options
809 - file names used by the redirect router (including filter files)
811 - paths used by single-key lookups
812 Previously this was permitted.
814 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
815 adjusted the size of a major service buffer; this failed because the
816 buffer was in use at the time. Change to a compile-time increase in the
817 buffer size, when this authenticator is compiled into exim.
819 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
820 previous fast-mode was untenable in the face of glibs using mmap to
821 support larger malloc requests.
823 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
824 New values supported, if defined on system where compiled:
825 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
826 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
828 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
829 running a limited number of proceses in parallel, a benefit is gained. The
830 amount varies with the platform hardware and load. The use of the option
831 queue_run_in_order means we cannot do this, as ordering becomes
834 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
835 had introduced a string-copy (for ensuring NUL-termination) which was not
836 appropriate for that case, which can include embedded NUL bytes in the
837 block of data. Investigation showed the copy to actually be needless, the
838 data being length-specified.
840 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
841 done during a receiving connection, and both used TLS, global info was
842 used rather than per-connection info for tracking the state of data
843 queued for transmission. This could result in a connection hang.
845 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
846 Previously, when delivering serveral messages down a single connection
847 only the first would provide a SIZE. This was due to the size information
848 not being properly tracked.
850 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
851 TAI (at 37 seconds currently), pretend to be in UTC for time-related
852 expansion and logging. Previously, spurious values such as a future
853 minute could be seen.
855 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
856 it could crash from a null-deref. This could also affect the
857 ${addresses: } operator and ${readsock } item.
859 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
860 message following a mime one, the variable was not reset.
862 JH/30 When an pipelined-connect fails at the first response, assume incorrect
863 cached capability (perhaps the peer reneged?) and immediately retry in
866 JH/31 Fix spurious detection of timeout while writing to transport filter.
868 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
869 an attempt to copy the string was made before checking it.
871 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
872 taint of the lookup key was maintained; we now regard the presence in the
873 filesystem as sufficient validation.
875 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
876 argument is supplied.
878 JH/35 The dsearch lookup now requires that the directory is an absolute path.
879 Previously this was not checked, and nonempty relative paths made an
880 access under Exim's current working directory.
882 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
883 Previously no event was raised.
885 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
886 parameter supplied by the sender MAIL FROM command. Previously it was
887 ignored, and only the check_spool_space option value for the required
890 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
891 the size of the signing public-key. Previously it was instead giving
892 the size of the signature hash.
894 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
895 the default. See the (new) dkim_verify_min_keysizes option.
897 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
898 and an ACL use a lookup for checking either the local_part or domain,
899 stale data could be accessed. Ensure that variable references are
900 dropped between messages.
902 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
903 by the client was not checked as pointing within response data before
904 being used. A malicious client could thus cause an out-of-bounds read and
905 possibly gain authentication. Fix by adding the check.
907 JH/42 Internationalisation: change the default for downconversion in the smtp
908 transport to be "if needed". Previously it was "as previously set" for
909 the message, which usually meant "if needed" for message-submission but
910 "no" for everything else. However, MTAs have been seen using SMTPUTF8
911 even when the envelope addresses did not need it, resulting in forwarding
912 failures to non-supporting MTAs. A downconvert in such cases will be
913 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
914 transport. The change does mean that addresses needing conversion will
915 be converted when previously a delivery failure would occur.
917 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
918 response was received it would be used unchecked in a fail-DSN, violating
919 standards on line-length limits. Truncate if needed.
921 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
922 doesn't work. (Additionally add a new main config option to configure the
929 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
930 side, to decode a TLS 1.3 packet capture.
932 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
933 Previously the default library behaviour applied, sending two, each in
936 JH/03 Debug output for ACL now gives the config file name and line number for
939 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
941 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
943 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
944 buffer overrun for (non-chunking) other transports.
946 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
947 TLS1.3, means that a server rejecting a client certificate is not visible
948 to the client until the first read of encrypted data (typically the
949 response to EHLO). Add detection for that case and treat it as a failed
950 TLS connection attempt, so that the normal retry-in-clear can work (if
951 suitably configured).
953 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
954 and/or domain. Found and fixed by Jason Betts.
956 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
957 configuration). If a CNAME target was not a wellformed name pattern, a
960 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
961 the OS reports them interleaved with other addresses.
963 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
964 used both for input and for a verify callout, both encrypted, SMTP
965 responses being sent by the server could be lost. This resulted in
966 dropped connections and sometimes bounces generated by a peer sending
969 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
970 string. Previously it was possible to cause undefined behaviour in a
971 library routine (usually a crash). Found by "zerons".
973 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
976 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
977 API was removed, so update to use the newer ones.
979 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
980 any timeout set, is taking a long time. Previously we would hang on to a
981 rotated logfile "forever" if the input was arriving with long gaps
982 (a previous attempt to fix addressed lack, for a long time, of initial
985 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
986 shared (NFS) environment. The length of the tempfile name is now
987 4 + 16 ("hdr.$message_exim_id") which might break on file
988 systems which restrict the file name length to lower values.
991 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
992 shared (NFS) environment.
994 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
995 did for all versions <4.90). Notably -M, -m, --invert, -I may be
998 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
999 on some platforms for bit 31.
1001 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1002 to changes apparently associated with TLS1.3 handling some of the APIs
1003 previously used were either nonfunctional or inappropriate. Strings
1004 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1005 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1006 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1007 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1008 and the use of specific cipher names in the encrypted= ACL condition.
1010 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1012 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1013 verification result was not updated unless hosts_require_ocsp applied.
1015 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1016 queue_list_requires_admin set to false, non-admin users were denied the
1019 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1020 directory-of-certs mode. Previously they were advertised despite the
1023 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1024 A single TCP connection by a client will now hold a TLS connection open
1025 for multiple message deliveries, by default. Previously the default was to
1028 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1029 default. If built with the facility, DANE will be used. The facility
1030 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1032 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1033 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1034 must be defined and you must still, unless you define DISABLE_TLS, manage
1035 the the include-dir and library-file requirements that go with that
1036 choice. Non-TLS builds are still supported.
1038 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1041 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1042 default. If the platform supports and has the facility enabled, it will
1043 be requested on all coneections.
1045 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1046 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1048 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1050 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1051 success-DSN messages. Previously the From: header was always the default
1052 one for these; the option was ignored.
1054 JH/28 Fix the timeout on smtp response to apply to the whole response.
1055 Previously it was reset for every read, so a teergrubing peer sending
1056 single bytes within the time limit could extend the connection for a
1057 long time. Credit to Qualsys Security Advisory Team for the discovery.
1059 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1060 delivery address, which leaked information of the results of local
1061 forwarding. Change to the original envelope recipient address, per
1064 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1065 requested. Previously not bounce was generated and a log entry of
1066 error ignored was made.
1068 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1070 JH/32 Introduce a general tainting mechanism for values read from the input
1071 channel, and values derived from them. Refuse to expand any tainted
1072 values, to catch one form of exploit.
1074 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1075 was unused and the unexpanded text used for the test. Found and
1076 fixed by Ruben Jenster.
1078 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1079 an attempt to use a TLS library read routine dereffed a nul pointer,
1082 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1083 them in our smtp response.
1085 JH/36 Have the general router option retry_use_local_part default to true when
1086 any of the restrictive preconditions are set (to anything). Previously it
1087 was only for check_local user. The change removes one item of manual
1088 configuration which is required for proper retries when a remote router
1089 handles a subset of addresses for a domain.
1091 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1092 link count into consideration.
1094 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1095 caused the extension of big_buffer, the following lines were ignored.
1097 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1098 accordance with RFC 2308. Previously there was no expiry, so a longlived
1099 receive process (eg. due to ACL delays) versus a short SOA value could
1102 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1104 JH/39 Promote DMARC support to mainline.
1106 JH/40 Bug 2452: Add a References: header to DSNs.
1108 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1109 parameters. The relevant library call is documented as "Deprecated: This
1110 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1111 3.6.0, DH parameters are negotiated following RFC7919."
1113 HS/06 Change the default of dnssec_request_domains to "*"
1115 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1116 carried on and emitted a BDAT command, even when PIPELINING was not
1119 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1120 buffer was used for the filename, resulting in a trap when tainted
1121 arguments (eg. $domain) were used.
1123 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1124 recommended to avoid a possible server-load attack. The feature can be
1125 re-enabled via the openssl_options main cofiguration option.
1127 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1128 for version 4.90 - adding a "more data" boolean to the arguments.
1129 Bumped the ABI version number also, this having been missed previously;
1130 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1131 respect of usage of smtp_printf() by either local_scan code or libraries
1132 accessed via the ${dlfunc } expansion item. Both will need coding
1133 adjustment for any calls to smtp_printf() to match the new function
1134 signature; a FALSE value for the new argument is always safe.
1136 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1137 the file-offset (which the Linux syscall does, and exim expects); this
1138 resulted in an indefinite loop.
1140 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1141 to do ARC verification. The Authentication-Results: header line added
1142 by the configuration then had no ARC item.
1144 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1145 an incorrect mode was used for reading the result, resulting in it being
1152 JH/01 Remove code calling the customisable local_scan function, unless a new
1153 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1155 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1156 non-signal-safe functions being used.
1158 JH/03 Bug 2269: When presented with a received message having a stupidly large
1159 number of DKIM-Signature headers, disable DKIM verification to avoid
1160 a resource-consumption attack. The limit is set at twenty.
1162 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1163 report of oldest_pass in ${authres } in consequence, and separate out
1164 some descriptions of reasons for verification fail.
1166 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1167 files in the spool were present and unlocked. A queue-runner could spot
1168 them, resulting in a duplicate delivery. Fix that by doing the unlock
1169 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1170 add more error-checking on spoolfile handling while that code is being
1173 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1174 No known attacks, no CVE, this is defensive hardening.
1176 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1177 a queue-runner could start a delivery while other operations were ongoing.
1178 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1179 Found and investigated by Tim Stewart. Fix by using the open message data
1180 file handle rather than opening another, and not locally closing it (which
1181 releases a lock) for that case, while creating the temporary .eml format
1182 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1184 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1185 $sender_verify_failure/$recipient_verify_failure to "random".
1187 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1190 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1191 Previously this would segfault.
1193 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1196 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1197 like zero, since the resolver should be doing this for us, But we need one
1198 as a CNAME but no MX presence gets the CNAME returned; we need to check
1199 that doesn't point to an MX to declare it "no MX returned" rather than
1200 "error, loop". A new main option is added so the older capability of
1201 following some limited number of chain links is maintained.
1203 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1205 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1206 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1207 now-deprecated earlier definitions used only specified the range up to TLS
1208 1.2 (in the older-version library docs).
1210 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1212 JH/15 Rework TLS client-side context management. Stop using a global, and
1213 explicitly pass a context around. This enables future use of TLS for
1214 connections to service-daemons (eg. malware scanning) while a client smtp
1215 connection is using TLS; with cutthrough connections this is quite likely.
1217 JH/16 Fix ARC verification to do AS checks in reverse order.
1219 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1221 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1222 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1223 Previously the "utf8" would be re-prepended for every additional message.
1225 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1226 Previously thery were accepted, resulting in issues when attempting to
1227 forward messages to a non-supporting MTA.
1229 PP/02 Let -n work with printing macros too, not just options.
1231 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1232 one parent address was copied, and bogus data was used at delivery-logging
1233 time. Either a crash (after delivery) or bogus log data could result.
1234 Discovery and analysis by Tim Stewart.
1236 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1237 Previously if the string ended mid-character, we did not insert the
1238 promised '?' replacement.
1240 PP/04 Documentation: current string operators work on bytes, not codepoints.
1242 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1243 should pack well giving a smaller memory footprint so better caching and
1244 therefore performance. Group the declarations where this can't be done so
1245 that the byte-sized flag variables are not interspersed among pointer
1246 variables, giving a better chance of good packing by the compiler.
1248 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1249 non-null, to avoid issues with sites running BATV. Previously reports were
1250 sent with an empty envelope sender so looked like bounces.
1252 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1253 The ignore_error flag wasn't being returned from the filter subprocess so
1254 was not set for later routers. Investigation and fix by Matthias Kurz.
1256 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1257 and a msg:complete for the whole, when a message is manually removed using
1258 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1260 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1261 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1262 hope that is portable enough.
1264 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1265 requiring a known-CA anchor certificate; make it now rely entirely on the
1266 TLSA as an anchor. Checking the name on the leaf cert against the name
1267 on the A-record for the host is still done for TA (but not for EE mode).
1269 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1270 included in delivery lines for non-proxied connections, when compiled with
1271 SUPPORT_SOCKS and running with proxy logging enabled.
1273 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1274 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1275 move the existing event to fire before the normal logging of message
1276 failure so that custom logging is bracketed by normal logging.
1278 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1279 msg:fail:internal event. Developement by Matthias Kurz.
1281 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1282 far too small for todays use of crypto signatures stored there. Go all
1283 the way to the max DNS message size of 64kB, even though this might be
1284 overmuch for IOT constrained device use.
1286 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1287 copy a string over itself. The library routine is documented as not
1288 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1290 JH/32 For main options check_spool_space and check_inode_space, where the
1291 platform supports 64b integers, support more than the previous 2^31 kB
1292 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1293 the previous G, M, k.
1295 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1296 $authenticated_fail_id variable on authentication failure. Previously
1299 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1300 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1301 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1302 GNUTLS_SEC_PARAM_MEDIUM.
1304 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1305 side. Previously we would continue as if no SNI had been received.
1307 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1308 "<" character, which should be a list-separator specification, we walked
1309 off past the nul-terimation.
1311 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1312 causes) even when the retry time is not yet met. Previously they were
1313 not, meaning that when (say) an account was over-quota and temp-rejecting,
1314 and multiple senders' messages were queued, only one sender would get
1315 notified on each configured delay_warning cycle.
1317 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1319 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1320 cipher-suites, an error can be left on the stack even for a succeeding
1321 accept; this results in impossible error messages when a later operation
1324 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1325 return error codes indicating retry. Under TLS1.3 this becomes required.
1327 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1328 it only wrote the new authenticators, resulting in a lack of tracking of
1329 peer changes of ESMTP extensions until the next cache flush.
1331 JH/41 Fix the loop reading a message header line to check for integer overflow,
1332 and more-often against header_maxsize. Previously a crafted message could
1333 induce a crash of the recive process; now the message is cleanly rejected.
1335 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1336 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1342 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1343 When redis_servers is set to a list of > 1 element, and the Redis servers
1344 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1345 case of MOVED into a DEFER case instead, thus moving the query onto the
1346 next server in the list. For a cluster of N elements, all N servers must
1347 be defined in redis_servers.
1349 GF/02 Catch and remove uninitialized value warning in exiqsumm
1350 Check for existence of @ARGV before looking at $ARGV[0]
1352 JH/01 Replace the store_release() internal interface with store_newblock(),
1353 which internalises the check required to safely use the old one, plus
1354 the allocate and data copy operations duplicated in both (!) of the
1355 extant use locations.
1357 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1358 modifier. This matches the restriction on the commandline.
1360 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1361 Previously only the last row was returned.
1363 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1364 we assumed that tags in the header were well-formed, and parsed the
1365 element content after inspecting only the first char of the tag.
1366 Assumptions at that stage could crash the receive process on malformed
1369 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1370 While running the DKIM ACL we operate on the Permanent memory pool so that
1371 variables created with "set" persist to the DATA ACL. Also (at any time)
1372 DNS lookups that fail create cache records using the Permanent pool. But
1373 expansions release any allocations made on the current pool - so a dnsdb
1374 lookup expansion done in the DKIM ACL releases the memory used for the
1375 DNS negative-cache, and bad things result. Solution is to switch to the
1376 Main pool for expansions.
1377 While we're in that code, add checks on the DNS cache during store_reset,
1378 active in the testsuite.
1379 Problem spotted, and debugging aided, by Wolfgang Breyha.
1381 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1382 When none of the hosts presented to a transport match an already-open
1383 connection, close it and proceed with the list. Previously we would
1384 queue the message. Spotted by Lena with Yahoo, probably involving
1387 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1388 Previously a spurious "250 OK id=" response was appended to the proper
1391 JH/08 The "support for" informational output now, which built with Content
1392 Scanning support, has a line for the malware scanner interfaces compiled
1393 in. Interface can be individually included or not at build time.
1395 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1396 by the template makefile "src/EDITME". The "STREAM" support for an older
1397 ClamAV interface method is removed.
1399 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1400 rows affected is given instead).
1402 JH/11 The runtime Berkeley DB library version is now additionally output by
1403 "exim -d -bV". Previously only the compile-time version was shown.
1405 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1406 SMTP connection. Previously, when one had more recipients than the
1407 first, an abortive onward connection was made. Move to full support for
1408 multiple onward connections in sequence, handling cutthrough connection
1409 for all multi-message initiating connections.
1411 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1412 routers. Previously, a multi-recipient message would fail to match the
1413 onward-connection opened for the first recipient, and cause its closure.
1415 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1416 a timeout on read on a GnuTLS initiating connection, resulting in the
1417 initiating connection being dropped. This mattered most when the callout
1418 was marked defer_ok. Fix to keep the two timeout-detection methods
1421 JH/15 Relax results from ACL control request to enable cutthrough, in
1422 unsupported situations, from error to silently (except under debug)
1423 ignoring. This covers use with PRDR, frozen messages, queue-only and
1426 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1428 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1429 metadata, resulting in a crash in free().
1431 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1432 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1433 Broken also in d185889f4, with init system revamp.
1435 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1436 Previously we abruptly closed the connection after reading a malware-
1437 found indication; now we go on to read the "scan ok" response line,
1440 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1441 ACL. Previously, a crash would result.
1443 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1444 macro text after a replacement (previously it was only once per line) and
1445 by skipping builtin macros when searching for an uppercase lead character.
1447 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1448 for the build is renamed.
1450 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1451 was allocated for every new TLS startup, meaning one per message. Fix
1452 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1454 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1455 reported the original. Fix to report (as far as possible) the ACL
1456 result replacing the original.
1458 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1459 OpenSSL. Certificate information is loaded for every new TLS startup,
1460 and the resources needed to be freed.
1462 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1464 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1467 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1468 DATA response info to the (existing) per-recipient response info for
1469 the "C=" log element. It can have useful tracking info from the
1470 destination system. Patch from Simon Arlott.
1472 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1473 length value. Previously this would segfault.
1475 HS/02 Support Avast multiline protoocol, this allows passing flags to
1476 newer versions of the scanner.
1478 JH/28 Ensure that variables possibly set during message acceptance are marked
1479 dead before release of memory in the daemon loop. This stops complaints
1480 about them when the debug_store option is enabled. Discovered specifically
1481 for sender_rate_period, but applies to a whole set of variables.
1482 Do the same for the queue-runner and queue-list loops, for variables set
1483 from spool message files. Do the same for the SMTP per-message loop, for
1484 certain variables indirectly set in ACL operations.
1486 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1487 as a multi-recipient message from a mailinglist manager). The coding had
1488 an arbitrary cutoff number of characters while checking for more input;
1489 enforced by writing a NUL into the buffer. This corrupted long / fast
1490 input. The problem was exposed more widely when more pipelineing of SMTP
1491 responses was introduced, and one Exim system was feeding another.
1492 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1493 receiving system, and refused recipients seen by the sending system
1494 (propating to people being dropped from mailing lists).
1495 Discovered and pinpointed by David Carter.
1497 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1498 replaced by the ${authresults } expansion.
1500 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1502 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1503 allows proper process termination in container environments.
1505 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1506 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1508 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1509 and "err_perm", deprecated since 4.83 when the RFC-defined words
1510 "temperror" and "permerror" were introduced.
1512 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1513 transport-filters or DKIM-signing. The restriction was lost in the
1514 consolidation of verify-callout and delivery SMTP handling.
1515 Extend the restriction to also cover ARC-signing.
1517 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1518 in defer=pass mode supply a 450 to the initiator. Previously the message
1521 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1522 tls_require_ciphers is used as before.
1524 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1525 "pass_unscanned". Only tmpfails from the scanner are written to
1526 the paniclog, as they may require admin intervention (permission
1527 denied, license issues). Other scanner errors (like decompression
1528 bombs) do not cause a paniclog entry.
1530 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1531 Previously it was possible to log spurious information in receive log
1534 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1535 triggered odd behaviour from Outlook Express clients.
1537 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1540 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1541 since the IETF WG has not yet settled on that versus the original
1542 "bare" representation.
1544 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1545 Previously the millisecond value corrupted the output.
1546 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1547 corrupted the output.
1553 JH/01 Rework error string handling in TLS interface so that the caller in
1554 more cases is responsible for logging. This permits library-sourced
1555 string to be attached to addresses during delivery, and collapses
1556 pairs of long lines into single ones.
1558 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1559 during configuration. Wildcards are allowed and expanded.
1561 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1562 This permits better logging.
1564 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1565 do not close it down and have the child transport start it up again on
1566 the passed-on TCP connection. Instead, proxy the child (and any
1567 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1568 affected: the continued delivery log lines do not have any DNSSEC, TLS
1569 Certificate or OCSP information. TLS cipher information is still logged.
1571 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1572 identical IP addresses on different listening ports. Will also affect
1575 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1576 add noisy ifdef guards to special-case this sillyness.
1577 Patch from Bernd Kuhls.
1579 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1580 than 255 are no longer allowed.
1582 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1583 Disable the session-cache too, which might reduce our load. Since we
1584 currrectly use a new context for every connection, both as server and
1585 client, there is no benefit for these.
1586 GnuTLS appears to not support tickets server-side by default (we don't
1587 call gnutls_session_ticket_enable_server()) but client side is enabled
1588 by default on recent versions (3.1.3 +) unless the PFS priority string
1591 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1592 <https://reproducible-builds.org/specs/source-date-epoch/>.
1594 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1595 the check for any unsuccessful recipients did not notice the limit, and
1596 erroneously found still-pending ones.
1598 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1599 MSG_MORE. Only in-clear (not on TLS connections).
1601 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1602 transport-filter is involved we can buffer the headers in memory for
1603 creating the signature, and read the spool data file once for the
1604 signature and again for transmission.
1606 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1607 4.77 as the kernel support then wasn't solid, having issues in 64bit
1608 mode. Now, it's been long enough. Add support for FreeBSD also.
1610 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1611 case where the routing stage had gathered several addresses to send to
1612 a host before calling the transport for the first, we previously failed
1613 to close down TLS in the old transport process before passing the TCP
1614 connection to the new process. The new one sent a STARTTLS command
1615 which naturally failed, giving a failed delivery and bloating the retry
1616 database. Investigation and fix prototype from Wolfgang Breyha.
1618 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1619 false-negatives in the check that the sender had not preempted a response
1620 or prompt from Exim (running as a server), due to that code's lack of
1621 awareness of the SMTP input buffering.
1623 PP/04 Add commandline_checks_require_admin option.
1624 Exim drops privileges sanely, various checks such as -be aren't a
1625 security problem, as long as you trust local users with access to their
1626 own account. When invoked by services which pass untrusted data to
1627 Exim, this might be an issue. Set this option in main configuration
1628 AND make fixes to the calling application, such as using `--` to stop
1631 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1632 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1633 responses to those, into a single TLS record each way (this usually means
1634 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1637 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1638 affects you only if you're dancing at the edge of the param size limits.
1639 If you are, and this message makes sense to you, then: raise the
1640 configured limit or use OpenSSL 1.1. Nothing we can do for older
1643 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1644 cmdline element to get the documented default one. Previously it was
1647 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1648 are unsuable for non-message-specific reasons - eg. connection timeout,
1649 banner-time rejection.
1651 JH/16 Fix logging of delivery remote port, when specified by router, under
1654 PP/06 Repair manualroute's ability to take options in any order, even if one
1655 is the name of a transport.
1658 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1660 JH/17 Change the list-building routines interface to use the expanding-string
1661 triplet model, for better allocation and copying behaviour.
1663 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1664 Previously it was constructed the first time a possibly-matching string
1665 was met in the configuration file input during startup; now it is done
1668 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1669 compatible one, to avoid the (poorly documented) possibility of a config
1670 file in the working directory redirecting the DB files, possibly correpting
1671 some existing file. CVE-2017-10140 assigned for BDB.
1673 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1674 cache-hot. Previously, although the result was properly cached, the
1675 initial verify call returned a defer.
1677 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1678 the main verify for receipient in uncached-mode.
1680 JH/22 Retire historical build files to an "unsupported" subdir. These are
1681 defined as "ones for which we have no current evidence of testing".
1683 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1684 if present. Previously it was ignored.
1686 JH/24 Start using specified-initialisers in C structure init coding. This is
1687 a C99 feature (it's 2017, so now considered safe).
1689 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1690 if was a fixed-sized field and bitmask ops via macros; it is now more
1693 PP/07 GitHub PR 56: Apply MariaDB build fix.
1694 Patch provided by Jaroslav Škarvada.
1696 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1697 during Coverity cleanups [4.87 JH/47]
1698 Diagnosis and fix provided by Michael Fischer v. Mollard.
1700 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1701 the right size to place the terminating semicolon on its own folded
1702 line, the header hash was calculated to an incorrect value thanks to
1703 the (relaxed) space the fold became.
1705 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1706 and confused the parent.
1708 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1709 which could crash as a result. This could lead to undeliverable messages.
1711 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1714 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1715 matters on fast-turnover and PID-randomising systems, which were getting
1716 out-of-order delivery.
1718 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1719 a possibly-overlapping copy. The symptom was that "Remote host closed
1720 connection in response to HELO" was logged instead of the actual 4xx
1723 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1724 Previously only that bufferd was discarded, resulting in SYMTP command
1727 JH/32 DKIM: when a message has multiple signatures matching an identity given
1728 in dkim_verify_signers, run the dkim acl once for each. Previously only
1729 one run was done. Bug 2189.
1731 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1732 "panic the current process" to "deliberately defer". The panic log is
1733 still written with the problem list name; the mail and reject logs now
1734 get a temp-reject line for the message that was being handled, saying
1735 something like "domains check lookup or other defer". The SMTP 451
1736 message is still "Temporary local problem".
1738 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1739 A crafted sequence of BDAT commands could result in in-use memory beeing
1740 freed. CVE-2017-16943.
1742 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1743 from SMTP input. Previously it was always done; now only done for DATA
1744 and not BDAT commands. CVE-2017-16944.
1746 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1747 to the message (such as an overlong header line). Previously this was
1748 not done and we did not exit BDAT mode. Followon from the previous item
1749 though a different problem.
1755 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1756 than -2003 did; needs libidn2 in addition to libidn.
1758 JH/02 The path option on a pipe transport is now expanded before use.
1760 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1761 Patch provided by "Björn", documentation fix added too.
1763 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1764 missing a wire-to-host endian conversion.
1766 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1767 close after a BDAT command line could be taken as a following command,
1768 giving a synch failure. Fix by only checking for synch immediately
1769 before acknowledging the chunk.
1771 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1772 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1773 macro. Patches provided by Josh Soref.
1775 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1776 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1777 by IANA as a service extension. However, John Klensin suggests that we
1780 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1781 the dkim code may be unix-mode line endings rather than smtp wire-format
1782 CRLF, so prepend a CR to any bare LF.
1784 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1785 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1787 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1788 the first were themselves being wrongly included in the feed into dkim
1789 processing; with most chunk sizes in use this resulted in an incorrect
1790 body hash calculated value.
1792 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1793 DKIM signature block, for verification. Although advised against by
1794 standards it is specifically not ruled illegal.
1796 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1798 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1799 missing a body hash (the bh= tag).
1801 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1802 It seems that HAProxy sends the Proxy Protocol information in clear and
1803 only then does a TLS startup, so do the same.
1805 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1806 TCP connections (such as for Spamd) unless the daemon successfully set
1807 Fast Open mode on its listening sockets. This fixes breakage seen on
1808 too-old kernels or those not configured for Fast Open, at the cost of
1809 requiring both directions being enabled for TFO, and TFO never being used
1810 by non-daemon-related Exim processes.
1812 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1813 endings, at least on the first header line. Try to canonify any that get
1814 past that check, despite the cost.
1816 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1817 now limited to an arbitrary five deep, while parsing addresses with the
1818 strip_excess_angle_brackets option enabled.
1820 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1821 instead leave the unprompted TLS handshake in socket buffer for the
1822 TLS library to consume.
1824 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1826 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1828 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1829 region in one operation, for speed, and this leaves assigned pointers
1830 dangling. Add checks run only under the testsuite which checks all
1831 variables at a store-reset and panics on a dangling pointer; add code
1832 explicitly nulling out all the variables discovered. Fixes one known
1833 bug: a transport crash, where a dangling pointer for $sending_ip_address
1834 originally assigned in a verify callout, is re-used.
1836 PP/06 Drop '.' from @INC in various Perl scripts.
1838 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1840 PP/08 Reduce a number of compilation warnings under clang; building with
1841 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1842 should be warning-free.
1844 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1846 HS/01 Fix portability problems introduced by PP/08 for platforms where
1847 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1849 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1850 chunk. This allows us to accept broken chunked messages. We need a more
1851 general solution here.
1853 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1854 already-broken messages in the queue.
1856 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1858 JH/19 Fix reference counting bug in routing-generated-address tracking.
1864 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1865 supports it and a size is available (ie. the sending peer gave us one).
1867 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1868 years of being deprecated). The replacements are the ACLs
1869 acl_smtp_mime and acl_not_smtp_mime.
1871 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1872 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1873 or even an in-clear connection were permitted. Now, if the host lookup
1874 was dnssec and dane was requested then the host is only used if the
1875 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1876 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1877 if one fails this test.
1878 This means that a poorly-configured remote DNS will make it incommunicado;
1879 but it protects against a DNS-interception attack on it.
1881 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1882 non-noisy when a race steals the message being considered.
1884 JH/05 If main configuration option tls_certificate is unset, generate a
1885 self-signed certificate for inbound TLS connections.
1887 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1888 in rewrites and routers.
1890 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1891 and logged a warning sing 4.83; now they are a configuration file error.
1893 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1894 (lacking @domain). Apply the same qualification processing as RCPT.
1896 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1898 JH/10 Support ${sha256:} applied to a string (as well as the previous
1901 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1902 a cutthrough deliver is pending, as we always want to make a connection.
1903 This also avoids re-routing the message when later placing the cutthrough
1904 connection after a verify cache hit.
1905 Do not update it with the verify result either.
1907 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1908 when routing results in more than one destination address.
1910 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1911 signing (which inhibits the cutthrough capability). Previously only
1912 the presence of an option was tested; now an expansion evaluating as
1913 empty is permissible (obviously it should depend only on data available
1914 when the cutthrough connection is made).
1916 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1917 the relevant preceding SMTP command did not note the pipelining mode.
1919 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1920 Previously they were not counted.
1922 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1923 as one having no matching records. Previously we deferred the message
1924 that needed the lookup.
1926 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1927 distinguished as "(=".
1929 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1930 for missing MX records. Previously it only worked for missing A records.
1932 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1934 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1935 after the data-go-ahead and data-ack. Patch from Jason Betts.
1937 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1938 even for a "none" policy. Patch from Tony Meyer.
1940 JH/22 Fix continued use of a connection for further deliveries. If a port was
1941 specified by a router, it must also match for the delivery to be
1944 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1945 When one of the recipients of a message was unsuitable for the connection
1946 (has no matching addresses), we lost track of needing to mark it
1947 deferred. As a result mail would be lost.
1949 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1951 JH/25 Decoding ACL controls is now done using a binary search; the source code
1952 takes up less space and should be simpler to maintain. Merge the ACL
1953 condition decode tables also, with similar effect.
1955 JH/26 Fix problem with one_time used on a redirect router which returned the
1956 parent address unchanged. A retry would see the parent address marked as
1957 delivered, so not attempt the (identical) child. As a result mail would
1960 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1961 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1962 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1965 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1966 Main config options check_{log,spool}_{inodes,space} are now
1967 100 inodes, 10MB unless set otherwise in the configuration.
1969 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1970 Previously it only applied to the main-section connection policy
1973 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1975 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1976 by me. Added RFC7919 DH primes as an alternative.
1978 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1979 are not in the system include path.
1981 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1982 GnuTLS, when a session startup failed (eg because the client disconnected)
1983 Exim did stdio operations after fclose. This was exposed by a recent
1984 change which nulled out the file handle after the fclose.
1986 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1987 signed directly by the cert-signing cert, rather than an intermediate
1988 OCSP-signing cert. This is the model used by LetsEncrypt.
1990 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1992 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1993 an incoming connection.
1995 HS/02 Bug 1802: Do not half-close the connection after sending a request
1998 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1999 fallback to "prime256v1".
2001 JH/34 SECURITY: Use proper copy of DATA command in error message.
2002 Could leak key material. Remotely exploitable. CVE-2016-9963.
2008 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2009 and 3.4.4 - once the server is enabled to respond to an OCSP request
2010 it does even when not requested, resulting in a stapling non-aware
2011 client dropping the TLS connection.
2013 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2014 support variable-length bit vectors. No functional change.
2016 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2017 The I= interface field on outgoing lines is now after the H= remote
2018 host field, same as incoming lines. There is a separate
2019 outgoing_interface log selector which allows you to disable the
2022 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2023 If not running log_selector +smtp_connection the mainlog would be held
2024 open indefinitely after a "too many connections" event, including to a
2025 deleted file after a log rotate. Leave the per net connection logging
2026 leaving it open for efficiency as that will be quickly detected by the
2027 check on the next write.
2029 HS/01 Bug 1671: Fix post transport crash.
2030 Processing the wait-<transport> messages could crash the delivery
2031 process if the message IDs didn't exist for some reason. When
2032 using 'split_spool_directory=yes' the construction of the spool
2033 file name failed already, exposing the same netto behaviour.
2035 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2036 mime_regex ACL conditions.
2038 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2039 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2040 to HELO, local diagnostic string.
2042 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2043 log line to debug. Even when configured with a tls authenticator many
2044 client connections are expected to not authenticate in this way, so
2045 an authenticate fail is not an error.
2047 HS/02 Add the Exim version string to the process info. This way exiwhat
2048 gives some more detail about the running daemon.
2050 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2051 matter for fast-change records such as DNSBLs.
2053 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2054 retry record, even if constant. There may be multiple transports with
2055 different interface settings and the retry behaviour needs to be kept
2058 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2060 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2062 JH/10 Bug 840: fix log_defer_output option of pipe transport
2064 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2065 hosts_randomize. This matters a lot when combined with mua_wrapper.
2067 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2068 ${quote_pgsql:<string>} operator.
2070 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2072 JH/14 Change header file name for freeradius-client. Relevant if compiling
2073 with Radius support; from the Gentoo tree and checked under Fedora.
2075 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2077 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2078 option transport_filter, meaning no filtering.
2080 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2082 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2083 defaults to "*" (all hosts). The variable is now available when not built
2084 with TLS, default unset, mainly to enable keeping the testsuite sane.
2085 If a server certificate is not supplied (via tls_certificate) an error is
2086 logged, and clients will find TLS connections fail on startup. Presumably
2087 they will retry in-clear.
2088 Packagers of Exim are strongly encouraged to create a server certificate
2089 at installation time.
2091 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2092 with the $config_file variable.
2094 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2095 in transport context, after the attempt, and per-recipient. The latter type
2096 is per host attempted. The event data is the error message, and the errno
2097 information encodes the lookup type (A vs. MX) used for the (first) host,
2098 and the trailing two digits of the smtp 4xx response.
2100 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2101 to write to mainlog (or rejectlog, paniclog) in the window between file
2102 creation and permissions/ownership being changed. Particularly affects
2103 installations where exicyclog is run as root, rather than exim user;
2104 result is that the running daemon panics and dies.
2106 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2108 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2109 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2110 "pri" and "weight". Note that the previous implicit priority given by the
2111 list order is no longer honoured.
2113 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2114 for DKIM processing.
2116 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2117 by defining SUPPORT_SOCKS.
2119 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2120 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2121 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2122 variables are renamed to proxy_{local,external}_{address,port}.
2124 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2125 for a build by defining SUPPORT_I18N
2127 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2128 of the query string, and make ${quote_redis:} do that quoting.
2130 JH/29 Move Events support from Experimental to mainline, enabled by default
2131 and removable for a build by defining DISABLE_EVENT.
2133 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2135 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2136 cached by the daemon.
2138 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2139 by defining LOOKUP_REDIS. The libhiredis library is required.
2141 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2142 keys are given for lookup.
2144 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2145 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2146 only supported when built with TLS support. The PolarSSL SHA routines
2147 are still used when the TLS library is too old for convenient support.
2149 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2150 openssl_options), for security. OpenSSL forces this from version 1.1.0
2151 server-side so match that on older versions.
2153 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2154 allocation for $value could be released as the expansion processing
2155 concluded, but leaving the global pointer active for it.
2157 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2158 and to use the domains and local_parts ACL conditions.
2160 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2161 incorrectly not doubled on cutthrough transmission, hence seen as a
2162 body-termination at the receiving system - resulting in truncated mails.
2163 Commonly the sender saw a TCP-level error, and retransmitted the message
2164 via the normal store-and-forward channel. This could result in duplicates
2165 received - but deduplicating mailstores were liable to retain only the
2166 initial truncated version.
2168 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2170 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2172 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2173 we're in there, support oversigning also; bug 1309.
2175 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2177 HS/04 Add support for keep_environment and add_environment options.
2179 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2180 either intentional arithmetic overflow during PRNG, or testing config-
2183 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2184 delivery resulted in actual delivery. Cancel cutthrough before DATA
2187 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2188 rejects a recipient: pass the reject to the originator.
2190 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2191 Many were false-positives and ignorable, but it's worth fixing the
2194 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2195 for the new environment-manipulation done at startup. Move the routines
2196 from being local to tls.c to being global via the os.c file.
2198 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2199 an extract embedded as result-arg for a map, the first arg for extract
2200 is unavailable so we cannot tell if this is a numbered or keyed
2201 extraction. Accept either.
2207 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2210 JH/02 The smtp transport option "multi_domain" is now expanded.
2212 JH/03 The smtp transport now requests PRDR by default, if the server offers
2215 JH/04 Certificate name checking on server certificates, when exim is a client,
2216 is now done by default. The transport option tls_verify_cert_hostnames
2217 can be used to disable this per-host. The build option
2218 EXPERIMENTAL_CERTNAMES is withdrawn.
2220 JH/05 The value of the tls_verify_certificates smtp transport and main options
2221 default to the word "system" to access the system default CA bundle.
2222 For GnuTLS, only version 3.0.20 or later.
2224 JH/06 Verification of the server certificate for a TLS connection is now tried
2225 (but not required) by default. The verification status is now logged by
2226 default, for both outbound TLS and client-certificate supplying inbound
2229 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2232 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2233 Status Notification (bounce) messages are now MIME format per RFC 3464.
2234 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2235 under the control of the dsn_advertise_hosts option, and routers may
2236 have a dsn_lasthop option.
2238 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2239 default, modifiable by a malware= option. The list separator for
2240 the options can now be changed in the usual way. Bug 68.
2242 JH/10 The smtp_receive_timeout main option is now expanded before use.
2244 JH/11 The incoming_interface log option now also enables logging of the
2245 local interface on delivery outgoing connections.
2247 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2248 if the interface and destination host and port all match.
2250 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2253 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2254 Patch from Andrew Lewis.
2256 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2257 now supports optional time-restrictions, weighting, and priority
2258 modifiers per server. Patch originally by <rommer@active.by>.
2260 JH/16 The spamd_address main option now supports a mixed list of local
2261 and remote servers. Remote servers can be IPv6 addresses, and
2262 specify a port-range.
2264 JH/17 Bug 68: The spamd_address main option now supports an optional
2265 timeout value per server.
2267 JH/18 Bug 1581: Router and transport options headers_add/remove can
2268 now have the list separator specified.
2270 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2273 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2276 JH/21 Support for the A6 type of dns record is withdrawn.
2278 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2279 rather than the verbs used.
2281 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2282 from 255 to 1024 chars.
2284 JH/24 Verification callouts now attempt to use TLS by default.
2286 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2287 are generic router options now. The defaults didn't change.
2289 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2290 Original patch from Alexander Shikoff, worked over by JH.
2292 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2293 files if ZCAT_COMMAND is not executable.
2295 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2297 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2299 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2300 Normally benign, it bites when the pair was led to by a CNAME;
2301 modern usage is to not canonicalize the domain to a CNAME target
2302 (and we were inconsistent anyway for A-only vs AAAA+A).
2304 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2306 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2307 when evaluating $sender_host_dnssec.
2309 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2310 $sender_helo_dnssec variable.
2312 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2314 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2316 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2318 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2319 documented as working, but never had. Support all but $spam_report.
2321 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2322 added for tls authenticator.
2324 HS/03 Add perl_taintmode main config option
2330 TL/01 When running the test suite, the README says that variables such as
2331 no_msglog_check are global and can be placed anywhere in a specific
2332 test's script, however it was observed that placement needed to be near
2333 the beginning for it to behave that way. Changed the runtest perl
2334 script to read through the entire script once to detect and set these
2335 variables, reset to the beginning of the script, and then run through
2336 the script parsing/test process like normal.
2338 TL/02 The BSD's have an arc4random API. One of the functions to induce
2339 adding randomness was arc4random_stir(), but it has been removed in
2340 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2341 function when detected.
2343 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2344 cause callback expansion.
2346 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2347 syntax errors in an expansion can be treated as a string instead of
2348 logging or causing an error, due to the internal use of bool_lax
2349 instead of bool when processing it.
2351 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2352 server certificates when making smtp deliveries.
2354 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2356 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2358 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2360 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2361 Merged patch from Sebastian Wiedenroth.
2363 JH/05 Fix results-pipe from transport process. Several recipients, combined
2364 with certificate use, exposed issues where response data items split
2365 over buffer boundaries were not parsed properly. This eventually
2366 resulted in duplicates being sent. This issue only became common enough
2367 to notice due to the introduction of connection certificate information,
2368 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2370 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2371 size buffer was used, resulting in syntax errors when an expansion
2374 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2375 version 3.3.6 or later.
2377 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2378 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2379 both become "event_action", the variables become $event_name, $event_data
2380 and $event_defer_errno. There is a new variable $verify_mode, usable in
2381 routers, transports and related events. The tls:cert event is now also
2382 raised for inbound connections, if the main configuration event_action
2385 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2386 early OCSP support, but no stapling (appears to be less than 1.0.0).
2388 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2389 server certificate names available under the smtp transport option
2390 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2393 JH/10 Time-related extraction expansions from certificates now use the main
2394 option "timezone" setting for output formatting, and are consistent
2395 between OpenSSL and GnuTLS compilations. Bug 1541.
2397 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2398 encoded parameter in the incoming message. Bug 1558.
2400 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2401 include certificate info, eximon was claiming there were spoolfile
2404 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2406 JH/14 Log delivery-related information more consistently, using the sequence
2407 "H=<name> [<ip>]" wherever possible.
2409 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2410 are problematic for Debian distribution, omit them from the release
2413 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2415 JH/16 Fix string representation of time values on 64bit time_t architectures.
2418 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2419 output list separator was used.
2424 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2425 checkers that were complaining about end of non-void function with no
2428 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2429 This was a regression introduced in 4.83 by another bugfix.
2431 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2433 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2434 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2440 TF/01 Correctly close the server side of TLS when forking for delivery.
2442 When a message was received over SMTP with TLS, Exim failed to clear up
2443 the incoming connection properly after forking off the child process to
2444 deliver the message. In some situations the subsequent outgoing
2445 delivery connection happened to have the same fd number as the incoming
2446 connection previously had. Exim would try to use TLS and fail, logging
2447 a "Bad file descriptor" error.
2449 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2450 utilities have not been installed.
2452 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2453 temporary space as the ACL may create new global variables.
2455 TL/01 LDAP support uses per connection or global context settings, depending
2456 upon the detected version of the libraries at build time.
2458 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2459 to extract and use the src ip:port in logging and expansions as if it
2460 were a direct connection from the outside internet. PPv2 support was
2461 updated based on HAProxy spec change in May 2014.
2463 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2465 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2466 Properly escape header and check for NULL return.
2468 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2471 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2473 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2474 characters in header names, implemented as a verify condition.
2475 Contributed by Michael Fischer v. Mollard.
2477 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2478 results permerror and temperror. Previous values are deprecated but
2479 still accepted. In a future release, err_perm and err_temp will be
2480 completely removed, which will be a backward incompatibility if the
2481 ACL tests for either of these two old results. Patch contributed by
2482 user bes-internal on the mailing list.
2484 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2486 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2487 selectors, in both main and reject logs.
2489 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2492 JH/07 Add malware type "sock" for talking to simple daemon.
2494 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2496 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2497 routers/transports under cutthrough routing.
2499 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2500 numbers. Touch up "bool" conditional to keep the same definition.
2502 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2504 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2506 JH/12 Expand items in router/transport headers_add or headers_remove lists
2507 individually rather than the list as a whole. Bug 1452.
2509 Required for reasonable handling of multiple headers_ options when
2510 they may be empty; requires that headers_remove items with embedded
2511 colons must have them doubled (or the list-separator changed).
2513 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2514 view the policy declared in the DMARC record. Currently, $dmarc_status
2515 is a combined value of both the record presence and the result of the
2518 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2520 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2521 dnslookup router and the smtp transport (applying to the forward
2524 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2525 of ldap servers used for a specific lookup. Patch provided by Heiko
2528 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2529 New variable $lookup_dnssec_authenticated for observability.
2531 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2532 Patch submitted by Lars Timman.
2534 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2536 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2537 Requires trusted mode and valid format message id, aborts otherwise.
2538 Patch contributed by Heiko Schlichting.
2540 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2541 certextract with support for various fields. Bug 1358.
2543 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2544 is requested by default, modifiable by smtp transport option
2547 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2548 operate on certificate variables to give certificate fingerprints
2549 Also new ${sha256:cert_variable}.
2551 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2553 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2554 Christian Aistleitner.
2556 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2558 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2559 file. Patch from Wolfgang Breyha.
2561 JH/25 Expand the coverage of the delivery $host and $host_address to
2562 client authenticators run in verify callout. Bug 1476.
2564 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2565 align with daemon_smtp_ports. Bug 72.
2567 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2568 support and error reporting did not work properly.
2570 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2571 and is readable. Patch from Andrew Colin Kissa.
2573 TL/14 Enhance documentation of ${run expansion and how it parses the
2574 commandline after expansion, particularly in the case when an
2575 unquoted variable expansion results in an empty value.
2577 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2579 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2580 by renaming away from it. Addresses GH issue 15, reported by
2583 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2585 TL/15 SECURITY: prevent double expansion in math comparison functions
2586 (can expand unsanitized data). Not remotely exploitable.
2593 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2595 PP/02 Make -n do something, by making it not do something.
2596 When combined with -bP, the name of an option is not output.
2598 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2601 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2602 $sender_host_name and config options to manage this, and basic check
2605 PP/05 DSCP support for outbound connections and control modifier for inbound.
2607 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2608 (Only plugin which currently uses this is kerberos4, which nobody should
2609 be using, but we should make it available and other future plugins might
2610 conceivably use it, even though it would break NAT; stuff *should* be
2611 using channel bindings instead).
2613 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2614 name; added for Sendmail compatibility; requires admin caller.
2615 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2616 just ignore it); requires trusted caller.
2617 Also parse but ignore: -Ac -Am -X<logfile>
2620 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2622 TL/02 Add +smtp_confirmation as a default logging option.
2624 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2625 Patch by Magnus Holmgren from 2007-02-20.
2627 TL/04 Bugzilla 1281 - Spec typo.
2628 Bugzilla 1283 - Spec typo.
2629 Bugzilla 1290 - Spec grammar fixes.
2631 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2633 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2635 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2636 mailing list by by Dmitry Isaikin.
2638 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2640 JH/02 Support "G" suffix to numbers in ${if comparisons.
2642 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2644 NM/01 Bugzilla 1197 - Spec typo
2645 Bugzilla 1196 - Spec examples corrections
2647 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2649 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2650 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2653 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2654 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2656 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2657 "acl {{name}{arg}...}", and optional args on acl condition
2660 JH/05 Permit multiple router/transport headers_add/remove lines.
2662 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2664 JH/07 Avoid using a waiting database for a single-message-only transport.
2665 Performance patch from Paul Fisher. Bugzilla 1262.
2667 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2670 JH/09 Add $headers_added variable, with content from use of ACL modifier
2671 add_header (but not yet added to the message). Bugzilla 199.
2673 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2674 Pulled from Bugzilla 817 by Wolfgang Breyha.
2676 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2678 (nb: this is the same fix as in Exim 4.80.1)
2680 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2683 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2686 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2688 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2689 advertise SMTP AUTH mechanism to us, instead of a generic
2690 protocol violation error. Also, make Exim more robust to bad
2691 data from the Dovecot auth socket.
2693 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2695 When a queue runner is handling a message, Exim first routes the
2696 recipient addresses, during which it prunes them based on the retry
2697 hints database. After that it attempts to deliver the message to
2698 any remaining recipients. It then updates the hints database using
2701 So if a recipient address works intermittently, it can get repeatedly
2702 deferred at routing time. The retry hints record remains fresh so the
2703 address never reaches the final cutoff time.
2705 This is a fairly common occurrence when a user is bumping up against
2706 their storage quota. Exim had some logic in its local delivery code
2707 to deal with this. However it did not apply to per-recipient defers
2708 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2710 This change adds a proper retry rule check during routing so that the
2711 final cutoff time is checked against the message's age. We only do
2712 this check if there is an address retry record and there is not a
2713 domain retry record; this implies that previous attempts to handle
2714 the address had the retry_use_local_parts option turned on. We use
2715 this as an approximation for the destination being like a local
2716 delivery, as in LMTP.
2718 I suspect this new check makes the old local delivery cutoff check
2719 redundant, but I have not verified this so I left the code in place.
2721 TF/02 Correct gecos expansion when From: is a prefix of the username.
2723 Test 0254 submits a message to Exim with the header
2727 When I ran the test suite under the user fanf2, Exim expanded
2728 the header to contain my full name, whereas it should have added
2729 a Resent-Sender: header. It erroneously treats any prefix of the
2730 username as equal to the username.
2732 This change corrects that bug.
2734 GF/01 DCC debug and logging tidyup
2735 Error conditions log to paniclog rather than rejectlog.
2736 Debug lines prefixed by "DCC: " to remove any ambiguity.
2738 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2740 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2741 Bug spotted by Jeremy Harris; was flawed since initial commit.
2742 Would have resulted in OCSP responses post-SNI triggering an Exim
2743 NULL dereference and crash.
2745 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2747 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2748 Bug detection, analysis and fix by Samuel Thibault.
2749 Bugzilla 1331, Debian bug #698092.
2751 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2753 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2754 Server implementation by Todd Lyons, client by JH.
2755 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2756 config variable "prdr_enable" controls whether the server
2757 advertises the facility. If the client requests PRDR a new
2758 acl_data_smtp_prdr ACL is called once for each recipient, after
2759 the body content is received and before the acl_smtp_data ACL.
2760 The client is controlled by both of: a hosts_try_prdr option
2761 on the smtp transport, and the server advertisement.
2762 Default client logging of deliveries and rejections involving
2763 PRDR are flagged with the string "PRDR".
2765 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2766 fclose(). Diagnosis by Todd Lyons.
2768 PP/17 Update configure.default to handle IPv6 localhost better.
2769 Patch by Alain Williams (plus minor tweaks).
2772 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2773 This is now consistent with GnuTLS, and is now documented: the
2774 previous undocumented portable approach to treating the option as
2775 unset was to force an expansion failure. That still works, and
2776 an empty string is now equivalent.
2778 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2779 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2780 not performing validation itself.
2782 PP/20 Added force_command boolean option to pipe transport.
2783 Patch from Nick Koston, of cPanel Inc.
2785 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2788 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2790 PP/21 Fix eximon continuous updating with timestamped log-files.
2791 Broken in a format-string cleanup in 4.80, missed when I repaired the
2792 other false fix of the same issue.
2793 Report and fix from Heiko Schlichting.
2796 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2797 Report from Prashanth Katuri.
2799 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2800 It's SecureTransport, so affects any MacOS clients which use the
2801 system-integrated TLS libraries, including email clients.
2803 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2804 using a MIME ACL for non-SMTP local injection.
2805 Report and assistance in diagnosis by Warren Baker.
2807 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2809 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2811 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2812 last id that failed so it may be referenced in subsequent ACL's.
2814 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2817 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2818 ldap library initialization, allowing self-signed CA's to be
2819 used. Also properly sets require_cert option later in code by
2820 using NULL (global ldap config) instead of ldap handle (per
2821 session). Bug diagnosis and testing by alxgomz.
2823 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2824 the src/util/ subdirectory.
2826 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2827 renamed to Transport Post Delivery Action by Jeremy Harris, as
2830 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2831 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2832 redis_servers = needs to be configured which will be used by the redis
2833 lookup. Patch from Warren Baker, of The Packet Hub.
2835 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2837 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2838 hostname or reverse DNS when processing a host list. Used suggestions
2839 from multiple comments on this bug.
2841 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2843 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2844 Missed a few lines, added it to make the runtest require no keyboard
2847 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2848 contains upper case chars. Make router use caseful_local_part.
2850 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2851 support when GnuTLS has been built with p11-kit.
2857 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2859 This, or similar/improved, will also be change PP/11 of 4.82.
2865 PP/01 Handle short writes when writing local log-files.
2866 In practice, only affects FreeBSD (8 onwards).
2867 Bugzilla 1053, with thanks to Dmitry Isaikin.
2869 NM/01 Bugzilla 949 - Documentation tweak
2871 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2874 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2876 PP/02 Implemented gsasl authenticator.
2878 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2880 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2881 `pkg-config foo` for cflags/libs.
2883 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2884 with rest of GSASL and with heimdal_gssapi.
2886 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2887 `pkg-config foo` for cflags/libs for the TLS implementation.
2889 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2890 properties get this fed in as external SSF. A number of robustness
2891 and debugging improvements to the cyrus_sasl authenticator.
2893 PP/08 cyrus_sasl server now expands the server_realm option.
2895 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2896 Patch by Jeremy Harris.
2898 PP/10 Added dbmjz lookup type.
2900 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2902 PP/12 MAIL args handles TAB as well as SP, for better interop with
2903 non-compliant senders.
2904 Analysis and variant patch by Todd Lyons.
2906 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2907 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2908 Patch from Dirk Mueller <dmueller@suse.com>
2910 PP/13 tls_peerdn now print-escaped for spool files.
2911 Observed some $tls_peerdn in wild which contained \n, which resulted
2912 in spool file corruption.
2914 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2915 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2916 or write after TLS renegotiation, which otherwise led to messages
2919 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2920 as a tracking header (ie: a signed header comes before the signature).
2921 Patch from Wolfgang Breyha.
2923 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2924 comma-sep list; embedded commas doubled.
2926 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2928 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2930 Report and patch from Dmitry Banschikov.
2932 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2933 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2934 failures. We appear to now support TLS1.1+ with Exim.
2936 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2937 lets Exim select keys and certificates based upon TLS SNI from client.
2938 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2939 before an outbound SMTP session. New log_selector, +tls_sni.
2941 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2942 NULL dereference. Report and patch from Alun Jones.
2944 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2945 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2946 Not seeing resolver debug output on NetBSD, but suspect this is a
2947 resolver implementation change.
2949 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2950 Left warnings. Added "eximon gdb" invocation mode.
2952 PP/21 Defaulting "accept_8bitmime" to true, not false.
2954 PP/22 Added -bw for inetd wait mode support.
2956 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2957 locate the relevant includes and libraries. Made this the default.
2959 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2960 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2962 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2963 This may cause build issues on older platforms.
2965 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2966 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2967 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2968 Added SNI support via GnuTLS too.
2969 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2971 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2973 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2974 Applied second patch from Janne, implementing suggestion to default
2975 multiple-strings-in-record handling to match SPF spec.
2977 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2979 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2980 read-only, out of scope).
2981 Patch from Wolfgang Breyha, report from Stuart Northfield.
2983 PP/29 Fix three issues highlighted by clang analyser static analysis.
2984 Only crash-plausible issue would require the Cambridge-specific
2985 iplookup router and a misconfiguration.
2986 Report from Marcin Mirosław.
2988 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2990 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2991 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2992 As part of this, removing so much warning spew let me fix some minor
2993 real issues in debug logging.
2995 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2996 assignment on my part. Fixed.
2998 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2999 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3000 Janne Snabb (who went above and beyond: thank you).
3002 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3003 string otherwise requires a connection and a bunch more work and it's
3004 relatively easy to get wrong. Should also expose TLS library linkage
3007 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3008 64-bit ${eval} (JH/03).
3010 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3011 GNU libc to support some of the 64-bit stuff, should not lead to
3012 conflicts. Defined before os.h is pulled in, so if a given platform
3013 needs to override this, it can.
3015 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3016 protection layer was required, which is not implemented.
3017 Bugzilla 1254, patch from Wolfgang Breyha.
3019 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3020 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3021 tls_dhparam take prime identifiers. Also unbreak combination of
3022 OpenSSL+DH_params+TLSSNI.
3024 PP/39 Disable SSLv2 by default in OpenSSL support.
3030 PP/01 Solaris build fix for Oracle's LDAP libraries.
3031 Bugzilla 1109, patch from Stephen Usher.
3033 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3035 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3038 TF/02 Fix a couple more cases where we did not log the error message
3039 when unlink() failed. See also change 4.74-TF/03.
3041 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3042 lock up or crash if it happened to be inside a call to libc when it
3043 got a SIGUSR1 from exiwhat.
3045 The SIGUSR1 handler appends the current process status to the process
3046 log which is later printed by exiwhat. It used to use the general
3047 purpose logging code to do this, but several functions it calls are
3048 not safe for signals.
3050 The new output code in the SIGUSR1 handler is specific to the process
3051 log, and simple enough that it's easy to inspect for signal safety.
3052 Removing some special cases also simplifies the general logging code.
3053 Removing the spurious timestamps from the process log simplifies
3056 TF/04 Improved ratelimit ACL condition.
3058 The /noupdate option has been deprecated in favour of /readonly which
3059 has clearer semantics. The /leaky, /strict, and /readonly update modes
3060 are mutually exclusive. The update mode is no longer included in the
3061 database key; it just determines when the database is updated. (This
3062 means that when you upgrade Exim will forget old rate measurements.)
3064 Exim now checks that the per_* options are used with an update mode that
3065 makes sense for the current ACL. For example, when Exim is processing a
3066 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3067 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3068 must specify per_mail/readonly. If you omit the update mode it defaults to
3069 /leaky where that makes sense (as before) or /readonly where required.
3071 The /noupdate option is now undocumented but still supported for
3072 backwards compatibility. It is equivalent to /readonly except that in
3073 ACLs where /readonly is required you may specify /leaky/noupdate or
3074 /strict/noupdate which are treated the same as /readonly.
3076 A useful new feature is the /count= option. This is a generalization
3077 of the per_byte option, so that you can measure the throughput of other
3078 aggregate values. For example, the per_byte option is now equivalent
3079 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3081 The per_rcpt option has been generalized using the /count= mechanism
3082 (though it's more complicated than the per_byte equivalence). When it is
3083 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3084 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3085 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3086 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3087 Note that using per_rcpt with a non-readonly update mode in more than
3088 one ACL will cause the recipients to be double-counted. (The per_mail
3089 and per_byte options don't have this problem.)
3091 The handling of very low rates has changed slightly. If the computed rate
3092 is less than the event's count (usually one) then this event is the first
3093 after a long gap. In this case the rate is set to the same as this event's
3094 count, so that the first message of a spam run is counted properly.
3096 The major new feature is a mechanism for counting the rate of unique
3097 events. The new per_addr option counts the number of different
3098 recipients that someone has sent messages to in the last time period. It
3099 behaves like per_rcpt if all the recipient addresses are different, but
3100 duplicate recipient addresses do not increase the measured rate. Like
3101 the /count= option this is a general mechanism, so the per_addr option
3102 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3103 example, measure the rate that a client uses different sender addresses
3104 with the options per_mail/unique=$sender_address. There are further
3105 details in the main documentation.
3107 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3109 TF/06 Removed a few PCRE remnants.
3111 TF/07 Automatically extract Exim's version number from tags in the git
3112 repository when doing development or release builds.
3114 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3115 Bugzilla 879. Patch from Paul Fisher.
3117 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3118 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3121 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3123 PP/05 Variable $av_failed, true if the AV scanner deferred.
3124 Bugzilla 1078. Patch from John Horne.
3126 PP/06 Stop make process more reliably on build failure.
3127 Bugzilla 1087. Patch from Heiko Schlittermann.
3129 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3130 Bugzilla 1089. Patch from Heiko Schlittermann.
3132 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3133 Bugzilla 1131. Patch from Holger Weiß.
3135 PP/09 Handle IPv6 addresses with SPF.
3136 Bugzilla 860. Patch from Wolfgang Breyha.
3138 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3140 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3143 PP/11 match_* no longer expand right-hand-side by default.
3144 New compile-time build option, EXPAND_LISTMATCH_RHS.
3145 New expansion conditions, "inlist", "inlisti".
3147 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3149 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3151 PP/14 fix log_write() format string regression from TF/03.
3152 Bugzilla 1152. Patch from Dmitry Isaikin.
3158 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3160 PP/02 Harmonised TLS library version reporting; only show if debugging.
3161 Layout now matches that introduced for other libraries in 4.74 PP/03.
3163 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3165 PP/04 New "dns_use_edns0" global option.
3167 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3170 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3171 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3173 TK/01 Updated PolarSSL code to 0.14.2.
3174 Bugzilla 1097. Patch from Andreas Metzler.
3176 PP/07 Catch divide-by-zero in ${eval:...}.
3177 Fixes bugzilla 1102.
3179 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3182 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3183 format-string attack -- SECURITY: remote arbitrary code execution.
3185 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3186 time unintentionally subject to list matching rules, letting the header
3187 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3188 arbitrary string expansion). This allowed for information disclosure.
3190 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3191 INT_MIN/-1 -- value coerced to INT_MAX.
3197 NM/01 Workaround for PCRE version dependency in version reporting
3200 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3201 This fixes portability to compilers other than gcc, notably
3202 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3204 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3205 makefiles for portability to HP-UX and POSIX correctness.
3207 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3208 Also via indented variable definition in the Makefile.
3209 (Debugging by Oliver Heesakkers).
3211 PP/02 Restore caching of spamd results with expanded spamd_address.
3212 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3214 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3215 Improves build reliability. Fix from: Frank Elsner
3217 NM/02 Fix wide character breakage in the rfc2047 coding
3218 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3220 NM/03 Allow underscore in dnslist lookups
3221 Fixes bug 1026. Patch from Graeme Fowler
3223 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3224 Code patches from Adam Ciarcinski of NetBSD.
3226 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3229 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3230 is logged, to avoid truncation. Patch from John Horne.
3232 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3233 Patch from Jakob Hirsch.
3235 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3236 SQL string expansion failure details.
3237 Patch from Andrey Oktyabrski.
3239 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3240 Patch from Simon Arlott.
3242 PP/09 New lookups functionality failed to compile on old gcc which rejects
3243 extern declarations in function scope.
3244 Patch from Oliver Fleischmann
3246 PP/10 Use sig_atomic_t for flags set from signal handlers.
3247 Check getgroups() return and improve debugging.
3248 Fixed developed for diagnosis in bug 927 (which turned out to be
3251 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3252 Patch from Mark Zealey.
3254 PP/12 Bugzilla 1056: Improved spamd server selection.
3255 Patch from Mark Zealey.
3257 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3258 Based on patch from Heiko Schlittermann.
3260 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3261 Patch from Uwe Doering, sign-off by Michael Haardt.
3263 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3264 variable declaration deep within a block. Bug and patch from
3267 PP/15 lookups-Makefile IRIX compatibility coercion.
3269 PP/16 Make DISABLE_DKIM build knob functional.
3271 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3272 Patch by Simon Arlott
3274 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3275 variable argument macros. Our copy now differs from upstream.
3281 TF/01 Failure to get a lock on a hints database can have serious
3282 consequences so log it to the panic log.
3284 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3285 controlled using the smtp_confirmation log selector.
3287 TF/03 Include the error message when we fail to unlink a spool file.
3289 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3290 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3291 for maintaining out-of-tree patches for some time.
3293 PP/01 Bugzilla 139: Documentation and portability issues.
3294 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3295 Handle per-OS dynamic-module compilation flags.
3297 PP/02 Let /dev/null have normal permissions.
3298 The 4.73 fixes were a little too stringent and complained about the
3299 permissions on /dev/null. Exempt it from some checks.
3300 Reported by Andreas M. Kirchwitz.
3302 PP/03 Report version information for many libraries, including
3303 Exim version information for dynamically loaded libraries. Created
3304 version.h, now support a version extension string for distributors
3305 who patch heavily. Dynamic module ABI change.
3307 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3308 privilege escalation vulnerability whereby the Exim run-time user
3309 can cause root to append content of the attacker's choosing to
3312 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3315 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3316 If dropping privileges for untrusted macros, we disabled normal logging
3317 on the basis that it would fail; for the Exim run-time user, this is not
3318 the case, and it resulted in successful deliveries going unlogged.
3319 Fixed. Reported by Andreas Metzler.
3325 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3326 only prepend for the Resent-* case. Fixes regression introduced in
3327 Exim 4.70 by NM/22 for Bugzilla 607.
3329 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3330 increasing numbers of administrators be bitten by this.
3332 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3334 PP/03 Bugzilla 994: added openssl_options main configuration option.
3336 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3338 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3340 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3342 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3343 without a peer certificate, leading to a segfault because of an
3344 assumption that peers always have certificates. Be a little more
3345 paranoid. Problem reported by Martin Tscholak.
3347 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3348 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3349 NB: ClamAV planning to remove STREAM in "middle of 2010".
3350 CL also introduces -bmalware, various -d+acl logging additions and
3351 more caution in buffer sizes.
3353 PP/09 Implemented reverse_ip expansion operator.
3355 PP/10 Bugzilla 937: provide a "debug" ACL control.
3357 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3359 PP/12 Bugzilla 973: Implement --version.
3361 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3363 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3365 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3367 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3368 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3369 ignore trailing whitespace.
3371 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3373 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3376 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3377 Notification from Dr Andrew Aitchison.
3379 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3380 ExtendedDetectionInfo response format.
3381 Notification from John Horne.
3383 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3386 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3387 XSL and documented dependency on system catalogs, with examples of how
3390 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3393 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3394 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3395 configuration file which is writeable by the Exim user or group.
3397 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3398 of configuration files to cover files specified with the -C option if
3399 they are going to be used with root privileges, not just the default
3402 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3403 option (effectively making it always true).
3405 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3406 files to be used while preserving root privileges.
3408 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3409 that rogue child processes cannot use them.
3411 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3412 run-time user, instead of root.
3414 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3415 Exim run-time user without dropping privileges.
3417 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3418 result string, instead of calling string_vformat() twice with the same
3421 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3422 for other users. Others should always drop root privileges if they use
3423 -C on the command line, even for a whitelisted configure file.
3425 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3427 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3433 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3434 $data_path, and $header_path variables; fixed documentation bugs and
3437 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3438 exipick to access non-standard spools, including the "frozen" queue
3441 NM/01 Bugzilla 965: Support mysql stored procedures.
3442 Patch from Alain Williams
3444 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3446 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3447 Patch from Andreas Metzler
3449 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3450 Patch from Kirill Miazine
3452 NM/05 Bugzilla 671: Added umask to procmail example.
3454 JJ/03 installed exipick 20100323.0, fixing doc bug
3456 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3457 directory. Notification and patch from Dan Rosenberg.
3459 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3461 TK/02 Improve log output when DKIM signing operation fails.
3463 MH/01 Treat the transport option dkim_domain as a colon separated
3464 list, not as a single string, and sign the message with each element,
3465 omitting multiple occurences of the same signer.
3467 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3468 Bugzilla 985, 986. Patch by Simon Arlott
3470 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3471 Patch by Simon Arlott
3473 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3474 MBX locking. Notification from Dan Rosenberg.
3480 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3482 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3484 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3486 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3488 NM/04 Bugzilla 915: Flag broken perl installation during build.
3494 TK/01 Added patch by Johannes Berg that expands the main option
3495 "spamd_address" if it starts with a dollar sign.
3497 TK/02 Write list of recipients to X-Envelope-Sender header when building
3498 the mbox-format spool file for content scanning (suggested by Jakob
3501 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3502 (http://www.dcc-servers.net/) support via dccifd. Activated by
3503 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3505 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3506 by Mark Daniel Reidel <mr@df.eu>.
3508 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3509 When building exim an external PCRE library is now needed -
3510 PCRE is a system library on the majority of modern systems.
3511 See entry on PCRE_LIBS in EDITME file.
3513 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3514 conversation. Added nologin parameter to request.
3515 Patch contributed by Kirill Miazine.
3517 TF/01 Do not log submission mode rewrites if they do not change the address.
3519 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3521 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3522 log files in place. Contributed by Roberto Lima.
3524 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3526 TF/03 Bugzilla 615: When checking the local_parts router precondition
3527 after a local_part_suffix or local_part_prefix option, Exim now
3528 does not use the address's named list lookup cache, since this
3529 contains cached lookups for the whole local part.
3531 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3532 Robert Millan. Documentation is in experimental-spec.txt.
3534 TF/04 Bugzilla 668: Fix parallel build (make -j).
3536 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3538 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3539 Patch provided by Jan Srzednicki.
3541 TF/05 Leading white space used to be stripped from $spam_report which
3542 wrecked the formatting. Now it is preserved.
3544 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3545 that they are available at delivery time.
3547 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3549 TF/08 TLS error reporting now respects the incoming_interface and
3550 incoming_port log selectors.
3552 TF/09 Produce a more useful error message if an SMTP transport's hosts
3553 setting expands to an empty string.
3555 NM/06 Bugzilla 744: EXPN did not work under TLS.
3556 Patch provided by Phil Pennock.
3558 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3559 Patch provided by Richard Godbee.
3561 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3562 acl_smtp_notquit, added index entry.
3564 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3565 Patch provided by Eugene Bujak.
3567 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3568 accept(). Patch provided by Maxim Dounin.
3570 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3571 Patch provided by Phil Pennock.
3573 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3575 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3576 Patch provided by Brad "anomie" Jorsch.
3578 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3579 Patch provided by Dean Brooks.
3581 TK/05 Add native DKIM support (does not depend on external libraries).
3583 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3584 Patch provided by Graeme Fowler.
3586 NM/16 Bugzilla 851: Documentation example syntax fix.
3588 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3590 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3593 NM/19 Bugzilla 745: TLS version reporting.
3594 Patch provided by Phil Pennock.
3596 NM/20 Bugzilla 167: bool: condition support.
3597 Patch provided by Phil Pennock.
3599 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3600 clients. Patch provided by Phil Pennock.
3602 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3603 Patch provided by Brad "anomie" Jorsch.
3605 NM/23 Bugzilla 687: Fix misparses in eximstats.
3606 Patch provided by Heiko Schlittermann.
3608 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3609 Patch provided by Heiko Schlittermann.
3611 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3612 plus update to original patch.
3614 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3616 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3617 Patch provided by David Brownlee.
3619 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3621 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3623 NM/30 Bugzilla 888: TLS documentation bugfixes.
3625 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3627 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3628 Unlike the original bugzilla I have changed all shell scripts in src tree.
3630 NM/33 Bugzilla 898: Transport filter timeout fix.
3631 Patch by Todd Rinaldo.
3633 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3634 Patch by Serge Demonchaux.
3636 NM/35 Bugzilla 39: Base64 decode bug fixes.
3637 Patch by Jakob Hirsch.
3639 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3641 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3643 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3645 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3651 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3652 ALT-N's libdkim that I have put here:
3653 http://duncanthrax.net/exim-experimental/
3655 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3656 were called 'true' and it seems that C99 defines that as a reserved
3657 keyword to be used with 'bool' variable types. That means you could
3658 not include C99-style headers which use bools without triggering
3659 build errors in sieve.c.
3661 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3662 as mailq or other aliases. Changed the --help handling significantly
3663 to do whats expected. exim_usage() emits usage/help information.
3665 SC/01 Added the -bylocaldomain option to eximstats.
3667 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3669 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3671 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3677 PH/01 Another patch from the Sieve maintainer.
3679 PH/02 When an IPv6 address is converted to a string for single-key lookup
3680 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3681 dots are used instead of colons so that keys in lsearch files need not
3682 contain colons. This was done some time before quoting was made available
3683 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3684 (notated using the quote facility) so as to distinguish them from IPv4
3685 keys. This meant that lookups for IP addresses in host lists did not work
3686 for iplsearch lookups.
3688 This has been fixed by arranging for IPv6 addresses to be expressed with
3689 colons if the lookup type is iplsearch. This is not incompatible, because
3690 previously such lookups could never work.
3692 The situation is now rather anomalous, since one *can* have colons in
3693 ordinary lsearch keys. However, making the change in all cases is
3694 incompatible and would probably break a number of configurations.
3696 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3699 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3700 conversion specification without a maximum field width, thereby enabling
3701 a rogue spamd server to cause a buffer overflow. While nobody in their
3702 right mind would setup Exim to query an untrusted spamd server, an
3703 attacker that gains access to a server running spamd could potentially
3704 exploit this vulnerability to run arbitrary code as the Exim user.
3706 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3707 $primary_hostname instead of what libspf2 thinks the hosts name is.
3709 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3710 a directory entry by the name of the lookup key. Previously, if a
3711 symlink pointed to a non-existing file or a file in a directory that
3712 Exim lacked permissions to read, a lookup for a key matching that
3713 symlink would fail. Now it is enough that a matching directory entry
3714 exists, symlink or not. (Bugzilla 503.)
3716 PH/03 The body_linecount and body_zerocount variables are now exported in the
3719 PH/04 Added the $dnslist_matched variable.
3721 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3722 This means they are set thereafter only if the connection becomes
3725 PH/06 Added the client_condition to authenticators so that some can be skipped
3726 by clients under certain conditions.
3728 PH/07 The error message for a badly-placed control=no_multiline_responses left
3729 "_responses" off the end of the name.
3731 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3733 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3734 (without spaces) instead of just copying the configuration text.
3736 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3738 PH/11 Added $max_received_linelength.
3740 PH/12 Added +ignore_defer and +include_defer to host lists.
3742 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3743 way in which PCRE > 7.0 is built.
3745 PH/14 Implemented queue_only_load_latch.
3747 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3748 MAIL command. The effect was to mangle the value on 64-bit systems.
3750 PH/16 Another patch from the Sieve maintainer.
3752 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3754 PH/18 If a system quota error occurred while trying to create the file for
3755 a maildir delivery, the message "Mailbox is full" was not appended to the
3756 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3757 only to a quota excession during the actual writing of the file.
3759 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3760 characters?) which causes problems in log lines. The DN values are now
3761 passed through string_printing() before being added to log lines.
3763 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3764 and InterBase are left for another time.)
3766 PH/21 Added message_body_newlines option.
3768 PH/22 Guard against possible overflow in moan_check_errorcopy().
3770 PH/23 POSIX allows open() to be a macro; guard against that.
3772 PH/24 If the recipient of an error message contained an @ in the local part
3773 (suitably quoted, of course), incorrect values were put in $domain and
3774 $local_part during the evaluation of errors_copy.
3780 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3781 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3784 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3785 issue a MAIL command.
3787 PH/02 In an ACL statement such as
3789 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3791 if a client was not listed at all, or was listed with a value other than
3792 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3793 the condition was not true (as it should be), so access was not denied.
3794 The bug was that the ! inversion was incorrectly passed on to the second
3795 item. This has been fixed.
3797 PH/03 Added additional dnslists conditions == and =& which are different from
3798 = and & when the dns lookup returns more than one IP address.
3800 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3801 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3803 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3804 FSYNC, which compiles an option called disable_fsync that allows for
3805 bypassing fsync(). The documentation is heavily laced with warnings.
3807 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3809 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3810 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3811 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3812 including adding "make clean"; (3) Added -fPIC when compiling the test
3813 dynamically loaded module, to get rid of a warning.
3815 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3816 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3817 = 0s. The bug is otherwise harmless.
3819 PH/07 There was a bug in the dovecot authenticator such that the value of
3820 $auth1 could be overwritten, and so not correctly preserved, after a
3821 successful authentication. This usually meant that the value preserved by
3822 the server_setid option was incorrect.
3824 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3826 PH/09 Installed PCRE release 7.0.
3828 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3829 run for batched SMTP input. It is now run at the start of every message
3830 in the batch. While fixing this I discovered that the process information
3831 (output by running exiwhat) was not always getting set for -bs and -bS
3832 input. This is fixed, and it now also says "batched" for BSMTP.
3834 PH/11 Added control=no_pipelining.
3836 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3837 patch, slightly modified), and move the expansion of helo_data till after
3838 the connection is made in the smtp transport (so it can use these
3841 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3843 PH/14 Added log_selector = +pid.
3845 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3847 PH/16 Add ${if forany and ${if forall.
3849 PH/17 Added dsn_from option to vary the From: line in DSNs.
3851 PH/18 Flush SMTP output before performing a callout, unless control =
3852 no_callout_flush is set.
3854 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3855 was true (the default) a successful delivery failed to delete the retry
3856 item, thus causing premature timeout of the address. The bug is now
3859 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3861 PH/21 Long custom messages for fakedefer and fakereject are now split up
3862 into multiline responses in the same way that messages for "deny" and
3863 other ACL rejections are.
3865 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3866 with slight modification.
3868 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3869 draft, changing the syntax and factoring some duplicate code".
3871 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3872 for deliveries of the second and subsequent messages over the same SMTP
3875 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3876 ${reduce, with only minor "tidies".
3878 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3880 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3881 expansion side effects.
3883 PH/27 When a message times out after an over-quota error from an Exim-imposed
3884 quota, the bounce message says "mailbox is full". This message was not
3885 being given when it was a system quota that was exceeded. It now should
3888 MH/03 Made $recipients available in local_scan(). local_scan() already has
3889 better access to the recipient list through recipients_list[], but
3890 $recipients can be useful in postmaster-provided expansion strings.
3892 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3893 in the case of a MAIL command with additional options following the
3894 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3895 were accidentally chopped off.
3897 PH/29 SMTP synchronization checks are implemented when a command is read -
3898 there is a check that no more input is waiting when there shouldn't be
3899 any. However, for some commands, a delay in an ACL can mean that it is
3900 some time before the response is written. In this time, more input might
3901 arrive, invalidly. So now there are extra checks after an ACL has run for
3902 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3903 pipelining has not been advertised.
3905 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3907 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3908 returned in the SMTP response when smtp_return_error_details was false.
3909 This has been fixed.
3911 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3912 instead of the C I/O that was originally supplied, because problems were
3913 reported on Solaris.
3915 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3916 Exim which did not show up earlier: it was assuming that a call to
3917 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3918 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3919 was a macro that became an assignment, so it seemed to work. This has
3920 changed to a proper function call with a void return, hence the compile
3921 error. Exim's code has been fixed.
3923 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3926 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3928 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3930 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3931 because it turns out to be too big or there is a timeout) there is no
3932 "Completed" line in the log. When some messages of this type were
3933 selected by exigrep, they were listed as "not completed". Others were
3934 picked up by some special patterns. I have improved the selection
3935 criteria to be more general.
3937 PH/38 The host_find_failed option in the manualroute router can now be set
3938 to "ignore", to completely ignore a host whose IP address cannot be
3939 found. If all hosts are ignored, the behaviour is controlled by the new
3940 host_all_ignored option.
3942 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3943 homing or because of multiple MX records with /mx) generated more than
3944 one IP address, and the following item turned out to be the local host,
3945 all the secondary addresses of the first item were incorrectly removed
3946 from the list, along with the local host and any following hosts (which
3947 is what is supposed to happen).
3949 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3950 whoever called Exim into the -H file. In the case of the daemon it was
3951 behaving confusingly. When first started, it used values for whoever
3952 started the daemon, but after a SIGHUP it used the Exim user (because it
3953 calls itself on a restart). I have changed the code so that it now always
3956 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3957 message are rejected with the same error (e.g. no authentication or bad
3958 sender address), and a DATA command is nevertheless sent (as can happen
3959 with PIPELINING or a stupid MUA), the error message that was given to the
3960 RCPT commands is included in the rejection of the DATA command. This is
3961 intended to be helpful for MUAs that show only the final error to their
3964 PH/42 Another patch from the Sieve maintainer.
3966 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3967 Eximstats - Fixed some broken HTML links and added missing column headers
3969 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3970 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3972 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3974 PH/43 Yet another patch from the Sieve maintainer.
3976 PH/44 I found a way to check for a TCP/IP connection going away before sending
3977 the response to the final '.' that terminates a message, but only in the
3978 case where the client has not sent further data following the '.'
3979 (unfortunately, this is allowed). However, in many cases there won't be
3980 any further data because there won't be any more messages to send. A call
3981 to select() can be used: if it shows that the input is "ready", there is
3982 either input waiting, or the socket has been closed. An attempt to read
3983 the next input character can distinguish the two cases. Previously, Exim
3984 would have sent an OK response which the client would never have see.
3985 This could lead to message repetition. This fix should cure that, at
3986 least in a lot of common cases.
3988 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3989 advertised in response to EHLO.
3995 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3996 fixed by 4.65/MH/01 (is this a record?) are fixed:
3998 (i) An empty string was always treated as zero by the numeric comparison
3999 operators. This behaviour has been restored.
4001 (ii) It is documented that the numeric comparison operators always treat
4002 their arguments as decimal numbers. This was broken in that numbers
4003 starting with 0 were being interpreted as octal.
4005 While fixing these problems I realized that there was another issue that
4006 hadn't been noticed. Values of message_size_limit (both the global option
4007 and the transport option) were treated as octal if they started with 0.
4008 The documentation was vague. These values are now always treated as
4009 decimal, and I will make that clear in the documentation.
4015 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4016 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4019 MH/01 Don't check that the operands of numeric comparison operators are
4020 integers when their expansion is in "skipping" mode (fixes bug
4021 introduced by 4.64-PH/07).
4023 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4024 child addresses, Exim now panics and dies. Previously, because the count
4025 is held in a short int, deliveries were likely to be lost. As such a
4026 large number of recipients for a single message is ridiculous
4027 (performance will be very, very poor), I have chosen to impose a limit
4028 rather than extend the field.
4034 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4035 leftover -K file (the existence of which was triggered by #402).
4036 While we were at it, introduced process PID as part of the -K
4037 filename. This should rule out race conditions when creating
4040 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4041 processing considerably. Previous code took too long for large mails,
4042 triggering a timeout which in turn triggers #401.
4044 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4045 in the DK code in transports.c. sendfile() is not really portable,
4046 hence the _LINUX specificness.
4048 TF/01 In the add_headers option to the mail command in an Exim filter,
4049 there was a bug that Exim would claim a syntax error in any
4050 header after the first one which had an odd number of characters
4053 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4054 callout verification, Exim cached a "reject" for the entire domain. This
4055 is correct for most verifications, but it is not correct for a recipient
4056 verification with use_sender or use_postmaster set, because in that case
4057 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4058 case of MAIL FROM:<> rejection from other early rejections (e.g.
4059 rejection of HELO). When verifying a recipient using a non-null MAIL
4060 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4061 Whatever the result of the callout, the value of the domain cache is
4062 left unchanged (for any other kind of callout, getting as far as trying
4063 RCPT means that the domain itself is ok).
4065 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4068 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4069 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4070 session, the daemon ignores ECONNECT errors and logs others; it now
4071 ignores EPIPE as well.
4073 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4074 (quoted-printable decoding).
4076 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4077 later the small subsequent patch to fix an introduced bug.
4079 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4081 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4083 PH/08 An error is now given if message_size_limit is specified negative.
4085 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4086 to be given (somewhat) arbitrary names.
4088 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4091 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4092 miscellaneous code fixes
4094 PH/10 Added the log_reject_target ACL modifier to specify where to log
4097 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4098 hostname. This is wrong, because it relates to the incoming message (and
4099 probably the interface on which it is arriving) and not to the outgoing
4100 callout (which could be using a different interface). This has been
4101 changed to use the value of the helo_data option from the smtp transport
4102 instead - this is what is used when a message is actually being sent. If
4103 there is no remote transport (possible with a router that sets up host
4104 addresses), $smtp_active_hostname is used.
4106 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4107 tweaks were necessary in order to get it to work (see also 21 below):
4108 (a) The code assumed that strncpy() returns a negative number on buffer
4109 overflow, which isn't the case. Replaced with Exim's string_format()
4111 (b) There were several signed/unsigned issues. I just did the minimum
4112 hacking in of casts. There is scope for a larger refactoring.
4113 (c) The code used strcasecmp() which is not a standard C function.
4114 Replaced with Exim's strcmpic() function.
4115 (d) The code set only $1; it now sets $auth1 as well.
4116 (e) A simple test gave the error "authentication client didn't specify
4117 service in request". It would seem that Dovecot has changed its
4118 interface. Fortunately there's a specification; I followed it and
4119 changed what the client sends and it appears to be working now.
4121 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4124 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4125 address A is aliased to B and C, where B exists and C does not. Without
4126 -v the output is "A verified" because verification stops after a
4127 successful redirection if more than one address is generated. However,
4128 with -v the child addresses are also verified. Exim was outputting "A
4129 failed to verify" and then showing the successful verification for C,
4130 with its parentage. It now outputs "B failed to verify", showing B's
4131 parentage before showing the successful verification of C.
4133 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4134 look up a TXT record in a specific list after matching in a combined
4137 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4138 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4139 they consult the DNS. I had assumed they would set it the way they
4140 wanted; and indeed my experiments on Linux seem to show that in some
4141 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4142 To be on the safe side, however, I have now made the interface to
4143 host_find_byname() similar to host_find_bydns(), with an argument
4144 containing the DNS resolver options. The host_find_byname() function now
4145 sets these options at its start, just as host_find_bydns() does. The smtp
4146 transport options dns_qualify_single and dns_search_parents are passed to
4147 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4148 of host_find_byname() use the default settings of RES_DEFNAMES
4149 (qualify_single) but not RES_DNSRCH (search_parents).
4151 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4152 spool_read_header() do less string testing, by means of a preliminary
4153 switch on the second character of optional "-foo" lines. (This is
4154 overdue, caused by the large number of possibilities that now exist.
4155 Originally there were few.) While I was there, I also converted the
4156 str(n)cmp tests so they don't re-test the leading "-" and the first
4157 character, in the hope this might squeeze out yet more improvement.
4159 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4160 flag allowing group syntax was set by the header_syntax check but not
4161 turned off, possible causing trouble later; (2) The flag was not being
4162 set at all for the header_verify test, causing "group"-style headers to
4163 be rejected. I have now set it in this case, and also caused header_
4164 verify to ignore an empty address taken from a group. While doing this, I
4165 came across some other cases where the code for allowing group syntax
4166 while scanning a header line wasn't quite right (mostly, not resetting
4167 the flag correctly in the right place). These bugs could have caused
4168 trouble for malformed header lines. I hope it is now all correct.
4170 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4171 with the "reply" argument non-NULL. The code, however (which originally
4172 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4173 but it didn't always do it. This confused somebody who was copying the
4174 code for some other use. I have removed all the tests.
4176 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4177 feature that was used to support insecure browsers during the U.S. crypto
4178 embargo. It requires special client support, and Exim is probably the
4179 only MTA that supported it -- and would never use it because real RSA is
4180 always available. This code has been removed, because it had the bad
4181 effect of slowing Exim down by computing (never used) parameters for the
4182 RSA_EXPORT functionality.
4184 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4185 authenticator to fail if there's a tab character in the incoming data
4186 (there should never be unless someone is messing about, as it's supposed
4187 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4188 if the connection is using TLS or if the remote IP is the same as the
4189 local IP, and the "valid-client-cert option" if a client certificate has
4192 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4193 authenticators. This can be used for authorization after authentication
4194 succeeds. (In the case of plaintext, it servers for both authentication
4197 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4198 if any retry times were supplied.
4200 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4201 connection, where there is no client IP address to check. In this
4202 situation, the verify now always succeeds.
4204 PH/25 Applied John Jetmore's -Mset patch.
4206 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4208 PH/27 In a string expansion for a processed (not raw) header when multiple
4209 headers of the same name were present, leading whitespace was being
4210 removed from all of them, but trailing whitespace was being removed only
4211 from the last one. Now trailing whitespace is removed from each header
4212 before concatenation. Completely empty headers in a concatenation (as
4213 before) are ignored.
4215 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4216 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4218 PH/29 [Removed. This was a change that I later backed out, and forgot to
4219 correct the ChangeLog entry (that I had efficiently created) before
4220 committing the later change.]
4222 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4223 address errors (4xx response to RCPT) over the same connection as other
4224 messages routed to the same hosts. Such deliveries are always "forced",
4225 so retry times are not inspected. This resulted in far too many retries
4226 for the affected addresses. The effect occurred only when there were more
4227 hosts than the hosts_max_try setting in the smtp transport when it had
4228 the 4xx errors. Those hosts that it had tried were not added to the list
4229 of hosts for which the message was waiting, so if all were tried, there
4230 was no problem. Two fixes have been applied:
4232 (i) If there are any address or message errors in an SMTP delivery, none
4233 of the hosts (tried or untried) are now added to the list of hosts
4234 for which the message is waiting, so the message should not be a
4235 candidate for sending over the same connection that was used for a
4236 successful delivery of some other message. This seems entirely
4237 reasonable: after all the message is NOT "waiting for some host".
4238 This is so "obvious" that I'm not sure why it wasn't done
4239 previously. Hope I haven't missed anything, but it can't do any
4240 harm, as the worst effect is to miss an optimization.
4242 (ii) If, despite (i), such a delivery is accidentally attempted, the
4243 routing retry time is respected, so at least it doesn't keep
4244 hammering the server.
4246 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4247 in ${readsocket because some servers need this prod.
4249 PH/32 Added some extra debug output when updating a wait-xxx database.
4251 PH/33 The hint "could be header name not terminated by colon", which has been
4252 given for certain expansion errors for a long time, was not being given
4253 for the ${if def:h_colon_omitted{... case.
4255 PH/34 The spec says: "With one important exception, whenever a domain list is
4256 being scanned, $domain contains the subject domain." There was at least
4257 one case where this was not true.
4259 PH/35 The error "getsockname() failed: connection reset by peer" was being
4260 written to the panic log as well as the main log, but it isn't really
4261 panic-worthy as it just means the connection died rather early on. I have
4262 removed the panic log writing for the ECONNRESET error when getsockname()
4265 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4266 runs only) independently of the message's sender address. This meant
4267 that, if the 4xx error was in fact related to the sender, a different
4268 message to the same recipient with a different sender could confuse
4269 things. In particular, this can happen when sending to a greylisting
4270 server, but other circumstances could also provoke similar problems.
4271 I have changed the default so that the retry time for these errors is now
4272 based a combination of the sender and recipient addresses. This change
4273 can be overridden by setting address_retry_include_sender=false in the
4276 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4277 remote server are returned as part of bounce messages. This was not
4278 happening for LMTP over a pipe (the lmtp transport), but now it is the
4279 same for both kinds of LMTP.
4281 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4282 in header lines that were in fact CNAMEs. This is no longer the case.
4284 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4285 and queue runs started by the daemon processed all messages. This has
4286 been fixed so that -R and -S can now usefully be given with -q<time>.
4288 PH/40 Import PCRE release 6.7 (fixes some bugs).
4290 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4292 PH/42 Give an error if -q is specified more than once.
4294 PH/43 Renamed the variables $interface_address and $interface_port as
4295 $received_ip_address and $received_port, to make it clear that these
4296 values apply to message reception, and not to the outgoing interface when
4297 a message is delivered. (The old names remain recognized, of course.)
4299 PH/44 There was no timeout on the connect() call when using a Unix domain
4300 socket in the ${readsocket expansion. There now is.
4302 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4303 be meaningful with "accept".
4305 SC/01 Eximstats V1.43
4306 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4308 SC/02 Eximstats V1.44
4309 Use a glob alias rather than an array ref in the generated
4310 parser. This improves both readability and performance.
4312 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4313 Collect SpamAssassin and rejection statistics.
4314 Don't display local sender or destination tables unless
4315 there is data to show.
4316 Added average volumes into the top table text output.
4318 SC/04 Eximstats V1.46
4319 Collect data on the number of addresses (recipients)
4320 as well as the number of messages.
4322 SC/05 Eximstats V1.47
4323 Added 'Message too big' to the list of mail rejection
4324 reasons (thanks to Marco Gaiarin).
4326 SC/06 Eximstats V1.48
4327 Mainlog lines which have GMT offsets and are too short to
4328 have a flag are now skipped.
4330 SC/07 Eximstats V1.49 (Alain Williams)
4331 Added the -emptyok flag.
4333 SC/08 Eximstats V1.50
4334 Fixes for obtaining the IP address from reject messages.
4336 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4337 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4338 whitespace changes from 4.64-PH/27
4340 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4343 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4344 are found, allow negative numbers in numeric criteria)
4346 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4348 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4349 to match changes made in 4.64-PH/43
4351 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4353 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4354 "domain" is actually an IP address, and force a failure. This locks out
4355 those revolvers/nameservers that support "A-for-A" lookups, in
4356 contravention of the specifications.
4358 PH/48 When a host name was looked up from an IP address, and the subsequent
4359 forward lookup of the name timed out, the host name was left in
4360 $sender_host_name, contrary to the specification.
4362 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4363 restricted to single-key lookups, Exim was not diagnosing an error if
4364 * or *@ was used with a query-style lookup.
4366 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4368 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4369 long ago, but noone interested enough thought of it. Let's just say that
4370 the "1.1" means that there are some new functions that weren't there at
4371 some point in the past.
4373 PH/51 Error processing for expansion failure of helo_data from an smtp
4374 transport during callout processing was broken.
4376 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4377 tested/used via the -bh/-bhc/-bs options.
4379 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4380 bug, fixed in subsequent PCRE releases).
4382 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4383 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4385 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4391 SC/01 Use a glob alias rather than an array ref in eximstats generated
4392 parser. This improves both readability and performance.
4394 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4395 Don't display local sender or destination tables in eximstats unless
4396 there is data to show.
4397 Added average volumes into the eximstats top table text output.
4399 SC/03 Collect data on the number of addresses (recipients) as well
4400 as the number of messages in eximstats.
4402 TF/01 Correct an error in the documentation for the redirect router. Exim
4403 does (usually) call initgroups() when daemonizing.
4405 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4406 with consistent privilege compared to when running as a daemon.
4408 TF/03 Note in the spec that $authenticated_id is not set for local
4409 submissions from trusted users.
4411 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4412 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4414 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4415 by adding some example configuration directives to the default
4416 configuration file. A little bit of work is required to uncomment the
4417 directives and define how usernames and passwords are checked, but
4418 there is now a framework to start from.
4420 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4421 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4422 without this. I don't know how relevant this is to other LDAP libraries.
4424 PH/02 Add the verb name to the "unknown ACL verb" error.
4426 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4428 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4430 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4431 directory not expanded when it should be if an expanded home directory
4432 was set for the address (which is overridden by the transport).
4434 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4437 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4438 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4439 because it is too late at that time, and has no effect.
4441 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4442 security issue with \' (bugzilla #107). I could not use the
4443 PQescapeStringConn() function, because it needs a PGconn value as one of
4446 PH/08 When testing addresses using -bt, indicate those final addresses that
4447 are duplicates that would not cause an additional delivery. At least one
4448 person was confused, thinking that -bt output corresponded to deliveries.
4449 (Suppressing duplicates isn't a good idea as you lose the information
4450 about possibly different redirections that led to the duplicates.)
4452 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4453 systems where poll() doesn't work, in particular OS X.
4455 PH/10 Added more information to debugging output for retry time not reached.
4457 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4458 operations in malware.c.
4460 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4463 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4464 syslog_duplication set false, log lines that would normally be written
4465 both the the main log and to the reject log were not written to syslog at
4468 PH/14 In the default configuration, change the use of "message" in ACL warn
4469 statements to "add_header".
4471 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4472 not followed by a command (e.g. "seen endif").
4474 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4475 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4478 PH/17 Added extra conditions to the default value of delay_warning_condition
4482 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4483 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4484 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4487 The Auto-Submitted: and various List- headers are standardised, whereas I
4488 don't think Precedence: ever was.
4490 PH/18 Refactored debugging code in route_finduser() to show more information,
4491 in particular, the error code if getpwnam() issues one.
4493 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4494 This is apparently needed in addition to the PH/07 change above to avoid
4495 any possible encoding problems.
4497 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4498 but not after initializing Perl.
4500 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4501 output them only if debugging. By default they are written stderr,
4502 apparently, which is not desirable.
4504 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4507 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4510 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4512 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4513 authenticated or an ident call has been made. Suppress the default
4514 values for $authenticated_id and $authenticated_sender (but permit -oMai
4515 and -oMas) when testing with -bh.
4517 PH/24 Re-jigged the order of the tests in the default configuration so that the
4518 tests for valid domains and recipients precede the DNS black list and CSA
4519 tests, on the grounds that those ones are more expensive.
4521 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4522 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4523 This bug exists in every version of Exim that I still have, right back to
4526 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4527 However, an attempt to turn on case-sensitivity in a regex key by
4528 including (?-i) didn't work because the subject string was already
4529 lowercased, and the effects were non-intuitive. It turns out that a
4530 one-line patch can be used to allow (?-i) to work as expected.
4536 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4537 other effects) broke the use of negated acl sub-conditions.
4539 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4542 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4543 "Deny" causes Exim to reject the incoming connection with a 554 error.
4544 Unfortunately, if there is a major crisis, such as a disk failure,
4545 tcp-wrappers gives "deny", whereas what one would like would be some
4546 kind of temporary error. A kludge has been added to help with this.
4547 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4548 554 error is used if errno is still zero or contains ENOENT (which occurs
4549 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4552 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4554 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4555 errors. Otherwise a message that provokes a temporary error (when other
4556 messages do not) can cause a whole host to time out.
4558 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4559 addresses were routed directly to files or pipes from a redirect router.
4560 File deliveries just didn't batch; pipe deliveries might have suffered
4563 PH/06 A failure to get a lock for a hints database would erroneously always say
4564 "Failed to get write lock", even when it was really a read lock.
4566 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4567 of 0600. This has been changed to use the value of the lockfile_mode
4568 option (which defaults to 0600).
4570 PH/08 Applied small patch from the Sieve maintainer.
4572 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4573 folder from quota calculations, a direct delivery into this folder messed
4574 up the contents of the maildirsize file. This was because the regex was
4575 used only to exclude .Trash (or whatever) when the size of the mailbox
4576 was calculated. There was no check that a delivery was happening into an
4577 excluded directory. This bug has been fixed by ignoring all quota
4578 processing for deliveries into excluded directories.
4580 PH/10 Added the maildirfolder_create_regex option to appendfile.
4586 PH/01 The code for finding all the local interface addresses on a FreeBSD
4587 system running IPv6 was broken. This may well have applied to all BSD
4588 systems, as well as to others that have similar system calls. The broken
4589 code found IPv4 interfaces correctly, but gave incorrect values for the
4590 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4591 that it would not match correctly against @[] and not recognize the IPv6
4594 PH/02 The ipliteral router was not recognizing addresses of the form user@
4595 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4597 PH/03 Added disable_ipv6.
4599 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4600 decoded form, because it is most often used to construct To: headers
4601 lines in autoreplies, and the decoded form may well be syntactically
4602 invalid. However, $reply_address has leading white space removed, and all
4603 newlines turned into spaces so that the autoreply transport does not
4606 PH/05 If group was specified without a user on a router, and no group or user
4607 was specified on a transport, the group from the router was ignored.
4609 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4610 for visible compile-time settings that can be used to change these
4611 numbers, for those that want even more. Backwards compatibility with old
4612 spool files has been maintained. However, going back to a previous Exim
4613 release will lost any variables that are in spool files.
4615 PH/07 Two small changes when running in the test harness: increase delay when
4616 passing a TCP/IP connection to a new process, in case the original
4617 process has to generate a bounce, and remove special handling of
4618 127.0.0.2 (sic), which is no longer necessary.
4620 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4621 be the same on different OS.
4623 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4626 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4627 whether --show-vars was specified or not
4629 JJ/02 exipick: Added support for new ACL variable spool format introduced
4632 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4633 syntactically invalid From: or Reply-to: line, and a filter used this to
4634 generate an autoreply, and therefore failed to obtain an address for the
4635 autoreply, Exim could try to deliver to a non-existent relative file
4636 name, causing unrelated and misleading errors. What now happens is that
4637 it logs this as a hard delivery error, but does not attempt to create a
4640 PH/11 The exinext utility has a -C option for testing purposes, but although
4641 the given file was scanned by exinext itself; it wasn't being passed on
4642 when Exim was called.
4644 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4645 an end-of-file indication when reading a command response.
4647 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4648 compiled. In many other places in Exim, IPv6 addresses are always
4649 recognized, so I have changed this. It also means that IPv4 domain
4650 literals of the form [IPV4:n.n.n.n] are now always recognized.
4652 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4653 used if the router is not running as root, for example, when verifying at
4654 ACL time, or when using -bh. The debugging output from this situation was
4655 non-existent - all you got was a failure to exec. I have made two
4658 (a) Failures to set uid/gid, the current directory, or a process leader
4659 in a subprocess such as that created by queryprogram now generate
4660 suitable debugging output when -d is set.
4662 (b) The queryprogram router detects when it is not running as root,
4663 outputs suitable debugging information if -d is set, and then runs
4664 the subprocess without attempting to change uid/gid.
4666 PH/15 Minor change to Makefile for building test_host (undocumented testing
4669 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4670 additional section of a DNS packet that returns MX or SRV records.
4671 Instead, it always explicitly searches for A/AAAA records. This avoids
4672 major problems that occur when a DNS server includes only records of one
4673 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4674 fixed another bug: if SRV records were looked up and the corresponding
4675 address records were *not* found in the additional section, the port
4676 values from the SRV records were lost.
4678 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4679 using the correct key (the original address) when searching the retry
4680 rules in order to find which one to use for generating the retry hint.
4682 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4683 adding the default one. Similarly, if it contains a Reply-To: header, the
4684 errors_reply_to option, if set, is not used.
4686 PH/19 When calculating a retry time, Exim used to measure the "time since
4687 failure" by looking at the "first failed" field in the retry record. Now
4688 it does not use this if it is later than than the arrival time of the
4689 message. Instead it uses the arrival time. This makes for better
4690 behaviour in cases where some deliveries succeed, thus re-setting the
4691 "first failed" field. An example is a quota failure for a huge message
4692 when small messages continue to be delivered. Without this change, the
4693 "time since failure" will always be short, possible causing more frequent
4694 delivery attempts for the huge message than are intended.
4695 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4697 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4698 $1, $2, $3) because the numerical variables can be reset during some
4699 expansion items (e.g. "match"), thereby losing the authentication data.
4701 PH/21 Make -bV show the size of off_t variables so that the test suite can
4702 decide whether to run tests for quotas > 2G.
4704 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4705 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4706 filecount value is greater than 2G or if a quota value is greater than 2G
4707 on a system where the size of off_t is not greater than 4, a panic error
4710 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4711 never match. The debug and -bh output now contains an explicit error
4712 message indicating a malformed IPv4 address or mask.
4714 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4715 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4716 PH/23 above applies.
4718 PH/25 Do not write to syslog when running in the test harness. The only
4719 occasion when this arises is a failure to open the main or panic logs
4720 (for which there is an explicit test).
4722 PH/26 Added the /no_tell option to "control=freeze".
4724 PH/27 If a host name lookup failed very early in a connection, for example, if
4725 the IP address matched host_lookup and the reverse lookup yielded a name
4726 that did not have a forward lookup, an error message of the form "no IP
4727 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4728 could be logged. Now it outputs the IP address instead of "NULL".
4730 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4731 allows the sender and the authenticated sender to be set when
4732 submitting a message from within Exim. Since child_open_exim() is
4733 documented for local_scan(), the new function should be too.
4735 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4736 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4737 results in an empty string is now treated as unset.
4739 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4741 PH/31 Added sender_verify_fail logging option.
4743 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4744 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4745 tidied the source and removed it altogether.
4747 PH/33 When a queue run was abandoned because the load average was too high, a
4748 log line was always written; now it is written only if the queue_run log
4749 selector is set. In addition, the log line for abandonment now contains
4750 information about the queue run such as the pid. This is always present
4751 in "start" and "stop" lines but was omitted from the "abandon" line.
4753 PH/34 Omit spaces between a header name and the colon in the error message that
4754 is given when verify = headers_syntax fails (if there are lots of them,
4755 the message gets confusing).
4757 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4758 names, as there are now some PTR records that contain slashes. This check
4759 is only to protect against broken name servers that fall over on strange
4760 characters, so the fact that it applies to all lookups doesn't matter.
4762 PH/36 Now that the new test suite is complete, we can remove some of the
4763 special code in Exim that was needed for the old test suite. For example,
4764 sorting DNS records because real resolvers return them in an arbitrary
4765 order. The new test suite's fake resolver always returns records in the
4768 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4769 bounces) except when queue_only is set, to avoid logging races between
4770 the different processes.
4772 PH/38 Panic-die if .include specifies a non-absolute path.
4774 PH/39 A tweak to the "H" retry rule from its user.
4776 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4777 a label. They prevented compilation on older perls.
4779 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4780 a warning to be raised on newish perls.
4782 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4783 on queue. Changes to match documented behaviour of showing count of
4784 messages matching specified criteria.
4786 PH/40 Changed the default ident timeout from 30s to 5s.
4788 PH/41 Added support for the use of login_cap features, on those BSD systems
4789 that have them, for controlling the resources used by pipe deliveries.
4791 PH/42 The content-scanning code uses fopen() to create files in which to put
4792 message data. Previously it was not paying any attention to the mode of
4793 the files. Exim runs with umask(0) because the rest of the code creates
4794 files with open(), and sets the required mode explicitly. Thus, these
4795 files were ending up world-writeable. This was not a big issue, because,
4796 being within the spool directory, they were not world-accessible. I have
4797 created a function called modefopen, which takes an additional mode
4798 argument. It sets umask(777), creates the file, chmods it to the required
4799 mode, then resets the umask. All the relevant calls to fopen() in the
4800 content scanning code have been changed to use this function.
4802 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4803 to 24 hours. This avoids potential overflow problems when processing G
4804 and H retry rules. I suspect nobody ever tinkers with this value.
4806 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4808 PH/45 When the plaintext authenticator is running as a client, the server's
4809 challenges are checked to ensure they are valid base64 strings. By
4810 default, the authentication attempt is cancelled if an invalid string is
4811 received. Setting client_ignore_invalid_base64 true ignores these errors.
4812 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4813 they are received. Thus, the responses can be made to depend on the
4814 challenges. If an invalid string is ignored, an empty string is placed in
4817 PH/46 Messages that are created by the autoreply transport now contains a
4818 References: header, in accordance with RFCs 2822 and 3834.
4820 PH/47 Added authenticated_sender_force to the smtp transport.
4822 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4824 PH/49 Installed latest patch from the Sieve maintainer.
4826 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4827 was also set, the appendfile transport was unnecessarily scanning a
4828 directory of message files (e.g. for maildir delivery) to find the count
4829 of files (along with the size), even though it did not need this
4830 information. It now does the scan only if it needs to find either the
4831 size of the count of files.
4833 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4835 PH/52 Two bugs concerned with error handling when the smtp transport is
4838 (i) Exim was not creating retry information for temporary errors given
4839 for individual recipients after the DATA command when the smtp transport
4840 was used in LMTP mode. This meant that they could be retried too
4841 frequently, and not timed out correctly.
4843 (ii) Exim was setting the flag that allows error details to be returned
4844 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4845 recipients that were returned after the DATA command.
4847 PH/53 This is related to PH/52, but is more general: for any failing address,
4848 when detailed error information was permitted to be returned to the
4849 sender, but the error was temporary, then after the final timeout, only
4850 "retry timeout exceeded" was returned. Now it returns the full error as
4851 well as "retry timeout exceeded".
4853 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4854 do this, and (what is worse) MTAs that accept it.
4856 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4857 will now be deprecated.
4859 PH/56 New os.c-cygwin from the Cygwin maintainer.
4861 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4862 formats (previously only available in exim formats via -bpr, -bpru,
4863 and -bpra. Now also available in native and exiqgrep formats)
4865 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4866 with very large, slow to parse queues
4868 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4870 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4872 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4873 responses to authentication challenges, though it was showing the
4874 challenges; (ii) I've removed the CR characters from the debug output for
4877 PH/58 Allow for the insertion of a newline as well as a space when a string
4878 is turned into more than one encoded-word during RFC 2047 encoding. The
4879 Sieve code now uses this.
4881 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4882 data_4xx, lost_connection, tls_required.
4884 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4885 message was being sent as an SMTP response.
4887 PH/61 Add -l and -k options to exicyclog.
4889 PH/62 When verifying, if an address was redirected to one new address, so that
4890 verification continued, and the new address failed or deferred after
4891 having set something in $address_data, the value of $address_data was not
4892 passed back to the ACL. This was different to the case when no
4893 redirection occurred. The value is now passed back in both cases.
4895 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4896 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4897 use login_cap.h, so on its own it isn't the distinguishing feature. The
4898 new name refers directly to the setclassresources() function.
4900 PH/65 Added configuration files for NetBSD3.
4902 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4904 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4905 is preferred over IPv4.
4907 PH/68 The bounce_return_message and bounce_return_body options were not being
4908 honoured for bounces generated during the reception of non-SMTP messages.
4909 In particular, this applied to messages rejected by the ACL. This bug has
4910 been fixed. However, if bounce_return_message is true and bounce_return_
4911 body is false, the headers that are returned for a non-SMTP message
4912 include only those that have been read before the error was detected.
4913 (In the case of an ACL rejection, they have all been read.)
4915 PH/69 The HTML version of the specification is now built in a directory called
4916 spec_html instead of spec.html, because the latter looks like a path with
4917 a MIME-type, and this confuses some software.
4919 PH/70 Catch two compiler warnings in sieve.c.
4921 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4922 function verify_get_ident() calls ip_connect() to connect a socket, but
4923 if the "connect()" function timed out, ip_connect() used to close the
4924 socket. However, verify_get_ident() also closes the socket later, and in
4925 between Exim writes to the log, which may get opened at this point. When
4926 the socket was closed in ip_connect(), the log could get the same file
4927 descriptor number as the socket. This naturally causes chaos. The fix is
4928 not to close the socket in ip_connect(); the socket should be closed by
4929 the function that creates it. There was only one place in the code where
4930 this was missing, in the iplookup router, which I don't think anybody now
4931 uses, but I've fixed it anyway.
4933 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4934 well as to direct DNS lookups. Otherwise the handling of names in host
4935 lists is inconsistent and therefore confusing.
4941 PH/01 Two changes to the default runtime configuration:
4943 (1) Move the checks for relay_from_hosts and authenticated clients from
4944 after to before the (commented out) DNS black list checks.
4946 (2) Add control=submission to the relay_from_hosts and authenticated
4947 clients checks, on the grounds that messages accepted by these
4948 statements are most likely to be submissions.
4950 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4952 (1) Generate an error if the third argument for the ${prvs expansion is
4955 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4958 (3) Reset the variables that are obtained from the first argument of
4959 ${prvscheck and used in the second argument before leaving the code,
4960 because their memory is reclaimed, so using them afterwards may do
4963 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4964 one (it's much easier than Tom thought :-).
4966 (5) Because of (4), we can now allow for the use of $prvscheck_result
4967 inside the third argument.
4969 PH/03 For some reason, the default setting of PATH when running a command from
4970 a pipe transport was just "/usr/bin". I have changed it to
4973 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4974 anything to be listed in the output from -bV.
4976 PH/05 When a filter generated an autoreply, the entire To: header line was
4977 quoted in the delivery log line, like this:
4979 => >A.N.Other <ano@some.domain> <original@ddress> ...
4981 This has been changed so that it extracts the operative address. There
4982 may be more than one such address. If so, they are comma-separated, like
4985 => >ano@some.domain,ona@other.domain <original@ddress> ...
4987 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4988 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4989 not being looked up in the rDNS to get a host name, Exim was showing the
4990 IP address twice in Received: lines, even though the IP addresses were
4991 identical. For example:
4993 Received: from [1.2.3.4] (helo=[1.2.3.4])
4995 However, if the real host name was known, it was omitting the HELO data
4996 if it matched the actual IP address. This has been tidied up so that it
4997 doesn't show the same IP address twice.
4999 PH/07 When both +timestamp and +memory debugging was on, the value given by
5000 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5001 called by the debug printing, thereby overwriting the timestamp buffer.
5002 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5004 PH/08 When the original message was included in an autoreply transport, it
5005 always said "this is a copy of the message, including all the headers",
5006 even if body_only or headers_only was set. It now gives an appropriate
5009 PH/09 Applied a patch from the Sieve maintainer which:
5011 o fixes some comments
5012 o adds the (disabled) notify extension core
5013 o adds some debug output for the result of if/elsif tests
5014 o points to the current vacation draft in the documentation
5015 and documents the missing references header update
5019 o fixes a bug in processing the envelope test (when testing
5020 multiple envelope elements, the last element determined the
5023 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5024 Electronic Mail") by including:
5026 Auto-submitted: auto-generated
5028 in the messages that it generates (bounce messages and others, such as
5029 warnings). In the case of bounce messages for non-SMTP messages, there was
5030 also a typo: it was using "Auto_submitted" (underscore instead of
5031 hyphen). Since every message generated by Exim is necessarily in response
5032 to another message, thes have all been changed to:
5034 Auto-Submitted: auto-replied
5036 in accordance with these statements in the RFC:
5038 The auto-replied keyword:
5040 - SHOULD be used on messages sent in direct response to another
5041 message by an automatic process,
5043 - MUST NOT be used on manually-generated messages,
5045 - MAY be used on Delivery Status Notifications (DSNs) and Message
5046 Disposition Notifications (MDNs),
5048 - MUST NOT be used on messages generated by automatic or periodic
5049 processes, except for messages which are automatic responses to
5052 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5053 to the default Received: header definition.
5055 PH/12 Added log selector acl_warn_skipped (default on).
5057 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5058 variables because (a) they are in the wrong storage pool and (b) even if
5059 they were copied, it wouldn't work properly because of the caching.
5061 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5062 checking when decoding. Apparently there are clients that generate
5063 overlong encoded strings. Why am I not surprised?
5065 PH/15 If the first argument of "${if match_address" was not empty, but did not
5066 contain an "@" character, Exim crashed. Now it writes a panic log message
5067 and treats the condition as false.
5069 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5071 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5072 extension "envelope-auth". The code is finished and in agreement with
5073 other implementations, but there is no documentation so far and in fact,
5074 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5075 not changing the active code.
5077 Print executed "if" and "elsif" statements when debugging is used. This
5078 helps a great deal to understand what a filter does.
5080 Document more things not specified clearly in RFC3028. I had all this
5081 sorted out, when out of a sudden new issues came to my mind. Oops."
5083 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5086 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5087 canonical form (as documented). However, after a host name lookup from
5088 the IP address, check_host() was doing a simple string comparison with
5089 addresses acquired from the DNS when checking that the found name did
5090 have the original IP as one of its addresses. Since any found IPv6
5091 addresses are likely to be in abbreviated form, the comparison could
5092 fail. Luckily, there already exists a function for doing the comparison
5093 by converting both addresses to binary, so now that is used instead of
5094 the text comparison.
5096 PH/20 There was another similar case to PH/19, when a complete host name was
5097 given in a host list; looking up its IP address could give an abbreviated
5098 form, whereas the current host's name might or might not be abbreviated.
5099 The same fix has been applied.
5105 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5106 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5109 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5110 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5112 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5114 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5115 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5116 Also, for "Auto-Submitted", treat anything other than "no" as
5117 non-personal, in accordance with RFC 3834. (Previously it treated
5118 anything starting "auto-" as non-personal.)
5120 TF/01 The control=submission/name=... option had a problem with syntax
5121 errors if the name included a slash character. The /name= option
5122 now slurps the rest of the string, so it can include any characters
5123 but it must come last in the list of options (after /sender_retain
5126 PH/05 Some modifications to the interface to the fake nameserver for the new
5134 TK/01 Added the "success_on_redirect" address verification option. See
5135 NewStuff for rationale and an example.
5137 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5139 PH/02 Patch to exigrep to allow it to work on syslog lines.
5141 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5142 fread() to skip over the body file's header line, because in Cygwin the
5143 header line is locked and is inaccessible.
5145 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5146 co-exist for some time) to make it clear that it is the Exim ID that is
5147 referenced, not the Message-ID: header line.
5149 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5150 string_format() function, because snprintf() does not exist on all
5153 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5154 ${stat: expansion item.
5156 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5157 protocol synchronization error", to keep the pedants happy.
5159 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5160 well as for IRIX systems, when gcc is being used. See the host.c source
5163 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5165 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5168 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5169 transport and to the smtp transport in LMTP mode.
5171 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5173 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5174 run a filter in a subprocess. This could lead to confusion in subsequent
5175 lookups in the parent process. There should also be a search_tidyup() at
5176 the end of the subprocess.
5178 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5179 only if the host matched helo_try_verify_hosts, which caused the
5180 verification to occur when the EHLO/HELO command was issued. The ACL just
5181 tested the remembered result. Now, if a previous verification attempt has
5182 not happened, "verify = helo" does it there and then.
5184 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5186 TK/03 Fix log output including CR from clamd.
5188 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5189 exist provoked a memory error which could cause a segfault.
5191 PH/15 Installed PCRE 6.2
5193 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5195 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5196 of the problem. Specifically, suggested +O2 rather than +O1 for the
5199 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5201 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5202 record, the port was not correctly logged when the outgoing_port log
5203 selector was set (it logged the transort's default port).
5205 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5206 fallback_hosts, and "hosts" in the smtp transport.
5208 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5209 host errors such as "Connection refused".
5211 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5212 authentication with radiusclient 0.4.9:
5214 - Error returned from rc_read_config was caught wrongly
5215 - Username/password not passed on to radius server due to wrong length.
5217 The presumption is that some radiusclient API changes for 4.51/PH/17
5218 were not taken care of correctly. The code is still untested by me (my
5219 Linux distribution still has 0.3.2 of radiusclient), but it was
5220 contributed by a Radius user.
5222 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5223 expanding the "port" option of the smtp transport.
5225 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5226 while reading a MIME header. Thanks to Tom Hughes for a patch.
5228 PH/24 Include config.h inside local_scan.h so that configuration settings are
5231 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5232 that in an ACL for RCPT (for example), you can examine exactly what was
5235 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5236 commands, but it was not correctly comparing the address with the actual
5237 client host address. Thus, it would show the EHLO address in Received:
5238 header lines when this was not necessary.
5240 PH/27 Added the % operator to ${eval:}.
5242 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5243 it should be ignoring failures (because with -C, for example, it has lost
5244 privilege). It wasn't ignoring creation failures other than "already
5247 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5250 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5251 because an input line was too long, either on its own, or by virtue of
5252 too many continuations, the temporary file was not being removed, and the
5253 return code was incorrect.
5255 PH/31 Missing "BOOL" in function definition in filtertest.c.
5257 PH/32 Applied Sieve patches from the maintainer.
5259 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5261 PH/33 Added "verify = not_blind".
5263 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5264 Local/Makefile (with some defaults set). These are used in built scripts
5265 such as exicyclog, but they have never been used in the exim_install
5266 script (though there are many overriding facilities there). I have
5267 arranged that the exim_install script now takes note of these two
5270 PH/35 Installed configuration files for Dragonfly.
5272 PH/36 When a locally submitted message by a trusted user did not contain a
5273 From: header, and the sender address was obtained from -f or from an SMTP
5274 MAIL command, and the trusted user did not use -F to supply a sender
5275 name, $originator_name was incorrectly used when constructing a From:
5276 header. Furthermore, $originator_name was used for submission mode
5277 messages from external hosts without From: headers in a similar way,
5278 which is clearly wrong.
5280 PH/37 Added control=suppress_local_fixups.
5282 PH/38 When log_selector = +received_sender was set, and the addition of the
5283 sender made the log line's construction buffer exactly full, or one byte
5284 less than full, an overflow happened when the terminating "\n" was
5287 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5288 when the result of a list match is failure because a DNS lookup failed.
5290 PH/40 RM_COMMAND is now used in the building process.
5292 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5293 the "build-* directories that it finds.
5295 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5296 address in a domain literal was a prefix of an interface address.
5298 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5299 when verifying a sender address, unless rewrite_headers is false.
5301 PH/44 Wrote a long comment about why errors_to addresses are verified as
5302 recipients, not senders.
5304 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5305 the ratelimit ACL was added.
5307 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5309 PH/46 Added extra information about PostgreSQL errors to the error string.
5311 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5312 avoiding the need to install special zones in a real server. This is
5313 backwards compatible; if it can't find the fake resolver, it drops back.
5314 Thus, both old and new test suites can be run.
5316 TF/02 Added util/ratelimit.pl
5318 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5319 clock is set back in time.
5321 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5322 Candler <B.Candler@pobox.com>.
5324 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5325 OK for addresses that are the result of redirections.
5327 PH/48 A number of further additions for the benefit of the new test suite,
5328 including a fake gethostbyname() that interfaces to the fake DNS resolver
5331 TF/06 The fix for widen_domains has also been applied to qualify_single and
5332 search_parents which are the other dnslookup options that can cause
5335 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5338 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5340 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5341 service" if it encountered an MX record with an empty target hostname.
5342 The message is now "an MX or SRV record indicated no SMTP service".
5344 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5345 if the DNS of the sending site is misconfigured. This is quite a
5346 common situation. This change restores the behaviour of treating a
5347 helo verification defer as a failure.
5349 PH/51 If self=fail was set on a router, the bounce message did not include the
5350 actual error message.
5356 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5358 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5359 command itself ended in error, the underlying message about the transport
5360 filter timeout was being overwritten with the pipe command error. Now the
5361 underlying error message should be appended to the second error message.
5363 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5365 PH/02 Reduce the amount of output that "make" produces by default. Full output
5366 can still be requested.
5368 PH/03 The warning log line about a condition test deferring for a "warn" verb
5369 was being output only once per connection, rather than after each
5370 occurrence (because it was using the same function as for successful
5371 "warn" verbs). This seems wrong, so I have changed it.
5373 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5374 it should not have, which might have caused a crash in the right
5375 circumstances, but probably never did.
5377 PH/04 Installed a modified version of Tony Finch's patch to make submission
5378 mode fix the return path as well as the Sender: header line, and to
5379 add a /name= option so that you can make the user's friendly name appear
5382 TF/03 Added the control = fakedefer ACL modifier.
5384 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5385 Mark Lowes for thorough testing.
5387 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5389 TK/03 Merged latest SRS patch from Miles Wilton.
5391 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5392 with the definition in sysexits.h (which is #included earlier).
5393 Fortunately, Exim does not actually use EX_OK. The code used to try to
5394 preserve the sysexits.h value, by assuming that macro definitions were
5395 scanned for macro replacements. I have been disabused of this notion,
5396 so now the code just undefines EX_OK before #including unistd.h.
5398 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5399 in the smtp transport. When a block could not be written in a single
5400 write() function, the timeout was being re-applied to each part-write.
5401 This seems wrong - if the receiver was accepting one byte at a time it
5402 would take for ever. The timeout is now adjusted when this happens. It
5403 doesn't have to be particularly precise.
5405 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5406 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5408 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5409 without a domain if the check to <postmaster@domain> fails.
5411 SC/01 Eximstats: added -xls and the ability to specify output files
5412 (patch written by Frank Heydlauf).
5414 SC/02 Eximstats: use FileHandles for outputting results.
5416 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5418 SC/04 Eximstats: fixed display of large numbers with -nvr option
5420 SC/05 Eximstats: fixed merging of reports with empty tables.
5422 SC/06 Eximstats: added the -include_original_destination flag
5424 SC/07 Eximstats: removed tabs and trailing whitespace.
5426 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5428 TK/06 MBOX spool code: Add real "From " MBOX separator line
5429 so the .eml file is really in mbox format (even though
5430 most programs do not really care). Patch from Alex Miller.
5432 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5433 The latter is generated from $received_to and is only set if the
5434 message has one envelope recipient. SA can use these headers,
5435 obviously out-of-the-box. Patch from Alex Miller.
5437 PH/08 The ${def test on a variable was returning false if the variable's
5438 value was "0", contrary to what the specification has always said!
5439 The result should be true unless the variable is empty.
5441 PH/09 The syntax error of a character other than { following "${if
5442 def:variable_name" (after optional whitespace) was not being diagnosed.
5443 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5444 accidental colon was present, for example, could give incorrect results.
5446 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5447 result is used (not including appendfile, where other changes are about
5450 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5451 This involved changing a lot of size variables from int to off_t. It
5452 should work with maildirs and everything.
5454 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5455 spamd dying while we are connected to it.
5457 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5460 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5461 function for BDB 4.3.
5463 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5465 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5466 the sources for PCRE have been re-arranged and more files are now
5469 PH/15 The code I had for printing potentially long long variables in PH/11
5470 above was not the best (it lost precision). The length of off_t variables
5471 is now inspected at build time, and an appropriate printing format (%ld
5472 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5473 to be "long long int" or "long int". This is needed for the internal
5474 formatting function string_vformat().
5476 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5477 the configuration file to be ":syslog", then the script "guesses" where
5478 the logs files are, rather than using the compiled in default. In our
5479 case the guess is not the same as the compiled default, so the script
5480 suddenly stopped working when I started to use syslog. The patch checks
5481 to see if log_file_path is "". If so, it attempts to read it from exim
5482 with no configuration file to get the compiled in version, before it
5483 falls back to the previous guessing code."
5485 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5486 implementing BATV in an Exim configuration. See NewStuff for the gory
5489 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5490 Makefile that are specific to HP-UX.
5492 PH/18 If the "use_postmaster" option was set for a recipient callout together
5493 with the "random" option, the postmaster address was used as the MAIL
5494 FROM address for the random test, but not for the subsequent recipient
5495 test. It is now used for both.
5497 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5498 patch removes a few documentation additions to RFC 3028, because the
5499 latest draft now contains them. It adds the new en;ascii-case comparator
5500 and a new error check for 8bit text in MIME parts. Comparator and
5501 require names are now matched exactly. I enabled the subaddress
5502 extension, but it is not well tested yet (read: it works for me)."
5504 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5505 rework some of the code of TK/09 above to avoid the hardwired use of
5506 "%lld" and "long long". Replaced the call to snprintf() with a call to
5509 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5510 records point to non-existent hosts", "retry timeout exceeded", and
5511 "retry time not reached for any host after a long failure period".
5513 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5514 experimental DomainKeys support:
5516 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5517 (2) On an error such as an illegally used "control", the wrong name for
5518 the control was given.
5520 These problems did NOT occur unless DomainKeys support was compiled.
5522 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5524 PH/24 Added ${if match_ip condition.
5526 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5527 databases so that it will be absolutely obvious if a crash occurs in the
5528 DB library. This is a regular occurrence (often caused by mis-matched
5531 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5532 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5533 picked up on a user's system that detects such things. There doesn't seem
5534 to be a gcc warning option for this - only an attribute that has to be
5535 put on the function's prototype. It seems that in Fedora Core 4 they have
5536 set this on a number of new functions. No doubt there will be more in due
5539 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5540 specify a transport. However, if an address that was verified by such a
5541 router was the subject of a callout, Exim crashed because it tried to
5542 read the rcpt_include_affixes from the non-existent transport. Now it
5543 just assumes that the setting of that option is false. This bug was
5544 introduced by 4.51/PH/31.
5546 PH/28 Changed -d+all to exclude +memory, because that information is very
5547 rarely of interest, but it makes the output a lot bigger. People tend to
5548 do -d+all out of habit.
5550 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5551 code in os-type was giving problems when libc.so lives in lib64, like on
5554 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5555 aren't the modern standard, and it seems that some systems' include files
5556 don't always have them. Exim was already checking for some of the newer
5557 ones like T_AAAA, and defining it itself. I've added checks for all the
5558 record types that Exim uses.
5560 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5561 not automatically generating a new one, as it is supposed to. This
5562 prevented TLS from working. If the file did exist, but contained invalid
5563 data, a new version was generated, as expected. It was only the case of a
5564 non-existent file that was broken.
5566 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5567 with a change in libdomainkeys > 0.64.
5569 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5570 from DNS. If the selector record carries the flag, it now has
5571 precedence over the domain-wide flag.
5573 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5575 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5576 the use of a port name that isn't defined in /etc/services) occurred, the
5577 message was deferred as in a normal delivery, and thus remained on the
5578 spool, instead of being failed because of the mua_wrapper setting. This
5579 is now fixed, and I tidied up some of the mua_wrapper messages at the
5582 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5583 the messages in a hash of arrays rather than using individual hashes.
5584 This is a bit cleaner and results in dramatic memory savings, albeit
5585 at a slight CPU cost.
5587 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5588 as requested by Marc Sherman.
5590 SC/10 Eximstats: added histograms for user specified patterns as requested
5593 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5595 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5596 fopen() in the content-scanning modules that did not already have it.
5602 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5603 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5605 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5607 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5609 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5610 to test Sieve filters that use "vacation".
5612 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5613 that changes the way the GnuTLS parameters are stored in the cache file.
5614 The new format can be generated externally. For backward compatibility,
5615 if the data in the cache doesn't make sense, Exim assumes it has read an
5616 old-format file, and it generates new data and writes a new file. This
5617 means that you can't go back to an older release without removing the
5620 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5621 work if there are any delivery delays because "one_time" forces the
5622 parent to be marked "delivered", so its unseen clone is never tried
5623 again. For this reason, Exim now forbids the simultaneous setting of
5626 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5627 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5628 entry if you want to know the details. The fix, however, neglected to
5629 consider the case when local delivery batching is involved. The test for
5630 "previously delivered" was not happening when checking to see if an
5631 address could be batched with a previous (undelivered) one; under
5632 certain circumstances this could lead to multiple deliveries to the same
5635 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5636 in its include files, and this causes problems building Exim.
5638 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5639 header_syntax) but Exim was just ignoring anything given after a slash.
5640 In particular, this caused confusion with an attempt to use "verify =
5641 reverse_host_lookup/defer_ok". An error is now given when options are
5642 supplied for verify items that do not have them. (Maybe reverse_host_
5643 lookup should have a defer_ok option, but that's a different point.)
5645 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5646 defined by RFC 821) to 2048, because there were problems with some AUTH
5647 commands, and RFC 1869 says the size should be increased for extended
5648 SMTP commands that take arguments.
5650 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5653 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5654 "unknown" error; now it says that the functionality isn't in the binary.
5656 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5657 an address' error message when a string expansion fails (syntax or
5658 whatever). Otherwise the password may appear in the log. Following change
5659 PH/42 below, there is no longer a chance of it appearing in a bounce
5662 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5664 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5665 of its addresses was ever tried. (Bugzilla bug #2.)
5667 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5668 the result incorrectly in the debug output. (It correctly added a newline
5669 to what was transported.)
5671 TF/01 Added $received_time.
5673 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5674 commented out examples of how to interface to a virus scanner and to
5675 SpamAssassin. Also added commented examples of av_scanner and
5676 spamd_address settings.
5678 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5679 and controls are allowed in which ACLs. There were a couple of minor
5680 errors. Some of the entries in the conditions table (which is a table of
5681 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5682 negation of where the condition IS allowed.
5684 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5686 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5687 header file does not have a version number, so I've had to invent a new
5688 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5689 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5690 radiusclient), but it was contributed by a Radius user.
5692 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5693 files or directories, for OpenSSL.
5695 PH/19 When an Exim process that is running as root has to create an Exim log
5696 file, it does so in a subprocess that runs as exim:exim so as to get the
5697 ownership right at creation (otherwise, other Exim processes might see
5698 the file with the wrong ownership). There was no test for failure of this
5699 fork() call, which would lead to the process getting stuck as it waited
5700 for a non-existent subprocess. Forks do occasionally fail when resources
5701 run out. I reviewed all the other calls to fork(); they all seem to check
5704 PH/20 When checking for unexpected SMTP input at connect time (before writing
5705 the banner), Exim was not dealing correctly with a non-positive return
5706 from the read() function. If the client had disconnected by this time,
5707 the result was a log entry for a synchronization error with an empty
5708 string after "input=" when read() returned zero. If read() returned -1
5709 (an event I could not check), uninitialized data bytes were printed.
5710 There were reports of junk text (parts of files, etc) appearing after
5713 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5715 PH/22 Added support for macro redefinition, and (re)definition in between
5716 driver and ACL definitions.
5718 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5719 forgetting to use the resulting value; it was using the unexpanded value.
5721 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5722 hadn't been configured. The fix is from Juergen Kreileder, who
5723 understands it better than I do:
5725 "Here's what I see happening with three configured cyrus_sasl
5726 authenticators configured (plain, login, cram-md5):
5728 On startup auth_cyrus_sasl_init() gets called for each of these.
5729 This means three calls to sasl_listmech() without a specified mech_list.
5730 => SASL tests which mechs of all available mechs actually work
5731 => three warnings about OTP not working
5732 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5734 With the patch, sasl_listmech() also gets called three times. But now
5735 SASL's mech_list option is set to the server_mech specified in the the
5736 authenticator. Or in other words, the answer from sasl_listmech()
5737 gets limited to just the mech you're testing for (which is different
5739 => the return list contains just 'plain' or 'login', 'cram-md5' or
5740 nothing depending on the value of ob->server_mech.
5742 I've just tested the patch: Authentication still works fine,
5743 unavailable mechs specified in the exim configuration are still
5744 caught, and the auth.log warnings about OTP are gone."
5746 PH/25 When debugging is enabled, the contents of the command line are added
5747 to the debugging output, even when log_selector=+arguments is not
5750 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5751 answer is "GNU", and only if the return is "GNU/something" is the answer
5754 PH/27 $acl_verify_message is now set immediately after the failure of a
5755 verification in an ACL, and so is available in subsequent modifiers. In
5756 particular, the message can be preserved by coding like this:
5758 warn !verify = sender
5759 set acl_m0 = $acl_verify_message
5761 Previously, $acl_verify_message was set only while expanding "message"
5762 and "log_message" when a very denied access.
5764 PH/28 Modified OS/os.c-Linux with
5766 -#ifndef OS_LOAD_AVERAGE
5767 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5769 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5770 nomenclature these days.)
5772 PH/29 Installed patch from the Sieve maintainer that adds the options
5773 sieve_useraddress and sieve_subaddress to the redirect router.
5775 PH/30 In these circumstances:
5776 . Two addresses routed to the same list of hosts;
5777 . First host does not offer TLS;
5778 . First host accepts first address;
5779 . First host gives temporary error to second address;
5780 . Second host offers TLS and a TLS session is established;
5781 . Second host accepts second address.
5782 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5783 and peerdn, if requested) that were in fact used only for the second
5786 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5787 attention to any local part prefix or suffix that was matched by the
5788 router that accepted the address. It now behaves in the same way as it
5789 does for delivery: the affixes are removed from the local part unless
5790 rcpt_include_affixes is set on the transport.
5792 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5793 timeout during the DATA phase of an incoming message.
5795 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5796 applied a patch sanctioned by the Sieve maintainer.
5798 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5799 the uid or gid is negative. A case of a negative gid caused this to be
5800 noticed. The fix allows for either to be negative.
5802 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5803 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5806 PH/36 The change PH/12 above was broken. Fixed it.
5808 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5809 the grounds that routing the same address twice would always produce the
5810 same answer. This might have been true once, but it is certainly no
5811 longer true now. Routing a child address may depend on the previous
5812 routing that produced that child. Some complicated redirection strategies
5813 went wrong when messages had multiple recipients, and made Exim's
5814 behaviour dependent on the order in which the addresses were given.
5816 I have moved the duplicate checking until after the routing is complete.
5817 Exim scans the addresses that are assigned to local and remote
5818 transports, and removes any duplicates. This means that more work will be
5819 done, as duplicates will always all be routed, but duplicates are
5820 presumably rare, so I don't expect this is of any significance.
5822 For deliveries to pipes, files, and autoreplies, the duplicate checking
5823 still happens during the routing process, since they are not going to be
5826 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5827 It corrects a timeout issue with spamd. This is Ian's comment: "The
5828 background is that sometimes spamd either never reads data from a
5829 connection it has accepted, or it never writes response data. The exiscan
5830 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5831 blindly assumes that writes won't block so it may never time out."
5833 PH/39 Allow G after quota size as well as K and M.
5835 PH/40 The value set for $authenticated_id in an authenticator may not contain
5836 binary zeroes or newlines because the value is written to log lines and
5837 to spool files. There was no check on this. Now the value is run through
5838 the string_printing() function so that such characters are converted to
5839 printable escape sequences.
5841 PH/41 $message_linecount is a new variable that contains the total number of
5842 lines in the message. Compare $body_linecount, which is the count for the
5845 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5846 bounce and delay warning messages, except in certain special cases, which
5849 (a) An SMTP error message from a remote host;
5850 (b) A message specified in a :fail: redirection;
5851 (c) A message specified in a "fail" command in a system filter;
5852 (d) A message specified in a FAIL return from the queryprogram router;
5853 (e) A message specified by the cannot_route_message router option.
5855 In these cases only, Exim does include the error details in bounce and
5856 warning messages. There are also a few cases where bland messages such
5857 as "unrouteable address" or "local delivery error" are given.
5859 PH/43 $value is now also set for the "else" part of a ${run expansion.
5861 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5862 being worked on, but at least Exim now implements the latest version to
5865 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5866 process to complete was treated as a delivery failure, a timeout while
5867 writing the message to the pipe was logged, but erroneously treated as a
5868 successful delivery. Such timeouts include transport filter timeouts. For
5869 consistency with the overall process timeout, these timeouts are now
5870 treated as errors, giving rise to delivery failures by default. However,
5871 there is now a new Boolean option for the pipe transport called
5872 timeout_defer, which, if set TRUE, converts the failures into defers for
5873 both kinds of timeout. A transport filter timeout is now identified in
5876 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5877 systems where "make" and "gmake" are different, calling "gmake" at top
5878 level broke things. I've arranged for the value of $(MAKE) to be passed
5879 from the Makefile to this script so that it can call the same version of
5883 A note about Exim versions 4.44 and 4.50
5884 ----------------------------------------
5886 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5887 changes of various kinds. As a consequence, a big documentation update was
5888 needed. This delayed the release for rather longer than seemed good, especially
5889 in the light of a couple of (minor) security issues. Therefore, the changes
5890 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5891 release. So 4.44 and 4.50 are in effect two different branches that both start
5894 I have left the 4.50 change log unchanged; it contains all the changes since
5895 4.43. The change log for 4.44 is below; many of its items are identical to
5896 those for 4.50. This seems to be the most sensible way to preserve the
5897 historical information.
5903 1. Minor wording change to the doc/README.SIEVE file.
5905 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5906 computation of the current number of files was incorrect.
5908 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5909 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5912 4. Give more explanation in the error message when the command for a transport
5913 filter fails to execute.
5915 5. There are several places where Exim runs a non-Exim command in a
5916 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5917 was being done only for the command run by the queryprogram router. It is
5918 now done for all such subprocesses. The other cases are: ${run, transport
5919 filters, and the commands run by the lmtp and pipe transports.
5921 6. Added CONFIGURE_GROUP build-time option.
5923 7. Some older OS have a limit of 256 on the maximum number of file
5924 descriptors. Exim was using setrlimit() to set 1000 as a large value
5925 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5926 systems. I've change it so that if it can't get 1000, it tries for 256.
5928 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5929 was an oversight, and furthermore, ever since the addition of extra
5930 controls (e.g. 4.43/32), the checks on when to allow different forms of
5931 "control" were broken. There should now be diagnostics for all cases when a
5932 control that does not make sense is encountered.
5934 9. Added the /retain_sender option to "control=submission".
5936 10. $recipients is now available in the predata ACL (oversight).
5938 11. Tidy the search cache before the fork to do a delivery from a message
5939 received from the command line. Otherwise the child will trigger a lookup
5940 failure and thereby defer the delivery if it tries to use (for example) a
5941 cached ldap connection that the parent has called unbind on.
5943 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5944 of $address_data from the recipient verification was clobbered by the
5945 sender verification.
5947 13. The value of address_data from a sender verification is now available in
5948 $sender_address_data in subsequent conditions in the ACL statement.
5950 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5952 15. Added a new option "connect=<time>" to callout options, to set a different
5955 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5956 was its contents. (It was OK if the option was not defined at all.)
5958 17. A "Completed" log line is now written for messages that are removed from
5959 the spool by the -Mrm option.
5961 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5962 information about exactly what failed.
5964 19. Added -dd to debug only the daemon process.
5966 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5967 handles timeouts, both on the server side and network timeouts. Renamed the
5968 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5970 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5971 It is now set to "smtps".
5973 22. $host_address is now set to the target address during the checking of
5974 ignore_target_hosts.
5976 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5977 being passed; this would have caused $sender_host_name to have been used if
5978 matching the list had actually called for a host name (not very likely,
5979 since this list is usually IP addresses). A host name is now passed as
5982 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5983 code that specifies a non-restarting handler (typically sigaction() in
5984 modern systems) in an attempt to fix a rare and obscure crash bug.
5986 25. Narrowed the window for a race in the daemon that could cause it to ignore
5987 SIGCHLD signals. This is not a major problem, because they are used only to
5988 wake it up if nothing else does.
5990 26. A malformed maildirsize file could cause Exim to calculate negative values
5991 for the mailbox size or file count. Odd effects could occur as a result.
5992 The maildirsize information is now recalculated if the size or filecount
5995 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5996 support for a long time. Removed HAVE_SYS_VFS_H.
5998 28. Installed the latest version of exipick from John Jetmore.
6000 29. In an address list, if the pattern was not a regular expression, an empty
6001 subject address (from a bounce message) matched only if the pattern was an
6002 empty string. Non-empty patterns were not even tested. This was the wrong
6003 because it is perfectly reasonable to use an empty address as part of a
6004 database query. An empty address is now tested by patterns that are
6005 lookups. However, all the other forms of pattern expect the subject to
6006 contain a local part and a domain, and therefore, for them, an empty
6007 address still always fails if the pattern is not itself empty.
6009 30. Exim went into a mad DNS loop when attempting to do a callout where the
6010 host was specified on an smtp transport, and looking it up yielded more
6011 than one IP address.
6013 31. Re-factored the code for checking spool and log partition space into a
6014 function that finds that data and another that does the check. The former
6015 is then used to implement four new variables: $spool_space, $log_space,
6016 $spool_inodes, and $log_inodes.
6018 32. The RFC2047 encoding function was originally intended for short strings
6019 such as real names; it was not keeping to the 75-character limit for
6020 encoded words that the RFC imposes. It now respects the limit, and
6021 generates multiple encoded words if necessary. To be on the safe side, I
6022 have increased the buffer size for the ${rfc2047: expansion operator from
6025 33. It is now permitted to omit both strings after an "if" condition; if the
6026 condition is true, the result is "true". As before, when the second string
6027 is omitted, a false condition yields an empty string. This makes it less
6028 cumbersome to write custom ACL and router conditions.
6030 34. Failure to deliver a bounce message always caused it to be frozen, even if
6031 there was an errors_to setting on the router. The errors_to setting is now
6034 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6035 canonical form (fully expanded) before being placed in
6036 $sender_host_address.
6038 36. The table in the code that translates DNS record types into text (T_A to
6039 "A" for instance) was missing entries for NS and CNAME. It is just possible
6040 that this could have caused confusion if both these types were looked up
6041 for the same domain, because the text type is used as part of Exim's
6042 per-process caching. But the chance of anyone hitting this buglet seems
6045 37. The dnsdb lookup has been extended in a number of ways.
6047 (1) There is a new type, "zns", which walks up the domain tree until it
6048 finds some nameserver records. It should be used with care.
6050 (2) There is a new type, "mxh", which is like "mx" except that it returns
6051 just the host names, not the priorities.
6053 (3) It is now possible to give a list of domains (or IP addresses) to be
6054 looked up. The behaviour when one of the lookups defers can be
6055 controlled by a keyword.
6057 (4) It is now possible to specify the separator character for use when
6058 multiple records are returned.
6060 38. The dnslists ACL condition has been extended: it is now possible to supply
6061 a list of IP addresses and/or domains to be looked up in a particular DNS
6064 39. Added log_selector=+queue_time_overall.
6066 40. When running the queue in the test harness, wait just a tad after forking a
6067 delivery process, to get repeatability of debugging output.
6069 41. Include certificate and key file names in error message when GnuTLS fails
6070 to set them up, because the GnuTLS error message doesn't include the name
6071 of the failing file when there is a problem reading it.
6073 42. Allow both -bf and -bF in the same test run.
6075 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6077 44. The "Exiscan patch" is now merged into the mainline Exim source.
6079 45. Sometimes the final signoff response after QUIT could fail to get
6080 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6081 < 0 before doing a fflush(). This bug looks as though it goes back to the
6082 introduction of TLS in release 3.20, but "sometimes" must have been rare
6083 because the tests only now provoked it.
6085 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6086 (this can affect the format of dates).
6088 47. exim_tidydb, when checking for the continued existence of a message for
6089 which it has found a message-specific retry record, was not finding
6090 messages that were in split spool directories. Consequently, it was
6091 deleting retry records that should have stayed in existence.
6093 48. Steve fixed some bugs in eximstats.
6095 49. The SPA authentication driver was not abandoning authentication and moving
6096 on to the next authenticator when an expansion was forced to fail,
6097 contradicting the general specification for all authenticators. Instead it
6098 was generating a temporary error. It now behaves as specified.
6100 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6101 (the order specifies the preference for clients). The order is now AES256,
6102 AES128, 3DES, ARCFOUR128.
6104 51. Small patch to Sieve code - explicitly set From: when generating an
6107 52. Exim crashed if a remote delivery caused a very long error message to be
6108 recorded - for instance if somebody sent an entire SpamAssassin report back
6109 as a large number of 550 error lines. This bug was coincidentally fixed by
6110 increasing the size of one of Exim's internal buffers (big_buffer) that
6111 happened as part of the Exiscan merge. However, to be on the safe side, I
6112 have made the code more robust (and fixed the comments that describe what
6115 53. Now that there can be additional text after "Completed" in log lines (if
6116 the queue_time_overall log selector is set), a one-byte patch to exigrep
6117 was needed to allow it to recognize "Completed" as not the last thing in
6120 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6121 patch that reportedly fixes this has been added. I am not expert enough to
6122 create a test for it. This is what the patch creator wrote:
6124 "I found a little strange behaviour of ldap code when working with
6125 Windows 2003 AD Domain, where users was placed in more than one
6126 Organization Units. When I tried to give exim partial DN, the exit code
6127 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6128 But simultaneously result of request was absolutely normal ldap result,
6129 so I produce this patch..."
6131 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6132 REFERENCE, so I have modified the code to exclude the patch when that macro
6135 55. Some experimental protocols are using DNS PTR records for new purposes. The
6136 keys for these records are domain names, not reversed IP addresses. The
6137 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6138 leaves it alone. Component reversal etc. now happens only for IP addresses.
6141 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6143 57. Double the size of the debug message buffer (to 2048) so that more of very
6144 long debug lines gets shown.
6146 58. The exicyclog utility now does better if the number of log files to keep
6147 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6149 59. Two changes related to the smtp_active_hostname option:
6151 (1) $smtp_active_hostname is now available as a variable.
6152 (2) The default for smtp_banner uses $smtp_active_hostname instead
6153 of $primary_hostname.
6155 60. The host_aton() function is supposed to be passed a string that is known
6156 to be a valid IP address. However, in the case of IPv6 addresses, it was
6157 not checking this. This is a hostage to fortune. Exim now panics and dies
6158 if the condition is not met. A case was found where this could be provoked
6159 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6160 components; fortuitously, this particular loophole had already been fixed
6161 by change 4.50/55 above.
6163 If there are any other similar loopholes, the new check in host_aton()
6164 itself should stop them being exploited. The report I received stated that
6165 data on the command line could provoke the exploit when Exim was running as
6166 exim, but did not say which command line option was involved. All I could
6167 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6168 running as the user.
6171 61. There was a buffer overflow vulnerability in the SPA authentication code
6172 (which came originally from the Samba project). I have added a test to the
6173 spa_base64_to_bits() function which I hope fixes it.
6176 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6177 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6179 63. The daemon start-up calls getloadavg() while still root for those OS that
6180 need the first call to be done as root, but it missed one case: when
6181 deliver_queue_load_max is set with deliver_drop_privilege. This is
6182 necessary for the benefit of the queue runner, because there is no re-exec
6183 when deliver_drop_privilege is set.
6185 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6186 This has been fixed.
6188 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6189 was in use, was not putting the data itself into the right store pool;
6190 consequently, it could be overwritten for a subsequent message in the same
6191 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6194 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6196 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6197 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6198 the function were treating the return as a boolean value, which happened to
6199 work because 0=false and not-0=true, but is not correct code.
6201 68. The host_aton() function was not handling scoped IPv6 addresses (those
6202 with, for example, "%eth0" on the end) correctly.
6204 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6205 negated items (that is, ~something) in unsigned ints. Some compilers
6206 apparently mutter when there is no cast.
6208 70. If an address verification called from an ACL failed, and did not produce a
6209 user-specific message (i.e. there was only a "system" message), nothing was
6210 put in $acl_verify_message. In this situation, it now puts the system
6213 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6214 session; change 4.31/43 added the unwanted input to the log line - except
6215 that it did not do this in the start of session case. It now does.
6217 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6218 This is wrong and can cause the other end to generate a synchronization
6219 error if it is another Exim or anything else that does the synchronization
6220 check. A QUIT command is no longer sent after a timeout.
6222 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6223 during host lookups.
6225 74. The defer_ok option of callout verification was not working if it was used
6226 when verifying addresses in header lines, that is, for this case:
6228 verify = header_sender/callout=defer_ok
6230 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6231 those file descriptors could be used for SMTP connections. If anything
6232 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6233 could be sent to the SMTP client, causing chaos. The daemon now opens
6234 stdin, stdout, and stderr to /dev/null when it puts itself into the
6237 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6238 log by default. The user can override this with suitable Perl magic.
6240 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6241 the log message when discard triggers, was not working for the DATA ACL or
6242 for the non-SMTP ACL.
6244 78. Error message wording change in sieve.c.
6246 79. If smtp_accept_max_per_host was set, the number of connections could be
6247 restricted to fewer than expected, because the daemon was trying to set up
6248 a new connection before checking whether the processes handling previous
6249 connections had finished. The check for completed processes is now done
6250 earlier. On busy systems, this bug wouldn't be noticed because something
6251 else would have woken the daemon, and it would have reaped the completed
6254 80. If a message was submitted locally by a user whose login name contained one
6255 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6256 It caused a spool format error. I have fixed the spool reading code. A
6257 related problem was that the "from" clause in the Received: line became
6258 illegal because of the space(s). It is now covered by ${quote_local_part.
6260 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6263 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6265 83. Patch from Sieve maintainer to fix unterminated string problem in
6266 "vacation" handling.
6268 84. Some minor changes to the Linux configuration files to help with other
6269 OS variants using glibc.
6271 85. One more patch for Sieve to update vacation handling to latest spec.
6274 ----------------------------------------------------
6275 See the note above about the 4.44 and 4.50 releases.
6276 ----------------------------------------------------
6282 1. Change 4.43/35 introduced a bug that caused file counts to be
6283 incorrectly computed when quota_filecount was set in an appendfile
6286 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6287 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6290 3. Give more explanation in the error message when the command for a transport
6291 filter fails to execute.
6293 4. There are several places where Exim runs a non-Exim command in a
6294 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6295 was being done only for the command run by the queryprogram router. It is
6296 now done for all such subprocesses. The other cases are: ${run, transport
6297 filters, and the commands run by the lmtp and pipe transports.
6299 5. Some older OS have a limit of 256 on the maximum number of file
6300 descriptors. Exim was using setrlimit() to set 1000 as a large value
6301 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6302 systems. I've change it so that if it can't get 1000, it tries for 256.
6304 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6305 was an oversight, and furthermore, ever since the addition of extra
6306 controls (e.g. 4.43/32), the checks on when to allow different forms of
6307 "control" were broken. There should now be diagnostics for all cases when a
6308 control that does not make sense is encountered.
6310 7. $recipients is now available in the predata ACL (oversight).
6312 8. Tidy the search cache before the fork to do a delivery from a message
6313 received from the command line. Otherwise the child will trigger a lookup
6314 failure and thereby defer the delivery if it tries to use (for example) a
6315 cached ldap connection that the parent has called unbind on.
6317 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6318 of $address_data from the recipient verification was clobbered by the
6319 sender verification.
6321 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6322 was its contents. (It was OK if the option was not defined at all.)
6324 11. A "Completed" log line is now written for messages that are removed from
6325 the spool by the -Mrm option.
6327 12. $host_address is now set to the target address during the checking of
6328 ignore_target_hosts.
6330 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6331 being passed; this would have caused $sender_host_name to have been used if
6332 matching the list had actually called for a host name (not very likely,
6333 since this list is usually IP addresses). A host name is now passed as
6336 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6337 code that specifies a non-restarting handler (typically sigaction() in
6338 modern systems) in an attempt to fix a rare and obscure crash bug.
6340 15. Narrowed the window for a race in the daemon that could cause it to ignore
6341 SIGCHLD signals. This is not a major problem, because they are used only to
6342 wake it up if nothing else does.
6344 16. A malformed maildirsize file could cause Exim to calculate negative values
6345 for the mailbox size or file count. Odd effects could occur as a result.
6346 The maildirsize information is now recalculated if the size or filecount
6349 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6350 support for a long time. Removed HAVE_SYS_VFS_H.
6352 18. Updated exipick to current release from John Jetmore.
6354 19. Allow an empty sender to be matched against a lookup in an address list.
6355 Previously the only cases considered were a regular expression, or an
6358 20. Exim went into a mad DNS lookup loop when doing a callout where the
6359 host was specified on the transport, if the DNS lookup yielded more than
6362 21. The RFC2047 encoding function was originally intended for short strings
6363 such as real names; it was not keeping to the 75-character limit for
6364 encoded words that the RFC imposes. It now respects the limit, and
6365 generates multiple encoded words if necessary. To be on the safe side, I
6366 have increased the buffer size for the ${rfc2047: expansion operator from
6369 22. Failure to deliver a bounce message always caused it to be frozen, even if
6370 there was an errors_to setting on the router. The errors_to setting is now
6373 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6374 canonical form (fully expanded) before being placed in
6375 $sender_host_address.
6377 24. Updated eximstats to version 1.33
6379 25. Include certificate and key file names in error message when GnuTLS fails
6380 to set them up, because the GnuTLS error message doesn't include the name
6381 of the failing file when there is a problem reading it.
6383 26. Expand error message when OpenSSL has problems setting up cert/key files.
6386 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6387 (this can affect the format of dates).
6389 28. exim_tidydb, when checking for the continued existence of a message for
6390 which it has found a message-specific retry record, was not finding
6391 messages that were in split spool directories. Consequently, it was
6392 deleting retry records that should have stayed in existence.
6394 29. eximstats updated to version 1.35
6395 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6396 1.35 - bugfix such that pie charts by volume are generated correctly
6398 30. The SPA authentication driver was not abandoning authentication and moving
6399 on to the next authenticator when an expansion was forced to fail,
6400 contradicting the general specification for all authenticators. Instead it
6401 was generating a temporary error. It now behaves as specified.
6403 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6404 (the order specifies the preference for clients). The order is now AES256,
6405 AES128, 3DES, ARCFOUR128.
6407 31. Small patch to Sieve code - explicitly set From: when generating an
6410 32. Exim crashed if a remote delivery caused a very long error message to be
6411 recorded - for instance if somebody sent an entire SpamAssassin report back
6412 as a large number of 550 error lines. This bug was coincidentally fixed by
6413 increasing the size of one of Exim's internal buffers (big_buffer) that
6414 happened as part of the Exiscan merge. However, to be on the safe side, I
6415 have made the code more robust (and fixed the comments that describe what
6418 33. Some experimental protocols are using DNS PTR records for new purposes. The
6419 keys for these records are domain names, not reversed IP addresses. The
6420 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6421 leaves it alone. Component reversal etc. now happens only for IP addresses.
6424 34. The host_aton() function is supposed to be passed a string that is known
6425 to be a valid IP address. However, in the case of IPv6 addresses, it was
6426 not checking this. This is a hostage to fortune. Exim now panics and dies
6427 if the condition is not met. A case was found where this could be provoked
6428 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6429 components; fortuitously, this particular loophole had already been fixed
6430 by change 4.50/55 or 4.44/33 above.
6432 If there are any other similar loopholes, the new check in host_aton()
6433 itself should stop them being exploited. The report I received stated that
6434 data on the command line could provoke the exploit when Exim was running as
6435 exim, but did not say which command line option was involved. All I could
6436 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6437 running as the user.
6440 35. There was a buffer overflow vulnerability in the SPA authentication code
6441 (which came originally from the Samba project). I have added a test to the
6442 spa_base64_to_bits() function which I hope fixes it.
6445 36. The daemon start-up calls getloadavg() while still root for those OS that
6446 need the first call to be done as root, but it missed one case: when
6447 deliver_queue_load_max is set with deliver_drop_privilege. This is
6448 necessary for the benefit of the queue runner, because there is no re-exec
6449 when deliver_drop_privilege is set.
6451 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6452 was in use, was not putting the data itself into the right store pool;
6453 consequently, it could be overwritten for a subsequent message in the same
6454 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6457 38. Sometimes the final signoff response after QUIT could fail to get
6458 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6459 < 0 before doing a fflush(). This bug looks as though it goes back to the
6460 introduction of TLS in release 3.20, but "sometimes" must have been rare
6461 because the tests only now provoked it.
6467 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6468 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6469 Now it returns an integer. A number of calls were still expecting a T/F
6470 return. Fortuitously, in all cases, the tests worked in OK situations,
6471 which is the norm. However, things would have gone wrong on any write
6472 failures on the smtp file descriptor. This function is used when sending
6473 messages over SMTP and also when doing verify callouts.
6475 2. When Exim is called to do synchronous delivery of a locally submitted
6476 message (the -odf or -odi options), it no longer closes stderr before doing
6479 3. Implemented the mua_wrapper option.
6481 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6483 5. Implemented the functions header_remove(), header_testname(),
6484 header_add_at_position(), and receive_remove_recipient(), and exported them
6487 6. If an ACL "warn" statement specified the addition of headers, Exim already
6488 inserted X-ACL-Warn: at the start if there was no header name. However, it
6489 was not making this test for the second and subsequent header lines if
6490 there were newlines in the string. This meant that an invalid header could
6491 be inserted if Exim was badly configured.
6493 7. Allow an ACL "warn" statement to add header lines at the start or after all
6494 the Received: headers, as well as at the end.
6496 8. Added the rcpt_4xx retry error code.
6498 9. Added postmaster_mailfrom=xxx to callout verification option.
6500 10. Added mailfrom=xxxx to the callout verification option, for verify=
6503 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6504 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6506 12. Inserted some casts to stop certain compilers warning when using pointer
6507 differences as field lengths or precisions in printf-type calls (mostly
6508 affecting debugging statements).
6510 13. Added optional readline() support for -be (dynamically loaded).
6512 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6513 same clock tick as a message's arrival, so that its received time was the
6514 same as the "first fail" time on the retry record, and that message
6515 remained on the queue past the ultimate address timeout, every queue runner
6516 would try a delivery (because it was past the ultimate address timeout) but
6517 after another failure, the ultimate address timeout, which should have then
6518 bounced the address, did not kick in. This was a "< instead of <=" error;
6519 in most cases the first failure would have been in the next clock tick
6520 after the received time, and all would be well.
6522 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6523 being recognized when the domain list was tested by the match_domain
6524 condition in an expansion string.
6526 16. Added the ${str2b64: operator.
6528 17. Exim was always calling setrlimit() to set a large limit for the number of
6529 processes, without checking whether the existing limit was already
6530 adequate. (It did check for the limit on file descriptors.) Furthermore,
6531 errors from getrlimit() and setrlimit() were being ignored. Now they are
6532 logged to the main and panic logs, but Exim does carry on, to try to do its
6533 job under whatever limits there are.
6535 18. Imported PCRE 5.0.
6537 19. Trivial typo in log message " temporarily refused connection" (the leading
6540 20. If the log selector return_path_on_delivery was set and an address was
6541 redirected to /dev/null, the delivery process crashed because it assumed
6542 that a return path would always be set for a "successful" delivery. In this
6543 case, the whole delivery is bypassed as an optimization, and therefore no
6546 21. Internal re-arrangement: the function for sending a challenge and reading
6547 a response while authentication was assuming a zero-terminated challenge
6548 string. It's now changed to take a pointer and a length, to allow for
6549 binary data in such strings.
6551 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6553 23. Exim was not respecting finduser_retries when seeking the login of the
6554 uid under which it was called; it was always trying 10 times. (The default
6555 setting of finduser_retries is zero.) Also, it was sleeping after the final
6556 failure, which is pointless.
6558 24. Implemented tls_on_connect_ports.
6560 25. Implemented acl_smtp_predata.
6562 26. If the domain in control=submission is set empty, Exim assumes that the
6563 authenticated id is a complete email address when it generates From: or
6564 Sender: header lines.
6566 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6567 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6568 chown and chgrp in /bin and hostname in /usr/bin.
6570 28. Exim was keeping the "process log" file open after each use, just as it
6571 does for the main log. This opens the possibility of it remaining open for
6572 long periods when the USR1 signal hits a daemon. Occasional processlog
6573 errors were reported, that could have been caused by this. Anyway, it seems
6574 much more sensible not to leave this file open at all, so that is what now
6577 29. The long-running daemon process does not normally write to the log once it
6578 has entered its main loop, and it closes the log before doing so. This is
6579 so that log files can straightforwardly be renamed and moved. However,
6580 there are a couple of unusual error situations where the daemon does write
6581 log entries, and I had neglected to close the log afterwards.
6583 30. The text of an SMTP error response that was received during a remote
6584 delivery was being truncated at 512 bytes. This is too short for some of
6585 the long messages that one sometimes sees. I've increased the limit to
6588 31. It is now possible to make retry rules that apply only when a message has a
6589 specific sender, in particular, an empty sender.
6591 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6592 it possible to be selective about when SMTP synchronization is enforced.
6594 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6596 32. Implemented hosts_connection_nolog.
6598 33. Added an ACL for QUIT.
6600 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6603 35. Added mailbox_size and mailbox_filecount to appendfile.
6605 36. Added control = no_multiline_responses to ACLs.
6607 37. There was a bug in the logic of the code that waits for the clock to tick
6608 in the case where the clock went backwards by a substantial amount such
6609 that the microsecond fraction of "now" was more than the microsecond
6610 fraction of "then" (but the whole seconds number was less).
6612 38. Added support for the libradius Radius client library this is found on
6613 FreeBSD (previously only the radiusclient library was supported).
6619 1. When certain lookups returned multiple values in the form name=value, the
6620 quoting of the values was not always being done properly. Specifically:
6621 (a) If the value started with a double quote, but contained no whitespace,
6623 (b) If the value contained whitespace other than a space character (i.e.
6624 tabs or newlines or carriage returns) it was not quoted.
6625 This fix has been applied to the mysql and pgsql lookups by writing a
6626 separate quoting function and calling it from the lookup code. The fix
6627 should probably also be applied to nisplus, ibase and oracle lookups, but
6628 since I cannot test any of those, I have not disturbed their existing code.
6630 2. A hit in the callout cache for a specific address caused a log line with no
6631 reason for rejecting RCPT. Now it says "Previous (cached) callout
6632 verification failure".
6634 3. There was an off-by-one bug in the queryprogram router. An over-long
6635 return line was truncated at 256 instead of 255 characters, thereby
6636 overflowing its buffer with the terminating zero. As well as fixing this, I
6637 have increased the buffer size to 1024 (and made a note to document this).
6639 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6640 when Exim is waiting for an SMTP response from a remote server, Exim
6641 restarts its select() call on the socket, thereby resetting its timeout.
6642 This is not a problem when such interrupts are rare. Somebody set up a cron
6643 job to run exiwhat every 2 minutes, which is less than the normal select()
6644 timeout (5 or 10 minutes). This meant that the select() timeout never
6645 kicked in because it was always reset. I have fixed this by comparing the
6646 time when an interrupt arrives with the time at the start of the first call
6647 to select(). If more time than the timeout has elapsed, the interrupt is
6648 treated as a timeout.
6650 5. Some internal re-factoring in preparation for the addition of Sieve
6651 extensions (by MH). In particular, the "personal" test is moved to a
6652 separate function, and given an option for scanning Cc: and Bcc: (which is
6653 not set for Exim filters).
6655 6. When Exim created an email address using the login of the caller as the
6656 local part (e.g. when creating a From: or Sender: header line), it was not
6657 quoting the local part when it contained special characters such as @.
6659 7. Installed new OpenBSD configuration files.
6661 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6662 try to make them clearer.
6664 9. Callout options, other than the timeout value, were being ignored when
6665 verifying sender addresses in header lines. For example, when using
6667 verify = header_sender/callout=no_cache
6669 the cache was (incorrectly) being used.
6671 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6672 only the Cygwin environment.
6674 11. When return_path_on_delivery was set as a log selector, if different remote
6675 addresses in the same message used different return paths and parallel
6676 remote delivery occurred, the wrong values would sometimes be logged.
6677 (Whenever a remote delivery process finished, the return path value from
6678 the most recently started remote delivery process was logged.)
6680 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6681 lines when AUTH and/or TLS are in use. This is the "received protocol"
6682 field. Exim used to use "asmtp" for authenticated SMTP, without any
6683 indication (in the protocol name) for TLS use. Now it follows the RFC and
6684 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6685 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6686 names appear in log lines as well as in Received: header lines.
6688 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6689 extensions, and comparison tests, and to fix some bugs.
6691 14. Changes to the "personal" filter test:
6693 (1) The test was buggy in that it was just doing the equivalent of
6694 "contains" tests on header lines. For example, if a user's address was
6695 anne@some.where, the "personal" test would incorrectly be true for
6697 To: susanne@some.where
6699 This test is now done by extracting each address from the header in turn,
6700 and checking the entire address. Other tests that are part of "personal"
6701 are now done using regular expressions (for example, to check local parts
6702 of addresses in From: header lines).
6704 (2) The list of non-personal local parts in From: addresses has been
6705 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6706 taken from the Sieve specification recommendations.
6708 (3) If the message contains any header line starting with "List-" it is
6709 treated as non-personal.
6711 (4) The test for "circular" in the Subject: header line has been removed
6712 because it now seems ill-conceived.
6714 15. Minor typos in src/EDITME comments corrected.
6716 16. Installed latest exipick from John Jetmore.
6718 17. If headers_add on a router specified a text string that was too long for
6719 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6720 of string_sprintf() is now avoided.
6722 18. $message_body_size was not set (it was always zero) when running the DATA
6723 ACL and the local_scan() function.
6725 19. For the "mail" command in an Exim filter, no default was being set for
6726 the once_repeat time, causing a random time value to be used if "once" was
6727 specified. (If the value happened to be <= 0, no repeat happened.) The
6728 default is now 0s, meaning "never repeat". The "vacation" command was OK
6729 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6730 (I found it when inspecting the code).
6732 20. There is now an overall timeout for performing a callout verification. It
6733 defaults to 4 times the callout timeout, which applies to individual SMTP
6734 commands during the callout. The overall timeout applies when there is more
6735 than one host that can be tried. The timeout is checked before trying the
6736 next host. This prevents very long delays if there are a large number of
6737 hosts and all are timing out (e.g. when the network connections are timing
6738 out). The value of the overall timeout can be changed by specifying an
6739 additional sub-option for "callout", called "maxwait". For example:
6741 verify = sender/callout=5s,maxwait=20s
6743 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6744 to the end before writing, but this should make it even safer).
6746 22. Exim was forgetting that it had advertised PIPELINING for the second and
6747 subsequent messages on an SMTP connection. It was also not resetting its
6748 memory on STARTTLS and an internal HELO.
6750 23. When Exim logs an SMTP synchronization error within a session, it now
6751 records whether PIPELINING has been advertised or not.
6753 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6754 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6755 rather than long int.
6757 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6759 26. Added the never_mail option to autoreply.
6765 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6766 crash if the getsockname() call failed; this can happen if a connection is
6767 closed very soon after it is established. The problem was simply in the
6768 order in which certain operations were done, causing Exim to try to write
6769 to the SMTP stream before it had set up the file descriptor. The bug has
6770 been fixed by making things happen in the correct order.
6776 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6777 before the connection was closed, thus losing the rejection response.
6779 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6780 some early Solaris releases, but causes trouble in current releases where
6781 socklen_t is defined.
6783 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6786 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6789 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6790 that releasing the top of it at the end releases what was used for sub-
6791 expansions (unless the block got too big). However, discard this block if
6792 the first thing is a variable or header, so that we can use its block when
6793 it is dynamic (useful for very large $message_headers, for example).
6795 6. Lookups now cache *every* query, not just the most recent. A new, separate
6796 store pool is used for this. It can be recovered when all lookup caches are
6797 flushed. Lookups now release memory at the end of their result strings.
6798 This has involved some general refactoring of the lookup sources.
6800 7. Some code has been added to the store_xxx() functions to reduce the amount
6801 of flapping under certain conditions.
6803 8. log_incoming_interface used to affect only the <= reception log lines. Now
6804 it causes the local interface and port to be added to several more SMTP log
6805 lines, for example "SMTP connection from", and rejection lines.
6807 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6809 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6811 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6812 could be overwritten at the end of the current message (or the start of a
6813 new message if it was set in a HELO ACL). The value is now preserved for
6814 the duration of the SMTP connection.
6816 12. If a transport had a headers_rewrite setting, and a matching header line
6817 contained an unqualified address, that address was qualified, even if it
6818 did not match any rewriting rules. The underlying bug was that the values
6819 of the flags that permit the existence of unqualified sender and recipient
6820 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6821 non-local messages, and by -bnq for local messages) were not being
6822 preserved with the message after it was received.
6824 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6825 "next input=" as part of the text comprising the host identity instead of
6826 the correct text. The code was using the same buffer for two different
6827 strings. However, depending on which order the printing function evaluated
6828 its arguments, the bug did not always show up. Under Linux, for example, my
6829 test suite worked just fine.
6831 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6832 doesn't work with some older versions of Perl. It has been changed to "my",
6833 which in any case is probably the better facility to use.
6835 15. A really picky compiler found some instances of statements for creating
6836 error messages that either had too many or two few arguments for the format
6839 16. The size of the buffer for calls to the DNS resolver has been increased
6840 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6841 for addresses that have a lot of PTR records. This alleviates a problem; it
6842 does not fully solve it.
6844 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6845 buffer now truncates the list and logs the incident, which is the same
6846 action as happens when Exim is looking up a host name and its aliases.
6847 Previously in this situation something unpredictable would happen;
6848 sometimes it was "internal error: store_reset failed".
6850 18. If a server dropped the connection unexpectedly when an Exim client was
6851 using GnuTLS and trying to read a response, the client delivery process
6852 crashed while trying to generate an error log message.
6854 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6855 string, for example:
6857 warn message = H1: something\nH2: something
6859 the text was added as a single header line from Exim's point of view
6860 though it ended up OK in the delivered message. However, searching for the
6861 second and subsequent header lines using $h_h2: did not work. This has been
6862 fixed. Similarly, if a system filter added multiple headers in this way,
6863 the routers could not see them.
6865 20. Expanded the error message when iplsearch is called with an invalid key to
6866 suggest using net-iplsearch in a host list.
6868 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6869 ACLs are no longer actually imposed (and a message to that effect is
6872 22. If a "gecos" field in a passwd entry contained escaped characters, in
6873 particular, if it contained a \" sequence, Exim got it wrong when building
6874 a From: or a Sender: header from that name. A second bug also caused
6875 incorrect handling when an unquoted " was present following a character
6876 that needed quoting.
6878 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6879 was not being matched caselessly.
6881 24. Arranged for all hyphens in the exim.8 source to be escaped with
6884 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6885 properly. Recipient callout cache records were still being keyed to include
6886 the sender, even when use_sender was set false. This led to far more
6887 callouts that were necessary. The sender is no longer included in the key
6888 when use_sender is false.
6890 26. Added "control = submission" modifier to ACLs.
6892 27. Added the ${base62d: operator to decode base 62 numbers.
6894 28. dnsdb lookups can now access SRV records.
6896 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6897 the configuration file.
6899 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6900 (-v) mode. This makes the output for a verbose queue run more intelligible.
6902 31. Added a use_postmaster feature to recipient callouts.
6904 32. Added the $body_zerocount variable, containing the number of binary zero
6905 bytes in the message body.
6907 33. The time of last modification of the "new" subdirectory is now used as the
6908 "mailbox time last read" when there is a quota error for a maildir
6911 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6913 35. Added +ignore_unknown as a special item in host lists.
6915 36. Code for decoding IPv6 addresses in host lists is now included, even if
6916 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6917 address was recognized as an IP address, but was then not correctly decoded
6918 into binary, causing unexpected and incorrect effects when compared with
6925 1. Very minor rewording of debugging text in manualroute to say "list of
6926 hosts" instead of "hostlist".
6928 2. If verify=header_syntax was set, and a header line with an unqualified
6929 address (no domain) and a large number of spaces between the end of the
6930 name and the colon was received, the reception process suffered a buffer
6931 overflow, and (when I tested it) crashed. This was caused by some obsolete
6932 code that should have been removed. The fix is to remove it!
6934 3. When running in the test harness, delay a bit after writing a bounce
6935 message to get a bit more predictability in the log output.
6937 4. Added a call to search_tidyup() just before forking a reception process. In
6938 theory, someone could use a lookup in the expansion of smtp_accept_max_
6939 per_host which, without the tidyup, could leave open a database connection.
6941 5. Added the variables $recipient_data and $sender_data which get set from a
6942 lookup success in an ACL "recipients" or "senders" condition, or a router
6943 "senders" option, similar to $domain_data and $local_part_data.
6945 6. Moved the writing of debug_print from before to after the "senders" test
6948 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6949 problems for message scanning, either using a data ACL, or using
6950 local_scan() because the Received: header was not generated till after they
6951 were called (in order to set the time as the time of reception completion).
6952 I have revised the way this works. The header is now generated after the
6953 body is received, but before the ACL or local_scan() are called. After they
6954 are run, the timestamp in the header is updated.
6960 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6961 before starting a queue runner without re-exec. This happened only when
6962 deliver_drop_privilege was set or when the Exim user was set to root. The
6963 effect of the bug was that timeouts during subsequent deliveries caused
6964 crashes instead of being properly handled. The handler is now left at its
6965 default (and expected) setting.
6967 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6968 message, again when deliver_drop_privilege is set or Exim is run as root.
6969 The bug described in (1) was not present in this case, but the tidying up
6970 of the other signals was missing. I have made the two cases consistent.
6972 3. The ignore_target_hosts setting on a manualroute router was being ignored
6973 for hosts that were looked up using the /MX notation.
6975 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6978 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6979 operated on the sender address. After changing the $sender_address to <>
6980 for the sender address verify, Exim was re-instated it as the original
6981 (before rewriting) address, but remembering that it had rewritten it, so it
6982 wasn't rewriting it again. This bug also had the effect of breaking the
6983 sender address verification caching when the sender address was rewritten.
6985 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6986 This has been changed so that if the ip literal address matches
6987 ignore_target_hosts, the router declines.
6989 7. Added expansion conditions match_domain, match_address, and match_local_
6990 part (NOT match_host).
6992 8. The placeholder for the Received: header didn't have a length field set.
6994 9. Added code to Exim itself and to exim_lock to test for a specific race
6995 condition that could lead to file corruption when using MBX delivery. The
6996 issue is with the lockfile that is created in /tmp. If this file is removed
6997 after a process has opened it but before that process has acquired a lock,
6998 there is the potential for a second process to recreate the file and also
6999 acquire a lock. This could lead to two Exim processes writing to the file
7000 at the same time. The added code performs the same test as UW imapd; it
7001 checks after acquiring the lock that its file descriptor still refers to
7002 the same named file.
7004 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7005 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7008 11. Added the smtp_active_hostname option. If used, this will typically be made
7009 to depend on the incoming interface address. Because $interface_address is
7010 not set up until the daemon has forked a reception process, error responses
7011 that can happen earlier (such as "too many connections") no longer contain
7014 12. If an expansion in a condition on a "warn" statement fails because a lookup
7015 defers, the "warn" statement is abandoned, and the next ACL statement is
7016 processed. Previously this caused the whole ACL to be aborted.
7018 13. Added the iplsearch lookup type.
7020 14. Added ident_timeout as a log selector.
7022 15. Added tls_certificate_verified as a log selector.
7024 16. Added a global option tls_require_ciphers (compare the smtp transport
7025 option of the same name). This controls incoming TLS connections.
7027 17. I finally figured out how to make tls_require_ciphers do a similar thing
7028 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7029 before starting the TLS session.
7031 18. Tabs are now shown as \t in -bP output.
7033 19. If the log selector return_path_on_delivery was set, Exim crashed when
7034 bouncing a message because it had too many Received: header lines.
7036 20. If two routers both had headers_remove settings, and the first one included
7037 a superfluous trailing colon, the final name in the first list and the
7038 first name in the second list were incorrectly joined into one item (with a
7039 colon in the middle).
7045 1. Added -C and -D options to the exinext utility, mainly to make it easier
7046 to include in the automated testing, but these could be helpful when
7047 multiple configurations are in use.
7049 2. The exinext utility was not formatting the output nicely when there was
7050 an alternate port involved in the retry record key, nor when there was a
7051 message id as well (for retries that were specific to a specific message
7052 and a specific host). It was also confused by IPv6 addresses, because of
7053 the additional colons they contain. I have fixed the IPv4 problem, and
7054 patched it up to do a reasonable job for IPv6.
7056 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7057 delivery, the log line now contains "pipelined" if PIPELINING was used.
7059 4. An SMTP transport process used to panic and die if the bind() call to set
7060 an explicit outgoing interface failed. This has been changed; it is now
7061 treated in the same way as a connect() failure.
7063 5. A reference to $sender_host_name in the part of a conditional expansion
7064 that was being skipped was still causing a DNS lookup. This no longer
7067 6. The def: expansion condition was not recognizing references to header lines
7068 that used bh_ and bheader_.
7070 7. Added the _cache feature to named lists.
7072 8. The code for checking quota_filecount in the appendfile transport was
7073 allowing one more file than it should have been.
7075 9. For compatibility with Sendmail, the command line option
7083 and sets the incoming protocol and host name (for trusted callers). The
7084 host name and its colon can be omitted when only the protocol is to be set.
7085 Note the Exim already has two private options, -pd and -ps, that refer to
7086 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7087 "s", but I don't think that's a major issue.
7089 10. A number of refactoring changes to the code, none of which should affect
7092 (a) The number of logging options was getting close to filling up the
7093 32-bit word that was used as a bit map. I have split them into two classes:
7094 those that are passed in the argument to log_write(), and those that are
7095 only ever tested independently outside of that function. These are now in
7096 separate 32-bit words, so there is plenty of room for expansion again.
7097 There is no change in the user interface or the logging behaviour.
7099 (b) When building, for example, log lines, the code previously used a
7100 macro that called string_cat() twice, in order to add two strings. This is
7101 not really sufficiently general. Furthermore, there was one instance where
7102 it was actually wrong because one of the argument was used twice, and in
7103 one call a function was used. (As it happened, calling the function twice
7104 did not affect the overall behaviour.) The macro has been replaced by a
7105 function that can join an arbitrary number of extra strings onto a growing
7108 (c) The code for expansion conditions now uses a table and a binary chop
7109 instead of a serial search (which was left over from when there were very
7110 few conditions). Also, it now recognizes conditions like "pam" even when
7111 the relevant support is not compiled in: a suitably worded error message is
7112 given if an attempt is made to use such a condition.
7114 11. Added ${time_interval:xxxxx}.
7116 12. A bug was causing one of the ddress fields not to be passed back correctly
7117 from remote delivery subprocesses. The field in question was not being
7118 subsequently used, so this caused to problems in practice.
7120 13. Added new log selectors queue_time and deliver_time.
7122 14. Might have fixed a bug in maildirsizefile handling that threw up
7123 "unexpected character" debug warnings, and recalculated the data
7124 unnecessarily. In any case, I expanded the warning message to give more
7127 15. Added the message "Restricted characters in address" to the statements in
7128 the default ACL that block characters like @ and % in local parts.
7130 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7131 Three changes have been made:
7133 (a) There was a serious bug; a negative response to MAIL caused the whole
7134 recipient domain to be cached as invalid, thereby blocking all messages
7135 to all local parts at the same domain, from all senders. This bug has
7136 been fixed. The domain is no longer cached after a negative response to
7137 MAIL if the sender used is not empty.
7139 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7142 (c) A new callout option, "use_sender" has been added for people who want
7143 the modified behaviour.
7149 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7152 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7153 indeed breaks things for older releases.
7155 3. Added additional logging to the case where there is a problem reading data
7156 from a filter that is running in a subprocess using a pipe, in order to
7157 try to track down a specific problem.
7159 4. Testing facility fudge: when running in the test harness and attempting
7160 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7161 getting "No route to host". Convert this to a timeout.
7163 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7166 6. Some OS don't have socklen_t but use size_t instead. This affects the
7167 fifth argument of getsockopt() amongst other things. This is now
7168 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7169 can be set for individual OS. I have set it for SunOS5, OSF1, and
7170 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7171 some earlier ones do not.
7173 7. Change 4.30/15 was not doing the test caselessly.
7175 8. The standard form for an IPv6 address literal was being rejected by address
7176 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7177 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7178 this, as well as the form without the "IPv6" on the front (but only when
7179 address literals are enabled, of course).
7181 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7183 10. Exim crashed if a message with an empty sender address specified by -f
7184 encountered a router with an errors_to setting. This could be provoked only
7185 by a command such as
7189 where an empty string was supplied; "<>" did not hit this bug.
7191 11. Installed PCRE release 4.5.
7193 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7194 remained set. It is now erased.
7196 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7197 times from message ids (which are base 36 rather than the normal 62).
7199 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7200 were being counted as actual protocol errors, and logged if the log
7201 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7202 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7203 and DATA following a set of rejected RCPTs do not count as protocol errors.
7204 In other words, Exim assumes they were pipelined, though this may not
7205 actually be the case. Of course, in all cases the client gets an
7206 appropriate error code.
7208 15. If a lookup fails in an ACL condition, a message about the failure may
7209 be available; it is used if testing the ACL cannot continue, because most
7210 such messages specify what the cause of the deferral is. However, some
7211 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7212 that caused an old message to be retained and used if a later statement
7213 caused a defer, replacing the real cause of the deferral.
7215 16. If an IP address had so many PTR records that the DNS lookup buffer
7216 was not large enough to hold them, Exim could crash while trying to process
7217 the truncated data. It now detects and logs this case.
7219 17. Further to 4.21/58, another change has been made: if (and only if) the
7220 first line of a message (the first header line) ends with CRLF, a bare LF
7221 in a subsequent header line has a space inserted after it, so as not to
7222 terminate the header.
7224 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7225 unnecessarily, used atoi() instead of strtol(), and didn't check the
7226 termination when getting file sizes from file names by regex.
7228 19. Completely re-implemented the support for maildirsize files, in the light
7229 of a number of problems with the previous contributed implementation
7230 (4.30/29). In particular:
7232 . If the quota is zero, the maildirsize file is maintained, but no quota is
7235 . If the maildir directory does not exist, it is created before any attempt
7236 to write a maildirsize file.
7238 . The quota value in the file is just a cache; if the quota is changed in
7239 the transport, the new value overrides.
7241 . A regular expression is available for excluding directories from the
7244 20. The autoreply transport checks the characters in options that define the
7245 message's headers; it allows continued headers, but it was checking with
7246 isspace() after an embedded newline instead of explicitly looking for a
7249 21. If all the "regular" hosts to which an address was routed had passed their
7250 expiry times, and had not reached their retry times, the address was
7251 bounced, even if fallback hosts were defined. Now Exim should go on to try
7254 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7255 equivalent code in the SMTP transport. Some hosts send humungous responses
7256 to HELO/EHLO, more than 1024 it seems.
7258 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7259 gives compiler warnings in some environments. I've now done it "properly",
7262 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7263 (because of problems with the built-in one) was declared to return uschar *
7264 instead of char *, causing compiler failure.
7266 25. Fixed a file descriptor leak when processing alias/forward files.
7268 26. Fixed a minor format string issue in dbfn.c.
7270 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7272 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7273 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7276 29. When a sender address is verified, it is cached, to save repeating the test
7277 when there is more than one recipient in a message. However, when the
7278 verification involves a callout, it is possible for different callout
7279 options to be set for different recipients. It is too complicated to keep
7280 track of this in the cache, so now Exim always runs a verification when a
7281 callout is required, relying on the callout cache for the optimization.
7282 The overhead is duplication of the address routing, but this should not be
7285 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7286 to be verified with callout=postmaster, and the main callout worked but the
7287 postmaster check failed, the verification correctly failed. However, if a
7288 subsequent RCPT command asked for sender verification *without* the
7289 postmaster check, incorrect caching caused this verification also to fail,
7292 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7293 it was not caching the DNS options (qualify_single, search_parents) that
7294 were used when the lookup failed. A subsequent lookup with different
7295 options therefore always gave the same answer, though there were cases
7296 where it should not have. (Example: a "domains = !$mx_any" option on a
7297 dnslookup router: the "domains" option is always processed without any
7298 widening, but the router might have qualify_single set.) Now Exim uses the
7299 cached value only when the same options are set.
7301 32. Added John Jetmore's "exipick" utility to the distribution.
7303 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7304 than a timeout (e.g. a certificate is required, and is not provided), an
7305 Exim server now closes the connection immediately. Previously it waited for
7306 the client to close - but if the client is SSL, it seems that they each
7307 wait for each other, leading to a delay before one of them times out.
7309 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7310 maintained 0.8.x compatibility because I don't think many are using it, and
7311 it is clearly obsolete.
7313 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7316 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7317 client certificate was expired. A simple patch fixes this, though I don't
7318 understand the full logic of why the verify callback is called multiple
7321 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7322 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7323 which causes problems with some clients (such as the Certicom SSL Plus
7324 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7325 disables the coutermeasure allowing Eudora to connect."
7327 38. Exim was not checking that a write() to a log file succeeded. This could
7328 lead to Bad Things if a log got too big, in particular if it hit a file
7329 size limit. Exim now panics and dies if it cannot write to a log file, just
7330 as it does if it cannot open a log file.
7332 39. Modified OS/Makefile-Linux so that it now contains
7334 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7336 The two -D definitions ensure that Exim is compiled with large file
7337 support, which makes it possible to handle log files that are bigger than
7340 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7341 instance) a domain was checked against a named list that involved a lookup,
7342 causing $domain_data to be set, then another domain was checked against the
7343 same list, then the first domain was re-checked, the value of $domain_data
7344 after the final check could be wrong. In particular, if the second check
7345 failed, it could be set empty. This bug probably also applied to
7348 41. The strip_trailing_dot option was not being applied to the address given
7349 with the -f command-line option.
7351 42. The code for reading a message's header from the spool was incrementing
7352 $received_count, but never initializing it. This meant that the value was
7353 incorrect (doubled) while delivering a message in the same process in which
7354 it was received. In the most common configuration of Exim, this never
7355 happens - a fresh exec is done - but it can happen when
7356 deliver_drop_privilege is set.
7358 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7359 it now includes up to 150 characters of the unexpected data in the log
7362 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7363 and building data strings. The size of both of these buffers was 10 000
7364 bytes - far larger than anybody would *ever* want, thought I. Needless to
7365 say, somebody hit the limit. I have increased the maximum line length to
7366 20 000 and the maximum data length of concatenated lines to 100 000. I have
7367 also fixed two bugs, because there was no checking on these buffers. Tsk,
7368 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7369 buffer is too small.
7371 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7372 lsearch lookups. Now it does.
7374 46. When parsing a route_list item in a manualroute router, a fixed-length
7375 buffer was used for the list of hosts. I made this 1024 bytes long,
7376 thinking that nobody would ever have a list of hosts that long. Wrong.
7377 Somebody had a whole pile of complicated expansion conditions, and the
7378 string was silently truncated, leading to an expansion error. It turns out
7379 that it is easier to change to an unlimited length (owing to other changes
7380 that have happened since this code was originally written) than to build
7381 structure for giving a limitation error. The length of the item that
7382 expands into the list of hosts is now unlimited.
7384 47. The lsearch lookup could not handle data where the length of text line was
7385 more than 4095 characters. Such lines were truncated, leading to shortened
7386 data being returned. It should now handle lines of any length.
7388 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7389 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7392 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7393 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7394 no longer happens in comment lines. A list of replacements is now placed
7395 at the head of all of the source files, except those whose only change is
7396 to replace PERL_COMMAND in the very first #! line.
7398 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7399 messages on the queue, with a bottom-up merge sort, using code contributed
7400 by Michael Haardt. This should make operations like -bp somewhat faster on
7401 large queues. It won't affect queue runners, except when queue_run_in_order
7404 51. Installed eximstats 1.31 in the distribution.
7406 52. Added support for SRV lookups to the dnslookup router.
7408 53. If an ACL referred to $message_body or $message_body_end, the value was not
7409 reset for any messages that followed in the same SMTP session.
7411 54. The store-handling optimization for building very long strings was not
7412 differentiating between the different store pools. I don't think this
7413 actually made any difference in practice, but I've tidied it.
7415 55. While running the routers to verify a sender address, $sender_address
7416 was still set to the sender address. This is wrong, because when routing to
7417 send a bounce to the sender, it would be empty. Therefore, I have changed
7418 it so that, while verifying a sender address, $sender_address is set to <>.
7419 (There is no change to what happens when verifying a recipient address.)
7421 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7422 target A or AAAA records (if not already returned) without resetting the
7423 qualify_single or search_parents options of the DNS resolver. These are
7424 inappropriate in this case because the targets of MX and SRV records must
7425 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7426 target that, when qualified, matched something in the local domain. These
7427 two options are now turned off when doing these lookups.
7429 57. It seems that at least some releases of Reiserfs (which does not have the
7430 concept of a fixed number of inodes) returns zero and not -1 for the
7431 number of available inodes. This interacted badly with check_spool_inodes,
7432 which assumed that -1 was the "no such thing" setting. What I have done is
7433 to check that the total number of inodes is greater than zero before doing
7434 the test of how many are available.
7436 58. When a "warn" ACL statement has a log_message modifier, the message is
7437 remembered, and not repeated. This is to avoid a lot of repetition when a
7438 message has many recipients that cause the same warning to be written.
7439 However, Exim was preserving the list of already written lines for an
7440 entire SMTP session, which doesn't seem right. The memory is now reset if a
7441 new message is started.
7443 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7444 debugging output unless log_rewrite was also set.
7446 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7447 of (int)(handle) when we know that handle contains (void *)(-1).
7449 61. The Exim daemon panic-logs an error return when it closes the incoming
7450 connection. However "connection reset by peer" seems to be common, and
7451 isn't really an error worthy of noting specially, so that particular error
7454 62. When Exim is trying to find all the local interfaces, it used to panic and
7455 die if the ioctl to get the interface flags failed. However, it seems that
7456 on at least one OS (Solaris 9) it is possible to have an interface that is
7457 included in the list of interfaces, but for which you get a failure error
7458 for this call. This happens when the interface is not "plumbed" into a
7459 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7460 failure of the "get flags" call assumes that the interface is down.
7462 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7463 makes life easier for people who are doing arithmetic on fields extracted
7464 from dates, where you often get leading zeros that should not be
7465 interpreted as octal.
7467 64. Added qualify_domain to the redirect router, to override the global
7470 65. If a pathologically long header line contained very many addresses (the
7471 report of this problem mentioned 10 000) and each of them was rewritten,
7472 Exim could use up a very large amount of memory. (It kept on making new
7473 copies of the header line as it rewrote, and never released the old ones.)
7474 At the expense of a bit more processing, the header rewriting function has
7475 been changed so that it no longer eats memory in this way.
7477 66. The generation of the Received: header has been moved from the time that a
7478 message starts to be received, to the time that it finishes. The timestamp
7479 in the Received: header should now be very close to that of the <= log
7480 line. There are two side-effects of this change:
7482 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7483 logged header lines no longer include the local Received: line, because
7484 it has not yet been created. The same applies to a copy of the message
7485 that is returned to a non-SMTP sender when a message is rejected.
7487 (b) When a filter file is tested using -bf, no additional Received: header
7488 is added to the test message. After some thought, I decided that this
7491 This change does not affect the value of $received_for. It is still set
7492 after address rewriting, but before local_scan() is called.
7494 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7496 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7497 gave an unhelpful panic error message, and a defer error. I have managed to
7498 change this behaviour so that it now rejects any supplied certificate,
7499 which seems right, as the list of acceptable certificates is empty.
7501 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7502 gave an unhelpful defer error. I have not managed to make this reject any
7503 supplied certificates, but the error message it gives is "no certificate
7504 supplied", which is not helpful.
7506 70. exigrep's output now also includes lines that are not associated with any
7507 message, but which match the given pattern. Implemented by a patch from
7508 Martin Sluka, which also tidied up the Perl a bit.
7510 71. Recipient callout verification, like sender verification, was using <> in
7511 the MAIL FROM command. This isn't really the right thing, since the actual
7512 sender may affect whether the remote host accepts the recipient or not. I
7513 have changed it to use the actual sender in the callout; this means that
7514 the cache record is now keyed on a recipient/sender pair, not just the
7515 recipient address. There doesn't seem to be a real danger of callout loops,
7516 since a callout by the remote host to check the sender would use <>.
7517 [SEE ABOVE: changed after hitting problems.]
7519 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7520 temporary errors. However, in the case of such a code being given after
7521 the end of a data transmission (i.e. after ".") Exim was failing to write
7522 a retry record for the message. (Yes, there was some broken host that was
7523 actually sending 8xx at this point.)
7525 73. An unknown lookup type in a host list could cause Exim to panic-die when
7526 the list was checked. (An example that provoked this was putting <; in the
7527 middle of a list instead of at the start.) If this happened during a DATA
7528 ACL check, a -D file could be left lying around. This kind of configuration
7529 error no longer causes Exim to die; instead it causes a defer error. The
7530 incident is still logged to the main and panic logs.
7532 74. Buglet left over from Exim 3 conversion. The message "too many messages
7533 in one connection" was written to the rejectlog but not the mainlog, except
7534 when address rewriting (yes!) was being logged.
7536 75. Added write_rejectlog option.
7538 76. When a system filter was run not as root (that is, when system_filter_user
7539 was set), the values of the $n variables were not being returned to the
7540 main process; thus, they were not subsequently available in the $sn
7543 77. Added +return_path_on_delivery log selector.
7545 78. A connection timeout was being treated differently from recipients deferred
7546 when testing hosts_max_try with a message that was older than the host's
7547 retry timeout. (The host should not be counted, thus allowing all hosts to
7548 be tried at least once before bouncing.) This may have been the cause of an
7549 occasionally reported bug whereby a message would remain on the queue
7550 longer than the retry timeout, but would be bounced if a delivery was
7551 forced. I say "may" because I never totally pinned down the problem;
7552 setting up timeout/retry tests is difficult. See also the next item.
7554 79. The ultimate address timeout was not being applied to errors that involved
7555 a combination of host plus message (for example, a timeout on a MAIL
7556 command). When an address resolved to a number of possible hosts, and they
7557 were not all tried for each delivery (e.g. because of hosts_max_try), a
7558 message could remain on the queue longer than the retry timeout.
7560 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7563 81. Fixed an obscure SMTP outgoing bug which required at least the following
7564 conditions: (a) there was another message waiting for the same server;
7565 (b) the server returned 5xx to all RCPT commands in the first message so
7566 that the message was not completed; (c) the server dropped the connection
7567 or gave a negative response to the RSET that Exim sends to abort the
7568 transaction. The observed case was a dropped connection after DATA that had
7569 been sent in pipelining mode. That is, the server had advertised PIPELINING
7570 but was not implementing it correctly. The effect of the bug was incorrect
7571 behaviour, such as trying another host, and this could lead to a crash.
7577 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7578 and daemon.c were passed as pointers to ints; they should have been
7579 pointers to socklen_t variables (which are typically unsigned ints).
7581 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7584 3. Fixed a really odd bug that affected only the testing scheme; patching a
7585 certain fixed string in the binary changed the value of another string that
7586 happened to be identical to the end of the original first string.
7588 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7589 name", it returns that address as the IP address. On some operating
7590 systems (e.g. Solaris), it also passes back the IP address string as the
7591 "host name". However, on others (e.g. Linux), it passes back an empty
7592 string. Exim wasn't checking for this, and was changing the host name to an
7593 empty string, assuming it had been canonicalized.
7595 5. Although rare, it is permitted to have more than one PTR record for a given
7596 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7597 all the names associated with an address, because they do in Solaris.
7598 However, it seems that they do not in Linux for data that comes from the
7599 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7600 I found this out when I moved to a new Linux workstation and tried to run
7601 the Exim test suite.
7603 To get round this problem I have changed the code so that it now does its
7604 own call to the DNS to look up PTR records when searching for a host name.
7605 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7606 addresses that are only in /etc/hosts are still found.
7608 This behaviour is, however, controlled by an option called host_lookup_
7609 order, which defaults to "bydns:byaddr". If people want to use the other
7610 order, or indeed, just use one or the other means of lookup, they can
7611 specify it in this variable.
7613 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7614 some operating systems, this comes back from gethostbyaddr() as an empty
7615 string, and this is what Exim used to test for. However, it seems that in
7616 other systems, "." is yielded. Exim now tests for this case too.
7618 7. The values of check_spool_space and check_log_space are now held internally
7619 as a number of kilobytes instead of an absolute number of bytes. If a
7620 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7621 kilobyte. This means that much larger values can be stored.
7623 8. Exim monitor: an attempt to get the action menu when not actually pointing
7624 at a message produces an empty menu entitled "No message selected". This
7625 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7626 no entries in it ("Shell widget menu has zero width and/or height"). So I
7627 have added a single, blank menu entry in this case.
7629 9. Added ${quote_local_part.
7631 10. MIME decoding is now applied to the contents of Subject: header lines when
7634 11. Now that a reference to $sender_host_address automatically causes a reverse
7635 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7636 host lookup before query-style lookups in lists that might use this
7637 variable. This has therefore been abolished, and the "net-" prefix is no
7638 longer necessary for query-style lookups.
7640 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7641 have been a typo for LFLAGS, so it has been changed.
7643 13. The install script calls Exim with "-C /dev/null" in order to find the
7644 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7645 to be output. However, since Exim outputs its version number before the
7646 error, it didn't break the script. It just looked ugly. I fixed this by
7647 always allowing "-C /dev/null" if the caller is root.
7649 14. Ignore overlarge ACL variable number when reading spool file - insurance
7650 against a later release with more variables having written the file.
7652 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7653 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7654 this, as well as the form without the "IPv6" on the front.
7656 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7657 OS/Makefile-Darwin file.
7659 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7660 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7662 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7663 variable. However, some OS use ints for inodes. I've added cast to long int
7664 to get rid of the compiler warning.
7666 19. I had forgotten to lock out "/../" in configuration file names when
7667 ALT_CONFIG_PREFIX was set.
7669 20. Routers used for verification do not need to specify transports. However,
7670 if such a router generated a host list, and callout was configured, Exim
7671 crashed, because it could not find a port number from the (non-existent)
7672 transport. It now assumes port 25 in this circumstance.
7674 21. Added the -t option to exigrep.
7676 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7677 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7678 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7679 them is set, LOOKUP_LSEARCH is forced.
7681 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7683 24. Added sender and host information to the "rejected by local_scan()" log
7684 line; previously there was no indication of these.
7686 25. Added .include_if_exists.
7688 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7689 Linux. It turns out that not all file systems support this. Apparently some
7690 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7691 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7692 not supported on the file descriptor, is now ignored when Exim is trying to
7693 sync a directory. This applies only to Linux.
7695 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7697 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7698 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7701 29. Added support for maildirsize files from supplied patch (modified a bit).
7703 30. The use of :fail: followed by an empty string could lead Exim to respond to
7704 sender verification failures with (e.g.):
7706 550 Verification failed for <xxx>
7707 550 Sender verify failed
7709 where the first response line was missing the '-' that indicates it is not
7710 the final line of the response.
7712 31. The loop for finding the name of the user that called Exim had a hardwired
7713 limit of 10; it now uses the value of finduser_retries, which is used for
7714 all other user lookups.
7716 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7719 33. Exim was neglecting to zero errno before one call of strtol() when
7720 expanding a string and expecting an integer value. On some systems this
7721 resulted in spurious "integer overflow" errors. Also, it was casting the
7722 result into an int without checking.
7724 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7725 did not work. The code looks as if it has *never* worked, though it appears
7726 to have been documented since at least release 1.62. I have made it work.
7728 35. The "timeout_DNS" error in retry rules, also documented since at least
7729 1.62, also never worked. As it isn't clear exactly what this means, and
7730 clearly it isn't a major issue, I have abolished the feature by treating it
7731 as "timeout", and writing a warning to the main and panic logs.
7733 36. The display of retry rules for -brt wasn't always showing the error code
7736 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7737 timeout_connect_A, timeout_connect_MX.
7739 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7740 to the empty sender.
7742 39. The daemon was not analysing the content of -oX till after it had closed
7743 stderr and disconnected from the controlling terminal. This meant that any
7744 syntax errors were only noted on the panic log, and the return code from
7745 the command was 0. By re-arranging the code a little, I've made the
7746 decoding happen first, so such errors now appear on stderr, and the return
7747 code is 1. However, the actual setting up of the sockets still happens in
7748 the disconnected process, so errors there are still only recorded on the
7751 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7752 connections (as happens on some IP stacks) was logged at start up time as
7753 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7754 it from "IPv6 and IPv4", which means that two separate sockets are being
7757 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7758 says whether AF_INET or AF_INET6 was passed as an argument.
7760 42. Exiwhat output was messed up when time zones were included in log
7763 43. Exiwhat now gives more information about the daemon's listening ports,
7764 and whether -tls-on-connect was used.
7766 44. The "port" option of the smtp transport is now expanded.
7768 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7769 silently ignored. Now an error message is written to the main and panic
7772 46. There's a new ACL modifier called "logwrite" which writes to a log file
7773 as soon as it is encountered.
7775 47. Added $local_user_uid and $local_user_gid at routing time.
7777 48. Exim crashed when trying to verify a sender address that was being
7780 49. Exim was recognizing only a space character after ".include". It now also
7781 recognizes a tab character.
7783 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7784 extracting the relevant information from the specification. The man page no
7785 longer contains scrambled data for the -d option, and I've added a section
7786 at the front about calling Exim under different names.
7788 51. Added "extra_headers" argument to the "mail" command in filter files.
7790 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7793 53. Installed eximstats 1.29.
7795 54. Added transport_filter_timeout as a generic transport option.
7797 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7798 Cc: header lines. This was required by RFC 822, but it not required by RFC
7801 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7802 incoming messages that did not have them. Now it does so only if the
7803 message originates locally, that is, if there is no associated remote host
7804 address. When Resent- header lines are present, this applies to the Resent-
7805 lines rather than the non-Resent- lines.
7807 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7808 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7810 58. Messages for configuration errors now include the name of the main
7811 configuration file - useful now that there may be more than one file in a
7812 list (.included file names were always shown).
7814 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7815 for those rare installations that do not start the daemon as root or run it
7816 setuid root. I've cut out the call to initgroups() if the daemon is not
7819 60. The Exim user and group can now be bound into the binary as text strings
7820 that are looked up at the start of Exim's processing.
7822 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7824 62. Added $mailstore_basename variable.
7826 63. Installed patch to sieve.c from Michael Haardt.
7828 64. When Exim failed to open the panic log after failing to open the main log,
7829 the original message it was trying to log was written to stderr and debug
7830 output, but if they were not available (the usual case in production), it
7831 was lost. Now it is written to syslog before the two lines that record the
7832 failures to open the logs.
7834 65. Users' Exim filters run in subprocesses under the user's uid. It is
7835 possible for a "deliver" command or an alias in a "personal" command to
7836 provoke an address rewrite. If logging of address rewriting is configured,
7837 this fails because the process is not running as root or exim. There may be
7838 a better way of dealing with this, but for the moment (because 4.30 needs
7839 to be released), I have disabled address rewrite logging when running a
7840 filter in a non-root, non-exim process.
7846 1. The buildconfig auxiliary program wasn't quoting the value set for
7847 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7848 not defined. This bug was masked in 4.22 by the effect that was fixed in
7851 2. Some messages that were rejected after a message id was allocated were
7852 shown as "incomplete" by exigrep. It no longer does this for messages that
7853 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7855 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7856 have allow_domain_literals set, the ID did not get logged in the <= line.
7857 Domain literals are now always recognized in Message-ID: header lines.
7859 4. The first argument for a ${extract expansion item is the key name or field
7860 number. Leading and trailing spaces in this item were not being ignored,
7861 causing some misleading effects.
7863 5. When deliver_drop_privilege was set, single queue runner processes started
7864 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7865 same command in the process it spins off) were not dropping privilege.
7867 6. When the daemon running as "exim" started a queue runner, it always
7868 re-executed Exim in the spun-off process. This is a waste of effort when
7869 deliver_drop_privilege is set. The new process now just calls the
7870 queue-runner function directly.
7876 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7879 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7880 ignored. Though the use of -oP was forcing the writing of a pid file, it
7881 was always written to the default place.
7883 3. If the message "no IP address found for host xxxx" is generated during
7884 incoming verification, it is now followed by identification of the incoming
7885 connection (so you can more easily find what provoked it).
7887 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7889 5. Added some features to "harden" Exim a bit more against certain attacks:
7891 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7892 be put in Local/Makefile. This is like the never_users runtime option,
7893 but it cannot be overridden. The default setting is "root".
7895 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7896 prefix string with which any file named in a -C command line option
7899 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7900 is retained for -C and -D only if the caller of Exim is root. Without
7901 it, the exim user may also use -C and -D and retain privilege.
7903 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7904 command line option is disabled.
7906 6. Macro names set by the -D option must start with an upper case letter, just
7907 like macro names defined in the configuration file.
7909 7. Added "dereference=" facility to LDAP.
7911 8. Two instances of the typo "uknown" in the source files are fixed.
7913 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7914 the Configure-Makefile script screwed up while processing it.
7916 10. Incorporated PCRE 4.4.
7918 11. The SMTP synchronization check was not operating right at the start of an
7919 SMTP session. For example, it could not catch a HELO sent before the client
7920 waited for the greeting. There is now a check for outstanding input at the
7921 point when the greeting is written. Because of the duplex, asynchronous
7922 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7923 way, but not yet received, when the check is performed.
7925 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7926 on TCP/IP sockets, because this apparently causes some broken clients to
7929 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7930 unchanged) from the Cygwin maintainer.
7932 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7933 was supported instead of testing for mbx. Effectively a typo.
7935 15. The spa authenticator server code was not checking that the input it
7936 received was valid base64.
7938 16. The debug output line for the "set" modifier in ACLs was not showing the
7939 name of the variable that was being set.
7941 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7943 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7944 lookup on its own. Something else was needed to trigger the lookup. For
7945 example, a match in host_lookup or the need for a host name in a host list.
7946 Now, if $sender_host_name is referenced and the host name has not yet been
7947 looked up, a lookup is performed. If the lookup fails, the variable remains
7948 empty, and $host_lookup_failed is set to "1".
7950 19. Added "eqi" as a case-independent comparison operator.
7952 20. The saslauthd authentication condition could segfault if neither service
7953 nor realm was specified.
7955 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7956 error message that was logged was misleading, and incoming SMTP
7957 connections were dropped. The message is now more accurate, and temporary
7958 errors are given to SMTP connections.
7960 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7961 (except RSET) with a 421 error, until QUIT is received. However, it was
7962 failing to send a response to QUIT.
7964 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7965 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7966 ACL is now run after the helo_try_verify_hosts code.
7968 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7969 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7970 case-independent, so other case variants are also recognized). Apparently
7971 some systems use these upper case variants.
7973 25. If more than two messages were waiting for the same host, and a transport
7974 filter was specified for the transport, Exim sent two messages over the
7975 same TCP/IP connection, and then failed with "socket operation on non-
7976 socket" when it tried to send the third.
7978 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7980 27. The extern definition of crypt16() in expand.c was not being excluded when
7981 the OS had its own crypt16() function.
7983 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7984 as a preferred synonym for return_size_limit, both as an option and as an
7987 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7989 30. Changed the default configuration ACL to relax the local part checking rule
7990 for addresses that are not in any local domains. For these addresses,
7991 slashes and pipe symbols are allowed within local parts, but the sequence
7992 /../ is explicitly forbidden.
7994 31. SPA server authentication was not clearing the challenge buffer before
7997 32. log_message in a "warn" ACL statement was writing to the reject log as
7998 well as to the main log, which contradicts the documentation and doesn't
7999 seem right (because no rejection is happening). So I have stopped it.
8001 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8002 However, I am unable to do any testing of this.
8004 34. Fixed an infelicity in the appendfile transport. When checking directories
8005 for a mailbox, to see if any needed to be created, it was accidentally
8006 using path names with one or more superfluous leading slashes; tracing
8007 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8009 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8010 contents are added to the log line that is written for every discarded
8011 recipient. (Previously a log_message setting was ignored.)
8013 36. The ${quote: operator now quotes the string if it is empty.
8015 37. The install script runs exim in order to find its version number. If for
8016 some reason other than non-existence or emptiness, which it checks, it
8017 could not run './exim', it was installing it with an empty version number,
8018 i.e. as "exim-". This error state is now caught, and the installation is
8021 38. An argument was missing from the function that creates an error message
8022 when Exim fails to connect to the socket for saslauthd authentication.
8023 This could cause Exim to crash, or give a corrupted message.
8025 39. Added isip, isip4, and isip6 to ${if conditions.
8027 40. The ACL variables $acl_xx are now saved with the message, and can be
8028 accessed later in routers, transports, and filters.
8030 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8031 strings in the file are not string-expanded.
8033 42. If a MAIL command specified a SIZE value that was too large to fit into an
8034 int variable, the check against message_size_limit failed. Such values are
8035 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8036 day this will have to be increased, but I don't think I want to be around
8037 when emails are that large.
8044 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8045 iconv() is not standard in FreeBSD.
8047 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8048 IPv6 enabled. The observed symptom was a segmentation fault on return from
8049 the function os_common_find_running_interfaces() in src/os.c.
8051 3. In the check_special_case() function in daemon.c I had used "errno" as an
8052 argument name, which causes warnings on some systems. This was basically a
8053 typo, since it was named "eno" in the comments!
8055 4. The code that waits for the clock to tick (at a resolution of some fraction
8056 of a second) so as to ensure message-id uniqueness was always waiting for
8057 at least one whole tick, when it could have waited for less. [This is
8058 almost certainly not relevant at current processor speeds, where it is
8059 unlikely to ever wait at all. But we try to future-proof.]
8061 5. The function that sleeps for a time interval that includes fractions of a
8062 second contained a race. It did not block SIGALRM between setting the
8063 timer, and suspending (a couple of lines later). If the interval was short
8064 and the sigsuspend() was delayed until after it had expired, the suspension
8065 never ended. On busy systems this could lead to processes getting stuck for
8068 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8069 process, before it forks any delivery processes. The open lookup caching
8070 mechanism meant that the open file or database connection was passed into
8071 the delivery process. The problem was that delivery processes always tidy
8072 up cached lookup data. This could cause a problem for the next delivery
8073 process started by the queue runner, because the external queue runner
8074 process does not know about the closure. So the next delivery process
8075 still has data in the lookup cache. In the case of a file lookup, there was
8076 no problem because closing a file descriptor in a subprocess doesn't affect
8077 the parent. However, if the lookup was caching a connection to a database,
8078 the connection was closed, and the second delivery process was likely to
8079 see errors such as "PGSQL: query failed: server closed the connection
8080 unexpectedly". The problem has been fixed by closing all cached lookups
8081 in a queue runner before running a delivery process.
8083 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8084 seem to have the "const" qualifier which it has on other OS. I've
8087 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8088 *of the same type* (client or server) with the same public name that an
8089 error should be diagnosed.
8091 9. When Exim looked up a host name for an IP address, but failed to find the
8092 original IP address when looking up the host name (a safety check), it
8093 output the message "<ip address> does not match any IP for NULL", which was
8094 confusing, to say the least. The bug was that the host name should have
8095 appeared instead of "NULL".
8097 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8098 user that is built into the binary, and the -C or -D options is used, root
8099 privilege is dropped before the configuration file is read. In addition,
8100 logging is switched to stderr instead of the normal log files. If the
8101 configuration then re-defines the Exim user, the unprivileged environment
8102 is probably not what is expected, so Exim logs a panic warning message (but
8105 However, if deliver_drop_privilege is set, the unprivileged state may well
8106 be exactly what is intended, so the warning has been cut out in that case,
8107 and Exim is allowed to try to write to its normal log files.
8113 1. smtp_return_error_details was not giving details for temporary sender
8114 or receiver verification errors.
8116 2. Diagnose a configuration error if two authenticators have the same public
8119 3. Exim used not to create the message log file for a message until the first
8120 delivery attempt. This could be confusing when incoming messages were held
8121 for policy or load reasons. The message log file is now created at the time
8122 the message is received, and an initial "Received" line is written to it.
8124 4. The automatically generated man page for command line options had a minor
8125 bug that caused no ill effects; however, a more serious problem was that
8126 the procedure for building the man page automatically didn't always
8127 operate. Consequently, release 4.20 contains an out-of-date version. This
8128 shouldn't happen again.
8130 5. When building Exim with embedded Perl support, the script that builds the
8131 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8132 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8134 6. The freeze_tell option was not being used for messages that were frozen on
8135 arrival, either by an ACL or by local_scan().
8137 7. Added the smtp_incomplete_transaction log selector.
8139 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8140 was accepting AUTH without a new EHLO.
8142 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8143 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8146 10. Logging of TCP/IP connections (when configured) now happens in the main
8147 daemon process instead of the child process, so that the TCP/IP connection
8148 count is more accurate (but it can never be perfect).
8150 11. The use of "drop" in a nested ACL was not being handled correctly in the
8151 outer ACL. Now, if condition failure induced by the nested "drop" causes
8152 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8153 or "require"), the connection is dropped.
8155 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8156 that yield "discard" can now be used with an "accept" or a "discard" verb,
8157 but an error is generated for any others (because I can't see a useful way
8158 to define what should happen).
8160 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8161 are now processed in the same way as lines in the Exim configuration file.
8162 In particular, continuation lines are supported.
8164 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8166 15. Added -ti meaning -t -i.
8168 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8169 domains, and warn by logging if others are found.
8171 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8172 returned from GIFCONF when Exim is trying to find the list of interfaces on
8173 a host. The code in os.c has been modified to copy each ifreq to an aligned
8174 structure in all cases.
8176 Also, in some cases, the returned ifreq's were being copied to a 'struct
8177 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8178 means the last couple of bytes of an IPv6 address could be chopped if the
8179 ifreq contained only a normal sockaddr (14 bytes storage).
8181 18. Named domain lists were not supported in the hosts_treat_as_local option.
8182 An entry such as +xxxx was not recognized, and was treated as a literal
8185 19. Ensure that header lines added by a DATA ACL are included in the reject log
8186 if the ACL subsequently rejects the message.
8188 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8189 MD5 (which is deprecated).
8191 21. When testing a filter file using -bf, Exim was writing a message when it
8192 took the sender from a "From " line in the message, but it was not doing so
8193 when it took $return_path from a Return-Path: header line. It now does.
8195 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8196 with a valid header line field name (a series of printing characters
8197 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8199 23. Changed "disc" in the source to "disk" to conform to the documentation and
8200 the book and for uniformity.
8202 24. Ignore Sendmail's -Ooption=value command line item.
8204 25. When execve() failed while trying to run a command in a pipe transport,
8205 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8206 could be confused with a return value of 69 from the command itself. This
8207 has been changed to 127, the value the shell returns if it is asked to run
8208 a non-existent command. The wording for the related log line suggests a
8209 non-existent command as the problem.
8211 26. If received_header_text expands to an empty string, do not add a Received:
8212 header line to the message. (Well, it adds a token one on the spool, but
8213 marks it "old" so that it doesn't get used or transmitted.)
8215 27. Installed eximstats 1.28 (addition of -nt option).
8217 28. There was no check for failure on the call to getsockname() in the daemon
8218 code. This can fail if there is a shortage of resources on the system, with
8219 ENOMEM, for example. A temporary error is now given on failure.
8221 29. Contrary to the C standard, it seems that in some environments, the
8222 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8223 program. Exim now does this explicitly; it affects the formatting of
8224 timestamps using strftime().
8226 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8227 complaints. I've now initialized all the variables, to avoid this.
8229 32. Header lines added by a system filter were not being "seen" during
8230 transport-time rewrites.
8232 33. The info_callback() function passed to OpenSSL is set up with type void
8233 (*)(SSL *, int, int), as described somewhere. However, when calling the
8234 function (actually a macro) that sets it up, the type void(*)() is
8235 expected. I've put in a cast to prevent warnings from picky compilers.
8237 34. If a DNS black list lookup found a CNAME record, but there were no A
8238 records associated with the domain it pointed at, Exim crashed.
8240 35. If a DNS black list lookup returned more than one A record, Exim ignored
8241 all but the first. It now scans all returned addresses if a particular IP
8242 value is being sought. In this situation, the contents of the
8243 $dnslist_value variable are a list of all the addresses, separated by a
8246 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8247 to accept a host name and all its aliases if the forward lookup for any of
8248 them yielded the IP address of the incoming connection. Now it accepts only
8249 those names whose forward lookup yields the correct IP address. Any other
8250 names are discarded. This closes a loophole whereby a rogue DNS
8251 administrator could create reverse DNS records to break through a
8252 wildcarded host restriction in an ACL.
8254 37. If a user filter or a system filter that ran in a subprocess used any of
8255 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8256 the wrong values were passed to the pipe command ($thisaddress had the
8257 value of $0, $0 had the value of $1, etc). This bug was introduced by
8258 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8260 38. Improved the line breaking for long SMTP error messages from ACLs.
8261 Previously, if there was no break point between 40 and 75 characters, Exim
8262 left the rest of the message alone. Two changes have been made: (a) I've
8263 reduced the minimum length to 35 characters; (b) if it can't find a break
8264 point between 35 and 75 characters, it looks ahead and uses the first one
8265 that it finds. This may give the occasional overlong line, but at least the
8266 remaining text gets split now.
8268 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8269 file descriptors might be low, and that setting 1000 would always raise it.
8270 It turns out that in some environments, the limit is already over 1000 and
8271 that lowering it causes trouble. So now Exim takes care not to decrease it.
8273 40. When delivering a message, the value of $return_path is set to $sender_
8274 address at the start of routing (routers may change the value). By an
8275 oversight, this default was not being set up when an address was tested by
8276 -bt or -bv, which affected the outcome if any router or filter referred to
8279 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8280 log only when "message" or "log_message" are set. However, if one of the
8281 conditions was an address verification, or a call to a nested ACL, the
8282 messages generated by the underlying test were being passed through. This
8283 no longer happens. The underlying message is available in $acl_verify_
8284 message for both "message" and "log_message" expansions, so it can be
8285 passed through if needed.
8287 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8288 new expansion $bh_ to give the encoded byte string without charset
8289 translation. Translation happens only if iconv() is available; HAVE_ICONV
8290 indicates this at build time. HEADERS_CHARSET gives the charset to
8291 translate to; headers_charset can change it in the configuration, and
8292 "headers charset" can change it in an individual filter file.
8294 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8295 that creates RFC 2047 encoded "words" labels them as that charset instead
8296 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8297 expansion operator; (ii) when Exim creates a From: line for a local
8298 message; (iii) when a header line is rewritten to include a "phrase" part.
8300 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8301 buggy, causing it to skip the first lines of messages whose message ID
8302 ended in 'D'. This would not have bitten before Exim release 4.14, because
8303 message IDs were unlikely to end in 'D' before then. The effect was to have
8304 incorrect size information for certain domains.
8306 45. #include "config.h" was missing at the start of the crypt16.c module. This
8307 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8310 46. If there was a timeout during a "random" callout check, Exim treated it as
8311 a failure of the random address, and carried on sending RSET and the real
8312 address. If the delay was just some slowness somewhere, the response to the
8313 original RCPT would be taken as a response to RSET and so on, causing
8314 mayhem of various kinds.
8316 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8317 when I implemented it. It didn't allow for the fact that some option values
8318 may legitimately be negative (e.g. size_addition), and it didn't even do
8319 the right test for positive values.
8321 48. Domain names in DNS records are case-independent. Exim always looks them up
8322 in lower case. Some resolvers return domain names in exactly the case they
8323 appear in the zone file, that is, they may contain uppercase letters. Not
8324 all resolvers do this - some return always lower case. Exim was treating a
8325 change of case by a resolver as a change of domain, similar to a widening
8326 of a domain abbreviation. This triggered its re-routing code and so it was
8327 trying to route what was effectively the same domain again. This normally
8328 caused routing to fail (because the router wouldn't handle the domain
8329 twice). Now Exim checks for this case specially, and just changes the
8330 casing of the domain that it ultimately uses when it transmits the message
8333 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8336 50. If a filter generated a file delivery with a non-absolute name (possible if
8337 no home directory exists for the router), the forbid_file option was not
8340 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8341 the existing equality matching.
8343 52. Exim was using ints instead of ino_t variables in some places where it was
8344 dealing with inode numbers.
8346 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8347 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8348 called TMPDIR, and if it finds it is different, it changes its value.
8350 54. The smtp_printf() function is now made available to local_scan() so
8351 additional output lines can be written before returning. There is also an
8352 smtp_fflush() function to enable the detection of a dropped connection.
8353 The variables smtp_input and smtp_batched_input are exported to
8356 55. Changed the default runtime configuration: the message "Unknown user"
8357 has been removed from the ACL, and instead placed on the localuser router,
8358 using the cannot_route_message feature. This means that any verification
8359 failures that generate their own messages won't get overridden. Similarly,
8360 the "Unrouteable address" message that was in the ACL for unverifiable
8361 relay addresses has also been removed.
8363 56. Added hosts_avoid_esmtp to the smtp transport.
8365 57. The exicyclog script was not checking for the esoteric option
8366 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8367 will work only if exicyclog is run under the appropriate euid.
8369 58. Following a discussion on the list, the rules by which Exim recognises line
8370 endings on incoming messages have been changed. The -dropcr and drop_cr
8371 options are now no-ops, retained only for backwards compatibility. The
8372 following line terminators are recognized: LF CRLF CR. However, special
8373 processing applies to CR:
8375 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8376 nor a local message in the state where . is a terminator.
8378 (ii) If a bare CR is encountered in a header line, an extra space is added
8379 after the line terminator so as not to end the header. The reasoning
8380 behind this is that bare CRs in header lines are most likely either
8381 to be mistakes, or people trying to play silly games.
8383 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8384 was being incorrectly given as 18 bytes larger than it should have been.
8385 This is a VOB (very old bug).
8387 60. This may never have affected anything current, but just in case it has:
8388 When the local host is found other than at the start of a list of hosts,
8389 the local host, those with the same MX, and any that follow, are discarded.
8390 When the list in question was part of a longer list of hosts, the following
8391 hosts (not currently being processed) were also being discarded. This no
8392 longer happens. I'm not sure if this situation could ever has previously
8395 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8398 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8399 header. This is something that is recommended in a new Internet Draft, and
8400 is something that is documented as being done by Sendmail. There are two
8401 possible values. For messages generated by the autoreply transport, Exim
8404 Auto-Submitted: auto-replied
8406 whereas for all other generated messages (e.g. bounces) it adds
8408 Auto-Submitted: auto-generated
8410 63. The "personal" condition in filters now includes a test for the
8411 Auto-Submitted: header. If it contains the string "auto-" the message it
8412 not considered personal.
8414 64. Added rcpt_include_affixes as a generic transport option.
8416 65. Added queue_only_override (default true).
8418 66. Added the syslog_duplication option.
8420 67. If what should have been the first header line of a message consisted of
8421 a space followed by a colon, Exim was mis-interpreting it as a header line.
8422 It isn't of course - it is syntactically invalid and should therefore be
8423 treated as the start of the message body. The misbehaviour could have
8424 caused a number of strange effects, including loss of data in subsequent
8425 header lines, and spool format errors.
8427 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8428 client host had authenticated. This control can now be exercised by an ACL
8429 for more flexibility.
8431 69. By default, callouts do not happen when testing with -bh. There is now a
8432 variant, -bhc, which does actually run the callout code, including
8433 consulting and updating the callout cache.
8435 70. Added support for saslauthd authentication, courtesy of Alexander
8438 71. If statvfs() failed on the spool or log directories while checking their
8439 size for availability, Exim confusingly gave the error "space shortage".
8440 Furthermore, in debugging mode it crashed with a floating point exception.
8441 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8442 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8443 serious problem, Exim now writes to the main and panic logs when this
8444 happens, with details of the failure. It then refuses to accept the
8445 incoming message, giving the message "spool directory problem" or "log
8446 directory problem" with a 421 code for SMTP messages.
8448 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8449 0, 1, and 2 exist, because some OS complain for execs without them (see
8450 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8451 descriptors. However, the code omitted to check that the open succeeded,
8452 causing mysterious errors if for some reason the permissions on /dev/null
8453 got screwed. Now Exim writes a message to the main and panic logs, and
8454 bombs out if it can't open /dev/null.
8456 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8457 interact so that it is all more flexible. It is supposed to remain
8458 backwards compatible. Also added extra_local_interfaces.
8460 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8461 to bomb out with an assertion failure - to the client this appears as a
8462 connection drop. This problem occurs in the part of the code that was taken
8463 from the Samba project. Fortunately, the assertion is in a very simple
8464 function, so I have fixed this by reproducing the function inline in the
8465 one place where it is called, and arranging for authentication to fail
8466 instead of killing the process with assert().
8468 75. The SPA client code was not working when the server requested OEM rather
8469 than Unicode encoding.
8471 76. Added code to make require_files with a specific uid setting more usable in
8472 the case where statting the file as root fails - usually a non-root-mounted
8473 NFS file system. When this happens and the failure is EACCES, Exim now
8474 forks a subprocess and does the per-uid checking as the relevant uid.
8476 77. Added process_log_path.
8478 78. If log_file_path was not explicitly set, a setting of check_log_space or
8479 check_log_inodes was ignored.
8481 79. If a space check for the spool or log partitions fails, the incident is now
8482 logged. Of course, in the latter case the data may get lost...
8484 80. Added the %p formatting code to string_format() so that it can be used to
8485 print addresses in debug_print(). Adjusted all the address printing in the
8486 debugging in store.c to use %p rather than %d.
8488 81. There was a concern that a line of code in smtp_in.c could overflow a
8489 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8490 initially expressed, the concern was not well-founded, because trailing
8491 spaces are removed early. However, if the trailing spaces were followed by
8492 a NULL, they did not get removed, so the overflow was possible. Two fixes
8495 (a) I re-wrote the offending code in a cleaner fashion.
8496 (b) If an incoming SMTP command contains a NULL character, it is rejected
8499 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8500 runs initgroups(), so that if the Exim user is in any additional groups,
8501 they will be used during message reception.
8507 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob