1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
52 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
53 SMTP connection" log lines.
55 JH/02 Option default value updates:
56 - queue_fast_ramp (main) true (was false)
57 - remote_max_parallel (main) 4 (was 2)
59 JH/03 Cache static regex pattern compilations, for use by ACLs.
61 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
62 Make the rewrite never match and keep the logging. Trust the
63 admin to be using verify=header-syntax (to actually reject the message).
65 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
66 (under Linux) we watch the dir containing the final file; previously
67 it would be the dir with the first symlink. We still do not monitor
70 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
71 dn_expand() is happy to pass them through.
73 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
74 the file with the proof had an unchanged name, the new proof(s) were
75 loaded on top of the old ones (and nover used; the old ones were stapled).
77 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
78 more than one message arrived in a single connection a reference from
79 the earlier message could be re-used. Often a sigsegv resulted.
80 These variables were introduced in Exim 4.87.
81 Debug help from Graeme Fowler.
83 JH/09 Fix ${filter } for conditions that modify $value. Previously the
84 modified version would be used in construction the result, and a memory
87 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
88 Find and fix by Jasen Betts.
90 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
91 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
92 the systemwide configuration to override the Exim config.
94 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
95 API changes in libopendmarc.
97 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
98 pid 1, in the normal "background daemon" mode, having to drop process-
99 group leadership also lost track of needing to create listener sockets.
101 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
102 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
104 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
105 a capture group which obtained no text (eg. "(abc)*" matching zero
106 occurrences) could cause a segfault if the corresponding $<n> was
109 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
110 included a close-brace character (eg. it itself used an expansion) an
113 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
114 starting TLS. Previously it was after, meaning that attackers on such
115 ports had to be screened using the host_reject_connection main config
116 option. The new sequence aligns better with the STARTTLS behaviour, and
117 permits defences against crypto-processing load attacks, even though it
118 is strictly an incompatible change.
119 Also, avoid sending any SMTP fail response for either the connect ACL
120 or host_reject_connection, for TLS-on-connect ports.
122 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
123 Previously this was not permitted, but it makes reasonable sense.
124 While there, restore a restriction on using it from a connect ACL; given
125 the change JH/16 it could only return false (and before 4.91 was not
128 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
129 was exactly sized compared to the log buffer, a crash occurred with the
130 misleading message "bad memory reference; pool not found".
131 Found and traced by Jasen Betts.
133 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
134 dns_again_means_nonexist included an element causing a DNS lookup which
135 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
136 included (though probably not limited to) a process crash from stack
137 memory limit, or from excessive open files. Replace this with a paniclog
138 whine (as this is likely a configuration error), and returning
141 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
142 this always failed, probably leading to the usual downgrade to in-clear
145 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
146 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
147 to not checking that list for these lookups.
149 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
152 JH/23 Fix crash in string expansions. Previously, if an empty variable was
153 immediately followed by an expansion operator, a null-indirection read
154 was done, killing the process.
156 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
157 include an SMTP response string which is longer than that supported
158 by the delivering transport. Alleviate by wrapping such lines before
161 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
162 chars (RFC limit). Previously a limit of 12 items was made, which with
163 a not-impossible References: in the message being bounced could still
164 be over-large and get stopped in the transport.
166 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
167 close. Previously a bare socket close was done.
169 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
172 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
173 message being considered for delivery was already being handled by
174 another process, and having an SMTP connection already open, the function
175 to close it tried to use an uninitialized variable. This would afftect
176 high-volume sites more, especially when running mailing-list-style loads.
177 Pollution of logs was the major effect, as the other process delivered
178 the message. Found and partly investigated by Graeme Fowler.
180 JH/29 Change format of the internal ID used for message identification. The old
181 version only supported 31 bits for a PID element; the new 64 (on systems
182 which can use Base-62 encoding, which is all currently supported ones
183 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
184 and must use Base-36). The new ID is 23 characters rather than 16, and is
185 visible in various places - notably logs, message headers, and spool file
186 names. Various of the ancillary utilities also have to know the format.
187 As well as the expanded PID portion, the sub-second part of the time
188 recorded in the ID is expanded to support finer precision. Theoretically
189 this permits a receive rate from a single comms channel of better than the
191 The major timestamp part of the ID is not changed; at 6 characters it is
192 usable until about year 3700.
193 Updating from previously releases is fully supported: old-format spool
194 files are still usable, and the utilities support both formats. New
195 message will use the new format. The one hints-DB file type which uses
196 message-IDs (the transport wait- DB) will be discarded if an old-format ID
197 is seen; new ones will be built with only new-format IDs.
198 Optionally, a utility can be used to convert spool files from old to new,
199 but this is only an efficiency measure not a requirement for operation
200 Downgrading from new to old requires running a provided utility, having
201 first stopped all operations. This will convert any spool files from new
202 back to old (losing time-precision and PID information) and remove any
203 wait- hints databases.
205 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
206 we treated them as item separators when parsing for a list item, but they
207 need to be protected by the doublequotes. While there, add handling for
210 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
211 Found and fixed by Jasen Betts. No testcase for this as my usual text
212 editor insists on emitting only valid UTF-8.
214 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
215 a null-indirection SIGSEGV for the receive process.
217 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
218 -bh use. Internal checking would spot this and take a panic.
220 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
221 In 4.96 this would expand to empty.
223 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
224 certificate. Find and fix by Andreas Metzler.
226 JH/36 Add ARC info to DMARC hostory records.
228 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
229 or fakedefer. Previously the sender could discover that the message
230 had in fact been accepted.
232 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
233 sequences. Previously the input was not noted as being tainted; notably
234 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
235 bad coding of authenticators.
237 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
238 and ${tr...}. Found and diagnosed by Heiko Schlichting.
240 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
241 could be triggered by externally-supplied input. Found by Trend Micro.
244 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
245 be triggered by externally-controlled input. Found by Trend Micro.
248 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
249 be triggered by externally-controlled input. Found by Trend Micro.
252 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
253 Make the rewrite never match and keep the logging. Trust the
254 admin to be using verify=header-syntax (to actually reject the message).
256 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
259 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
265 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
266 after reception to before a subsequent reception. This should
267 mean slightly faster delivery, and also confirmation of reception
270 JH/02 Move from using the pcre library to pcre2. The former is no longer
271 being developed or supported (by the original developer).
273 JH/03 Constification work in the filters module required a major version
274 bump for the local-scan API. Specifically, the "headers_charset"
275 global which is visible via the API is now const and may therefore
276 not be modified by local-scan code.
278 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
279 sendfile() didi not account for the way the ClamAV driver code called it.
281 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
282 time check was being done for every character; replace that with one
285 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
286 sent was prefixed with a length byte.
288 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
289 RFC 5321. Previously Dovecot (at least) would log errors during
292 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
293 in 4.61 without a reason logged. Binaries will be bigger, which might
294 matter on diskspace-constrained systems, but debug is easier.
296 JH/09 Fix macro-definition during "-be" expansion testing. The move to
297 write-protected store for macros had not accounted for these runtime
298 additions; fix by removing this protection for "-be" mode.
300 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
301 handing out large-numbered file descriptors, violating the usual Unix
302 assumption (and required by Posix) that the lowest possible number will be
303 allocated by the kernel when a new one is needed. In the daemon, and any
304 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
305 are not useable for FD_SET() [and hence select()] and overwrite the stack.
306 Assorted crashes happen.
308 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
309 main-section options or in a connect ACL, the value from the first ever
310 connection was never replaced for subsequent connections. Found by
313 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
314 though only once PCRE2 was introduced: the memory accounting used under
315 debug offset allocations by an int, giving a hard trap in early startup.
316 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
318 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
319 with underbars is given. The write-protection of configuration introduced
320 in 4.95 trapped when normalisation was applied to an option not needing
323 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
325 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
326 failing to close the certificates directory, every hour or any time it
329 JH/16 Debugging initiated by an ACL control now continues through into routing
330 and transport processes. Previously debugging stopped any time Exim
331 re-execs, or for processing a queued message.
333 JH/17 The "expand" debug selector now gives more detail, specifically on the
334 result of expansion operators and items.
336 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
337 bad comparison between the option value and the name of the file to
338 be included was done, and a mismatch was wrongly identified.
339 4.88 to 4.95 are affected.
341 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
343 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
344 supplied as the db file-pair basename. Previously, if a directory
345 path was given, for example via the autoreply "once" option, the DB
346 file.pag and file.dir files would be created in that directory's
349 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
350 "taint" log_selector. These were previously deprecated.
352 JH/22 Fix static address-list lookups to properly return the matched item.
353 Previously only the domain part was returned.
355 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
356 the call into OpenSSL to send a TLS Close was being repeated; this
357 resulted in the library waiting for the peer's Close. If that was never
358 sent we waited forever. Fix by tracking send calls.
360 JH/24 The ${run} expansion item now expands its command string elements after
361 splitting. Previously it was before; the new ordering makes handling
362 zero-length arguments simpler. The old ordering can be obtained by
363 appending a new option "preexpand", after a comma, to the "run".
365 JH/25 Taint-check exec arguments for transport-initiated external processes.
366 Previously, tainted values could be used. This affects "pipe", "lmtp" and
367 "queryprogram" transport, transport-filter, and ETRN commands.
368 The ${run} expansion is also affected: in "preexpand" mode no part of
369 the command line may be tainted, in default mode the executable name
372 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
373 the the facility was not passed across execs, and only the first message
374 passed over a connection could use BDAT; any further ones using DATA.
376 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
377 uses $sending_ip_address and an interface is specified.
378 Previously any use of the local address in the EHLO name disabled
379 PIPECONNECT, the common case being to use the rDNS of it.
381 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
382 resumption. Previously verify failed because no certificate status is
383 passed on the wire for the restarted session. Fix by using the recorded
384 ocsp status of the stored session for the new connection.
386 JH/29 TLS resumption: the key for session lookup in the client now includes
387 more info that a server could potentially use in configuring a TLS
388 session, avoiding oferring mismatching sessions to such a server.
389 Previously only the server IP was used.
391 JH/30 Fix string_copyn() for limit greater than actual string length.
392 Previously the copied amount was the limit, which could result in a
393 overlapping memcpy for newly allocated destination soon after a
394 source string shorter than the limit. Found/investigated by KM.
396 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
397 close; it may be needed for a subsequent connection. This caused a
398 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
400 JH/32 Fix CHUNKING for a second message on a connection when the first was
401 rejected. Previously we did not reset the chunking-offered state, and
402 erroneously rejected the BDAT command. Investigation help from
405 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
406 an empty address. Previously the expansion returned an error.
408 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
409 proxy. Previously these were misparsed, leading to paniclog entries.
415 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
416 related applications. Previously an "H" was used where available info
417 says that "M" should be, so change to match.
419 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
420 as arguments, so an implementation trying to copy these into a local
421 buffer was taking a taint-enforcement trap. Fix by using dynamically
422 created buffers. Similar fix for radius expansion condition.
424 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
425 reasonable, eg. to count headers. Fix by using dynamically created
426 buffers rather than a local. Do similar fixes for ACL actions "dcc",
427 "log_reject_target", "malware" and "spam"; the arguments are expanded
428 so could be handling tainted values.
430 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
431 broken the (no-op) support for this sendmail command. Restore it
432 to doing nothing, silently, and returning good status.
434 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
435 record path was given (or the default used) without a leading directory
436 path, an error occurred on trying to open it. Use the transport's working
439 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
440 transport. Previously it was the name on the DNS A-record; use instead
441 the head of the CNAME chain leading there (if there is one). This seems
442 to align better with RFC 6125.
444 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
445 smtp_accept_max_per_host allocated resources which were not released
446 when the limit was exceeded. This eventually crashed the daemon. Fix
447 by adding a release action in that path.
449 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
450 expanded; previously using tainted values was rejected. Fix by using
451 dynamically-created buffers.
453 JH/09 Relax restrictions on ACL verify condition needing access to message
454 headers. Previously they were only permitted in data and non-smtp ACLs;
455 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
456 not_blind, header_sender and header_names_ascii verification.
458 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
459 Previously a macro used one argument twice; when called with the
460 argument as an expression having side-effects, incorrect operation
461 resulted. Use an inlineable function.
463 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
464 held open for a verify callout. Previously this wan not accounted for
465 and a corrupt onward SMTP conversation resulted.
467 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
468 Previously they were embedded naively in the constructed address; when
469 needed, strip the quoting and quote the entire local_part.
470 Also make the inbound_srs expansion condition handle quoting.
472 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
473 excluded, not matching the documentation.
475 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
476 was given for the sqlite_dbfile a trap resulted.
478 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
479 "name" argument resulted in a trap. There is no reason to disallow such;
480 this was a coding error.
482 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
483 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
484 spent suspended, ignoring the POSIX definition. Previously we assumed
485 it did and a constant offset from real time could be used as a correction.
486 Change to using the same clock source for the start-of-message and the
487 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
488 exists, just to get a clock slightly more aligned to reality.
490 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
491 RFC says it is optional some validators care. The missing char was not
492 intended but triggered by a line-wrap alignment. Discovery and fix by
493 Guillaume Outters, hacked on by JH.
495 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
496 name being quoted was tainted a trap would be taken. Fix by using
497 dynamically created buffers. The routine could have been called by a
498 rewrite with the "h" flag, by using the "-F" command-line option, or
499 by using a "name=" option on a control=submission ACL modifier.
501 JH/19 SPF: change the Authentication-Results expansion component to give
502 smtp.helo when the sender domain is empty. Previously it gave
505 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
506 not in 127.0.0.0/8 to help in spotting list domains taken over by a
507 domain-parking registrar.
509 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
510 Previously when a whitespace character was specified it was not inserted
511 after removing the newline.
513 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
514 the domain part of the recipient address. This overrides any tls_sni
515 option set, which was previously used.
517 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
520 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
521 is_tainted() had an off-by-one error in the overenthusiastic direction.
522 Find and fix by Gavan. Although NetBSD is not a supported platform for
523 4.94 this bug could affect other platforms.
525 PP/01 Fix default prime selection to be consistent.
526 One path used ike23 still, instead of exim.dev.20160529.3; now both
527 execution flows will use the same DH primes (currently
528 exim.dev.20160529.3).
530 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
531 option in smtp transport, to match the documentation. Previously
532 verification was not being done in some cases where it should have been.
534 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
535 than one server was defined and depending on the platform memory layout
536 details, an internal consistency trap could be hit while walking the list
539 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
540 files. The value is used by the authresults expansion item. Previously
541 if this was used in a router or transport, a crash could result.
543 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
544 have a sufficient incidence of EINTR returns from select that an
545 interaction with other operations done by the main daemon loop exposed
546 a bug in the error-handling. This was benign apart from the log
549 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
550 for consistency with delivered "=>" and failed "**" lines. While we're
551 there, handle PRX and TFO.
553 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
554 applied. This resulted, if any header-line rewrite rules were configured,
555 in a panic-log triggerable by sending a message with a long address in
556 a header. Fix by increasing the arbitrary limit to larger than a single
557 (dewrapped) 5322 header line maximum size.
559 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
560 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
561 RFC 6648 which deprecates X- options in protocols as a general practice.
562 Changeover between the implementations is handled by the mechanisms
565 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
566 a remote which uses callout/hold. Previously the local was queued.
568 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
571 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
572 log line. The asterisk is now consistently appended to the remote IP
573 (and port, if given), and will also be provided on defer and fail log
574 lines. Previously it could be placed on the local IP if that was being
575 logged, and was only provided on delivery lines.
577 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
579 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
580 authenticator client_send option. Previously the next char, after a pair
581 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
582 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
583 documentation. There is still no way to get a leading ^ immediately
584 after a NUL (ie. for the password of a PLAIN method authenticator.
586 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
587 files. For bad sizes read, delete the record and whine to paniclog.
589 JH/38 When logging an AUTH failure, as server, do not include sensitive
590 information. Previously, the credentials would be included if given
591 as part of the AUTH command line and an ACL denied authentication.
593 JH/39 Bug 2691: fix $local_part_data. When the matching list element
594 referred to a file, bad data was returned. This likely also affected
597 JH/40 The gsasl authenticator now supports caching of the salted password
598 generated by the client-side implementation. This required the addition
599 of a new variable: $auth4.
601 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
602 left undeleted; the attempt to re-create it then failed - resulting in
603 the usual "SIGHUP tp have daemon reload configuration" to not work.
604 This affected any platform not supporting "abstract" Unix-domain
605 sockets (i.e. not Linux).
607 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
608 recipients" response to RCPT in a later response, with a 250. The
609 previous coding assumed this would not happen, and under PIPELINING
610 would result in both lost and duplicate recipients for a message.
612 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
613 Previously the weighting was incorrectly applied. Similar fix for socks
614 proxies. Found and fixed by Heiko Schlichting.
616 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
617 not handle sub-lists included using the +namedlist syntax. While
618 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
619 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
622 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
623 Previously the permanent pool was used, so the sore could not be freed.
624 This meant a connection with many messages would use continually-growing
627 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
628 per-pool so as not to waste too much space. Previously a constant size
629 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
630 DOS attacks harder. The cost is wasted memory use in the larger blocks.
632 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
633 same space repeatedly, and a smaller process footprint.
635 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
636 for GnuTLS. Previously the name started "%s" which, while not a bug,
637 looked as if if might be one.
639 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
640 "continued connection") make the $proxy_* variables available. Previously
641 the information was not passed across the exec() call for subsequent
642 transport executions. This also mean that the log lines for the
643 messages can show the proxy information.
645 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
646 receive time. With modern systems the difference is significant.
647 The historical behaviour can be restored by disabling (a new) log_selector
648 "queue_time_exclusive".
650 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
651 content) were specifically tested for. Now, also cover expansions
652 resulting in ACL names and inline ACL content.
654 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
655 making it unusable in complex expressions.
657 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
658 fail with a taint-check on the spoolfile name, and leave the message
661 HS/01 Enforce absolute PID file path name.
663 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
665 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
667 PP/02 Bug 2643: Correct TLS DH constants.
668 A missing NUL termination in our code-generation tool had led to some
669 incorrect Diffie-Hellman constants in the Exim source.
670 Reported by kylon94, code-gen tool fix by Simon Arlott.
672 PP/03 Impose security length checks on various command-line options.
673 Fixes CVE-2020-SPRSS reported by Qualys.
675 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
676 better. Reported by Qualys.
678 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
679 providing a particularly obnoxious sender full name.
682 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
684 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
687 PP/08 Change default for recipients_max from unlimited to 50,000.
689 PP/09 Fix security issue with too many recipients on a message (to remove a
690 known security problem if someone does set recipients_max to unlimited,
691 or if local additions add to the recipient list).
692 Fixes CVE-2020-RCPTL reported by Qualys.
694 PP/10 Fix security issue in SMTP verb option parsing
695 Fixes CVE-2020-EXOPT reported by Qualys.
697 PP/11 Fix security issue in BDAT state confusion.
698 Ensure we reset known-good where we know we need to not be reading BDAT
699 data, as a general case fix, and move the places where we switch to BDAT
700 mode until after various protocol state checks.
701 Fixes CVE-2020-BDATA reported by Qualys.
703 HS/03 Die on "/../" in msglog file names
705 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
706 the Exim runtime user are allowed to create files.
708 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
711 QS/03 When reading the output from interpreted forward files we do not
712 pass the pipe between the parent and the interpreting process to
713 executed child processes (if any).
715 QS/04 Always die if requested from internal logging, even is logging is
718 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
719 the API; compilation noo longer completes with DMARC support included.
720 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
721 been reported on other platforms.
723 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
725 JH/56 Make the majority of info read from config files readonly, for defence-in-
726 depth against exploits. Suggestion by Qualys.
727 Not supported on Solaris 10.
729 JH/57 Fix control=fakreject for a custom message containing tainted data.
730 Previously this resulted in a log complaint, due to a re-expansion present
731 since fakereject was originally introduced.
733 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
734 modifier was given, a loop resulted.
736 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
737 pipelined SMTP command followed the BDAT LAST then it would be
738 incorrectly treated as part of the message body, causing a verification
741 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
742 require looser validation rules than those for 821-level addresses,
743 which only permit IP addresses.
749 JH/01 Avoid costly startup code when not strictly needed. This reduces time
750 for some exim process initialisations. It does mean that the logging
751 of TLS configuration problems is only done for the daemon startup.
753 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
755 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
756 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
759 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
760 DKIM signing is being done. Previously a transport_filter would always
761 disable CHUNKING, falling back to traditional DATA.
763 JH/05 Regard command-line recipients as tainted.
765 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
767 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
768 PAM library frees one of the arguments given to it, despite the
769 documentation. Therefore a plain malloc must be used.
771 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
772 on-stack buffers were used, resulting in a taint trap when DSN information
773 copied from a received message was written into the buffer.
775 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
776 the ordering of its ARC headers. This caused a crash.
778 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
779 a new record was being constructed with information from the peer, a trap
782 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
783 installation would get error messages from DMARC verify, when it hit the
784 nonexistent file indicated by the default. Distros wanting DMARC enabled
785 should both provide the file and set the option.
786 Also enforce no DMARC verification for command-line sourced messages.
788 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
789 could, depending on the platform, hang at the STARTTLS response.
791 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
792 message on a connection. Previously if one message had ARC headers and
793 the following one did not, a crash could result when adding an
794 Authentication-Results: header.
796 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
797 Exim main code and Exim-related utities. The introduction of taint
798 tracking also did many adjustments to string handling. Since then, eximon
799 frequently terminated with an assert failure.
801 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
802 check for 452 responses. This slightly helps the inefficieny of doing
803 a large alias-expansion into a recipient-limited target. The max_rcpt
804 transport option still applies (and at the current default, will override
805 the new feature). The check is done for either cause of synch, and forces
806 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
807 connection. The new facility is not tunable at this time.
809 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
810 library live data was being used, so the results became garbage. Make
811 copies while it is still usable.
813 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
814 on delivery deferred (==) and failed (**) lines (if a delivery was
815 attemtped). Previously it was only on completion (=>) lines.
817 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
818 for the expansion of the server_scram_iter and server_scram_salt options.
820 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
821 are now specifically given a NO_DATA response without hitting the system
822 resolver. The library goes on to do the now-standard TXT lookup.
823 Use of dnsdb lookups is not affected.
825 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
826 only retrieve the errormessage once. Previously two calls to dlerror()
827 were used, and the second one (for mainlog/paniclog) retrieved null
830 JH/20 Taint checking: disallow use of tainted data for
831 - the appendfile transport file and directory options
832 - the pipe transport command
833 - the autoreply transport file, log and once options
834 - file names used by the redirect router (including filter files)
836 - paths used by single-key lookups
837 Previously this was permitted.
839 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
840 adjusted the size of a major service buffer; this failed because the
841 buffer was in use at the time. Change to a compile-time increase in the
842 buffer size, when this authenticator is compiled into exim.
844 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
845 previous fast-mode was untenable in the face of glibs using mmap to
846 support larger malloc requests.
848 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
849 New values supported, if defined on system where compiled:
850 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
851 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
853 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
854 running a limited number of proceses in parallel, a benefit is gained. The
855 amount varies with the platform hardware and load. The use of the option
856 queue_run_in_order means we cannot do this, as ordering becomes
859 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
860 had introduced a string-copy (for ensuring NUL-termination) which was not
861 appropriate for that case, which can include embedded NUL bytes in the
862 block of data. Investigation showed the copy to actually be needless, the
863 data being length-specified.
865 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
866 done during a receiving connection, and both used TLS, global info was
867 used rather than per-connection info for tracking the state of data
868 queued for transmission. This could result in a connection hang.
870 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
871 Previously, when delivering serveral messages down a single connection
872 only the first would provide a SIZE. This was due to the size information
873 not being properly tracked.
875 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
876 TAI (at 37 seconds currently), pretend to be in UTC for time-related
877 expansion and logging. Previously, spurious values such as a future
878 minute could be seen.
880 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
881 it could crash from a null-deref. This could also affect the
882 ${addresses: } operator and ${readsock } item.
884 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
885 message following a mime one, the variable was not reset.
887 JH/30 When an pipelined-connect fails at the first response, assume incorrect
888 cached capability (perhaps the peer reneged?) and immediately retry in
891 JH/31 Fix spurious detection of timeout while writing to transport filter.
893 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
894 an attempt to copy the string was made before checking it.
896 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
897 taint of the lookup key was maintained; we now regard the presence in the
898 filesystem as sufficient validation.
900 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
901 argument is supplied.
903 JH/35 The dsearch lookup now requires that the directory is an absolute path.
904 Previously this was not checked, and nonempty relative paths made an
905 access under Exim's current working directory.
907 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
908 Previously no event was raised.
910 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
911 parameter supplied by the sender MAIL FROM command. Previously it was
912 ignored, and only the check_spool_space option value for the required
915 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
916 the size of the signing public-key. Previously it was instead giving
917 the size of the signature hash.
919 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
920 the default. See the (new) dkim_verify_min_keysizes option.
922 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
923 and an ACL use a lookup for checking either the local_part or domain,
924 stale data could be accessed. Ensure that variable references are
925 dropped between messages.
927 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
928 by the client was not checked as pointing within response data before
929 being used. A malicious client could thus cause an out-of-bounds read and
930 possibly gain authentication. Fix by adding the check.
932 JH/42 Internationalisation: change the default for downconversion in the smtp
933 transport to be "if needed". Previously it was "as previously set" for
934 the message, which usually meant "if needed" for message-submission but
935 "no" for everything else. However, MTAs have been seen using SMTPUTF8
936 even when the envelope addresses did not need it, resulting in forwarding
937 failures to non-supporting MTAs. A downconvert in such cases will be
938 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
939 transport. The change does mean that addresses needing conversion will
940 be converted when previously a delivery failure would occur.
942 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
943 response was received it would be used unchecked in a fail-DSN, violating
944 standards on line-length limits. Truncate if needed.
946 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
947 doesn't work. (Additionally add a new main config option to configure the
954 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
955 side, to decode a TLS 1.3 packet capture.
957 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
958 Previously the default library behaviour applied, sending two, each in
961 JH/03 Debug output for ACL now gives the config file name and line number for
964 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
966 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
968 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
969 buffer overrun for (non-chunking) other transports.
971 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
972 TLS1.3, means that a server rejecting a client certificate is not visible
973 to the client until the first read of encrypted data (typically the
974 response to EHLO). Add detection for that case and treat it as a failed
975 TLS connection attempt, so that the normal retry-in-clear can work (if
976 suitably configured).
978 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
979 and/or domain. Found and fixed by Jason Betts.
981 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
982 configuration). If a CNAME target was not a wellformed name pattern, a
985 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
986 the OS reports them interleaved with other addresses.
988 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
989 used both for input and for a verify callout, both encrypted, SMTP
990 responses being sent by the server could be lost. This resulted in
991 dropped connections and sometimes bounces generated by a peer sending
994 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
995 string. Previously it was possible to cause undefined behaviour in a
996 library routine (usually a crash). Found by "zerons".
998 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1001 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1002 API was removed, so update to use the newer ones.
1004 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1005 any timeout set, is taking a long time. Previously we would hang on to a
1006 rotated logfile "forever" if the input was arriving with long gaps
1007 (a previous attempt to fix addressed lack, for a long time, of initial
1010 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1011 shared (NFS) environment. The length of the tempfile name is now
1012 4 + 16 ("hdr.$message_exim_id") which might break on file
1013 systems which restrict the file name length to lower values.
1014 (It was "hdr.$pid".)
1016 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1017 shared (NFS) environment.
1019 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1020 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1023 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1024 on some platforms for bit 31.
1026 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1027 to changes apparently associated with TLS1.3 handling some of the APIs
1028 previously used were either nonfunctional or inappropriate. Strings
1029 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1030 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1031 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1032 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1033 and the use of specific cipher names in the encrypted= ACL condition.
1035 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1037 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1038 verification result was not updated unless hosts_require_ocsp applied.
1040 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1041 queue_list_requires_admin set to false, non-admin users were denied the
1044 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1045 directory-of-certs mode. Previously they were advertised despite the
1048 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1049 A single TCP connection by a client will now hold a TLS connection open
1050 for multiple message deliveries, by default. Previously the default was to
1053 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1054 default. If built with the facility, DANE will be used. The facility
1055 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1057 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1058 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1059 must be defined and you must still, unless you define DISABLE_TLS, manage
1060 the the include-dir and library-file requirements that go with that
1061 choice. Non-TLS builds are still supported.
1063 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1066 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1067 default. If the platform supports and has the facility enabled, it will
1068 be requested on all coneections.
1070 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1071 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1073 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1075 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1076 success-DSN messages. Previously the From: header was always the default
1077 one for these; the option was ignored.
1079 JH/28 Fix the timeout on smtp response to apply to the whole response.
1080 Previously it was reset for every read, so a teergrubing peer sending
1081 single bytes within the time limit could extend the connection for a
1082 long time. Credit to Qualsys Security Advisory Team for the discovery.
1084 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1085 delivery address, which leaked information of the results of local
1086 forwarding. Change to the original envelope recipient address, per
1089 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1090 requested. Previously not bounce was generated and a log entry of
1091 error ignored was made.
1093 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1095 JH/32 Introduce a general tainting mechanism for values read from the input
1096 channel, and values derived from them. Refuse to expand any tainted
1097 values, to catch one form of exploit.
1099 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1100 was unused and the unexpanded text used for the test. Found and
1101 fixed by Ruben Jenster.
1103 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1104 an attempt to use a TLS library read routine dereffed a nul pointer,
1107 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1108 them in our smtp response.
1110 JH/36 Have the general router option retry_use_local_part default to true when
1111 any of the restrictive preconditions are set (to anything). Previously it
1112 was only for check_local user. The change removes one item of manual
1113 configuration which is required for proper retries when a remote router
1114 handles a subset of addresses for a domain.
1116 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1117 link count into consideration.
1119 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1120 caused the extension of big_buffer, the following lines were ignored.
1122 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1123 accordance with RFC 2308. Previously there was no expiry, so a longlived
1124 receive process (eg. due to ACL delays) versus a short SOA value could
1127 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1129 JH/39 Promote DMARC support to mainline.
1131 JH/40 Bug 2452: Add a References: header to DSNs.
1133 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1134 parameters. The relevant library call is documented as "Deprecated: This
1135 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1136 3.6.0, DH parameters are negotiated following RFC7919."
1138 HS/06 Change the default of dnssec_request_domains to "*"
1140 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1141 carried on and emitted a BDAT command, even when PIPELINING was not
1144 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1145 buffer was used for the filename, resulting in a trap when tainted
1146 arguments (eg. $domain) were used.
1148 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1149 recommended to avoid a possible server-load attack. The feature can be
1150 re-enabled via the openssl_options main cofiguration option.
1152 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1153 for version 4.90 - adding a "more data" boolean to the arguments.
1154 Bumped the ABI version number also, this having been missed previously;
1155 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1156 respect of usage of smtp_printf() by either local_scan code or libraries
1157 accessed via the ${dlfunc } expansion item. Both will need coding
1158 adjustment for any calls to smtp_printf() to match the new function
1159 signature; a FALSE value for the new argument is always safe.
1161 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1162 the file-offset (which the Linux syscall does, and exim expects); this
1163 resulted in an indefinite loop.
1165 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1166 to do ARC verification. The Authentication-Results: header line added
1167 by the configuration then had no ARC item.
1169 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1170 an incorrect mode was used for reading the result, resulting in it being
1177 JH/01 Remove code calling the customisable local_scan function, unless a new
1178 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1180 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1181 non-signal-safe functions being used.
1183 JH/03 Bug 2269: When presented with a received message having a stupidly large
1184 number of DKIM-Signature headers, disable DKIM verification to avoid
1185 a resource-consumption attack. The limit is set at twenty.
1187 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1188 report of oldest_pass in ${authres } in consequence, and separate out
1189 some descriptions of reasons for verification fail.
1191 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1192 files in the spool were present and unlocked. A queue-runner could spot
1193 them, resulting in a duplicate delivery. Fix that by doing the unlock
1194 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1195 add more error-checking on spoolfile handling while that code is being
1198 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1199 No known attacks, no CVE, this is defensive hardening.
1201 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1202 a queue-runner could start a delivery while other operations were ongoing.
1203 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1204 Found and investigated by Tim Stewart. Fix by using the open message data
1205 file handle rather than opening another, and not locally closing it (which
1206 releases a lock) for that case, while creating the temporary .eml format
1207 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1209 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1210 $sender_verify_failure/$recipient_verify_failure to "random".
1212 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1215 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1216 Previously this would segfault.
1218 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1221 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1222 like zero, since the resolver should be doing this for us, But we need one
1223 as a CNAME but no MX presence gets the CNAME returned; we need to check
1224 that doesn't point to an MX to declare it "no MX returned" rather than
1225 "error, loop". A new main option is added so the older capability of
1226 following some limited number of chain links is maintained.
1228 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1230 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1231 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1232 now-deprecated earlier definitions used only specified the range up to TLS
1233 1.2 (in the older-version library docs).
1235 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1237 JH/15 Rework TLS client-side context management. Stop using a global, and
1238 explicitly pass a context around. This enables future use of TLS for
1239 connections to service-daemons (eg. malware scanning) while a client smtp
1240 connection is using TLS; with cutthrough connections this is quite likely.
1242 JH/16 Fix ARC verification to do AS checks in reverse order.
1244 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1246 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1247 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1248 Previously the "utf8" would be re-prepended for every additional message.
1250 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1251 Previously thery were accepted, resulting in issues when attempting to
1252 forward messages to a non-supporting MTA.
1254 PP/02 Let -n work with printing macros too, not just options.
1256 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1257 one parent address was copied, and bogus data was used at delivery-logging
1258 time. Either a crash (after delivery) or bogus log data could result.
1259 Discovery and analysis by Tim Stewart.
1261 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1262 Previously if the string ended mid-character, we did not insert the
1263 promised '?' replacement.
1265 PP/04 Documentation: current string operators work on bytes, not codepoints.
1267 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1268 should pack well giving a smaller memory footprint so better caching and
1269 therefore performance. Group the declarations where this can't be done so
1270 that the byte-sized flag variables are not interspersed among pointer
1271 variables, giving a better chance of good packing by the compiler.
1273 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1274 non-null, to avoid issues with sites running BATV. Previously reports were
1275 sent with an empty envelope sender so looked like bounces.
1277 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1278 The ignore_error flag wasn't being returned from the filter subprocess so
1279 was not set for later routers. Investigation and fix by Matthias Kurz.
1281 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1282 and a msg:complete for the whole, when a message is manually removed using
1283 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1285 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1286 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1287 hope that is portable enough.
1289 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1290 requiring a known-CA anchor certificate; make it now rely entirely on the
1291 TLSA as an anchor. Checking the name on the leaf cert against the name
1292 on the A-record for the host is still done for TA (but not for EE mode).
1294 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1295 included in delivery lines for non-proxied connections, when compiled with
1296 SUPPORT_SOCKS and running with proxy logging enabled.
1298 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1299 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1300 move the existing event to fire before the normal logging of message
1301 failure so that custom logging is bracketed by normal logging.
1303 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1304 msg:fail:internal event. Developement by Matthias Kurz.
1306 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1307 far too small for todays use of crypto signatures stored there. Go all
1308 the way to the max DNS message size of 64kB, even though this might be
1309 overmuch for IOT constrained device use.
1311 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1312 copy a string over itself. The library routine is documented as not
1313 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1315 JH/32 For main options check_spool_space and check_inode_space, where the
1316 platform supports 64b integers, support more than the previous 2^31 kB
1317 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1318 the previous G, M, k.
1320 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1321 $authenticated_fail_id variable on authentication failure. Previously
1324 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1325 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1326 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1327 GNUTLS_SEC_PARAM_MEDIUM.
1329 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1330 side. Previously we would continue as if no SNI had been received.
1332 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1333 "<" character, which should be a list-separator specification, we walked
1334 off past the nul-terimation.
1336 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1337 causes) even when the retry time is not yet met. Previously they were
1338 not, meaning that when (say) an account was over-quota and temp-rejecting,
1339 and multiple senders' messages were queued, only one sender would get
1340 notified on each configured delay_warning cycle.
1342 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1344 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1345 cipher-suites, an error can be left on the stack even for a succeeding
1346 accept; this results in impossible error messages when a later operation
1349 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1350 return error codes indicating retry. Under TLS1.3 this becomes required.
1352 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1353 it only wrote the new authenticators, resulting in a lack of tracking of
1354 peer changes of ESMTP extensions until the next cache flush.
1356 JH/41 Fix the loop reading a message header line to check for integer overflow,
1357 and more-often against header_maxsize. Previously a crafted message could
1358 induce a crash of the recive process; now the message is cleanly rejected.
1360 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1361 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1367 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1368 When redis_servers is set to a list of > 1 element, and the Redis servers
1369 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1370 case of MOVED into a DEFER case instead, thus moving the query onto the
1371 next server in the list. For a cluster of N elements, all N servers must
1372 be defined in redis_servers.
1374 GF/02 Catch and remove uninitialized value warning in exiqsumm
1375 Check for existence of @ARGV before looking at $ARGV[0]
1377 JH/01 Replace the store_release() internal interface with store_newblock(),
1378 which internalises the check required to safely use the old one, plus
1379 the allocate and data copy operations duplicated in both (!) of the
1380 extant use locations.
1382 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1383 modifier. This matches the restriction on the commandline.
1385 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1386 Previously only the last row was returned.
1388 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1389 we assumed that tags in the header were well-formed, and parsed the
1390 element content after inspecting only the first char of the tag.
1391 Assumptions at that stage could crash the receive process on malformed
1394 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1395 While running the DKIM ACL we operate on the Permanent memory pool so that
1396 variables created with "set" persist to the DATA ACL. Also (at any time)
1397 DNS lookups that fail create cache records using the Permanent pool. But
1398 expansions release any allocations made on the current pool - so a dnsdb
1399 lookup expansion done in the DKIM ACL releases the memory used for the
1400 DNS negative-cache, and bad things result. Solution is to switch to the
1401 Main pool for expansions.
1402 While we're in that code, add checks on the DNS cache during store_reset,
1403 active in the testsuite.
1404 Problem spotted, and debugging aided, by Wolfgang Breyha.
1406 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1407 When none of the hosts presented to a transport match an already-open
1408 connection, close it and proceed with the list. Previously we would
1409 queue the message. Spotted by Lena with Yahoo, probably involving
1412 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1413 Previously a spurious "250 OK id=" response was appended to the proper
1416 JH/08 The "support for" informational output now, which built with Content
1417 Scanning support, has a line for the malware scanner interfaces compiled
1418 in. Interface can be individually included or not at build time.
1420 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1421 by the template makefile "src/EDITME". The "STREAM" support for an older
1422 ClamAV interface method is removed.
1424 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1425 rows affected is given instead).
1427 JH/11 The runtime Berkeley DB library version is now additionally output by
1428 "exim -d -bV". Previously only the compile-time version was shown.
1430 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1431 SMTP connection. Previously, when one had more recipients than the
1432 first, an abortive onward connection was made. Move to full support for
1433 multiple onward connections in sequence, handling cutthrough connection
1434 for all multi-message initiating connections.
1436 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1437 routers. Previously, a multi-recipient message would fail to match the
1438 onward-connection opened for the first recipient, and cause its closure.
1440 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1441 a timeout on read on a GnuTLS initiating connection, resulting in the
1442 initiating connection being dropped. This mattered most when the callout
1443 was marked defer_ok. Fix to keep the two timeout-detection methods
1446 JH/15 Relax results from ACL control request to enable cutthrough, in
1447 unsupported situations, from error to silently (except under debug)
1448 ignoring. This covers use with PRDR, frozen messages, queue-only and
1451 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1453 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1454 metadata, resulting in a crash in free().
1456 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1457 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1458 Broken also in d185889f4, with init system revamp.
1460 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1461 Previously we abruptly closed the connection after reading a malware-
1462 found indication; now we go on to read the "scan ok" response line,
1465 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1466 ACL. Previously, a crash would result.
1468 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1469 macro text after a replacement (previously it was only once per line) and
1470 by skipping builtin macros when searching for an uppercase lead character.
1472 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1473 for the build is renamed.
1475 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1476 was allocated for every new TLS startup, meaning one per message. Fix
1477 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1479 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1480 reported the original. Fix to report (as far as possible) the ACL
1481 result replacing the original.
1483 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1484 OpenSSL. Certificate information is loaded for every new TLS startup,
1485 and the resources needed to be freed.
1487 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1489 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1492 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1493 DATA response info to the (existing) per-recipient response info for
1494 the "C=" log element. It can have useful tracking info from the
1495 destination system. Patch from Simon Arlott.
1497 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1498 length value. Previously this would segfault.
1500 HS/02 Support Avast multiline protoocol, this allows passing flags to
1501 newer versions of the scanner.
1503 JH/28 Ensure that variables possibly set during message acceptance are marked
1504 dead before release of memory in the daemon loop. This stops complaints
1505 about them when the debug_store option is enabled. Discovered specifically
1506 for sender_rate_period, but applies to a whole set of variables.
1507 Do the same for the queue-runner and queue-list loops, for variables set
1508 from spool message files. Do the same for the SMTP per-message loop, for
1509 certain variables indirectly set in ACL operations.
1511 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1512 as a multi-recipient message from a mailinglist manager). The coding had
1513 an arbitrary cutoff number of characters while checking for more input;
1514 enforced by writing a NUL into the buffer. This corrupted long / fast
1515 input. The problem was exposed more widely when more pipelineing of SMTP
1516 responses was introduced, and one Exim system was feeding another.
1517 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1518 receiving system, and refused recipients seen by the sending system
1519 (propating to people being dropped from mailing lists).
1520 Discovered and pinpointed by David Carter.
1522 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1523 replaced by the ${authresults } expansion.
1525 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1527 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1528 allows proper process termination in container environments.
1530 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1531 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1533 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1534 and "err_perm", deprecated since 4.83 when the RFC-defined words
1535 "temperror" and "permerror" were introduced.
1537 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1538 transport-filters or DKIM-signing. The restriction was lost in the
1539 consolidation of verify-callout and delivery SMTP handling.
1540 Extend the restriction to also cover ARC-signing.
1542 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1543 in defer=pass mode supply a 450 to the initiator. Previously the message
1546 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1547 tls_require_ciphers is used as before.
1549 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1550 "pass_unscanned". Only tmpfails from the scanner are written to
1551 the paniclog, as they may require admin intervention (permission
1552 denied, license issues). Other scanner errors (like decompression
1553 bombs) do not cause a paniclog entry.
1555 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1556 Previously it was possible to log spurious information in receive log
1559 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1560 triggered odd behaviour from Outlook Express clients.
1562 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1565 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1566 since the IETF WG has not yet settled on that versus the original
1567 "bare" representation.
1569 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1570 Previously the millisecond value corrupted the output.
1571 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1572 corrupted the output.
1578 JH/01 Rework error string handling in TLS interface so that the caller in
1579 more cases is responsible for logging. This permits library-sourced
1580 string to be attached to addresses during delivery, and collapses
1581 pairs of long lines into single ones.
1583 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1584 during configuration. Wildcards are allowed and expanded.
1586 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1587 This permits better logging.
1589 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1590 do not close it down and have the child transport start it up again on
1591 the passed-on TCP connection. Instead, proxy the child (and any
1592 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1593 affected: the continued delivery log lines do not have any DNSSEC, TLS
1594 Certificate or OCSP information. TLS cipher information is still logged.
1596 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1597 identical IP addresses on different listening ports. Will also affect
1600 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1601 add noisy ifdef guards to special-case this sillyness.
1602 Patch from Bernd Kuhls.
1604 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1605 than 255 are no longer allowed.
1607 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1608 Disable the session-cache too, which might reduce our load. Since we
1609 currrectly use a new context for every connection, both as server and
1610 client, there is no benefit for these.
1611 GnuTLS appears to not support tickets server-side by default (we don't
1612 call gnutls_session_ticket_enable_server()) but client side is enabled
1613 by default on recent versions (3.1.3 +) unless the PFS priority string
1616 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1617 <https://reproducible-builds.org/specs/source-date-epoch/>.
1619 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1620 the check for any unsuccessful recipients did not notice the limit, and
1621 erroneously found still-pending ones.
1623 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1624 MSG_MORE. Only in-clear (not on TLS connections).
1626 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1627 transport-filter is involved we can buffer the headers in memory for
1628 creating the signature, and read the spool data file once for the
1629 signature and again for transmission.
1631 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1632 4.77 as the kernel support then wasn't solid, having issues in 64bit
1633 mode. Now, it's been long enough. Add support for FreeBSD also.
1635 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1636 case where the routing stage had gathered several addresses to send to
1637 a host before calling the transport for the first, we previously failed
1638 to close down TLS in the old transport process before passing the TCP
1639 connection to the new process. The new one sent a STARTTLS command
1640 which naturally failed, giving a failed delivery and bloating the retry
1641 database. Investigation and fix prototype from Wolfgang Breyha.
1643 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1644 false-negatives in the check that the sender had not preempted a response
1645 or prompt from Exim (running as a server), due to that code's lack of
1646 awareness of the SMTP input buffering.
1648 PP/04 Add commandline_checks_require_admin option.
1649 Exim drops privileges sanely, various checks such as -be aren't a
1650 security problem, as long as you trust local users with access to their
1651 own account. When invoked by services which pass untrusted data to
1652 Exim, this might be an issue. Set this option in main configuration
1653 AND make fixes to the calling application, such as using `--` to stop
1656 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1657 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1658 responses to those, into a single TLS record each way (this usually means
1659 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1662 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1663 affects you only if you're dancing at the edge of the param size limits.
1664 If you are, and this message makes sense to you, then: raise the
1665 configured limit or use OpenSSL 1.1. Nothing we can do for older
1668 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1669 cmdline element to get the documented default one. Previously it was
1672 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1673 are unsuable for non-message-specific reasons - eg. connection timeout,
1674 banner-time rejection.
1676 JH/16 Fix logging of delivery remote port, when specified by router, under
1679 PP/06 Repair manualroute's ability to take options in any order, even if one
1680 is the name of a transport.
1683 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1685 JH/17 Change the list-building routines interface to use the expanding-string
1686 triplet model, for better allocation and copying behaviour.
1688 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1689 Previously it was constructed the first time a possibly-matching string
1690 was met in the configuration file input during startup; now it is done
1693 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1694 compatible one, to avoid the (poorly documented) possibility of a config
1695 file in the working directory redirecting the DB files, possibly correpting
1696 some existing file. CVE-2017-10140 assigned for BDB.
1698 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1699 cache-hot. Previously, although the result was properly cached, the
1700 initial verify call returned a defer.
1702 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1703 the main verify for receipient in uncached-mode.
1705 JH/22 Retire historical build files to an "unsupported" subdir. These are
1706 defined as "ones for which we have no current evidence of testing".
1708 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1709 if present. Previously it was ignored.
1711 JH/24 Start using specified-initialisers in C structure init coding. This is
1712 a C99 feature (it's 2017, so now considered safe).
1714 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1715 if was a fixed-sized field and bitmask ops via macros; it is now more
1718 PP/07 GitHub PR 56: Apply MariaDB build fix.
1719 Patch provided by Jaroslav Škarvada.
1721 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1722 during Coverity cleanups [4.87 JH/47]
1723 Diagnosis and fix provided by Michael Fischer v. Mollard.
1725 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1726 the right size to place the terminating semicolon on its own folded
1727 line, the header hash was calculated to an incorrect value thanks to
1728 the (relaxed) space the fold became.
1730 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1731 and confused the parent.
1733 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1734 which could crash as a result. This could lead to undeliverable messages.
1736 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1739 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1740 matters on fast-turnover and PID-randomising systems, which were getting
1741 out-of-order delivery.
1743 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1744 a possibly-overlapping copy. The symptom was that "Remote host closed
1745 connection in response to HELO" was logged instead of the actual 4xx
1748 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1749 Previously only that bufferd was discarded, resulting in SYMTP command
1752 JH/32 DKIM: when a message has multiple signatures matching an identity given
1753 in dkim_verify_signers, run the dkim acl once for each. Previously only
1754 one run was done. Bug 2189.
1756 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1757 "panic the current process" to "deliberately defer". The panic log is
1758 still written with the problem list name; the mail and reject logs now
1759 get a temp-reject line for the message that was being handled, saying
1760 something like "domains check lookup or other defer". The SMTP 451
1761 message is still "Temporary local problem".
1763 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1764 A crafted sequence of BDAT commands could result in in-use memory beeing
1765 freed. CVE-2017-16943.
1767 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1768 from SMTP input. Previously it was always done; now only done for DATA
1769 and not BDAT commands. CVE-2017-16944.
1771 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1772 to the message (such as an overlong header line). Previously this was
1773 not done and we did not exit BDAT mode. Followon from the previous item
1774 though a different problem.
1780 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1781 than -2003 did; needs libidn2 in addition to libidn.
1783 JH/02 The path option on a pipe transport is now expanded before use.
1785 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1786 Patch provided by "Björn", documentation fix added too.
1788 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1789 missing a wire-to-host endian conversion.
1791 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1792 close after a BDAT command line could be taken as a following command,
1793 giving a synch failure. Fix by only checking for synch immediately
1794 before acknowledging the chunk.
1796 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1797 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1798 macro. Patches provided by Josh Soref.
1800 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1801 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1802 by IANA as a service extension. However, John Klensin suggests that we
1805 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1806 the dkim code may be unix-mode line endings rather than smtp wire-format
1807 CRLF, so prepend a CR to any bare LF.
1809 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1810 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1812 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1813 the first were themselves being wrongly included in the feed into dkim
1814 processing; with most chunk sizes in use this resulted in an incorrect
1815 body hash calculated value.
1817 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1818 DKIM signature block, for verification. Although advised against by
1819 standards it is specifically not ruled illegal.
1821 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1823 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1824 missing a body hash (the bh= tag).
1826 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1827 It seems that HAProxy sends the Proxy Protocol information in clear and
1828 only then does a TLS startup, so do the same.
1830 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1831 TCP connections (such as for Spamd) unless the daemon successfully set
1832 Fast Open mode on its listening sockets. This fixes breakage seen on
1833 too-old kernels or those not configured for Fast Open, at the cost of
1834 requiring both directions being enabled for TFO, and TFO never being used
1835 by non-daemon-related Exim processes.
1837 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1838 endings, at least on the first header line. Try to canonify any that get
1839 past that check, despite the cost.
1841 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1842 now limited to an arbitrary five deep, while parsing addresses with the
1843 strip_excess_angle_brackets option enabled.
1845 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1846 instead leave the unprompted TLS handshake in socket buffer for the
1847 TLS library to consume.
1849 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1851 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1853 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1854 region in one operation, for speed, and this leaves assigned pointers
1855 dangling. Add checks run only under the testsuite which checks all
1856 variables at a store-reset and panics on a dangling pointer; add code
1857 explicitly nulling out all the variables discovered. Fixes one known
1858 bug: a transport crash, where a dangling pointer for $sending_ip_address
1859 originally assigned in a verify callout, is re-used.
1861 PP/06 Drop '.' from @INC in various Perl scripts.
1863 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1865 PP/08 Reduce a number of compilation warnings under clang; building with
1866 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1867 should be warning-free.
1869 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1871 HS/01 Fix portability problems introduced by PP/08 for platforms where
1872 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1874 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1875 chunk. This allows us to accept broken chunked messages. We need a more
1876 general solution here.
1878 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1879 already-broken messages in the queue.
1881 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1883 JH/19 Fix reference counting bug in routing-generated-address tracking.
1889 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1890 supports it and a size is available (ie. the sending peer gave us one).
1892 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1893 years of being deprecated). The replacements are the ACLs
1894 acl_smtp_mime and acl_not_smtp_mime.
1896 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1897 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1898 or even an in-clear connection were permitted. Now, if the host lookup
1899 was dnssec and dane was requested then the host is only used if the
1900 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1901 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1902 if one fails this test.
1903 This means that a poorly-configured remote DNS will make it incommunicado;
1904 but it protects against a DNS-interception attack on it.
1906 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1907 non-noisy when a race steals the message being considered.
1909 JH/05 If main configuration option tls_certificate is unset, generate a
1910 self-signed certificate for inbound TLS connections.
1912 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1913 in rewrites and routers.
1915 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1916 and logged a warning sing 4.83; now they are a configuration file error.
1918 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1919 (lacking @domain). Apply the same qualification processing as RCPT.
1921 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1923 JH/10 Support ${sha256:} applied to a string (as well as the previous
1926 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1927 a cutthrough deliver is pending, as we always want to make a connection.
1928 This also avoids re-routing the message when later placing the cutthrough
1929 connection after a verify cache hit.
1930 Do not update it with the verify result either.
1932 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1933 when routing results in more than one destination address.
1935 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1936 signing (which inhibits the cutthrough capability). Previously only
1937 the presence of an option was tested; now an expansion evaluating as
1938 empty is permissible (obviously it should depend only on data available
1939 when the cutthrough connection is made).
1941 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1942 the relevant preceding SMTP command did not note the pipelining mode.
1944 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1945 Previously they were not counted.
1947 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1948 as one having no matching records. Previously we deferred the message
1949 that needed the lookup.
1951 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1952 distinguished as "(=".
1954 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1955 for missing MX records. Previously it only worked for missing A records.
1957 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1959 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1960 after the data-go-ahead and data-ack. Patch from Jason Betts.
1962 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1963 even for a "none" policy. Patch from Tony Meyer.
1965 JH/22 Fix continued use of a connection for further deliveries. If a port was
1966 specified by a router, it must also match for the delivery to be
1969 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1970 When one of the recipients of a message was unsuitable for the connection
1971 (has no matching addresses), we lost track of needing to mark it
1972 deferred. As a result mail would be lost.
1974 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1976 JH/25 Decoding ACL controls is now done using a binary search; the source code
1977 takes up less space and should be simpler to maintain. Merge the ACL
1978 condition decode tables also, with similar effect.
1980 JH/26 Fix problem with one_time used on a redirect router which returned the
1981 parent address unchanged. A retry would see the parent address marked as
1982 delivered, so not attempt the (identical) child. As a result mail would
1985 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1986 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1987 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1990 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1991 Main config options check_{log,spool}_{inodes,space} are now
1992 100 inodes, 10MB unless set otherwise in the configuration.
1994 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1995 Previously it only applied to the main-section connection policy
1998 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2000 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2001 by me. Added RFC7919 DH primes as an alternative.
2003 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2004 are not in the system include path.
2006 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2007 GnuTLS, when a session startup failed (eg because the client disconnected)
2008 Exim did stdio operations after fclose. This was exposed by a recent
2009 change which nulled out the file handle after the fclose.
2011 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2012 signed directly by the cert-signing cert, rather than an intermediate
2013 OCSP-signing cert. This is the model used by LetsEncrypt.
2015 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2017 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2018 an incoming connection.
2020 HS/02 Bug 1802: Do not half-close the connection after sending a request
2023 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2024 fallback to "prime256v1".
2026 JH/34 SECURITY: Use proper copy of DATA command in error message.
2027 Could leak key material. Remotely exploitable. CVE-2016-9963.
2033 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2034 and 3.4.4 - once the server is enabled to respond to an OCSP request
2035 it does even when not requested, resulting in a stapling non-aware
2036 client dropping the TLS connection.
2038 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2039 support variable-length bit vectors. No functional change.
2041 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2042 The I= interface field on outgoing lines is now after the H= remote
2043 host field, same as incoming lines. There is a separate
2044 outgoing_interface log selector which allows you to disable the
2047 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2048 If not running log_selector +smtp_connection the mainlog would be held
2049 open indefinitely after a "too many connections" event, including to a
2050 deleted file after a log rotate. Leave the per net connection logging
2051 leaving it open for efficiency as that will be quickly detected by the
2052 check on the next write.
2054 HS/01 Bug 1671: Fix post transport crash.
2055 Processing the wait-<transport> messages could crash the delivery
2056 process if the message IDs didn't exist for some reason. When
2057 using 'split_spool_directory=yes' the construction of the spool
2058 file name failed already, exposing the same netto behaviour.
2060 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2061 mime_regex ACL conditions.
2063 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2064 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2065 to HELO, local diagnostic string.
2067 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2068 log line to debug. Even when configured with a tls authenticator many
2069 client connections are expected to not authenticate in this way, so
2070 an authenticate fail is not an error.
2072 HS/02 Add the Exim version string to the process info. This way exiwhat
2073 gives some more detail about the running daemon.
2075 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2076 matter for fast-change records such as DNSBLs.
2078 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2079 retry record, even if constant. There may be multiple transports with
2080 different interface settings and the retry behaviour needs to be kept
2083 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2085 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2087 JH/10 Bug 840: fix log_defer_output option of pipe transport
2089 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2090 hosts_randomize. This matters a lot when combined with mua_wrapper.
2092 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2093 ${quote_pgsql:<string>} operator.
2095 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2097 JH/14 Change header file name for freeradius-client. Relevant if compiling
2098 with Radius support; from the Gentoo tree and checked under Fedora.
2100 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2102 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2103 option transport_filter, meaning no filtering.
2105 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2107 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2108 defaults to "*" (all hosts). The variable is now available when not built
2109 with TLS, default unset, mainly to enable keeping the testsuite sane.
2110 If a server certificate is not supplied (via tls_certificate) an error is
2111 logged, and clients will find TLS connections fail on startup. Presumably
2112 they will retry in-clear.
2113 Packagers of Exim are strongly encouraged to create a server certificate
2114 at installation time.
2116 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2117 with the $config_file variable.
2119 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2120 in transport context, after the attempt, and per-recipient. The latter type
2121 is per host attempted. The event data is the error message, and the errno
2122 information encodes the lookup type (A vs. MX) used for the (first) host,
2123 and the trailing two digits of the smtp 4xx response.
2125 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2126 to write to mainlog (or rejectlog, paniclog) in the window between file
2127 creation and permissions/ownership being changed. Particularly affects
2128 installations where exicyclog is run as root, rather than exim user;
2129 result is that the running daemon panics and dies.
2131 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2133 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2134 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2135 "pri" and "weight". Note that the previous implicit priority given by the
2136 list order is no longer honoured.
2138 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2139 for DKIM processing.
2141 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2142 by defining SUPPORT_SOCKS.
2144 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2145 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2146 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2147 variables are renamed to proxy_{local,external}_{address,port}.
2149 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2150 for a build by defining SUPPORT_I18N
2152 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2153 of the query string, and make ${quote_redis:} do that quoting.
2155 JH/29 Move Events support from Experimental to mainline, enabled by default
2156 and removable for a build by defining DISABLE_EVENT.
2158 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2160 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2161 cached by the daemon.
2163 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2164 by defining LOOKUP_REDIS. The libhiredis library is required.
2166 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2167 keys are given for lookup.
2169 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2170 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2171 only supported when built with TLS support. The PolarSSL SHA routines
2172 are still used when the TLS library is too old for convenient support.
2174 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2175 openssl_options), for security. OpenSSL forces this from version 1.1.0
2176 server-side so match that on older versions.
2178 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2179 allocation for $value could be released as the expansion processing
2180 concluded, but leaving the global pointer active for it.
2182 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2183 and to use the domains and local_parts ACL conditions.
2185 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2186 incorrectly not doubled on cutthrough transmission, hence seen as a
2187 body-termination at the receiving system - resulting in truncated mails.
2188 Commonly the sender saw a TCP-level error, and retransmitted the message
2189 via the normal store-and-forward channel. This could result in duplicates
2190 received - but deduplicating mailstores were liable to retain only the
2191 initial truncated version.
2193 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2195 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2197 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2198 we're in there, support oversigning also; bug 1309.
2200 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2202 HS/04 Add support for keep_environment and add_environment options.
2204 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2205 either intentional arithmetic overflow during PRNG, or testing config-
2208 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2209 delivery resulted in actual delivery. Cancel cutthrough before DATA
2212 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2213 rejects a recipient: pass the reject to the originator.
2215 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2216 Many were false-positives and ignorable, but it's worth fixing the
2219 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2220 for the new environment-manipulation done at startup. Move the routines
2221 from being local to tls.c to being global via the os.c file.
2223 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2224 an extract embedded as result-arg for a map, the first arg for extract
2225 is unavailable so we cannot tell if this is a numbered or keyed
2226 extraction. Accept either.
2232 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2235 JH/02 The smtp transport option "multi_domain" is now expanded.
2237 JH/03 The smtp transport now requests PRDR by default, if the server offers
2240 JH/04 Certificate name checking on server certificates, when exim is a client,
2241 is now done by default. The transport option tls_verify_cert_hostnames
2242 can be used to disable this per-host. The build option
2243 EXPERIMENTAL_CERTNAMES is withdrawn.
2245 JH/05 The value of the tls_verify_certificates smtp transport and main options
2246 default to the word "system" to access the system default CA bundle.
2247 For GnuTLS, only version 3.0.20 or later.
2249 JH/06 Verification of the server certificate for a TLS connection is now tried
2250 (but not required) by default. The verification status is now logged by
2251 default, for both outbound TLS and client-certificate supplying inbound
2254 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2257 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2258 Status Notification (bounce) messages are now MIME format per RFC 3464.
2259 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2260 under the control of the dsn_advertise_hosts option, and routers may
2261 have a dsn_lasthop option.
2263 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2264 default, modifiable by a malware= option. The list separator for
2265 the options can now be changed in the usual way. Bug 68.
2267 JH/10 The smtp_receive_timeout main option is now expanded before use.
2269 JH/11 The incoming_interface log option now also enables logging of the
2270 local interface on delivery outgoing connections.
2272 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2273 if the interface and destination host and port all match.
2275 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2278 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2279 Patch from Andrew Lewis.
2281 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2282 now supports optional time-restrictions, weighting, and priority
2283 modifiers per server. Patch originally by <rommer@active.by>.
2285 JH/16 The spamd_address main option now supports a mixed list of local
2286 and remote servers. Remote servers can be IPv6 addresses, and
2287 specify a port-range.
2289 JH/17 Bug 68: The spamd_address main option now supports an optional
2290 timeout value per server.
2292 JH/18 Bug 1581: Router and transport options headers_add/remove can
2293 now have the list separator specified.
2295 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2298 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2301 JH/21 Support for the A6 type of dns record is withdrawn.
2303 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2304 rather than the verbs used.
2306 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2307 from 255 to 1024 chars.
2309 JH/24 Verification callouts now attempt to use TLS by default.
2311 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2312 are generic router options now. The defaults didn't change.
2314 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2315 Original patch from Alexander Shikoff, worked over by JH.
2317 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2318 files if ZCAT_COMMAND is not executable.
2320 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2322 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2324 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2325 Normally benign, it bites when the pair was led to by a CNAME;
2326 modern usage is to not canonicalize the domain to a CNAME target
2327 (and we were inconsistent anyway for A-only vs AAAA+A).
2329 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2331 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2332 when evaluating $sender_host_dnssec.
2334 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2335 $sender_helo_dnssec variable.
2337 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2339 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2341 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2343 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2344 documented as working, but never had. Support all but $spam_report.
2346 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2347 added for tls authenticator.
2349 HS/03 Add perl_taintmode main config option
2355 TL/01 When running the test suite, the README says that variables such as
2356 no_msglog_check are global and can be placed anywhere in a specific
2357 test's script, however it was observed that placement needed to be near
2358 the beginning for it to behave that way. Changed the runtest perl
2359 script to read through the entire script once to detect and set these
2360 variables, reset to the beginning of the script, and then run through
2361 the script parsing/test process like normal.
2363 TL/02 The BSD's have an arc4random API. One of the functions to induce
2364 adding randomness was arc4random_stir(), but it has been removed in
2365 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2366 function when detected.
2368 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2369 cause callback expansion.
2371 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2372 syntax errors in an expansion can be treated as a string instead of
2373 logging or causing an error, due to the internal use of bool_lax
2374 instead of bool when processing it.
2376 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2377 server certificates when making smtp deliveries.
2379 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2381 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2383 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2385 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2386 Merged patch from Sebastian Wiedenroth.
2388 JH/05 Fix results-pipe from transport process. Several recipients, combined
2389 with certificate use, exposed issues where response data items split
2390 over buffer boundaries were not parsed properly. This eventually
2391 resulted in duplicates being sent. This issue only became common enough
2392 to notice due to the introduction of connection certificate information,
2393 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2395 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2396 size buffer was used, resulting in syntax errors when an expansion
2399 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2400 version 3.3.6 or later.
2402 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2403 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2404 both become "event_action", the variables become $event_name, $event_data
2405 and $event_defer_errno. There is a new variable $verify_mode, usable in
2406 routers, transports and related events. The tls:cert event is now also
2407 raised for inbound connections, if the main configuration event_action
2410 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2411 early OCSP support, but no stapling (appears to be less than 1.0.0).
2413 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2414 server certificate names available under the smtp transport option
2415 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2418 JH/10 Time-related extraction expansions from certificates now use the main
2419 option "timezone" setting for output formatting, and are consistent
2420 between OpenSSL and GnuTLS compilations. Bug 1541.
2422 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2423 encoded parameter in the incoming message. Bug 1558.
2425 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2426 include certificate info, eximon was claiming there were spoolfile
2429 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2431 JH/14 Log delivery-related information more consistently, using the sequence
2432 "H=<name> [<ip>]" wherever possible.
2434 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2435 are problematic for Debian distribution, omit them from the release
2438 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2440 JH/16 Fix string representation of time values on 64bit time_t architectures.
2443 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2444 output list separator was used.
2449 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2450 checkers that were complaining about end of non-void function with no
2453 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2454 This was a regression introduced in 4.83 by another bugfix.
2456 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2458 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2459 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2465 TF/01 Correctly close the server side of TLS when forking for delivery.
2467 When a message was received over SMTP with TLS, Exim failed to clear up
2468 the incoming connection properly after forking off the child process to
2469 deliver the message. In some situations the subsequent outgoing
2470 delivery connection happened to have the same fd number as the incoming
2471 connection previously had. Exim would try to use TLS and fail, logging
2472 a "Bad file descriptor" error.
2474 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2475 utilities have not been installed.
2477 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2478 temporary space as the ACL may create new global variables.
2480 TL/01 LDAP support uses per connection or global context settings, depending
2481 upon the detected version of the libraries at build time.
2483 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2484 to extract and use the src ip:port in logging and expansions as if it
2485 were a direct connection from the outside internet. PPv2 support was
2486 updated based on HAProxy spec change in May 2014.
2488 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2490 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2491 Properly escape header and check for NULL return.
2493 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2496 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2498 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2499 characters in header names, implemented as a verify condition.
2500 Contributed by Michael Fischer v. Mollard.
2502 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2503 results permerror and temperror. Previous values are deprecated but
2504 still accepted. In a future release, err_perm and err_temp will be
2505 completely removed, which will be a backward incompatibility if the
2506 ACL tests for either of these two old results. Patch contributed by
2507 user bes-internal on the mailing list.
2509 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2511 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2512 selectors, in both main and reject logs.
2514 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2517 JH/07 Add malware type "sock" for talking to simple daemon.
2519 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2521 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2522 routers/transports under cutthrough routing.
2524 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2525 numbers. Touch up "bool" conditional to keep the same definition.
2527 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2529 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2531 JH/12 Expand items in router/transport headers_add or headers_remove lists
2532 individually rather than the list as a whole. Bug 1452.
2534 Required for reasonable handling of multiple headers_ options when
2535 they may be empty; requires that headers_remove items with embedded
2536 colons must have them doubled (or the list-separator changed).
2538 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2539 view the policy declared in the DMARC record. Currently, $dmarc_status
2540 is a combined value of both the record presence and the result of the
2543 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2545 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2546 dnslookup router and the smtp transport (applying to the forward
2549 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2550 of ldap servers used for a specific lookup. Patch provided by Heiko
2553 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2554 New variable $lookup_dnssec_authenticated for observability.
2556 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2557 Patch submitted by Lars Timman.
2559 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2561 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2562 Requires trusted mode and valid format message id, aborts otherwise.
2563 Patch contributed by Heiko Schlichting.
2565 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2566 certextract with support for various fields. Bug 1358.
2568 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2569 is requested by default, modifiable by smtp transport option
2572 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2573 operate on certificate variables to give certificate fingerprints
2574 Also new ${sha256:cert_variable}.
2576 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2578 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2579 Christian Aistleitner.
2581 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2583 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2584 file. Patch from Wolfgang Breyha.
2586 JH/25 Expand the coverage of the delivery $host and $host_address to
2587 client authenticators run in verify callout. Bug 1476.
2589 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2590 align with daemon_smtp_ports. Bug 72.
2592 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2593 support and error reporting did not work properly.
2595 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2596 and is readable. Patch from Andrew Colin Kissa.
2598 TL/14 Enhance documentation of ${run expansion and how it parses the
2599 commandline after expansion, particularly in the case when an
2600 unquoted variable expansion results in an empty value.
2602 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2604 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2605 by renaming away from it. Addresses GH issue 15, reported by
2608 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2610 TL/15 SECURITY: prevent double expansion in math comparison functions
2611 (can expand unsanitized data). Not remotely exploitable.
2618 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2620 PP/02 Make -n do something, by making it not do something.
2621 When combined with -bP, the name of an option is not output.
2623 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2626 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2627 $sender_host_name and config options to manage this, and basic check
2630 PP/05 DSCP support for outbound connections and control modifier for inbound.
2632 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2633 (Only plugin which currently uses this is kerberos4, which nobody should
2634 be using, but we should make it available and other future plugins might
2635 conceivably use it, even though it would break NAT; stuff *should* be
2636 using channel bindings instead).
2638 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2639 name; added for Sendmail compatibility; requires admin caller.
2640 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2641 just ignore it); requires trusted caller.
2642 Also parse but ignore: -Ac -Am -X<logfile>
2645 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2647 TL/02 Add +smtp_confirmation as a default logging option.
2649 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2650 Patch by Magnus Holmgren from 2007-02-20.
2652 TL/04 Bugzilla 1281 - Spec typo.
2653 Bugzilla 1283 - Spec typo.
2654 Bugzilla 1290 - Spec grammar fixes.
2656 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2658 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2660 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2661 mailing list by by Dmitry Isaikin.
2663 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2665 JH/02 Support "G" suffix to numbers in ${if comparisons.
2667 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2669 NM/01 Bugzilla 1197 - Spec typo
2670 Bugzilla 1196 - Spec examples corrections
2672 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2674 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2675 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2678 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2679 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2681 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2682 "acl {{name}{arg}...}", and optional args on acl condition
2685 JH/05 Permit multiple router/transport headers_add/remove lines.
2687 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2689 JH/07 Avoid using a waiting database for a single-message-only transport.
2690 Performance patch from Paul Fisher. Bugzilla 1262.
2692 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2695 JH/09 Add $headers_added variable, with content from use of ACL modifier
2696 add_header (but not yet added to the message). Bugzilla 199.
2698 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2699 Pulled from Bugzilla 817 by Wolfgang Breyha.
2701 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2703 (nb: this is the same fix as in Exim 4.80.1)
2705 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2708 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2711 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2713 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2714 advertise SMTP AUTH mechanism to us, instead of a generic
2715 protocol violation error. Also, make Exim more robust to bad
2716 data from the Dovecot auth socket.
2718 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2720 When a queue runner is handling a message, Exim first routes the
2721 recipient addresses, during which it prunes them based on the retry
2722 hints database. After that it attempts to deliver the message to
2723 any remaining recipients. It then updates the hints database using
2726 So if a recipient address works intermittently, it can get repeatedly
2727 deferred at routing time. The retry hints record remains fresh so the
2728 address never reaches the final cutoff time.
2730 This is a fairly common occurrence when a user is bumping up against
2731 their storage quota. Exim had some logic in its local delivery code
2732 to deal with this. However it did not apply to per-recipient defers
2733 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2735 This change adds a proper retry rule check during routing so that the
2736 final cutoff time is checked against the message's age. We only do
2737 this check if there is an address retry record and there is not a
2738 domain retry record; this implies that previous attempts to handle
2739 the address had the retry_use_local_parts option turned on. We use
2740 this as an approximation for the destination being like a local
2741 delivery, as in LMTP.
2743 I suspect this new check makes the old local delivery cutoff check
2744 redundant, but I have not verified this so I left the code in place.
2746 TF/02 Correct gecos expansion when From: is a prefix of the username.
2748 Test 0254 submits a message to Exim with the header
2752 When I ran the test suite under the user fanf2, Exim expanded
2753 the header to contain my full name, whereas it should have added
2754 a Resent-Sender: header. It erroneously treats any prefix of the
2755 username as equal to the username.
2757 This change corrects that bug.
2759 GF/01 DCC debug and logging tidyup
2760 Error conditions log to paniclog rather than rejectlog.
2761 Debug lines prefixed by "DCC: " to remove any ambiguity.
2763 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2765 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2766 Bug spotted by Jeremy Harris; was flawed since initial commit.
2767 Would have resulted in OCSP responses post-SNI triggering an Exim
2768 NULL dereference and crash.
2770 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2772 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2773 Bug detection, analysis and fix by Samuel Thibault.
2774 Bugzilla 1331, Debian bug #698092.
2776 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2778 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2779 Server implementation by Todd Lyons, client by JH.
2780 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2781 config variable "prdr_enable" controls whether the server
2782 advertises the facility. If the client requests PRDR a new
2783 acl_data_smtp_prdr ACL is called once for each recipient, after
2784 the body content is received and before the acl_smtp_data ACL.
2785 The client is controlled by both of: a hosts_try_prdr option
2786 on the smtp transport, and the server advertisement.
2787 Default client logging of deliveries and rejections involving
2788 PRDR are flagged with the string "PRDR".
2790 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2791 fclose(). Diagnosis by Todd Lyons.
2793 PP/17 Update configure.default to handle IPv6 localhost better.
2794 Patch by Alain Williams (plus minor tweaks).
2797 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2798 This is now consistent with GnuTLS, and is now documented: the
2799 previous undocumented portable approach to treating the option as
2800 unset was to force an expansion failure. That still works, and
2801 an empty string is now equivalent.
2803 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2804 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2805 not performing validation itself.
2807 PP/20 Added force_command boolean option to pipe transport.
2808 Patch from Nick Koston, of cPanel Inc.
2810 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2813 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2815 PP/21 Fix eximon continuous updating with timestamped log-files.
2816 Broken in a format-string cleanup in 4.80, missed when I repaired the
2817 other false fix of the same issue.
2818 Report and fix from Heiko Schlichting.
2821 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2822 Report from Prashanth Katuri.
2824 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2825 It's SecureTransport, so affects any MacOS clients which use the
2826 system-integrated TLS libraries, including email clients.
2828 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2829 using a MIME ACL for non-SMTP local injection.
2830 Report and assistance in diagnosis by Warren Baker.
2832 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2834 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2836 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2837 last id that failed so it may be referenced in subsequent ACL's.
2839 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2842 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2843 ldap library initialization, allowing self-signed CA's to be
2844 used. Also properly sets require_cert option later in code by
2845 using NULL (global ldap config) instead of ldap handle (per
2846 session). Bug diagnosis and testing by alxgomz.
2848 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2849 the src/util/ subdirectory.
2851 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2852 renamed to Transport Post Delivery Action by Jeremy Harris, as
2855 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2856 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2857 redis_servers = needs to be configured which will be used by the redis
2858 lookup. Patch from Warren Baker, of The Packet Hub.
2860 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2862 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2863 hostname or reverse DNS when processing a host list. Used suggestions
2864 from multiple comments on this bug.
2866 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2868 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2869 Missed a few lines, added it to make the runtest require no keyboard
2872 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2873 contains upper case chars. Make router use caseful_local_part.
2875 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2876 support when GnuTLS has been built with p11-kit.
2882 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2884 This, or similar/improved, will also be change PP/11 of 4.82.
2890 PP/01 Handle short writes when writing local log-files.
2891 In practice, only affects FreeBSD (8 onwards).
2892 Bugzilla 1053, with thanks to Dmitry Isaikin.
2894 NM/01 Bugzilla 949 - Documentation tweak
2896 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2899 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2901 PP/02 Implemented gsasl authenticator.
2903 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2905 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2906 `pkg-config foo` for cflags/libs.
2908 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2909 with rest of GSASL and with heimdal_gssapi.
2911 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2912 `pkg-config foo` for cflags/libs for the TLS implementation.
2914 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2915 properties get this fed in as external SSF. A number of robustness
2916 and debugging improvements to the cyrus_sasl authenticator.
2918 PP/08 cyrus_sasl server now expands the server_realm option.
2920 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2921 Patch by Jeremy Harris.
2923 PP/10 Added dbmjz lookup type.
2925 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2927 PP/12 MAIL args handles TAB as well as SP, for better interop with
2928 non-compliant senders.
2929 Analysis and variant patch by Todd Lyons.
2931 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2932 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2933 Patch from Dirk Mueller <dmueller@suse.com>
2935 PP/13 tls_peerdn now print-escaped for spool files.
2936 Observed some $tls_peerdn in wild which contained \n, which resulted
2937 in spool file corruption.
2939 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2940 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2941 or write after TLS renegotiation, which otherwise led to messages
2944 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2945 as a tracking header (ie: a signed header comes before the signature).
2946 Patch from Wolfgang Breyha.
2948 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2949 comma-sep list; embedded commas doubled.
2951 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2953 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2955 Report and patch from Dmitry Banschikov.
2957 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2958 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2959 failures. We appear to now support TLS1.1+ with Exim.
2961 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2962 lets Exim select keys and certificates based upon TLS SNI from client.
2963 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2964 before an outbound SMTP session. New log_selector, +tls_sni.
2966 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2967 NULL dereference. Report and patch from Alun Jones.
2969 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2970 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2971 Not seeing resolver debug output on NetBSD, but suspect this is a
2972 resolver implementation change.
2974 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2975 Left warnings. Added "eximon gdb" invocation mode.
2977 PP/21 Defaulting "accept_8bitmime" to true, not false.
2979 PP/22 Added -bw for inetd wait mode support.
2981 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2982 locate the relevant includes and libraries. Made this the default.
2984 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2985 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2987 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2988 This may cause build issues on older platforms.
2990 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2991 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2992 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2993 Added SNI support via GnuTLS too.
2994 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2996 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2998 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2999 Applied second patch from Janne, implementing suggestion to default
3000 multiple-strings-in-record handling to match SPF spec.
3002 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3004 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3005 read-only, out of scope).
3006 Patch from Wolfgang Breyha, report from Stuart Northfield.
3008 PP/29 Fix three issues highlighted by clang analyser static analysis.
3009 Only crash-plausible issue would require the Cambridge-specific
3010 iplookup router and a misconfiguration.
3011 Report from Marcin Mirosław.
3013 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3015 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3016 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3017 As part of this, removing so much warning spew let me fix some minor
3018 real issues in debug logging.
3020 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3021 assignment on my part. Fixed.
3023 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3024 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3025 Janne Snabb (who went above and beyond: thank you).
3027 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3028 string otherwise requires a connection and a bunch more work and it's
3029 relatively easy to get wrong. Should also expose TLS library linkage
3032 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3033 64-bit ${eval} (JH/03).
3035 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3036 GNU libc to support some of the 64-bit stuff, should not lead to
3037 conflicts. Defined before os.h is pulled in, so if a given platform
3038 needs to override this, it can.
3040 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3041 protection layer was required, which is not implemented.
3042 Bugzilla 1254, patch from Wolfgang Breyha.
3044 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3045 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3046 tls_dhparam take prime identifiers. Also unbreak combination of
3047 OpenSSL+DH_params+TLSSNI.
3049 PP/39 Disable SSLv2 by default in OpenSSL support.
3055 PP/01 Solaris build fix for Oracle's LDAP libraries.
3056 Bugzilla 1109, patch from Stephen Usher.
3058 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3060 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3063 TF/02 Fix a couple more cases where we did not log the error message
3064 when unlink() failed. See also change 4.74-TF/03.
3066 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3067 lock up or crash if it happened to be inside a call to libc when it
3068 got a SIGUSR1 from exiwhat.
3070 The SIGUSR1 handler appends the current process status to the process
3071 log which is later printed by exiwhat. It used to use the general
3072 purpose logging code to do this, but several functions it calls are
3073 not safe for signals.
3075 The new output code in the SIGUSR1 handler is specific to the process
3076 log, and simple enough that it's easy to inspect for signal safety.
3077 Removing some special cases also simplifies the general logging code.
3078 Removing the spurious timestamps from the process log simplifies
3081 TF/04 Improved ratelimit ACL condition.
3083 The /noupdate option has been deprecated in favour of /readonly which
3084 has clearer semantics. The /leaky, /strict, and /readonly update modes
3085 are mutually exclusive. The update mode is no longer included in the
3086 database key; it just determines when the database is updated. (This
3087 means that when you upgrade Exim will forget old rate measurements.)
3089 Exim now checks that the per_* options are used with an update mode that
3090 makes sense for the current ACL. For example, when Exim is processing a
3091 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3092 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3093 must specify per_mail/readonly. If you omit the update mode it defaults to
3094 /leaky where that makes sense (as before) or /readonly where required.
3096 The /noupdate option is now undocumented but still supported for
3097 backwards compatibility. It is equivalent to /readonly except that in
3098 ACLs where /readonly is required you may specify /leaky/noupdate or
3099 /strict/noupdate which are treated the same as /readonly.
3101 A useful new feature is the /count= option. This is a generalization
3102 of the per_byte option, so that you can measure the throughput of other
3103 aggregate values. For example, the per_byte option is now equivalent
3104 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3106 The per_rcpt option has been generalized using the /count= mechanism
3107 (though it's more complicated than the per_byte equivalence). When it is
3108 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3109 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3110 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3111 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3112 Note that using per_rcpt with a non-readonly update mode in more than
3113 one ACL will cause the recipients to be double-counted. (The per_mail
3114 and per_byte options don't have this problem.)
3116 The handling of very low rates has changed slightly. If the computed rate
3117 is less than the event's count (usually one) then this event is the first
3118 after a long gap. In this case the rate is set to the same as this event's
3119 count, so that the first message of a spam run is counted properly.
3121 The major new feature is a mechanism for counting the rate of unique
3122 events. The new per_addr option counts the number of different
3123 recipients that someone has sent messages to in the last time period. It
3124 behaves like per_rcpt if all the recipient addresses are different, but
3125 duplicate recipient addresses do not increase the measured rate. Like
3126 the /count= option this is a general mechanism, so the per_addr option
3127 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3128 example, measure the rate that a client uses different sender addresses
3129 with the options per_mail/unique=$sender_address. There are further
3130 details in the main documentation.
3132 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3134 TF/06 Removed a few PCRE remnants.
3136 TF/07 Automatically extract Exim's version number from tags in the git
3137 repository when doing development or release builds.
3139 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3140 Bugzilla 879. Patch from Paul Fisher.
3142 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3143 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3146 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3148 PP/05 Variable $av_failed, true if the AV scanner deferred.
3149 Bugzilla 1078. Patch from John Horne.
3151 PP/06 Stop make process more reliably on build failure.
3152 Bugzilla 1087. Patch from Heiko Schlittermann.
3154 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3155 Bugzilla 1089. Patch from Heiko Schlittermann.
3157 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3158 Bugzilla 1131. Patch from Holger Weiß.
3160 PP/09 Handle IPv6 addresses with SPF.
3161 Bugzilla 860. Patch from Wolfgang Breyha.
3163 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3165 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3168 PP/11 match_* no longer expand right-hand-side by default.
3169 New compile-time build option, EXPAND_LISTMATCH_RHS.
3170 New expansion conditions, "inlist", "inlisti".
3172 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3174 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3176 PP/14 fix log_write() format string regression from TF/03.
3177 Bugzilla 1152. Patch from Dmitry Isaikin.
3183 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3185 PP/02 Harmonised TLS library version reporting; only show if debugging.
3186 Layout now matches that introduced for other libraries in 4.74 PP/03.
3188 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3190 PP/04 New "dns_use_edns0" global option.
3192 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3195 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3196 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3198 TK/01 Updated PolarSSL code to 0.14.2.
3199 Bugzilla 1097. Patch from Andreas Metzler.
3201 PP/07 Catch divide-by-zero in ${eval:...}.
3202 Fixes bugzilla 1102.
3204 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3207 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3208 format-string attack -- SECURITY: remote arbitrary code execution.
3210 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3211 time unintentionally subject to list matching rules, letting the header
3212 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3213 arbitrary string expansion). This allowed for information disclosure.
3215 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3216 INT_MIN/-1 -- value coerced to INT_MAX.
3222 NM/01 Workaround for PCRE version dependency in version reporting
3225 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3226 This fixes portability to compilers other than gcc, notably
3227 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3229 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3230 makefiles for portability to HP-UX and POSIX correctness.
3232 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3233 Also via indented variable definition in the Makefile.
3234 (Debugging by Oliver Heesakkers).
3236 PP/02 Restore caching of spamd results with expanded spamd_address.
3237 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3239 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3240 Improves build reliability. Fix from: Frank Elsner
3242 NM/02 Fix wide character breakage in the rfc2047 coding
3243 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3245 NM/03 Allow underscore in dnslist lookups
3246 Fixes bug 1026. Patch from Graeme Fowler
3248 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3249 Code patches from Adam Ciarcinski of NetBSD.
3251 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3254 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3255 is logged, to avoid truncation. Patch from John Horne.
3257 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3258 Patch from Jakob Hirsch.
3260 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3261 SQL string expansion failure details.
3262 Patch from Andrey Oktyabrski.
3264 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3265 Patch from Simon Arlott.
3267 PP/09 New lookups functionality failed to compile on old gcc which rejects
3268 extern declarations in function scope.
3269 Patch from Oliver Fleischmann
3271 PP/10 Use sig_atomic_t for flags set from signal handlers.
3272 Check getgroups() return and improve debugging.
3273 Fixed developed for diagnosis in bug 927 (which turned out to be
3276 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3277 Patch from Mark Zealey.
3279 PP/12 Bugzilla 1056: Improved spamd server selection.
3280 Patch from Mark Zealey.
3282 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3283 Based on patch from Heiko Schlittermann.
3285 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3286 Patch from Uwe Doering, sign-off by Michael Haardt.
3288 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3289 variable declaration deep within a block. Bug and patch from
3292 PP/15 lookups-Makefile IRIX compatibility coercion.
3294 PP/16 Make DISABLE_DKIM build knob functional.
3296 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3297 Patch by Simon Arlott
3299 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3300 variable argument macros. Our copy now differs from upstream.
3306 TF/01 Failure to get a lock on a hints database can have serious
3307 consequences so log it to the panic log.
3309 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3310 controlled using the smtp_confirmation log selector.
3312 TF/03 Include the error message when we fail to unlink a spool file.
3314 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3315 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3316 for maintaining out-of-tree patches for some time.
3318 PP/01 Bugzilla 139: Documentation and portability issues.
3319 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3320 Handle per-OS dynamic-module compilation flags.
3322 PP/02 Let /dev/null have normal permissions.
3323 The 4.73 fixes were a little too stringent and complained about the
3324 permissions on /dev/null. Exempt it from some checks.
3325 Reported by Andreas M. Kirchwitz.
3327 PP/03 Report version information for many libraries, including
3328 Exim version information for dynamically loaded libraries. Created
3329 version.h, now support a version extension string for distributors
3330 who patch heavily. Dynamic module ABI change.
3332 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3333 privilege escalation vulnerability whereby the Exim run-time user
3334 can cause root to append content of the attacker's choosing to
3337 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3340 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3341 If dropping privileges for untrusted macros, we disabled normal logging
3342 on the basis that it would fail; for the Exim run-time user, this is not
3343 the case, and it resulted in successful deliveries going unlogged.
3344 Fixed. Reported by Andreas Metzler.
3350 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3351 only prepend for the Resent-* case. Fixes regression introduced in
3352 Exim 4.70 by NM/22 for Bugzilla 607.
3354 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3355 increasing numbers of administrators be bitten by this.
3357 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3359 PP/03 Bugzilla 994: added openssl_options main configuration option.
3361 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3363 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3365 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3367 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3368 without a peer certificate, leading to a segfault because of an
3369 assumption that peers always have certificates. Be a little more
3370 paranoid. Problem reported by Martin Tscholak.
3372 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3373 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3374 NB: ClamAV planning to remove STREAM in "middle of 2010".
3375 CL also introduces -bmalware, various -d+acl logging additions and
3376 more caution in buffer sizes.
3378 PP/09 Implemented reverse_ip expansion operator.
3380 PP/10 Bugzilla 937: provide a "debug" ACL control.
3382 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3384 PP/12 Bugzilla 973: Implement --version.
3386 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3388 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3390 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3392 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3393 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3394 ignore trailing whitespace.
3396 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3398 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3401 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3402 Notification from Dr Andrew Aitchison.
3404 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3405 ExtendedDetectionInfo response format.
3406 Notification from John Horne.
3408 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3411 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3412 XSL and documented dependency on system catalogs, with examples of how
3415 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3418 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3419 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3420 configuration file which is writeable by the Exim user or group.
3422 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3423 of configuration files to cover files specified with the -C option if
3424 they are going to be used with root privileges, not just the default
3427 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3428 option (effectively making it always true).
3430 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3431 files to be used while preserving root privileges.
3433 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3434 that rogue child processes cannot use them.
3436 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3437 run-time user, instead of root.
3439 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3440 Exim run-time user without dropping privileges.
3442 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3443 result string, instead of calling string_vformat() twice with the same
3446 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3447 for other users. Others should always drop root privileges if they use
3448 -C on the command line, even for a whitelisted configure file.
3450 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3452 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3458 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3459 $data_path, and $header_path variables; fixed documentation bugs and
3462 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3463 exipick to access non-standard spools, including the "frozen" queue
3466 NM/01 Bugzilla 965: Support mysql stored procedures.
3467 Patch from Alain Williams
3469 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3471 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3472 Patch from Andreas Metzler
3474 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3475 Patch from Kirill Miazine
3477 NM/05 Bugzilla 671: Added umask to procmail example.
3479 JJ/03 installed exipick 20100323.0, fixing doc bug
3481 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3482 directory. Notification and patch from Dan Rosenberg.
3484 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3486 TK/02 Improve log output when DKIM signing operation fails.
3488 MH/01 Treat the transport option dkim_domain as a colon separated
3489 list, not as a single string, and sign the message with each element,
3490 omitting multiple occurences of the same signer.
3492 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3493 Bugzilla 985, 986. Patch by Simon Arlott
3495 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3496 Patch by Simon Arlott
3498 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3499 MBX locking. Notification from Dan Rosenberg.
3505 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3507 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3509 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3511 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3513 NM/04 Bugzilla 915: Flag broken perl installation during build.
3519 TK/01 Added patch by Johannes Berg that expands the main option
3520 "spamd_address" if it starts with a dollar sign.
3522 TK/02 Write list of recipients to X-Envelope-Sender header when building
3523 the mbox-format spool file for content scanning (suggested by Jakob
3526 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3527 (http://www.dcc-servers.net/) support via dccifd. Activated by
3528 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3530 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3531 by Mark Daniel Reidel <mr@df.eu>.
3533 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3534 When building exim an external PCRE library is now needed -
3535 PCRE is a system library on the majority of modern systems.
3536 See entry on PCRE_LIBS in EDITME file.
3538 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3539 conversation. Added nologin parameter to request.
3540 Patch contributed by Kirill Miazine.
3542 TF/01 Do not log submission mode rewrites if they do not change the address.
3544 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3546 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3547 log files in place. Contributed by Roberto Lima.
3549 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3551 TF/03 Bugzilla 615: When checking the local_parts router precondition
3552 after a local_part_suffix or local_part_prefix option, Exim now
3553 does not use the address's named list lookup cache, since this
3554 contains cached lookups for the whole local part.
3556 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3557 Robert Millan. Documentation is in experimental-spec.txt.
3559 TF/04 Bugzilla 668: Fix parallel build (make -j).
3561 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3563 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3564 Patch provided by Jan Srzednicki.
3566 TF/05 Leading white space used to be stripped from $spam_report which
3567 wrecked the formatting. Now it is preserved.
3569 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3570 that they are available at delivery time.
3572 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3574 TF/08 TLS error reporting now respects the incoming_interface and
3575 incoming_port log selectors.
3577 TF/09 Produce a more useful error message if an SMTP transport's hosts
3578 setting expands to an empty string.
3580 NM/06 Bugzilla 744: EXPN did not work under TLS.
3581 Patch provided by Phil Pennock.
3583 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3584 Patch provided by Richard Godbee.
3586 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3587 acl_smtp_notquit, added index entry.
3589 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3590 Patch provided by Eugene Bujak.
3592 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3593 accept(). Patch provided by Maxim Dounin.
3595 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3596 Patch provided by Phil Pennock.
3598 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3600 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3601 Patch provided by Brad "anomie" Jorsch.
3603 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3604 Patch provided by Dean Brooks.
3606 TK/05 Add native DKIM support (does not depend on external libraries).
3608 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3609 Patch provided by Graeme Fowler.
3611 NM/16 Bugzilla 851: Documentation example syntax fix.
3613 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3615 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3618 NM/19 Bugzilla 745: TLS version reporting.
3619 Patch provided by Phil Pennock.
3621 NM/20 Bugzilla 167: bool: condition support.
3622 Patch provided by Phil Pennock.
3624 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3625 clients. Patch provided by Phil Pennock.
3627 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3628 Patch provided by Brad "anomie" Jorsch.
3630 NM/23 Bugzilla 687: Fix misparses in eximstats.
3631 Patch provided by Heiko Schlittermann.
3633 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3634 Patch provided by Heiko Schlittermann.
3636 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3637 plus update to original patch.
3639 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3641 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3642 Patch provided by David Brownlee.
3644 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3646 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3648 NM/30 Bugzilla 888: TLS documentation bugfixes.
3650 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3652 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3653 Unlike the original bugzilla I have changed all shell scripts in src tree.
3655 NM/33 Bugzilla 898: Transport filter timeout fix.
3656 Patch by Todd Rinaldo.
3658 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3659 Patch by Serge Demonchaux.
3661 NM/35 Bugzilla 39: Base64 decode bug fixes.
3662 Patch by Jakob Hirsch.
3664 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3666 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3668 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3670 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3676 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3677 ALT-N's libdkim that I have put here:
3678 http://duncanthrax.net/exim-experimental/
3680 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3681 were called 'true' and it seems that C99 defines that as a reserved
3682 keyword to be used with 'bool' variable types. That means you could
3683 not include C99-style headers which use bools without triggering
3684 build errors in sieve.c.
3686 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3687 as mailq or other aliases. Changed the --help handling significantly
3688 to do whats expected. exim_usage() emits usage/help information.
3690 SC/01 Added the -bylocaldomain option to eximstats.
3692 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3694 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3696 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3702 PH/01 Another patch from the Sieve maintainer.
3704 PH/02 When an IPv6 address is converted to a string for single-key lookup
3705 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3706 dots are used instead of colons so that keys in lsearch files need not
3707 contain colons. This was done some time before quoting was made available
3708 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3709 (notated using the quote facility) so as to distinguish them from IPv4
3710 keys. This meant that lookups for IP addresses in host lists did not work
3711 for iplsearch lookups.
3713 This has been fixed by arranging for IPv6 addresses to be expressed with
3714 colons if the lookup type is iplsearch. This is not incompatible, because
3715 previously such lookups could never work.
3717 The situation is now rather anomalous, since one *can* have colons in
3718 ordinary lsearch keys. However, making the change in all cases is
3719 incompatible and would probably break a number of configurations.
3721 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3724 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3725 conversion specification without a maximum field width, thereby enabling
3726 a rogue spamd server to cause a buffer overflow. While nobody in their
3727 right mind would setup Exim to query an untrusted spamd server, an
3728 attacker that gains access to a server running spamd could potentially
3729 exploit this vulnerability to run arbitrary code as the Exim user.
3731 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3732 $primary_hostname instead of what libspf2 thinks the hosts name is.
3734 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3735 a directory entry by the name of the lookup key. Previously, if a
3736 symlink pointed to a non-existing file or a file in a directory that
3737 Exim lacked permissions to read, a lookup for a key matching that
3738 symlink would fail. Now it is enough that a matching directory entry
3739 exists, symlink or not. (Bugzilla 503.)
3741 PH/03 The body_linecount and body_zerocount variables are now exported in the
3744 PH/04 Added the $dnslist_matched variable.
3746 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3747 This means they are set thereafter only if the connection becomes
3750 PH/06 Added the client_condition to authenticators so that some can be skipped
3751 by clients under certain conditions.
3753 PH/07 The error message for a badly-placed control=no_multiline_responses left
3754 "_responses" off the end of the name.
3756 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3758 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3759 (without spaces) instead of just copying the configuration text.
3761 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3763 PH/11 Added $max_received_linelength.
3765 PH/12 Added +ignore_defer and +include_defer to host lists.
3767 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3768 way in which PCRE > 7.0 is built.
3770 PH/14 Implemented queue_only_load_latch.
3772 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3773 MAIL command. The effect was to mangle the value on 64-bit systems.
3775 PH/16 Another patch from the Sieve maintainer.
3777 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3779 PH/18 If a system quota error occurred while trying to create the file for
3780 a maildir delivery, the message "Mailbox is full" was not appended to the
3781 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3782 only to a quota excession during the actual writing of the file.
3784 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3785 characters?) which causes problems in log lines. The DN values are now
3786 passed through string_printing() before being added to log lines.
3788 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3789 and InterBase are left for another time.)
3791 PH/21 Added message_body_newlines option.
3793 PH/22 Guard against possible overflow in moan_check_errorcopy().
3795 PH/23 POSIX allows open() to be a macro; guard against that.
3797 PH/24 If the recipient of an error message contained an @ in the local part
3798 (suitably quoted, of course), incorrect values were put in $domain and
3799 $local_part during the evaluation of errors_copy.
3805 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3806 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3809 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3810 issue a MAIL command.
3812 PH/02 In an ACL statement such as
3814 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3816 if a client was not listed at all, or was listed with a value other than
3817 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3818 the condition was not true (as it should be), so access was not denied.
3819 The bug was that the ! inversion was incorrectly passed on to the second
3820 item. This has been fixed.
3822 PH/03 Added additional dnslists conditions == and =& which are different from
3823 = and & when the dns lookup returns more than one IP address.
3825 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3826 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3828 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3829 FSYNC, which compiles an option called disable_fsync that allows for
3830 bypassing fsync(). The documentation is heavily laced with warnings.
3832 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3834 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3835 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3836 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3837 including adding "make clean"; (3) Added -fPIC when compiling the test
3838 dynamically loaded module, to get rid of a warning.
3840 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3841 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3842 = 0s. The bug is otherwise harmless.
3844 PH/07 There was a bug in the dovecot authenticator such that the value of
3845 $auth1 could be overwritten, and so not correctly preserved, after a
3846 successful authentication. This usually meant that the value preserved by
3847 the server_setid option was incorrect.
3849 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3851 PH/09 Installed PCRE release 7.0.
3853 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3854 run for batched SMTP input. It is now run at the start of every message
3855 in the batch. While fixing this I discovered that the process information
3856 (output by running exiwhat) was not always getting set for -bs and -bS
3857 input. This is fixed, and it now also says "batched" for BSMTP.
3859 PH/11 Added control=no_pipelining.
3861 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3862 patch, slightly modified), and move the expansion of helo_data till after
3863 the connection is made in the smtp transport (so it can use these
3866 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3868 PH/14 Added log_selector = +pid.
3870 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3872 PH/16 Add ${if forany and ${if forall.
3874 PH/17 Added dsn_from option to vary the From: line in DSNs.
3876 PH/18 Flush SMTP output before performing a callout, unless control =
3877 no_callout_flush is set.
3879 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3880 was true (the default) a successful delivery failed to delete the retry
3881 item, thus causing premature timeout of the address. The bug is now
3884 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3886 PH/21 Long custom messages for fakedefer and fakereject are now split up
3887 into multiline responses in the same way that messages for "deny" and
3888 other ACL rejections are.
3890 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3891 with slight modification.
3893 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3894 draft, changing the syntax and factoring some duplicate code".
3896 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3897 for deliveries of the second and subsequent messages over the same SMTP
3900 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3901 ${reduce, with only minor "tidies".
3903 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3905 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3906 expansion side effects.
3908 PH/27 When a message times out after an over-quota error from an Exim-imposed
3909 quota, the bounce message says "mailbox is full". This message was not
3910 being given when it was a system quota that was exceeded. It now should
3913 MH/03 Made $recipients available in local_scan(). local_scan() already has
3914 better access to the recipient list through recipients_list[], but
3915 $recipients can be useful in postmaster-provided expansion strings.
3917 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3918 in the case of a MAIL command with additional options following the
3919 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3920 were accidentally chopped off.
3922 PH/29 SMTP synchronization checks are implemented when a command is read -
3923 there is a check that no more input is waiting when there shouldn't be
3924 any. However, for some commands, a delay in an ACL can mean that it is
3925 some time before the response is written. In this time, more input might
3926 arrive, invalidly. So now there are extra checks after an ACL has run for
3927 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3928 pipelining has not been advertised.
3930 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3932 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3933 returned in the SMTP response when smtp_return_error_details was false.
3934 This has been fixed.
3936 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3937 instead of the C I/O that was originally supplied, because problems were
3938 reported on Solaris.
3940 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3941 Exim which did not show up earlier: it was assuming that a call to
3942 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3943 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3944 was a macro that became an assignment, so it seemed to work. This has
3945 changed to a proper function call with a void return, hence the compile
3946 error. Exim's code has been fixed.
3948 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3951 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3953 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3955 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3956 because it turns out to be too big or there is a timeout) there is no
3957 "Completed" line in the log. When some messages of this type were
3958 selected by exigrep, they were listed as "not completed". Others were
3959 picked up by some special patterns. I have improved the selection
3960 criteria to be more general.
3962 PH/38 The host_find_failed option in the manualroute router can now be set
3963 to "ignore", to completely ignore a host whose IP address cannot be
3964 found. If all hosts are ignored, the behaviour is controlled by the new
3965 host_all_ignored option.
3967 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3968 homing or because of multiple MX records with /mx) generated more than
3969 one IP address, and the following item turned out to be the local host,
3970 all the secondary addresses of the first item were incorrectly removed
3971 from the list, along with the local host and any following hosts (which
3972 is what is supposed to happen).
3974 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3975 whoever called Exim into the -H file. In the case of the daemon it was
3976 behaving confusingly. When first started, it used values for whoever
3977 started the daemon, but after a SIGHUP it used the Exim user (because it
3978 calls itself on a restart). I have changed the code so that it now always
3981 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3982 message are rejected with the same error (e.g. no authentication or bad
3983 sender address), and a DATA command is nevertheless sent (as can happen
3984 with PIPELINING or a stupid MUA), the error message that was given to the
3985 RCPT commands is included in the rejection of the DATA command. This is
3986 intended to be helpful for MUAs that show only the final error to their
3989 PH/42 Another patch from the Sieve maintainer.
3991 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3992 Eximstats - Fixed some broken HTML links and added missing column headers
3994 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3995 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3997 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3999 PH/43 Yet another patch from the Sieve maintainer.
4001 PH/44 I found a way to check for a TCP/IP connection going away before sending
4002 the response to the final '.' that terminates a message, but only in the
4003 case where the client has not sent further data following the '.'
4004 (unfortunately, this is allowed). However, in many cases there won't be
4005 any further data because there won't be any more messages to send. A call
4006 to select() can be used: if it shows that the input is "ready", there is
4007 either input waiting, or the socket has been closed. An attempt to read
4008 the next input character can distinguish the two cases. Previously, Exim
4009 would have sent an OK response which the client would never have see.
4010 This could lead to message repetition. This fix should cure that, at
4011 least in a lot of common cases.
4013 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4014 advertised in response to EHLO.
4020 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4021 fixed by 4.65/MH/01 (is this a record?) are fixed:
4023 (i) An empty string was always treated as zero by the numeric comparison
4024 operators. This behaviour has been restored.
4026 (ii) It is documented that the numeric comparison operators always treat
4027 their arguments as decimal numbers. This was broken in that numbers
4028 starting with 0 were being interpreted as octal.
4030 While fixing these problems I realized that there was another issue that
4031 hadn't been noticed. Values of message_size_limit (both the global option
4032 and the transport option) were treated as octal if they started with 0.
4033 The documentation was vague. These values are now always treated as
4034 decimal, and I will make that clear in the documentation.
4040 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4041 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4044 MH/01 Don't check that the operands of numeric comparison operators are
4045 integers when their expansion is in "skipping" mode (fixes bug
4046 introduced by 4.64-PH/07).
4048 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4049 child addresses, Exim now panics and dies. Previously, because the count
4050 is held in a short int, deliveries were likely to be lost. As such a
4051 large number of recipients for a single message is ridiculous
4052 (performance will be very, very poor), I have chosen to impose a limit
4053 rather than extend the field.
4059 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4060 leftover -K file (the existence of which was triggered by #402).
4061 While we were at it, introduced process PID as part of the -K
4062 filename. This should rule out race conditions when creating
4065 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4066 processing considerably. Previous code took too long for large mails,
4067 triggering a timeout which in turn triggers #401.
4069 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4070 in the DK code in transports.c. sendfile() is not really portable,
4071 hence the _LINUX specificness.
4073 TF/01 In the add_headers option to the mail command in an Exim filter,
4074 there was a bug that Exim would claim a syntax error in any
4075 header after the first one which had an odd number of characters
4078 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4079 callout verification, Exim cached a "reject" for the entire domain. This
4080 is correct for most verifications, but it is not correct for a recipient
4081 verification with use_sender or use_postmaster set, because in that case
4082 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4083 case of MAIL FROM:<> rejection from other early rejections (e.g.
4084 rejection of HELO). When verifying a recipient using a non-null MAIL
4085 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4086 Whatever the result of the callout, the value of the domain cache is
4087 left unchanged (for any other kind of callout, getting as far as trying
4088 RCPT means that the domain itself is ok).
4090 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4093 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4094 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4095 session, the daemon ignores ECONNECT errors and logs others; it now
4096 ignores EPIPE as well.
4098 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4099 (quoted-printable decoding).
4101 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4102 later the small subsequent patch to fix an introduced bug.
4104 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4106 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4108 PH/08 An error is now given if message_size_limit is specified negative.
4110 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4111 to be given (somewhat) arbitrary names.
4113 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4116 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4117 miscellaneous code fixes
4119 PH/10 Added the log_reject_target ACL modifier to specify where to log
4122 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4123 hostname. This is wrong, because it relates to the incoming message (and
4124 probably the interface on which it is arriving) and not to the outgoing
4125 callout (which could be using a different interface). This has been
4126 changed to use the value of the helo_data option from the smtp transport
4127 instead - this is what is used when a message is actually being sent. If
4128 there is no remote transport (possible with a router that sets up host
4129 addresses), $smtp_active_hostname is used.
4131 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4132 tweaks were necessary in order to get it to work (see also 21 below):
4133 (a) The code assumed that strncpy() returns a negative number on buffer
4134 overflow, which isn't the case. Replaced with Exim's string_format()
4136 (b) There were several signed/unsigned issues. I just did the minimum
4137 hacking in of casts. There is scope for a larger refactoring.
4138 (c) The code used strcasecmp() which is not a standard C function.
4139 Replaced with Exim's strcmpic() function.
4140 (d) The code set only $1; it now sets $auth1 as well.
4141 (e) A simple test gave the error "authentication client didn't specify
4142 service in request". It would seem that Dovecot has changed its
4143 interface. Fortunately there's a specification; I followed it and
4144 changed what the client sends and it appears to be working now.
4146 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4149 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4150 address A is aliased to B and C, where B exists and C does not. Without
4151 -v the output is "A verified" because verification stops after a
4152 successful redirection if more than one address is generated. However,
4153 with -v the child addresses are also verified. Exim was outputting "A
4154 failed to verify" and then showing the successful verification for C,
4155 with its parentage. It now outputs "B failed to verify", showing B's
4156 parentage before showing the successful verification of C.
4158 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4159 look up a TXT record in a specific list after matching in a combined
4162 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4163 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4164 they consult the DNS. I had assumed they would set it the way they
4165 wanted; and indeed my experiments on Linux seem to show that in some
4166 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4167 To be on the safe side, however, I have now made the interface to
4168 host_find_byname() similar to host_find_bydns(), with an argument
4169 containing the DNS resolver options. The host_find_byname() function now
4170 sets these options at its start, just as host_find_bydns() does. The smtp
4171 transport options dns_qualify_single and dns_search_parents are passed to
4172 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4173 of host_find_byname() use the default settings of RES_DEFNAMES
4174 (qualify_single) but not RES_DNSRCH (search_parents).
4176 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4177 spool_read_header() do less string testing, by means of a preliminary
4178 switch on the second character of optional "-foo" lines. (This is
4179 overdue, caused by the large number of possibilities that now exist.
4180 Originally there were few.) While I was there, I also converted the
4181 str(n)cmp tests so they don't re-test the leading "-" and the first
4182 character, in the hope this might squeeze out yet more improvement.
4184 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4185 flag allowing group syntax was set by the header_syntax check but not
4186 turned off, possible causing trouble later; (2) The flag was not being
4187 set at all for the header_verify test, causing "group"-style headers to
4188 be rejected. I have now set it in this case, and also caused header_
4189 verify to ignore an empty address taken from a group. While doing this, I
4190 came across some other cases where the code for allowing group syntax
4191 while scanning a header line wasn't quite right (mostly, not resetting
4192 the flag correctly in the right place). These bugs could have caused
4193 trouble for malformed header lines. I hope it is now all correct.
4195 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4196 with the "reply" argument non-NULL. The code, however (which originally
4197 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4198 but it didn't always do it. This confused somebody who was copying the
4199 code for some other use. I have removed all the tests.
4201 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4202 feature that was used to support insecure browsers during the U.S. crypto
4203 embargo. It requires special client support, and Exim is probably the
4204 only MTA that supported it -- and would never use it because real RSA is
4205 always available. This code has been removed, because it had the bad
4206 effect of slowing Exim down by computing (never used) parameters for the
4207 RSA_EXPORT functionality.
4209 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4210 authenticator to fail if there's a tab character in the incoming data
4211 (there should never be unless someone is messing about, as it's supposed
4212 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4213 if the connection is using TLS or if the remote IP is the same as the
4214 local IP, and the "valid-client-cert option" if a client certificate has
4217 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4218 authenticators. This can be used for authorization after authentication
4219 succeeds. (In the case of plaintext, it servers for both authentication
4222 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4223 if any retry times were supplied.
4225 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4226 connection, where there is no client IP address to check. In this
4227 situation, the verify now always succeeds.
4229 PH/25 Applied John Jetmore's -Mset patch.
4231 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4233 PH/27 In a string expansion for a processed (not raw) header when multiple
4234 headers of the same name were present, leading whitespace was being
4235 removed from all of them, but trailing whitespace was being removed only
4236 from the last one. Now trailing whitespace is removed from each header
4237 before concatenation. Completely empty headers in a concatenation (as
4238 before) are ignored.
4240 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4241 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4243 PH/29 [Removed. This was a change that I later backed out, and forgot to
4244 correct the ChangeLog entry (that I had efficiently created) before
4245 committing the later change.]
4247 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4248 address errors (4xx response to RCPT) over the same connection as other
4249 messages routed to the same hosts. Such deliveries are always "forced",
4250 so retry times are not inspected. This resulted in far too many retries
4251 for the affected addresses. The effect occurred only when there were more
4252 hosts than the hosts_max_try setting in the smtp transport when it had
4253 the 4xx errors. Those hosts that it had tried were not added to the list
4254 of hosts for which the message was waiting, so if all were tried, there
4255 was no problem. Two fixes have been applied:
4257 (i) If there are any address or message errors in an SMTP delivery, none
4258 of the hosts (tried or untried) are now added to the list of hosts
4259 for which the message is waiting, so the message should not be a
4260 candidate for sending over the same connection that was used for a
4261 successful delivery of some other message. This seems entirely
4262 reasonable: after all the message is NOT "waiting for some host".
4263 This is so "obvious" that I'm not sure why it wasn't done
4264 previously. Hope I haven't missed anything, but it can't do any
4265 harm, as the worst effect is to miss an optimization.
4267 (ii) If, despite (i), such a delivery is accidentally attempted, the
4268 routing retry time is respected, so at least it doesn't keep
4269 hammering the server.
4271 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4272 in ${readsocket because some servers need this prod.
4274 PH/32 Added some extra debug output when updating a wait-xxx database.
4276 PH/33 The hint "could be header name not terminated by colon", which has been
4277 given for certain expansion errors for a long time, was not being given
4278 for the ${if def:h_colon_omitted{... case.
4280 PH/34 The spec says: "With one important exception, whenever a domain list is
4281 being scanned, $domain contains the subject domain." There was at least
4282 one case where this was not true.
4284 PH/35 The error "getsockname() failed: connection reset by peer" was being
4285 written to the panic log as well as the main log, but it isn't really
4286 panic-worthy as it just means the connection died rather early on. I have
4287 removed the panic log writing for the ECONNRESET error when getsockname()
4290 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4291 runs only) independently of the message's sender address. This meant
4292 that, if the 4xx error was in fact related to the sender, a different
4293 message to the same recipient with a different sender could confuse
4294 things. In particular, this can happen when sending to a greylisting
4295 server, but other circumstances could also provoke similar problems.
4296 I have changed the default so that the retry time for these errors is now
4297 based a combination of the sender and recipient addresses. This change
4298 can be overridden by setting address_retry_include_sender=false in the
4301 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4302 remote server are returned as part of bounce messages. This was not
4303 happening for LMTP over a pipe (the lmtp transport), but now it is the
4304 same for both kinds of LMTP.
4306 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4307 in header lines that were in fact CNAMEs. This is no longer the case.
4309 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4310 and queue runs started by the daemon processed all messages. This has
4311 been fixed so that -R and -S can now usefully be given with -q<time>.
4313 PH/40 Import PCRE release 6.7 (fixes some bugs).
4315 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4317 PH/42 Give an error if -q is specified more than once.
4319 PH/43 Renamed the variables $interface_address and $interface_port as
4320 $received_ip_address and $received_port, to make it clear that these
4321 values apply to message reception, and not to the outgoing interface when
4322 a message is delivered. (The old names remain recognized, of course.)
4324 PH/44 There was no timeout on the connect() call when using a Unix domain
4325 socket in the ${readsocket expansion. There now is.
4327 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4328 be meaningful with "accept".
4330 SC/01 Eximstats V1.43
4331 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4333 SC/02 Eximstats V1.44
4334 Use a glob alias rather than an array ref in the generated
4335 parser. This improves both readability and performance.
4337 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4338 Collect SpamAssassin and rejection statistics.
4339 Don't display local sender or destination tables unless
4340 there is data to show.
4341 Added average volumes into the top table text output.
4343 SC/04 Eximstats V1.46
4344 Collect data on the number of addresses (recipients)
4345 as well as the number of messages.
4347 SC/05 Eximstats V1.47
4348 Added 'Message too big' to the list of mail rejection
4349 reasons (thanks to Marco Gaiarin).
4351 SC/06 Eximstats V1.48
4352 Mainlog lines which have GMT offsets and are too short to
4353 have a flag are now skipped.
4355 SC/07 Eximstats V1.49 (Alain Williams)
4356 Added the -emptyok flag.
4358 SC/08 Eximstats V1.50
4359 Fixes for obtaining the IP address from reject messages.
4361 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4362 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4363 whitespace changes from 4.64-PH/27
4365 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4368 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4369 are found, allow negative numbers in numeric criteria)
4371 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4373 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4374 to match changes made in 4.64-PH/43
4376 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4378 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4379 "domain" is actually an IP address, and force a failure. This locks out
4380 those revolvers/nameservers that support "A-for-A" lookups, in
4381 contravention of the specifications.
4383 PH/48 When a host name was looked up from an IP address, and the subsequent
4384 forward lookup of the name timed out, the host name was left in
4385 $sender_host_name, contrary to the specification.
4387 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4388 restricted to single-key lookups, Exim was not diagnosing an error if
4389 * or *@ was used with a query-style lookup.
4391 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4393 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4394 long ago, but noone interested enough thought of it. Let's just say that
4395 the "1.1" means that there are some new functions that weren't there at
4396 some point in the past.
4398 PH/51 Error processing for expansion failure of helo_data from an smtp
4399 transport during callout processing was broken.
4401 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4402 tested/used via the -bh/-bhc/-bs options.
4404 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4405 bug, fixed in subsequent PCRE releases).
4407 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4408 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4410 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4416 SC/01 Use a glob alias rather than an array ref in eximstats generated
4417 parser. This improves both readability and performance.
4419 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4420 Don't display local sender or destination tables in eximstats unless
4421 there is data to show.
4422 Added average volumes into the eximstats top table text output.
4424 SC/03 Collect data on the number of addresses (recipients) as well
4425 as the number of messages in eximstats.
4427 TF/01 Correct an error in the documentation for the redirect router. Exim
4428 does (usually) call initgroups() when daemonizing.
4430 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4431 with consistent privilege compared to when running as a daemon.
4433 TF/03 Note in the spec that $authenticated_id is not set for local
4434 submissions from trusted users.
4436 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4437 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4439 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4440 by adding some example configuration directives to the default
4441 configuration file. A little bit of work is required to uncomment the
4442 directives and define how usernames and passwords are checked, but
4443 there is now a framework to start from.
4445 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4446 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4447 without this. I don't know how relevant this is to other LDAP libraries.
4449 PH/02 Add the verb name to the "unknown ACL verb" error.
4451 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4453 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4455 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4456 directory not expanded when it should be if an expanded home directory
4457 was set for the address (which is overridden by the transport).
4459 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4462 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4463 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4464 because it is too late at that time, and has no effect.
4466 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4467 security issue with \' (bugzilla #107). I could not use the
4468 PQescapeStringConn() function, because it needs a PGconn value as one of
4471 PH/08 When testing addresses using -bt, indicate those final addresses that
4472 are duplicates that would not cause an additional delivery. At least one
4473 person was confused, thinking that -bt output corresponded to deliveries.
4474 (Suppressing duplicates isn't a good idea as you lose the information
4475 about possibly different redirections that led to the duplicates.)
4477 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4478 systems where poll() doesn't work, in particular OS X.
4480 PH/10 Added more information to debugging output for retry time not reached.
4482 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4483 operations in malware.c.
4485 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4488 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4489 syslog_duplication set false, log lines that would normally be written
4490 both the the main log and to the reject log were not written to syslog at
4493 PH/14 In the default configuration, change the use of "message" in ACL warn
4494 statements to "add_header".
4496 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4497 not followed by a command (e.g. "seen endif").
4499 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4500 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4503 PH/17 Added extra conditions to the default value of delay_warning_condition
4507 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4508 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4509 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4512 The Auto-Submitted: and various List- headers are standardised, whereas I
4513 don't think Precedence: ever was.
4515 PH/18 Refactored debugging code in route_finduser() to show more information,
4516 in particular, the error code if getpwnam() issues one.
4518 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4519 This is apparently needed in addition to the PH/07 change above to avoid
4520 any possible encoding problems.
4522 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4523 but not after initializing Perl.
4525 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4526 output them only if debugging. By default they are written stderr,
4527 apparently, which is not desirable.
4529 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4532 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4535 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4537 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4538 authenticated or an ident call has been made. Suppress the default
4539 values for $authenticated_id and $authenticated_sender (but permit -oMai
4540 and -oMas) when testing with -bh.
4542 PH/24 Re-jigged the order of the tests in the default configuration so that the
4543 tests for valid domains and recipients precede the DNS black list and CSA
4544 tests, on the grounds that those ones are more expensive.
4546 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4547 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4548 This bug exists in every version of Exim that I still have, right back to
4551 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4552 However, an attempt to turn on case-sensitivity in a regex key by
4553 including (?-i) didn't work because the subject string was already
4554 lowercased, and the effects were non-intuitive. It turns out that a
4555 one-line patch can be used to allow (?-i) to work as expected.
4561 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4562 other effects) broke the use of negated acl sub-conditions.
4564 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4567 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4568 "Deny" causes Exim to reject the incoming connection with a 554 error.
4569 Unfortunately, if there is a major crisis, such as a disk failure,
4570 tcp-wrappers gives "deny", whereas what one would like would be some
4571 kind of temporary error. A kludge has been added to help with this.
4572 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4573 554 error is used if errno is still zero or contains ENOENT (which occurs
4574 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4577 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4579 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4580 errors. Otherwise a message that provokes a temporary error (when other
4581 messages do not) can cause a whole host to time out.
4583 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4584 addresses were routed directly to files or pipes from a redirect router.
4585 File deliveries just didn't batch; pipe deliveries might have suffered
4588 PH/06 A failure to get a lock for a hints database would erroneously always say
4589 "Failed to get write lock", even when it was really a read lock.
4591 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4592 of 0600. This has been changed to use the value of the lockfile_mode
4593 option (which defaults to 0600).
4595 PH/08 Applied small patch from the Sieve maintainer.
4597 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4598 folder from quota calculations, a direct delivery into this folder messed
4599 up the contents of the maildirsize file. This was because the regex was
4600 used only to exclude .Trash (or whatever) when the size of the mailbox
4601 was calculated. There was no check that a delivery was happening into an
4602 excluded directory. This bug has been fixed by ignoring all quota
4603 processing for deliveries into excluded directories.
4605 PH/10 Added the maildirfolder_create_regex option to appendfile.
4611 PH/01 The code for finding all the local interface addresses on a FreeBSD
4612 system running IPv6 was broken. This may well have applied to all BSD
4613 systems, as well as to others that have similar system calls. The broken
4614 code found IPv4 interfaces correctly, but gave incorrect values for the
4615 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4616 that it would not match correctly against @[] and not recognize the IPv6
4619 PH/02 The ipliteral router was not recognizing addresses of the form user@
4620 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4622 PH/03 Added disable_ipv6.
4624 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4625 decoded form, because it is most often used to construct To: headers
4626 lines in autoreplies, and the decoded form may well be syntactically
4627 invalid. However, $reply_address has leading white space removed, and all
4628 newlines turned into spaces so that the autoreply transport does not
4631 PH/05 If group was specified without a user on a router, and no group or user
4632 was specified on a transport, the group from the router was ignored.
4634 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4635 for visible compile-time settings that can be used to change these
4636 numbers, for those that want even more. Backwards compatibility with old
4637 spool files has been maintained. However, going back to a previous Exim
4638 release will lost any variables that are in spool files.
4640 PH/07 Two small changes when running in the test harness: increase delay when
4641 passing a TCP/IP connection to a new process, in case the original
4642 process has to generate a bounce, and remove special handling of
4643 127.0.0.2 (sic), which is no longer necessary.
4645 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4646 be the same on different OS.
4648 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4651 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4652 whether --show-vars was specified or not
4654 JJ/02 exipick: Added support for new ACL variable spool format introduced
4657 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4658 syntactically invalid From: or Reply-to: line, and a filter used this to
4659 generate an autoreply, and therefore failed to obtain an address for the
4660 autoreply, Exim could try to deliver to a non-existent relative file
4661 name, causing unrelated and misleading errors. What now happens is that
4662 it logs this as a hard delivery error, but does not attempt to create a
4665 PH/11 The exinext utility has a -C option for testing purposes, but although
4666 the given file was scanned by exinext itself; it wasn't being passed on
4667 when Exim was called.
4669 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4670 an end-of-file indication when reading a command response.
4672 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4673 compiled. In many other places in Exim, IPv6 addresses are always
4674 recognized, so I have changed this. It also means that IPv4 domain
4675 literals of the form [IPV4:n.n.n.n] are now always recognized.
4677 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4678 used if the router is not running as root, for example, when verifying at
4679 ACL time, or when using -bh. The debugging output from this situation was
4680 non-existent - all you got was a failure to exec. I have made two
4683 (a) Failures to set uid/gid, the current directory, or a process leader
4684 in a subprocess such as that created by queryprogram now generate
4685 suitable debugging output when -d is set.
4687 (b) The queryprogram router detects when it is not running as root,
4688 outputs suitable debugging information if -d is set, and then runs
4689 the subprocess without attempting to change uid/gid.
4691 PH/15 Minor change to Makefile for building test_host (undocumented testing
4694 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4695 additional section of a DNS packet that returns MX or SRV records.
4696 Instead, it always explicitly searches for A/AAAA records. This avoids
4697 major problems that occur when a DNS server includes only records of one
4698 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4699 fixed another bug: if SRV records were looked up and the corresponding
4700 address records were *not* found in the additional section, the port
4701 values from the SRV records were lost.
4703 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4704 using the correct key (the original address) when searching the retry
4705 rules in order to find which one to use for generating the retry hint.
4707 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4708 adding the default one. Similarly, if it contains a Reply-To: header, the
4709 errors_reply_to option, if set, is not used.
4711 PH/19 When calculating a retry time, Exim used to measure the "time since
4712 failure" by looking at the "first failed" field in the retry record. Now
4713 it does not use this if it is later than than the arrival time of the
4714 message. Instead it uses the arrival time. This makes for better
4715 behaviour in cases where some deliveries succeed, thus re-setting the
4716 "first failed" field. An example is a quota failure for a huge message
4717 when small messages continue to be delivered. Without this change, the
4718 "time since failure" will always be short, possible causing more frequent
4719 delivery attempts for the huge message than are intended.
4720 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4722 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4723 $1, $2, $3) because the numerical variables can be reset during some
4724 expansion items (e.g. "match"), thereby losing the authentication data.
4726 PH/21 Make -bV show the size of off_t variables so that the test suite can
4727 decide whether to run tests for quotas > 2G.
4729 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4730 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4731 filecount value is greater than 2G or if a quota value is greater than 2G
4732 on a system where the size of off_t is not greater than 4, a panic error
4735 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4736 never match. The debug and -bh output now contains an explicit error
4737 message indicating a malformed IPv4 address or mask.
4739 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4740 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4741 PH/23 above applies.
4743 PH/25 Do not write to syslog when running in the test harness. The only
4744 occasion when this arises is a failure to open the main or panic logs
4745 (for which there is an explicit test).
4747 PH/26 Added the /no_tell option to "control=freeze".
4749 PH/27 If a host name lookup failed very early in a connection, for example, if
4750 the IP address matched host_lookup and the reverse lookup yielded a name
4751 that did not have a forward lookup, an error message of the form "no IP
4752 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4753 could be logged. Now it outputs the IP address instead of "NULL".
4755 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4756 allows the sender and the authenticated sender to be set when
4757 submitting a message from within Exim. Since child_open_exim() is
4758 documented for local_scan(), the new function should be too.
4760 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4761 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4762 results in an empty string is now treated as unset.
4764 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4766 PH/31 Added sender_verify_fail logging option.
4768 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4769 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4770 tidied the source and removed it altogether.
4772 PH/33 When a queue run was abandoned because the load average was too high, a
4773 log line was always written; now it is written only if the queue_run log
4774 selector is set. In addition, the log line for abandonment now contains
4775 information about the queue run such as the pid. This is always present
4776 in "start" and "stop" lines but was omitted from the "abandon" line.
4778 PH/34 Omit spaces between a header name and the colon in the error message that
4779 is given when verify = headers_syntax fails (if there are lots of them,
4780 the message gets confusing).
4782 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4783 names, as there are now some PTR records that contain slashes. This check
4784 is only to protect against broken name servers that fall over on strange
4785 characters, so the fact that it applies to all lookups doesn't matter.
4787 PH/36 Now that the new test suite is complete, we can remove some of the
4788 special code in Exim that was needed for the old test suite. For example,
4789 sorting DNS records because real resolvers return them in an arbitrary
4790 order. The new test suite's fake resolver always returns records in the
4793 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4794 bounces) except when queue_only is set, to avoid logging races between
4795 the different processes.
4797 PH/38 Panic-die if .include specifies a non-absolute path.
4799 PH/39 A tweak to the "H" retry rule from its user.
4801 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4802 a label. They prevented compilation on older perls.
4804 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4805 a warning to be raised on newish perls.
4807 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4808 on queue. Changes to match documented behaviour of showing count of
4809 messages matching specified criteria.
4811 PH/40 Changed the default ident timeout from 30s to 5s.
4813 PH/41 Added support for the use of login_cap features, on those BSD systems
4814 that have them, for controlling the resources used by pipe deliveries.
4816 PH/42 The content-scanning code uses fopen() to create files in which to put
4817 message data. Previously it was not paying any attention to the mode of
4818 the files. Exim runs with umask(0) because the rest of the code creates
4819 files with open(), and sets the required mode explicitly. Thus, these
4820 files were ending up world-writeable. This was not a big issue, because,
4821 being within the spool directory, they were not world-accessible. I have
4822 created a function called modefopen, which takes an additional mode
4823 argument. It sets umask(777), creates the file, chmods it to the required
4824 mode, then resets the umask. All the relevant calls to fopen() in the
4825 content scanning code have been changed to use this function.
4827 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4828 to 24 hours. This avoids potential overflow problems when processing G
4829 and H retry rules. I suspect nobody ever tinkers with this value.
4831 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4833 PH/45 When the plaintext authenticator is running as a client, the server's
4834 challenges are checked to ensure they are valid base64 strings. By
4835 default, the authentication attempt is cancelled if an invalid string is
4836 received. Setting client_ignore_invalid_base64 true ignores these errors.
4837 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4838 they are received. Thus, the responses can be made to depend on the
4839 challenges. If an invalid string is ignored, an empty string is placed in
4842 PH/46 Messages that are created by the autoreply transport now contains a
4843 References: header, in accordance with RFCs 2822 and 3834.
4845 PH/47 Added authenticated_sender_force to the smtp transport.
4847 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4849 PH/49 Installed latest patch from the Sieve maintainer.
4851 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4852 was also set, the appendfile transport was unnecessarily scanning a
4853 directory of message files (e.g. for maildir delivery) to find the count
4854 of files (along with the size), even though it did not need this
4855 information. It now does the scan only if it needs to find either the
4856 size of the count of files.
4858 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4860 PH/52 Two bugs concerned with error handling when the smtp transport is
4863 (i) Exim was not creating retry information for temporary errors given
4864 for individual recipients after the DATA command when the smtp transport
4865 was used in LMTP mode. This meant that they could be retried too
4866 frequently, and not timed out correctly.
4868 (ii) Exim was setting the flag that allows error details to be returned
4869 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4870 recipients that were returned after the DATA command.
4872 PH/53 This is related to PH/52, but is more general: for any failing address,
4873 when detailed error information was permitted to be returned to the
4874 sender, but the error was temporary, then after the final timeout, only
4875 "retry timeout exceeded" was returned. Now it returns the full error as
4876 well as "retry timeout exceeded".
4878 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4879 do this, and (what is worse) MTAs that accept it.
4881 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4882 will now be deprecated.
4884 PH/56 New os.c-cygwin from the Cygwin maintainer.
4886 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4887 formats (previously only available in exim formats via -bpr, -bpru,
4888 and -bpra. Now also available in native and exiqgrep formats)
4890 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4891 with very large, slow to parse queues
4893 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4895 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4897 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4898 responses to authentication challenges, though it was showing the
4899 challenges; (ii) I've removed the CR characters from the debug output for
4902 PH/58 Allow for the insertion of a newline as well as a space when a string
4903 is turned into more than one encoded-word during RFC 2047 encoding. The
4904 Sieve code now uses this.
4906 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4907 data_4xx, lost_connection, tls_required.
4909 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4910 message was being sent as an SMTP response.
4912 PH/61 Add -l and -k options to exicyclog.
4914 PH/62 When verifying, if an address was redirected to one new address, so that
4915 verification continued, and the new address failed or deferred after
4916 having set something in $address_data, the value of $address_data was not
4917 passed back to the ACL. This was different to the case when no
4918 redirection occurred. The value is now passed back in both cases.
4920 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4921 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4922 use login_cap.h, so on its own it isn't the distinguishing feature. The
4923 new name refers directly to the setclassresources() function.
4925 PH/65 Added configuration files for NetBSD3.
4927 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4929 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4930 is preferred over IPv4.
4932 PH/68 The bounce_return_message and bounce_return_body options were not being
4933 honoured for bounces generated during the reception of non-SMTP messages.
4934 In particular, this applied to messages rejected by the ACL. This bug has
4935 been fixed. However, if bounce_return_message is true and bounce_return_
4936 body is false, the headers that are returned for a non-SMTP message
4937 include only those that have been read before the error was detected.
4938 (In the case of an ACL rejection, they have all been read.)
4940 PH/69 The HTML version of the specification is now built in a directory called
4941 spec_html instead of spec.html, because the latter looks like a path with
4942 a MIME-type, and this confuses some software.
4944 PH/70 Catch two compiler warnings in sieve.c.
4946 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4947 function verify_get_ident() calls ip_connect() to connect a socket, but
4948 if the "connect()" function timed out, ip_connect() used to close the
4949 socket. However, verify_get_ident() also closes the socket later, and in
4950 between Exim writes to the log, which may get opened at this point. When
4951 the socket was closed in ip_connect(), the log could get the same file
4952 descriptor number as the socket. This naturally causes chaos. The fix is
4953 not to close the socket in ip_connect(); the socket should be closed by
4954 the function that creates it. There was only one place in the code where
4955 this was missing, in the iplookup router, which I don't think anybody now
4956 uses, but I've fixed it anyway.
4958 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4959 well as to direct DNS lookups. Otherwise the handling of names in host
4960 lists is inconsistent and therefore confusing.
4966 PH/01 Two changes to the default runtime configuration:
4968 (1) Move the checks for relay_from_hosts and authenticated clients from
4969 after to before the (commented out) DNS black list checks.
4971 (2) Add control=submission to the relay_from_hosts and authenticated
4972 clients checks, on the grounds that messages accepted by these
4973 statements are most likely to be submissions.
4975 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4977 (1) Generate an error if the third argument for the ${prvs expansion is
4980 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4983 (3) Reset the variables that are obtained from the first argument of
4984 ${prvscheck and used in the second argument before leaving the code,
4985 because their memory is reclaimed, so using them afterwards may do
4988 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4989 one (it's much easier than Tom thought :-).
4991 (5) Because of (4), we can now allow for the use of $prvscheck_result
4992 inside the third argument.
4994 PH/03 For some reason, the default setting of PATH when running a command from
4995 a pipe transport was just "/usr/bin". I have changed it to
4998 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4999 anything to be listed in the output from -bV.
5001 PH/05 When a filter generated an autoreply, the entire To: header line was
5002 quoted in the delivery log line, like this:
5004 => >A.N.Other <ano@some.domain> <original@ddress> ...
5006 This has been changed so that it extracts the operative address. There
5007 may be more than one such address. If so, they are comma-separated, like
5010 => >ano@some.domain,ona@other.domain <original@ddress> ...
5012 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5013 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5014 not being looked up in the rDNS to get a host name, Exim was showing the
5015 IP address twice in Received: lines, even though the IP addresses were
5016 identical. For example:
5018 Received: from [1.2.3.4] (helo=[1.2.3.4])
5020 However, if the real host name was known, it was omitting the HELO data
5021 if it matched the actual IP address. This has been tidied up so that it
5022 doesn't show the same IP address twice.
5024 PH/07 When both +timestamp and +memory debugging was on, the value given by
5025 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5026 called by the debug printing, thereby overwriting the timestamp buffer.
5027 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5029 PH/08 When the original message was included in an autoreply transport, it
5030 always said "this is a copy of the message, including all the headers",
5031 even if body_only or headers_only was set. It now gives an appropriate
5034 PH/09 Applied a patch from the Sieve maintainer which:
5036 o fixes some comments
5037 o adds the (disabled) notify extension core
5038 o adds some debug output for the result of if/elsif tests
5039 o points to the current vacation draft in the documentation
5040 and documents the missing references header update
5044 o fixes a bug in processing the envelope test (when testing
5045 multiple envelope elements, the last element determined the
5048 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5049 Electronic Mail") by including:
5051 Auto-submitted: auto-generated
5053 in the messages that it generates (bounce messages and others, such as
5054 warnings). In the case of bounce messages for non-SMTP messages, there was
5055 also a typo: it was using "Auto_submitted" (underscore instead of
5056 hyphen). Since every message generated by Exim is necessarily in response
5057 to another message, thes have all been changed to:
5059 Auto-Submitted: auto-replied
5061 in accordance with these statements in the RFC:
5063 The auto-replied keyword:
5065 - SHOULD be used on messages sent in direct response to another
5066 message by an automatic process,
5068 - MUST NOT be used on manually-generated messages,
5070 - MAY be used on Delivery Status Notifications (DSNs) and Message
5071 Disposition Notifications (MDNs),
5073 - MUST NOT be used on messages generated by automatic or periodic
5074 processes, except for messages which are automatic responses to
5077 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5078 to the default Received: header definition.
5080 PH/12 Added log selector acl_warn_skipped (default on).
5082 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5083 variables because (a) they are in the wrong storage pool and (b) even if
5084 they were copied, it wouldn't work properly because of the caching.
5086 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5087 checking when decoding. Apparently there are clients that generate
5088 overlong encoded strings. Why am I not surprised?
5090 PH/15 If the first argument of "${if match_address" was not empty, but did not
5091 contain an "@" character, Exim crashed. Now it writes a panic log message
5092 and treats the condition as false.
5094 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5096 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5097 extension "envelope-auth". The code is finished and in agreement with
5098 other implementations, but there is no documentation so far and in fact,
5099 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5100 not changing the active code.
5102 Print executed "if" and "elsif" statements when debugging is used. This
5103 helps a great deal to understand what a filter does.
5105 Document more things not specified clearly in RFC3028. I had all this
5106 sorted out, when out of a sudden new issues came to my mind. Oops."
5108 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5111 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5112 canonical form (as documented). However, after a host name lookup from
5113 the IP address, check_host() was doing a simple string comparison with
5114 addresses acquired from the DNS when checking that the found name did
5115 have the original IP as one of its addresses. Since any found IPv6
5116 addresses are likely to be in abbreviated form, the comparison could
5117 fail. Luckily, there already exists a function for doing the comparison
5118 by converting both addresses to binary, so now that is used instead of
5119 the text comparison.
5121 PH/20 There was another similar case to PH/19, when a complete host name was
5122 given in a host list; looking up its IP address could give an abbreviated
5123 form, whereas the current host's name might or might not be abbreviated.
5124 The same fix has been applied.
5130 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5131 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5134 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5135 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5137 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5139 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5140 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5141 Also, for "Auto-Submitted", treat anything other than "no" as
5142 non-personal, in accordance with RFC 3834. (Previously it treated
5143 anything starting "auto-" as non-personal.)
5145 TF/01 The control=submission/name=... option had a problem with syntax
5146 errors if the name included a slash character. The /name= option
5147 now slurps the rest of the string, so it can include any characters
5148 but it must come last in the list of options (after /sender_retain
5151 PH/05 Some modifications to the interface to the fake nameserver for the new
5159 TK/01 Added the "success_on_redirect" address verification option. See
5160 NewStuff for rationale and an example.
5162 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5164 PH/02 Patch to exigrep to allow it to work on syslog lines.
5166 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5167 fread() to skip over the body file's header line, because in Cygwin the
5168 header line is locked and is inaccessible.
5170 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5171 co-exist for some time) to make it clear that it is the Exim ID that is
5172 referenced, not the Message-ID: header line.
5174 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5175 string_format() function, because snprintf() does not exist on all
5178 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5179 ${stat: expansion item.
5181 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5182 protocol synchronization error", to keep the pedants happy.
5184 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5185 well as for IRIX systems, when gcc is being used. See the host.c source
5188 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5190 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5193 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5194 transport and to the smtp transport in LMTP mode.
5196 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5198 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5199 run a filter in a subprocess. This could lead to confusion in subsequent
5200 lookups in the parent process. There should also be a search_tidyup() at
5201 the end of the subprocess.
5203 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5204 only if the host matched helo_try_verify_hosts, which caused the
5205 verification to occur when the EHLO/HELO command was issued. The ACL just
5206 tested the remembered result. Now, if a previous verification attempt has
5207 not happened, "verify = helo" does it there and then.
5209 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5211 TK/03 Fix log output including CR from clamd.
5213 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5214 exist provoked a memory error which could cause a segfault.
5216 PH/15 Installed PCRE 6.2
5218 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5220 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5221 of the problem. Specifically, suggested +O2 rather than +O1 for the
5224 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5226 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5227 record, the port was not correctly logged when the outgoing_port log
5228 selector was set (it logged the transort's default port).
5230 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5231 fallback_hosts, and "hosts" in the smtp transport.
5233 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5234 host errors such as "Connection refused".
5236 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5237 authentication with radiusclient 0.4.9:
5239 - Error returned from rc_read_config was caught wrongly
5240 - Username/password not passed on to radius server due to wrong length.
5242 The presumption is that some radiusclient API changes for 4.51/PH/17
5243 were not taken care of correctly. The code is still untested by me (my
5244 Linux distribution still has 0.3.2 of radiusclient), but it was
5245 contributed by a Radius user.
5247 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5248 expanding the "port" option of the smtp transport.
5250 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5251 while reading a MIME header. Thanks to Tom Hughes for a patch.
5253 PH/24 Include config.h inside local_scan.h so that configuration settings are
5256 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5257 that in an ACL for RCPT (for example), you can examine exactly what was
5260 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5261 commands, but it was not correctly comparing the address with the actual
5262 client host address. Thus, it would show the EHLO address in Received:
5263 header lines when this was not necessary.
5265 PH/27 Added the % operator to ${eval:}.
5267 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5268 it should be ignoring failures (because with -C, for example, it has lost
5269 privilege). It wasn't ignoring creation failures other than "already
5272 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5275 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5276 because an input line was too long, either on its own, or by virtue of
5277 too many continuations, the temporary file was not being removed, and the
5278 return code was incorrect.
5280 PH/31 Missing "BOOL" in function definition in filtertest.c.
5282 PH/32 Applied Sieve patches from the maintainer.
5284 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5286 PH/33 Added "verify = not_blind".
5288 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5289 Local/Makefile (with some defaults set). These are used in built scripts
5290 such as exicyclog, but they have never been used in the exim_install
5291 script (though there are many overriding facilities there). I have
5292 arranged that the exim_install script now takes note of these two
5295 PH/35 Installed configuration files for Dragonfly.
5297 PH/36 When a locally submitted message by a trusted user did not contain a
5298 From: header, and the sender address was obtained from -f or from an SMTP
5299 MAIL command, and the trusted user did not use -F to supply a sender
5300 name, $originator_name was incorrectly used when constructing a From:
5301 header. Furthermore, $originator_name was used for submission mode
5302 messages from external hosts without From: headers in a similar way,
5303 which is clearly wrong.
5305 PH/37 Added control=suppress_local_fixups.
5307 PH/38 When log_selector = +received_sender was set, and the addition of the
5308 sender made the log line's construction buffer exactly full, or one byte
5309 less than full, an overflow happened when the terminating "\n" was
5312 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5313 when the result of a list match is failure because a DNS lookup failed.
5315 PH/40 RM_COMMAND is now used in the building process.
5317 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5318 the "build-* directories that it finds.
5320 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5321 address in a domain literal was a prefix of an interface address.
5323 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5324 when verifying a sender address, unless rewrite_headers is false.
5326 PH/44 Wrote a long comment about why errors_to addresses are verified as
5327 recipients, not senders.
5329 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5330 the ratelimit ACL was added.
5332 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5334 PH/46 Added extra information about PostgreSQL errors to the error string.
5336 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5337 avoiding the need to install special zones in a real server. This is
5338 backwards compatible; if it can't find the fake resolver, it drops back.
5339 Thus, both old and new test suites can be run.
5341 TF/02 Added util/ratelimit.pl
5343 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5344 clock is set back in time.
5346 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5347 Candler <B.Candler@pobox.com>.
5349 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5350 OK for addresses that are the result of redirections.
5352 PH/48 A number of further additions for the benefit of the new test suite,
5353 including a fake gethostbyname() that interfaces to the fake DNS resolver
5356 TF/06 The fix for widen_domains has also been applied to qualify_single and
5357 search_parents which are the other dnslookup options that can cause
5360 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5363 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5365 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5366 service" if it encountered an MX record with an empty target hostname.
5367 The message is now "an MX or SRV record indicated no SMTP service".
5369 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5370 if the DNS of the sending site is misconfigured. This is quite a
5371 common situation. This change restores the behaviour of treating a
5372 helo verification defer as a failure.
5374 PH/51 If self=fail was set on a router, the bounce message did not include the
5375 actual error message.
5381 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5383 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5384 command itself ended in error, the underlying message about the transport
5385 filter timeout was being overwritten with the pipe command error. Now the
5386 underlying error message should be appended to the second error message.
5388 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5390 PH/02 Reduce the amount of output that "make" produces by default. Full output
5391 can still be requested.
5393 PH/03 The warning log line about a condition test deferring for a "warn" verb
5394 was being output only once per connection, rather than after each
5395 occurrence (because it was using the same function as for successful
5396 "warn" verbs). This seems wrong, so I have changed it.
5398 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5399 it should not have, which might have caused a crash in the right
5400 circumstances, but probably never did.
5402 PH/04 Installed a modified version of Tony Finch's patch to make submission
5403 mode fix the return path as well as the Sender: header line, and to
5404 add a /name= option so that you can make the user's friendly name appear
5407 TF/03 Added the control = fakedefer ACL modifier.
5409 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5410 Mark Lowes for thorough testing.
5412 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5414 TK/03 Merged latest SRS patch from Miles Wilton.
5416 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5417 with the definition in sysexits.h (which is #included earlier).
5418 Fortunately, Exim does not actually use EX_OK. The code used to try to
5419 preserve the sysexits.h value, by assuming that macro definitions were
5420 scanned for macro replacements. I have been disabused of this notion,
5421 so now the code just undefines EX_OK before #including unistd.h.
5423 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5424 in the smtp transport. When a block could not be written in a single
5425 write() function, the timeout was being re-applied to each part-write.
5426 This seems wrong - if the receiver was accepting one byte at a time it
5427 would take for ever. The timeout is now adjusted when this happens. It
5428 doesn't have to be particularly precise.
5430 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5431 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5433 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5434 without a domain if the check to <postmaster@domain> fails.
5436 SC/01 Eximstats: added -xls and the ability to specify output files
5437 (patch written by Frank Heydlauf).
5439 SC/02 Eximstats: use FileHandles for outputting results.
5441 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5443 SC/04 Eximstats: fixed display of large numbers with -nvr option
5445 SC/05 Eximstats: fixed merging of reports with empty tables.
5447 SC/06 Eximstats: added the -include_original_destination flag
5449 SC/07 Eximstats: removed tabs and trailing whitespace.
5451 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5453 TK/06 MBOX spool code: Add real "From " MBOX separator line
5454 so the .eml file is really in mbox format (even though
5455 most programs do not really care). Patch from Alex Miller.
5457 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5458 The latter is generated from $received_to and is only set if the
5459 message has one envelope recipient. SA can use these headers,
5460 obviously out-of-the-box. Patch from Alex Miller.
5462 PH/08 The ${def test on a variable was returning false if the variable's
5463 value was "0", contrary to what the specification has always said!
5464 The result should be true unless the variable is empty.
5466 PH/09 The syntax error of a character other than { following "${if
5467 def:variable_name" (after optional whitespace) was not being diagnosed.
5468 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5469 accidental colon was present, for example, could give incorrect results.
5471 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5472 result is used (not including appendfile, where other changes are about
5475 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5476 This involved changing a lot of size variables from int to off_t. It
5477 should work with maildirs and everything.
5479 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5480 spamd dying while we are connected to it.
5482 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5485 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5486 function for BDB 4.3.
5488 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5490 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5491 the sources for PCRE have been re-arranged and more files are now
5494 PH/15 The code I had for printing potentially long long variables in PH/11
5495 above was not the best (it lost precision). The length of off_t variables
5496 is now inspected at build time, and an appropriate printing format (%ld
5497 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5498 to be "long long int" or "long int". This is needed for the internal
5499 formatting function string_vformat().
5501 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5502 the configuration file to be ":syslog", then the script "guesses" where
5503 the logs files are, rather than using the compiled in default. In our
5504 case the guess is not the same as the compiled default, so the script
5505 suddenly stopped working when I started to use syslog. The patch checks
5506 to see if log_file_path is "". If so, it attempts to read it from exim
5507 with no configuration file to get the compiled in version, before it
5508 falls back to the previous guessing code."
5510 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5511 implementing BATV in an Exim configuration. See NewStuff for the gory
5514 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5515 Makefile that are specific to HP-UX.
5517 PH/18 If the "use_postmaster" option was set for a recipient callout together
5518 with the "random" option, the postmaster address was used as the MAIL
5519 FROM address for the random test, but not for the subsequent recipient
5520 test. It is now used for both.
5522 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5523 patch removes a few documentation additions to RFC 3028, because the
5524 latest draft now contains them. It adds the new en;ascii-case comparator
5525 and a new error check for 8bit text in MIME parts. Comparator and
5526 require names are now matched exactly. I enabled the subaddress
5527 extension, but it is not well tested yet (read: it works for me)."
5529 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5530 rework some of the code of TK/09 above to avoid the hardwired use of
5531 "%lld" and "long long". Replaced the call to snprintf() with a call to
5534 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5535 records point to non-existent hosts", "retry timeout exceeded", and
5536 "retry time not reached for any host after a long failure period".
5538 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5539 experimental DomainKeys support:
5541 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5542 (2) On an error such as an illegally used "control", the wrong name for
5543 the control was given.
5545 These problems did NOT occur unless DomainKeys support was compiled.
5547 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5549 PH/24 Added ${if match_ip condition.
5551 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5552 databases so that it will be absolutely obvious if a crash occurs in the
5553 DB library. This is a regular occurrence (often caused by mis-matched
5556 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5557 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5558 picked up on a user's system that detects such things. There doesn't seem
5559 to be a gcc warning option for this - only an attribute that has to be
5560 put on the function's prototype. It seems that in Fedora Core 4 they have
5561 set this on a number of new functions. No doubt there will be more in due
5564 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5565 specify a transport. However, if an address that was verified by such a
5566 router was the subject of a callout, Exim crashed because it tried to
5567 read the rcpt_include_affixes from the non-existent transport. Now it
5568 just assumes that the setting of that option is false. This bug was
5569 introduced by 4.51/PH/31.
5571 PH/28 Changed -d+all to exclude +memory, because that information is very
5572 rarely of interest, but it makes the output a lot bigger. People tend to
5573 do -d+all out of habit.
5575 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5576 code in os-type was giving problems when libc.so lives in lib64, like on
5579 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5580 aren't the modern standard, and it seems that some systems' include files
5581 don't always have them. Exim was already checking for some of the newer
5582 ones like T_AAAA, and defining it itself. I've added checks for all the
5583 record types that Exim uses.
5585 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5586 not automatically generating a new one, as it is supposed to. This
5587 prevented TLS from working. If the file did exist, but contained invalid
5588 data, a new version was generated, as expected. It was only the case of a
5589 non-existent file that was broken.
5591 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5592 with a change in libdomainkeys > 0.64.
5594 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5595 from DNS. If the selector record carries the flag, it now has
5596 precedence over the domain-wide flag.
5598 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5600 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5601 the use of a port name that isn't defined in /etc/services) occurred, the
5602 message was deferred as in a normal delivery, and thus remained on the
5603 spool, instead of being failed because of the mua_wrapper setting. This
5604 is now fixed, and I tidied up some of the mua_wrapper messages at the
5607 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5608 the messages in a hash of arrays rather than using individual hashes.
5609 This is a bit cleaner and results in dramatic memory savings, albeit
5610 at a slight CPU cost.
5612 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5613 as requested by Marc Sherman.
5615 SC/10 Eximstats: added histograms for user specified patterns as requested
5618 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5620 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5621 fopen() in the content-scanning modules that did not already have it.
5627 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5628 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5630 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5632 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5634 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5635 to test Sieve filters that use "vacation".
5637 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5638 that changes the way the GnuTLS parameters are stored in the cache file.
5639 The new format can be generated externally. For backward compatibility,
5640 if the data in the cache doesn't make sense, Exim assumes it has read an
5641 old-format file, and it generates new data and writes a new file. This
5642 means that you can't go back to an older release without removing the
5645 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5646 work if there are any delivery delays because "one_time" forces the
5647 parent to be marked "delivered", so its unseen clone is never tried
5648 again. For this reason, Exim now forbids the simultaneous setting of
5651 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5652 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5653 entry if you want to know the details. The fix, however, neglected to
5654 consider the case when local delivery batching is involved. The test for
5655 "previously delivered" was not happening when checking to see if an
5656 address could be batched with a previous (undelivered) one; under
5657 certain circumstances this could lead to multiple deliveries to the same
5660 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5661 in its include files, and this causes problems building Exim.
5663 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5664 header_syntax) but Exim was just ignoring anything given after a slash.
5665 In particular, this caused confusion with an attempt to use "verify =
5666 reverse_host_lookup/defer_ok". An error is now given when options are
5667 supplied for verify items that do not have them. (Maybe reverse_host_
5668 lookup should have a defer_ok option, but that's a different point.)
5670 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5671 defined by RFC 821) to 2048, because there were problems with some AUTH
5672 commands, and RFC 1869 says the size should be increased for extended
5673 SMTP commands that take arguments.
5675 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5678 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5679 "unknown" error; now it says that the functionality isn't in the binary.
5681 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5682 an address' error message when a string expansion fails (syntax or
5683 whatever). Otherwise the password may appear in the log. Following change
5684 PH/42 below, there is no longer a chance of it appearing in a bounce
5687 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5689 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5690 of its addresses was ever tried. (Bugzilla bug #2.)
5692 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5693 the result incorrectly in the debug output. (It correctly added a newline
5694 to what was transported.)
5696 TF/01 Added $received_time.
5698 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5699 commented out examples of how to interface to a virus scanner and to
5700 SpamAssassin. Also added commented examples of av_scanner and
5701 spamd_address settings.
5703 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5704 and controls are allowed in which ACLs. There were a couple of minor
5705 errors. Some of the entries in the conditions table (which is a table of
5706 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5707 negation of where the condition IS allowed.
5709 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5711 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5712 header file does not have a version number, so I've had to invent a new
5713 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5714 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5715 radiusclient), but it was contributed by a Radius user.
5717 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5718 files or directories, for OpenSSL.
5720 PH/19 When an Exim process that is running as root has to create an Exim log
5721 file, it does so in a subprocess that runs as exim:exim so as to get the
5722 ownership right at creation (otherwise, other Exim processes might see
5723 the file with the wrong ownership). There was no test for failure of this
5724 fork() call, which would lead to the process getting stuck as it waited
5725 for a non-existent subprocess. Forks do occasionally fail when resources
5726 run out. I reviewed all the other calls to fork(); they all seem to check
5729 PH/20 When checking for unexpected SMTP input at connect time (before writing
5730 the banner), Exim was not dealing correctly with a non-positive return
5731 from the read() function. If the client had disconnected by this time,
5732 the result was a log entry for a synchronization error with an empty
5733 string after "input=" when read() returned zero. If read() returned -1
5734 (an event I could not check), uninitialized data bytes were printed.
5735 There were reports of junk text (parts of files, etc) appearing after
5738 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5740 PH/22 Added support for macro redefinition, and (re)definition in between
5741 driver and ACL definitions.
5743 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5744 forgetting to use the resulting value; it was using the unexpanded value.
5746 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5747 hadn't been configured. The fix is from Juergen Kreileder, who
5748 understands it better than I do:
5750 "Here's what I see happening with three configured cyrus_sasl
5751 authenticators configured (plain, login, cram-md5):
5753 On startup auth_cyrus_sasl_init() gets called for each of these.
5754 This means three calls to sasl_listmech() without a specified mech_list.
5755 => SASL tests which mechs of all available mechs actually work
5756 => three warnings about OTP not working
5757 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5759 With the patch, sasl_listmech() also gets called three times. But now
5760 SASL's mech_list option is set to the server_mech specified in the the
5761 authenticator. Or in other words, the answer from sasl_listmech()
5762 gets limited to just the mech you're testing for (which is different
5764 => the return list contains just 'plain' or 'login', 'cram-md5' or
5765 nothing depending on the value of ob->server_mech.
5767 I've just tested the patch: Authentication still works fine,
5768 unavailable mechs specified in the exim configuration are still
5769 caught, and the auth.log warnings about OTP are gone."
5771 PH/25 When debugging is enabled, the contents of the command line are added
5772 to the debugging output, even when log_selector=+arguments is not
5775 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5776 answer is "GNU", and only if the return is "GNU/something" is the answer
5779 PH/27 $acl_verify_message is now set immediately after the failure of a
5780 verification in an ACL, and so is available in subsequent modifiers. In
5781 particular, the message can be preserved by coding like this:
5783 warn !verify = sender
5784 set acl_m0 = $acl_verify_message
5786 Previously, $acl_verify_message was set only while expanding "message"
5787 and "log_message" when a very denied access.
5789 PH/28 Modified OS/os.c-Linux with
5791 -#ifndef OS_LOAD_AVERAGE
5792 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5794 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5795 nomenclature these days.)
5797 PH/29 Installed patch from the Sieve maintainer that adds the options
5798 sieve_useraddress and sieve_subaddress to the redirect router.
5800 PH/30 In these circumstances:
5801 . Two addresses routed to the same list of hosts;
5802 . First host does not offer TLS;
5803 . First host accepts first address;
5804 . First host gives temporary error to second address;
5805 . Second host offers TLS and a TLS session is established;
5806 . Second host accepts second address.
5807 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5808 and peerdn, if requested) that were in fact used only for the second
5811 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5812 attention to any local part prefix or suffix that was matched by the
5813 router that accepted the address. It now behaves in the same way as it
5814 does for delivery: the affixes are removed from the local part unless
5815 rcpt_include_affixes is set on the transport.
5817 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5818 timeout during the DATA phase of an incoming message.
5820 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5821 applied a patch sanctioned by the Sieve maintainer.
5823 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5824 the uid or gid is negative. A case of a negative gid caused this to be
5825 noticed. The fix allows for either to be negative.
5827 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5828 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5831 PH/36 The change PH/12 above was broken. Fixed it.
5833 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5834 the grounds that routing the same address twice would always produce the
5835 same answer. This might have been true once, but it is certainly no
5836 longer true now. Routing a child address may depend on the previous
5837 routing that produced that child. Some complicated redirection strategies
5838 went wrong when messages had multiple recipients, and made Exim's
5839 behaviour dependent on the order in which the addresses were given.
5841 I have moved the duplicate checking until after the routing is complete.
5842 Exim scans the addresses that are assigned to local and remote
5843 transports, and removes any duplicates. This means that more work will be
5844 done, as duplicates will always all be routed, but duplicates are
5845 presumably rare, so I don't expect this is of any significance.
5847 For deliveries to pipes, files, and autoreplies, the duplicate checking
5848 still happens during the routing process, since they are not going to be
5851 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5852 It corrects a timeout issue with spamd. This is Ian's comment: "The
5853 background is that sometimes spamd either never reads data from a
5854 connection it has accepted, or it never writes response data. The exiscan
5855 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5856 blindly assumes that writes won't block so it may never time out."
5858 PH/39 Allow G after quota size as well as K and M.
5860 PH/40 The value set for $authenticated_id in an authenticator may not contain
5861 binary zeroes or newlines because the value is written to log lines and
5862 to spool files. There was no check on this. Now the value is run through
5863 the string_printing() function so that such characters are converted to
5864 printable escape sequences.
5866 PH/41 $message_linecount is a new variable that contains the total number of
5867 lines in the message. Compare $body_linecount, which is the count for the
5870 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5871 bounce and delay warning messages, except in certain special cases, which
5874 (a) An SMTP error message from a remote host;
5875 (b) A message specified in a :fail: redirection;
5876 (c) A message specified in a "fail" command in a system filter;
5877 (d) A message specified in a FAIL return from the queryprogram router;
5878 (e) A message specified by the cannot_route_message router option.
5880 In these cases only, Exim does include the error details in bounce and
5881 warning messages. There are also a few cases where bland messages such
5882 as "unrouteable address" or "local delivery error" are given.
5884 PH/43 $value is now also set for the "else" part of a ${run expansion.
5886 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5887 being worked on, but at least Exim now implements the latest version to
5890 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5891 process to complete was treated as a delivery failure, a timeout while
5892 writing the message to the pipe was logged, but erroneously treated as a
5893 successful delivery. Such timeouts include transport filter timeouts. For
5894 consistency with the overall process timeout, these timeouts are now
5895 treated as errors, giving rise to delivery failures by default. However,
5896 there is now a new Boolean option for the pipe transport called
5897 timeout_defer, which, if set TRUE, converts the failures into defers for
5898 both kinds of timeout. A transport filter timeout is now identified in
5901 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5902 systems where "make" and "gmake" are different, calling "gmake" at top
5903 level broke things. I've arranged for the value of $(MAKE) to be passed
5904 from the Makefile to this script so that it can call the same version of
5908 A note about Exim versions 4.44 and 4.50
5909 ----------------------------------------
5911 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5912 changes of various kinds. As a consequence, a big documentation update was
5913 needed. This delayed the release for rather longer than seemed good, especially
5914 in the light of a couple of (minor) security issues. Therefore, the changes
5915 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5916 release. So 4.44 and 4.50 are in effect two different branches that both start
5919 I have left the 4.50 change log unchanged; it contains all the changes since
5920 4.43. The change log for 4.44 is below; many of its items are identical to
5921 those for 4.50. This seems to be the most sensible way to preserve the
5922 historical information.
5928 1. Minor wording change to the doc/README.SIEVE file.
5930 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5931 computation of the current number of files was incorrect.
5933 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5934 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5937 4. Give more explanation in the error message when the command for a transport
5938 filter fails to execute.
5940 5. There are several places where Exim runs a non-Exim command in a
5941 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5942 was being done only for the command run by the queryprogram router. It is
5943 now done for all such subprocesses. The other cases are: ${run, transport
5944 filters, and the commands run by the lmtp and pipe transports.
5946 6. Added CONFIGURE_GROUP build-time option.
5948 7. Some older OS have a limit of 256 on the maximum number of file
5949 descriptors. Exim was using setrlimit() to set 1000 as a large value
5950 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5951 systems. I've change it so that if it can't get 1000, it tries for 256.
5953 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5954 was an oversight, and furthermore, ever since the addition of extra
5955 controls (e.g. 4.43/32), the checks on when to allow different forms of
5956 "control" were broken. There should now be diagnostics for all cases when a
5957 control that does not make sense is encountered.
5959 9. Added the /retain_sender option to "control=submission".
5961 10. $recipients is now available in the predata ACL (oversight).
5963 11. Tidy the search cache before the fork to do a delivery from a message
5964 received from the command line. Otherwise the child will trigger a lookup
5965 failure and thereby defer the delivery if it tries to use (for example) a
5966 cached ldap connection that the parent has called unbind on.
5968 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5969 of $address_data from the recipient verification was clobbered by the
5970 sender verification.
5972 13. The value of address_data from a sender verification is now available in
5973 $sender_address_data in subsequent conditions in the ACL statement.
5975 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5977 15. Added a new option "connect=<time>" to callout options, to set a different
5980 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5981 was its contents. (It was OK if the option was not defined at all.)
5983 17. A "Completed" log line is now written for messages that are removed from
5984 the spool by the -Mrm option.
5986 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5987 information about exactly what failed.
5989 19. Added -dd to debug only the daemon process.
5991 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5992 handles timeouts, both on the server side and network timeouts. Renamed the
5993 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5995 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5996 It is now set to "smtps".
5998 22. $host_address is now set to the target address during the checking of
5999 ignore_target_hosts.
6001 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6002 being passed; this would have caused $sender_host_name to have been used if
6003 matching the list had actually called for a host name (not very likely,
6004 since this list is usually IP addresses). A host name is now passed as
6007 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6008 code that specifies a non-restarting handler (typically sigaction() in
6009 modern systems) in an attempt to fix a rare and obscure crash bug.
6011 25. Narrowed the window for a race in the daemon that could cause it to ignore
6012 SIGCHLD signals. This is not a major problem, because they are used only to
6013 wake it up if nothing else does.
6015 26. A malformed maildirsize file could cause Exim to calculate negative values
6016 for the mailbox size or file count. Odd effects could occur as a result.
6017 The maildirsize information is now recalculated if the size or filecount
6020 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6021 support for a long time. Removed HAVE_SYS_VFS_H.
6023 28. Installed the latest version of exipick from John Jetmore.
6025 29. In an address list, if the pattern was not a regular expression, an empty
6026 subject address (from a bounce message) matched only if the pattern was an
6027 empty string. Non-empty patterns were not even tested. This was the wrong
6028 because it is perfectly reasonable to use an empty address as part of a
6029 database query. An empty address is now tested by patterns that are
6030 lookups. However, all the other forms of pattern expect the subject to
6031 contain a local part and a domain, and therefore, for them, an empty
6032 address still always fails if the pattern is not itself empty.
6034 30. Exim went into a mad DNS loop when attempting to do a callout where the
6035 host was specified on an smtp transport, and looking it up yielded more
6036 than one IP address.
6038 31. Re-factored the code for checking spool and log partition space into a
6039 function that finds that data and another that does the check. The former
6040 is then used to implement four new variables: $spool_space, $log_space,
6041 $spool_inodes, and $log_inodes.
6043 32. The RFC2047 encoding function was originally intended for short strings
6044 such as real names; it was not keeping to the 75-character limit for
6045 encoded words that the RFC imposes. It now respects the limit, and
6046 generates multiple encoded words if necessary. To be on the safe side, I
6047 have increased the buffer size for the ${rfc2047: expansion operator from
6050 33. It is now permitted to omit both strings after an "if" condition; if the
6051 condition is true, the result is "true". As before, when the second string
6052 is omitted, a false condition yields an empty string. This makes it less
6053 cumbersome to write custom ACL and router conditions.
6055 34. Failure to deliver a bounce message always caused it to be frozen, even if
6056 there was an errors_to setting on the router. The errors_to setting is now
6059 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6060 canonical form (fully expanded) before being placed in
6061 $sender_host_address.
6063 36. The table in the code that translates DNS record types into text (T_A to
6064 "A" for instance) was missing entries for NS and CNAME. It is just possible
6065 that this could have caused confusion if both these types were looked up
6066 for the same domain, because the text type is used as part of Exim's
6067 per-process caching. But the chance of anyone hitting this buglet seems
6070 37. The dnsdb lookup has been extended in a number of ways.
6072 (1) There is a new type, "zns", which walks up the domain tree until it
6073 finds some nameserver records. It should be used with care.
6075 (2) There is a new type, "mxh", which is like "mx" except that it returns
6076 just the host names, not the priorities.
6078 (3) It is now possible to give a list of domains (or IP addresses) to be
6079 looked up. The behaviour when one of the lookups defers can be
6080 controlled by a keyword.
6082 (4) It is now possible to specify the separator character for use when
6083 multiple records are returned.
6085 38. The dnslists ACL condition has been extended: it is now possible to supply
6086 a list of IP addresses and/or domains to be looked up in a particular DNS
6089 39. Added log_selector=+queue_time_overall.
6091 40. When running the queue in the test harness, wait just a tad after forking a
6092 delivery process, to get repeatability of debugging output.
6094 41. Include certificate and key file names in error message when GnuTLS fails
6095 to set them up, because the GnuTLS error message doesn't include the name
6096 of the failing file when there is a problem reading it.
6098 42. Allow both -bf and -bF in the same test run.
6100 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6102 44. The "Exiscan patch" is now merged into the mainline Exim source.
6104 45. Sometimes the final signoff response after QUIT could fail to get
6105 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6106 < 0 before doing a fflush(). This bug looks as though it goes back to the
6107 introduction of TLS in release 3.20, but "sometimes" must have been rare
6108 because the tests only now provoked it.
6110 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6111 (this can affect the format of dates).
6113 47. exim_tidydb, when checking for the continued existence of a message for
6114 which it has found a message-specific retry record, was not finding
6115 messages that were in split spool directories. Consequently, it was
6116 deleting retry records that should have stayed in existence.
6118 48. Steve fixed some bugs in eximstats.
6120 49. The SPA authentication driver was not abandoning authentication and moving
6121 on to the next authenticator when an expansion was forced to fail,
6122 contradicting the general specification for all authenticators. Instead it
6123 was generating a temporary error. It now behaves as specified.
6125 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6126 (the order specifies the preference for clients). The order is now AES256,
6127 AES128, 3DES, ARCFOUR128.
6129 51. Small patch to Sieve code - explicitly set From: when generating an
6132 52. Exim crashed if a remote delivery caused a very long error message to be
6133 recorded - for instance if somebody sent an entire SpamAssassin report back
6134 as a large number of 550 error lines. This bug was coincidentally fixed by
6135 increasing the size of one of Exim's internal buffers (big_buffer) that
6136 happened as part of the Exiscan merge. However, to be on the safe side, I
6137 have made the code more robust (and fixed the comments that describe what
6140 53. Now that there can be additional text after "Completed" in log lines (if
6141 the queue_time_overall log selector is set), a one-byte patch to exigrep
6142 was needed to allow it to recognize "Completed" as not the last thing in
6145 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6146 patch that reportedly fixes this has been added. I am not expert enough to
6147 create a test for it. This is what the patch creator wrote:
6149 "I found a little strange behaviour of ldap code when working with
6150 Windows 2003 AD Domain, where users was placed in more than one
6151 Organization Units. When I tried to give exim partial DN, the exit code
6152 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6153 But simultaneously result of request was absolutely normal ldap result,
6154 so I produce this patch..."
6156 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6157 REFERENCE, so I have modified the code to exclude the patch when that macro
6160 55. Some experimental protocols are using DNS PTR records for new purposes. The
6161 keys for these records are domain names, not reversed IP addresses. The
6162 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6163 leaves it alone. Component reversal etc. now happens only for IP addresses.
6166 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6168 57. Double the size of the debug message buffer (to 2048) so that more of very
6169 long debug lines gets shown.
6171 58. The exicyclog utility now does better if the number of log files to keep
6172 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6174 59. Two changes related to the smtp_active_hostname option:
6176 (1) $smtp_active_hostname is now available as a variable.
6177 (2) The default for smtp_banner uses $smtp_active_hostname instead
6178 of $primary_hostname.
6180 60. The host_aton() function is supposed to be passed a string that is known
6181 to be a valid IP address. However, in the case of IPv6 addresses, it was
6182 not checking this. This is a hostage to fortune. Exim now panics and dies
6183 if the condition is not met. A case was found where this could be provoked
6184 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6185 components; fortuitously, this particular loophole had already been fixed
6186 by change 4.50/55 above.
6188 If there are any other similar loopholes, the new check in host_aton()
6189 itself should stop them being exploited. The report I received stated that
6190 data on the command line could provoke the exploit when Exim was running as
6191 exim, but did not say which command line option was involved. All I could
6192 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6193 running as the user.
6196 61. There was a buffer overflow vulnerability in the SPA authentication code
6197 (which came originally from the Samba project). I have added a test to the
6198 spa_base64_to_bits() function which I hope fixes it.
6201 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6202 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6204 63. The daemon start-up calls getloadavg() while still root for those OS that
6205 need the first call to be done as root, but it missed one case: when
6206 deliver_queue_load_max is set with deliver_drop_privilege. This is
6207 necessary for the benefit of the queue runner, because there is no re-exec
6208 when deliver_drop_privilege is set.
6210 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6211 This has been fixed.
6213 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6214 was in use, was not putting the data itself into the right store pool;
6215 consequently, it could be overwritten for a subsequent message in the same
6216 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6219 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6221 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6222 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6223 the function were treating the return as a boolean value, which happened to
6224 work because 0=false and not-0=true, but is not correct code.
6226 68. The host_aton() function was not handling scoped IPv6 addresses (those
6227 with, for example, "%eth0" on the end) correctly.
6229 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6230 negated items (that is, ~something) in unsigned ints. Some compilers
6231 apparently mutter when there is no cast.
6233 70. If an address verification called from an ACL failed, and did not produce a
6234 user-specific message (i.e. there was only a "system" message), nothing was
6235 put in $acl_verify_message. In this situation, it now puts the system
6238 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6239 session; change 4.31/43 added the unwanted input to the log line - except
6240 that it did not do this in the start of session case. It now does.
6242 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6243 This is wrong and can cause the other end to generate a synchronization
6244 error if it is another Exim or anything else that does the synchronization
6245 check. A QUIT command is no longer sent after a timeout.
6247 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6248 during host lookups.
6250 74. The defer_ok option of callout verification was not working if it was used
6251 when verifying addresses in header lines, that is, for this case:
6253 verify = header_sender/callout=defer_ok
6255 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6256 those file descriptors could be used for SMTP connections. If anything
6257 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6258 could be sent to the SMTP client, causing chaos. The daemon now opens
6259 stdin, stdout, and stderr to /dev/null when it puts itself into the
6262 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6263 log by default. The user can override this with suitable Perl magic.
6265 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6266 the log message when discard triggers, was not working for the DATA ACL or
6267 for the non-SMTP ACL.
6269 78. Error message wording change in sieve.c.
6271 79. If smtp_accept_max_per_host was set, the number of connections could be
6272 restricted to fewer than expected, because the daemon was trying to set up
6273 a new connection before checking whether the processes handling previous
6274 connections had finished. The check for completed processes is now done
6275 earlier. On busy systems, this bug wouldn't be noticed because something
6276 else would have woken the daemon, and it would have reaped the completed
6279 80. If a message was submitted locally by a user whose login name contained one
6280 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6281 It caused a spool format error. I have fixed the spool reading code. A
6282 related problem was that the "from" clause in the Received: line became
6283 illegal because of the space(s). It is now covered by ${quote_local_part.
6285 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6288 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6290 83. Patch from Sieve maintainer to fix unterminated string problem in
6291 "vacation" handling.
6293 84. Some minor changes to the Linux configuration files to help with other
6294 OS variants using glibc.
6296 85. One more patch for Sieve to update vacation handling to latest spec.
6299 ----------------------------------------------------
6300 See the note above about the 4.44 and 4.50 releases.
6301 ----------------------------------------------------
6307 1. Change 4.43/35 introduced a bug that caused file counts to be
6308 incorrectly computed when quota_filecount was set in an appendfile
6311 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6312 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6315 3. Give more explanation in the error message when the command for a transport
6316 filter fails to execute.
6318 4. There are several places where Exim runs a non-Exim command in a
6319 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6320 was being done only for the command run by the queryprogram router. It is
6321 now done for all such subprocesses. The other cases are: ${run, transport
6322 filters, and the commands run by the lmtp and pipe transports.
6324 5. Some older OS have a limit of 256 on the maximum number of file
6325 descriptors. Exim was using setrlimit() to set 1000 as a large value
6326 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6327 systems. I've change it so that if it can't get 1000, it tries for 256.
6329 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6330 was an oversight, and furthermore, ever since the addition of extra
6331 controls (e.g. 4.43/32), the checks on when to allow different forms of
6332 "control" were broken. There should now be diagnostics for all cases when a
6333 control that does not make sense is encountered.
6335 7. $recipients is now available in the predata ACL (oversight).
6337 8. Tidy the search cache before the fork to do a delivery from a message
6338 received from the command line. Otherwise the child will trigger a lookup
6339 failure and thereby defer the delivery if it tries to use (for example) a
6340 cached ldap connection that the parent has called unbind on.
6342 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6343 of $address_data from the recipient verification was clobbered by the
6344 sender verification.
6346 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6347 was its contents. (It was OK if the option was not defined at all.)
6349 11. A "Completed" log line is now written for messages that are removed from
6350 the spool by the -Mrm option.
6352 12. $host_address is now set to the target address during the checking of
6353 ignore_target_hosts.
6355 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6356 being passed; this would have caused $sender_host_name to have been used if
6357 matching the list had actually called for a host name (not very likely,
6358 since this list is usually IP addresses). A host name is now passed as
6361 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6362 code that specifies a non-restarting handler (typically sigaction() in
6363 modern systems) in an attempt to fix a rare and obscure crash bug.
6365 15. Narrowed the window for a race in the daemon that could cause it to ignore
6366 SIGCHLD signals. This is not a major problem, because they are used only to
6367 wake it up if nothing else does.
6369 16. A malformed maildirsize file could cause Exim to calculate negative values
6370 for the mailbox size or file count. Odd effects could occur as a result.
6371 The maildirsize information is now recalculated if the size or filecount
6374 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6375 support for a long time. Removed HAVE_SYS_VFS_H.
6377 18. Updated exipick to current release from John Jetmore.
6379 19. Allow an empty sender to be matched against a lookup in an address list.
6380 Previously the only cases considered were a regular expression, or an
6383 20. Exim went into a mad DNS lookup loop when doing a callout where the
6384 host was specified on the transport, if the DNS lookup yielded more than
6387 21. The RFC2047 encoding function was originally intended for short strings
6388 such as real names; it was not keeping to the 75-character limit for
6389 encoded words that the RFC imposes. It now respects the limit, and
6390 generates multiple encoded words if necessary. To be on the safe side, I
6391 have increased the buffer size for the ${rfc2047: expansion operator from
6394 22. Failure to deliver a bounce message always caused it to be frozen, even if
6395 there was an errors_to setting on the router. The errors_to setting is now
6398 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6399 canonical form (fully expanded) before being placed in
6400 $sender_host_address.
6402 24. Updated eximstats to version 1.33
6404 25. Include certificate and key file names in error message when GnuTLS fails
6405 to set them up, because the GnuTLS error message doesn't include the name
6406 of the failing file when there is a problem reading it.
6408 26. Expand error message when OpenSSL has problems setting up cert/key files.
6411 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6412 (this can affect the format of dates).
6414 28. exim_tidydb, when checking for the continued existence of a message for
6415 which it has found a message-specific retry record, was not finding
6416 messages that were in split spool directories. Consequently, it was
6417 deleting retry records that should have stayed in existence.
6419 29. eximstats updated to version 1.35
6420 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6421 1.35 - bugfix such that pie charts by volume are generated correctly
6423 30. The SPA authentication driver was not abandoning authentication and moving
6424 on to the next authenticator when an expansion was forced to fail,
6425 contradicting the general specification for all authenticators. Instead it
6426 was generating a temporary error. It now behaves as specified.
6428 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6429 (the order specifies the preference for clients). The order is now AES256,
6430 AES128, 3DES, ARCFOUR128.
6432 31. Small patch to Sieve code - explicitly set From: when generating an
6435 32. Exim crashed if a remote delivery caused a very long error message to be
6436 recorded - for instance if somebody sent an entire SpamAssassin report back
6437 as a large number of 550 error lines. This bug was coincidentally fixed by
6438 increasing the size of one of Exim's internal buffers (big_buffer) that
6439 happened as part of the Exiscan merge. However, to be on the safe side, I
6440 have made the code more robust (and fixed the comments that describe what
6443 33. Some experimental protocols are using DNS PTR records for new purposes. The
6444 keys for these records are domain names, not reversed IP addresses. The
6445 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6446 leaves it alone. Component reversal etc. now happens only for IP addresses.
6449 34. The host_aton() function is supposed to be passed a string that is known
6450 to be a valid IP address. However, in the case of IPv6 addresses, it was
6451 not checking this. This is a hostage to fortune. Exim now panics and dies
6452 if the condition is not met. A case was found where this could be provoked
6453 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6454 components; fortuitously, this particular loophole had already been fixed
6455 by change 4.50/55 or 4.44/33 above.
6457 If there are any other similar loopholes, the new check in host_aton()
6458 itself should stop them being exploited. The report I received stated that
6459 data on the command line could provoke the exploit when Exim was running as
6460 exim, but did not say which command line option was involved. All I could
6461 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6462 running as the user.
6465 35. There was a buffer overflow vulnerability in the SPA authentication code
6466 (which came originally from the Samba project). I have added a test to the
6467 spa_base64_to_bits() function which I hope fixes it.
6470 36. The daemon start-up calls getloadavg() while still root for those OS that
6471 need the first call to be done as root, but it missed one case: when
6472 deliver_queue_load_max is set with deliver_drop_privilege. This is
6473 necessary for the benefit of the queue runner, because there is no re-exec
6474 when deliver_drop_privilege is set.
6476 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6477 was in use, was not putting the data itself into the right store pool;
6478 consequently, it could be overwritten for a subsequent message in the same
6479 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6482 38. Sometimes the final signoff response after QUIT could fail to get
6483 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6484 < 0 before doing a fflush(). This bug looks as though it goes back to the
6485 introduction of TLS in release 3.20, but "sometimes" must have been rare
6486 because the tests only now provoked it.
6492 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6493 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6494 Now it returns an integer. A number of calls were still expecting a T/F
6495 return. Fortuitously, in all cases, the tests worked in OK situations,
6496 which is the norm. However, things would have gone wrong on any write
6497 failures on the smtp file descriptor. This function is used when sending
6498 messages over SMTP and also when doing verify callouts.
6500 2. When Exim is called to do synchronous delivery of a locally submitted
6501 message (the -odf or -odi options), it no longer closes stderr before doing
6504 3. Implemented the mua_wrapper option.
6506 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6508 5. Implemented the functions header_remove(), header_testname(),
6509 header_add_at_position(), and receive_remove_recipient(), and exported them
6512 6. If an ACL "warn" statement specified the addition of headers, Exim already
6513 inserted X-ACL-Warn: at the start if there was no header name. However, it
6514 was not making this test for the second and subsequent header lines if
6515 there were newlines in the string. This meant that an invalid header could
6516 be inserted if Exim was badly configured.
6518 7. Allow an ACL "warn" statement to add header lines at the start or after all
6519 the Received: headers, as well as at the end.
6521 8. Added the rcpt_4xx retry error code.
6523 9. Added postmaster_mailfrom=xxx to callout verification option.
6525 10. Added mailfrom=xxxx to the callout verification option, for verify=
6528 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6529 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6531 12. Inserted some casts to stop certain compilers warning when using pointer
6532 differences as field lengths or precisions in printf-type calls (mostly
6533 affecting debugging statements).
6535 13. Added optional readline() support for -be (dynamically loaded).
6537 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6538 same clock tick as a message's arrival, so that its received time was the
6539 same as the "first fail" time on the retry record, and that message
6540 remained on the queue past the ultimate address timeout, every queue runner
6541 would try a delivery (because it was past the ultimate address timeout) but
6542 after another failure, the ultimate address timeout, which should have then
6543 bounced the address, did not kick in. This was a "< instead of <=" error;
6544 in most cases the first failure would have been in the next clock tick
6545 after the received time, and all would be well.
6547 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6548 being recognized when the domain list was tested by the match_domain
6549 condition in an expansion string.
6551 16. Added the ${str2b64: operator.
6553 17. Exim was always calling setrlimit() to set a large limit for the number of
6554 processes, without checking whether the existing limit was already
6555 adequate. (It did check for the limit on file descriptors.) Furthermore,
6556 errors from getrlimit() and setrlimit() were being ignored. Now they are
6557 logged to the main and panic logs, but Exim does carry on, to try to do its
6558 job under whatever limits there are.
6560 18. Imported PCRE 5.0.
6562 19. Trivial typo in log message " temporarily refused connection" (the leading
6565 20. If the log selector return_path_on_delivery was set and an address was
6566 redirected to /dev/null, the delivery process crashed because it assumed
6567 that a return path would always be set for a "successful" delivery. In this
6568 case, the whole delivery is bypassed as an optimization, and therefore no
6571 21. Internal re-arrangement: the function for sending a challenge and reading
6572 a response while authentication was assuming a zero-terminated challenge
6573 string. It's now changed to take a pointer and a length, to allow for
6574 binary data in such strings.
6576 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6578 23. Exim was not respecting finduser_retries when seeking the login of the
6579 uid under which it was called; it was always trying 10 times. (The default
6580 setting of finduser_retries is zero.) Also, it was sleeping after the final
6581 failure, which is pointless.
6583 24. Implemented tls_on_connect_ports.
6585 25. Implemented acl_smtp_predata.
6587 26. If the domain in control=submission is set empty, Exim assumes that the
6588 authenticated id is a complete email address when it generates From: or
6589 Sender: header lines.
6591 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6592 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6593 chown and chgrp in /bin and hostname in /usr/bin.
6595 28. Exim was keeping the "process log" file open after each use, just as it
6596 does for the main log. This opens the possibility of it remaining open for
6597 long periods when the USR1 signal hits a daemon. Occasional processlog
6598 errors were reported, that could have been caused by this. Anyway, it seems
6599 much more sensible not to leave this file open at all, so that is what now
6602 29. The long-running daemon process does not normally write to the log once it
6603 has entered its main loop, and it closes the log before doing so. This is
6604 so that log files can straightforwardly be renamed and moved. However,
6605 there are a couple of unusual error situations where the daemon does write
6606 log entries, and I had neglected to close the log afterwards.
6608 30. The text of an SMTP error response that was received during a remote
6609 delivery was being truncated at 512 bytes. This is too short for some of
6610 the long messages that one sometimes sees. I've increased the limit to
6613 31. It is now possible to make retry rules that apply only when a message has a
6614 specific sender, in particular, an empty sender.
6616 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6617 it possible to be selective about when SMTP synchronization is enforced.
6619 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6621 32. Implemented hosts_connection_nolog.
6623 33. Added an ACL for QUIT.
6625 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6628 35. Added mailbox_size and mailbox_filecount to appendfile.
6630 36. Added control = no_multiline_responses to ACLs.
6632 37. There was a bug in the logic of the code that waits for the clock to tick
6633 in the case where the clock went backwards by a substantial amount such
6634 that the microsecond fraction of "now" was more than the microsecond
6635 fraction of "then" (but the whole seconds number was less).
6637 38. Added support for the libradius Radius client library this is found on
6638 FreeBSD (previously only the radiusclient library was supported).
6644 1. When certain lookups returned multiple values in the form name=value, the
6645 quoting of the values was not always being done properly. Specifically:
6646 (a) If the value started with a double quote, but contained no whitespace,
6648 (b) If the value contained whitespace other than a space character (i.e.
6649 tabs or newlines or carriage returns) it was not quoted.
6650 This fix has been applied to the mysql and pgsql lookups by writing a
6651 separate quoting function and calling it from the lookup code. The fix
6652 should probably also be applied to nisplus, ibase and oracle lookups, but
6653 since I cannot test any of those, I have not disturbed their existing code.
6655 2. A hit in the callout cache for a specific address caused a log line with no
6656 reason for rejecting RCPT. Now it says "Previous (cached) callout
6657 verification failure".
6659 3. There was an off-by-one bug in the queryprogram router. An over-long
6660 return line was truncated at 256 instead of 255 characters, thereby
6661 overflowing its buffer with the terminating zero. As well as fixing this, I
6662 have increased the buffer size to 1024 (and made a note to document this).
6664 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6665 when Exim is waiting for an SMTP response from a remote server, Exim
6666 restarts its select() call on the socket, thereby resetting its timeout.
6667 This is not a problem when such interrupts are rare. Somebody set up a cron
6668 job to run exiwhat every 2 minutes, which is less than the normal select()
6669 timeout (5 or 10 minutes). This meant that the select() timeout never
6670 kicked in because it was always reset. I have fixed this by comparing the
6671 time when an interrupt arrives with the time at the start of the first call
6672 to select(). If more time than the timeout has elapsed, the interrupt is
6673 treated as a timeout.
6675 5. Some internal re-factoring in preparation for the addition of Sieve
6676 extensions (by MH). In particular, the "personal" test is moved to a
6677 separate function, and given an option for scanning Cc: and Bcc: (which is
6678 not set for Exim filters).
6680 6. When Exim created an email address using the login of the caller as the
6681 local part (e.g. when creating a From: or Sender: header line), it was not
6682 quoting the local part when it contained special characters such as @.
6684 7. Installed new OpenBSD configuration files.
6686 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6687 try to make them clearer.
6689 9. Callout options, other than the timeout value, were being ignored when
6690 verifying sender addresses in header lines. For example, when using
6692 verify = header_sender/callout=no_cache
6694 the cache was (incorrectly) being used.
6696 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6697 only the Cygwin environment.
6699 11. When return_path_on_delivery was set as a log selector, if different remote
6700 addresses in the same message used different return paths and parallel
6701 remote delivery occurred, the wrong values would sometimes be logged.
6702 (Whenever a remote delivery process finished, the return path value from
6703 the most recently started remote delivery process was logged.)
6705 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6706 lines when AUTH and/or TLS are in use. This is the "received protocol"
6707 field. Exim used to use "asmtp" for authenticated SMTP, without any
6708 indication (in the protocol name) for TLS use. Now it follows the RFC and
6709 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6710 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6711 names appear in log lines as well as in Received: header lines.
6713 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6714 extensions, and comparison tests, and to fix some bugs.
6716 14. Changes to the "personal" filter test:
6718 (1) The test was buggy in that it was just doing the equivalent of
6719 "contains" tests on header lines. For example, if a user's address was
6720 anne@some.where, the "personal" test would incorrectly be true for
6722 To: susanne@some.where
6724 This test is now done by extracting each address from the header in turn,
6725 and checking the entire address. Other tests that are part of "personal"
6726 are now done using regular expressions (for example, to check local parts
6727 of addresses in From: header lines).
6729 (2) The list of non-personal local parts in From: addresses has been
6730 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6731 taken from the Sieve specification recommendations.
6733 (3) If the message contains any header line starting with "List-" it is
6734 treated as non-personal.
6736 (4) The test for "circular" in the Subject: header line has been removed
6737 because it now seems ill-conceived.
6739 15. Minor typos in src/EDITME comments corrected.
6741 16. Installed latest exipick from John Jetmore.
6743 17. If headers_add on a router specified a text string that was too long for
6744 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6745 of string_sprintf() is now avoided.
6747 18. $message_body_size was not set (it was always zero) when running the DATA
6748 ACL and the local_scan() function.
6750 19. For the "mail" command in an Exim filter, no default was being set for
6751 the once_repeat time, causing a random time value to be used if "once" was
6752 specified. (If the value happened to be <= 0, no repeat happened.) The
6753 default is now 0s, meaning "never repeat". The "vacation" command was OK
6754 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6755 (I found it when inspecting the code).
6757 20. There is now an overall timeout for performing a callout verification. It
6758 defaults to 4 times the callout timeout, which applies to individual SMTP
6759 commands during the callout. The overall timeout applies when there is more
6760 than one host that can be tried. The timeout is checked before trying the
6761 next host. This prevents very long delays if there are a large number of
6762 hosts and all are timing out (e.g. when the network connections are timing
6763 out). The value of the overall timeout can be changed by specifying an
6764 additional sub-option for "callout", called "maxwait". For example:
6766 verify = sender/callout=5s,maxwait=20s
6768 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6769 to the end before writing, but this should make it even safer).
6771 22. Exim was forgetting that it had advertised PIPELINING for the second and
6772 subsequent messages on an SMTP connection. It was also not resetting its
6773 memory on STARTTLS and an internal HELO.
6775 23. When Exim logs an SMTP synchronization error within a session, it now
6776 records whether PIPELINING has been advertised or not.
6778 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6779 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6780 rather than long int.
6782 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6784 26. Added the never_mail option to autoreply.
6790 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6791 crash if the getsockname() call failed; this can happen if a connection is
6792 closed very soon after it is established. The problem was simply in the
6793 order in which certain operations were done, causing Exim to try to write
6794 to the SMTP stream before it had set up the file descriptor. The bug has
6795 been fixed by making things happen in the correct order.
6801 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6802 before the connection was closed, thus losing the rejection response.
6804 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6805 some early Solaris releases, but causes trouble in current releases where
6806 socklen_t is defined.
6808 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6811 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6814 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6815 that releasing the top of it at the end releases what was used for sub-
6816 expansions (unless the block got too big). However, discard this block if
6817 the first thing is a variable or header, so that we can use its block when
6818 it is dynamic (useful for very large $message_headers, for example).
6820 6. Lookups now cache *every* query, not just the most recent. A new, separate
6821 store pool is used for this. It can be recovered when all lookup caches are
6822 flushed. Lookups now release memory at the end of their result strings.
6823 This has involved some general refactoring of the lookup sources.
6825 7. Some code has been added to the store_xxx() functions to reduce the amount
6826 of flapping under certain conditions.
6828 8. log_incoming_interface used to affect only the <= reception log lines. Now
6829 it causes the local interface and port to be added to several more SMTP log
6830 lines, for example "SMTP connection from", and rejection lines.
6832 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6834 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6836 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6837 could be overwritten at the end of the current message (or the start of a
6838 new message if it was set in a HELO ACL). The value is now preserved for
6839 the duration of the SMTP connection.
6841 12. If a transport had a headers_rewrite setting, and a matching header line
6842 contained an unqualified address, that address was qualified, even if it
6843 did not match any rewriting rules. The underlying bug was that the values
6844 of the flags that permit the existence of unqualified sender and recipient
6845 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6846 non-local messages, and by -bnq for local messages) were not being
6847 preserved with the message after it was received.
6849 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6850 "next input=" as part of the text comprising the host identity instead of
6851 the correct text. The code was using the same buffer for two different
6852 strings. However, depending on which order the printing function evaluated
6853 its arguments, the bug did not always show up. Under Linux, for example, my
6854 test suite worked just fine.
6856 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6857 doesn't work with some older versions of Perl. It has been changed to "my",
6858 which in any case is probably the better facility to use.
6860 15. A really picky compiler found some instances of statements for creating
6861 error messages that either had too many or two few arguments for the format
6864 16. The size of the buffer for calls to the DNS resolver has been increased
6865 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6866 for addresses that have a lot of PTR records. This alleviates a problem; it
6867 does not fully solve it.
6869 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6870 buffer now truncates the list and logs the incident, which is the same
6871 action as happens when Exim is looking up a host name and its aliases.
6872 Previously in this situation something unpredictable would happen;
6873 sometimes it was "internal error: store_reset failed".
6875 18. If a server dropped the connection unexpectedly when an Exim client was
6876 using GnuTLS and trying to read a response, the client delivery process
6877 crashed while trying to generate an error log message.
6879 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6880 string, for example:
6882 warn message = H1: something\nH2: something
6884 the text was added as a single header line from Exim's point of view
6885 though it ended up OK in the delivered message. However, searching for the
6886 second and subsequent header lines using $h_h2: did not work. This has been
6887 fixed. Similarly, if a system filter added multiple headers in this way,
6888 the routers could not see them.
6890 20. Expanded the error message when iplsearch is called with an invalid key to
6891 suggest using net-iplsearch in a host list.
6893 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6894 ACLs are no longer actually imposed (and a message to that effect is
6897 22. If a "gecos" field in a passwd entry contained escaped characters, in
6898 particular, if it contained a \" sequence, Exim got it wrong when building
6899 a From: or a Sender: header from that name. A second bug also caused
6900 incorrect handling when an unquoted " was present following a character
6901 that needed quoting.
6903 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6904 was not being matched caselessly.
6906 24. Arranged for all hyphens in the exim.8 source to be escaped with
6909 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6910 properly. Recipient callout cache records were still being keyed to include
6911 the sender, even when use_sender was set false. This led to far more
6912 callouts that were necessary. The sender is no longer included in the key
6913 when use_sender is false.
6915 26. Added "control = submission" modifier to ACLs.
6917 27. Added the ${base62d: operator to decode base 62 numbers.
6919 28. dnsdb lookups can now access SRV records.
6921 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6922 the configuration file.
6924 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6925 (-v) mode. This makes the output for a verbose queue run more intelligible.
6927 31. Added a use_postmaster feature to recipient callouts.
6929 32. Added the $body_zerocount variable, containing the number of binary zero
6930 bytes in the message body.
6932 33. The time of last modification of the "new" subdirectory is now used as the
6933 "mailbox time last read" when there is a quota error for a maildir
6936 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6938 35. Added +ignore_unknown as a special item in host lists.
6940 36. Code for decoding IPv6 addresses in host lists is now included, even if
6941 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6942 address was recognized as an IP address, but was then not correctly decoded
6943 into binary, causing unexpected and incorrect effects when compared with
6950 1. Very minor rewording of debugging text in manualroute to say "list of
6951 hosts" instead of "hostlist".
6953 2. If verify=header_syntax was set, and a header line with an unqualified
6954 address (no domain) and a large number of spaces between the end of the
6955 name and the colon was received, the reception process suffered a buffer
6956 overflow, and (when I tested it) crashed. This was caused by some obsolete
6957 code that should have been removed. The fix is to remove it!
6959 3. When running in the test harness, delay a bit after writing a bounce
6960 message to get a bit more predictability in the log output.
6962 4. Added a call to search_tidyup() just before forking a reception process. In
6963 theory, someone could use a lookup in the expansion of smtp_accept_max_
6964 per_host which, without the tidyup, could leave open a database connection.
6966 5. Added the variables $recipient_data and $sender_data which get set from a
6967 lookup success in an ACL "recipients" or "senders" condition, or a router
6968 "senders" option, similar to $domain_data and $local_part_data.
6970 6. Moved the writing of debug_print from before to after the "senders" test
6973 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6974 problems for message scanning, either using a data ACL, or using
6975 local_scan() because the Received: header was not generated till after they
6976 were called (in order to set the time as the time of reception completion).
6977 I have revised the way this works. The header is now generated after the
6978 body is received, but before the ACL or local_scan() are called. After they
6979 are run, the timestamp in the header is updated.
6985 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6986 before starting a queue runner without re-exec. This happened only when
6987 deliver_drop_privilege was set or when the Exim user was set to root. The
6988 effect of the bug was that timeouts during subsequent deliveries caused
6989 crashes instead of being properly handled. The handler is now left at its
6990 default (and expected) setting.
6992 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6993 message, again when deliver_drop_privilege is set or Exim is run as root.
6994 The bug described in (1) was not present in this case, but the tidying up
6995 of the other signals was missing. I have made the two cases consistent.
6997 3. The ignore_target_hosts setting on a manualroute router was being ignored
6998 for hosts that were looked up using the /MX notation.
7000 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7003 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7004 operated on the sender address. After changing the $sender_address to <>
7005 for the sender address verify, Exim was re-instated it as the original
7006 (before rewriting) address, but remembering that it had rewritten it, so it
7007 wasn't rewriting it again. This bug also had the effect of breaking the
7008 sender address verification caching when the sender address was rewritten.
7010 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7011 This has been changed so that if the ip literal address matches
7012 ignore_target_hosts, the router declines.
7014 7. Added expansion conditions match_domain, match_address, and match_local_
7015 part (NOT match_host).
7017 8. The placeholder for the Received: header didn't have a length field set.
7019 9. Added code to Exim itself and to exim_lock to test for a specific race
7020 condition that could lead to file corruption when using MBX delivery. The
7021 issue is with the lockfile that is created in /tmp. If this file is removed
7022 after a process has opened it but before that process has acquired a lock,
7023 there is the potential for a second process to recreate the file and also
7024 acquire a lock. This could lead to two Exim processes writing to the file
7025 at the same time. The added code performs the same test as UW imapd; it
7026 checks after acquiring the lock that its file descriptor still refers to
7027 the same named file.
7029 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7030 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7033 11. Added the smtp_active_hostname option. If used, this will typically be made
7034 to depend on the incoming interface address. Because $interface_address is
7035 not set up until the daemon has forked a reception process, error responses
7036 that can happen earlier (such as "too many connections") no longer contain
7039 12. If an expansion in a condition on a "warn" statement fails because a lookup
7040 defers, the "warn" statement is abandoned, and the next ACL statement is
7041 processed. Previously this caused the whole ACL to be aborted.
7043 13. Added the iplsearch lookup type.
7045 14. Added ident_timeout as a log selector.
7047 15. Added tls_certificate_verified as a log selector.
7049 16. Added a global option tls_require_ciphers (compare the smtp transport
7050 option of the same name). This controls incoming TLS connections.
7052 17. I finally figured out how to make tls_require_ciphers do a similar thing
7053 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7054 before starting the TLS session.
7056 18. Tabs are now shown as \t in -bP output.
7058 19. If the log selector return_path_on_delivery was set, Exim crashed when
7059 bouncing a message because it had too many Received: header lines.
7061 20. If two routers both had headers_remove settings, and the first one included
7062 a superfluous trailing colon, the final name in the first list and the
7063 first name in the second list were incorrectly joined into one item (with a
7064 colon in the middle).
7070 1. Added -C and -D options to the exinext utility, mainly to make it easier
7071 to include in the automated testing, but these could be helpful when
7072 multiple configurations are in use.
7074 2. The exinext utility was not formatting the output nicely when there was
7075 an alternate port involved in the retry record key, nor when there was a
7076 message id as well (for retries that were specific to a specific message
7077 and a specific host). It was also confused by IPv6 addresses, because of
7078 the additional colons they contain. I have fixed the IPv4 problem, and
7079 patched it up to do a reasonable job for IPv6.
7081 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7082 delivery, the log line now contains "pipelined" if PIPELINING was used.
7084 4. An SMTP transport process used to panic and die if the bind() call to set
7085 an explicit outgoing interface failed. This has been changed; it is now
7086 treated in the same way as a connect() failure.
7088 5. A reference to $sender_host_name in the part of a conditional expansion
7089 that was being skipped was still causing a DNS lookup. This no longer
7092 6. The def: expansion condition was not recognizing references to header lines
7093 that used bh_ and bheader_.
7095 7. Added the _cache feature to named lists.
7097 8. The code for checking quota_filecount in the appendfile transport was
7098 allowing one more file than it should have been.
7100 9. For compatibility with Sendmail, the command line option
7108 and sets the incoming protocol and host name (for trusted callers). The
7109 host name and its colon can be omitted when only the protocol is to be set.
7110 Note the Exim already has two private options, -pd and -ps, that refer to
7111 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7112 "s", but I don't think that's a major issue.
7114 10. A number of refactoring changes to the code, none of which should affect
7117 (a) The number of logging options was getting close to filling up the
7118 32-bit word that was used as a bit map. I have split them into two classes:
7119 those that are passed in the argument to log_write(), and those that are
7120 only ever tested independently outside of that function. These are now in
7121 separate 32-bit words, so there is plenty of room for expansion again.
7122 There is no change in the user interface or the logging behaviour.
7124 (b) When building, for example, log lines, the code previously used a
7125 macro that called string_cat() twice, in order to add two strings. This is
7126 not really sufficiently general. Furthermore, there was one instance where
7127 it was actually wrong because one of the argument was used twice, and in
7128 one call a function was used. (As it happened, calling the function twice
7129 did not affect the overall behaviour.) The macro has been replaced by a
7130 function that can join an arbitrary number of extra strings onto a growing
7133 (c) The code for expansion conditions now uses a table and a binary chop
7134 instead of a serial search (which was left over from when there were very
7135 few conditions). Also, it now recognizes conditions like "pam" even when
7136 the relevant support is not compiled in: a suitably worded error message is
7137 given if an attempt is made to use such a condition.
7139 11. Added ${time_interval:xxxxx}.
7141 12. A bug was causing one of the ddress fields not to be passed back correctly
7142 from remote delivery subprocesses. The field in question was not being
7143 subsequently used, so this caused to problems in practice.
7145 13. Added new log selectors queue_time and deliver_time.
7147 14. Might have fixed a bug in maildirsizefile handling that threw up
7148 "unexpected character" debug warnings, and recalculated the data
7149 unnecessarily. In any case, I expanded the warning message to give more
7152 15. Added the message "Restricted characters in address" to the statements in
7153 the default ACL that block characters like @ and % in local parts.
7155 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7156 Three changes have been made:
7158 (a) There was a serious bug; a negative response to MAIL caused the whole
7159 recipient domain to be cached as invalid, thereby blocking all messages
7160 to all local parts at the same domain, from all senders. This bug has
7161 been fixed. The domain is no longer cached after a negative response to
7162 MAIL if the sender used is not empty.
7164 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7167 (c) A new callout option, "use_sender" has been added for people who want
7168 the modified behaviour.
7174 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7177 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7178 indeed breaks things for older releases.
7180 3. Added additional logging to the case where there is a problem reading data
7181 from a filter that is running in a subprocess using a pipe, in order to
7182 try to track down a specific problem.
7184 4. Testing facility fudge: when running in the test harness and attempting
7185 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7186 getting "No route to host". Convert this to a timeout.
7188 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7191 6. Some OS don't have socklen_t but use size_t instead. This affects the
7192 fifth argument of getsockopt() amongst other things. This is now
7193 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7194 can be set for individual OS. I have set it for SunOS5, OSF1, and
7195 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7196 some earlier ones do not.
7198 7. Change 4.30/15 was not doing the test caselessly.
7200 8. The standard form for an IPv6 address literal was being rejected by address
7201 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7202 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7203 this, as well as the form without the "IPv6" on the front (but only when
7204 address literals are enabled, of course).
7206 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7208 10. Exim crashed if a message with an empty sender address specified by -f
7209 encountered a router with an errors_to setting. This could be provoked only
7210 by a command such as
7214 where an empty string was supplied; "<>" did not hit this bug.
7216 11. Installed PCRE release 4.5.
7218 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7219 remained set. It is now erased.
7221 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7222 times from message ids (which are base 36 rather than the normal 62).
7224 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7225 were being counted as actual protocol errors, and logged if the log
7226 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7227 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7228 and DATA following a set of rejected RCPTs do not count as protocol errors.
7229 In other words, Exim assumes they were pipelined, though this may not
7230 actually be the case. Of course, in all cases the client gets an
7231 appropriate error code.
7233 15. If a lookup fails in an ACL condition, a message about the failure may
7234 be available; it is used if testing the ACL cannot continue, because most
7235 such messages specify what the cause of the deferral is. However, some
7236 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7237 that caused an old message to be retained and used if a later statement
7238 caused a defer, replacing the real cause of the deferral.
7240 16. If an IP address had so many PTR records that the DNS lookup buffer
7241 was not large enough to hold them, Exim could crash while trying to process
7242 the truncated data. It now detects and logs this case.
7244 17. Further to 4.21/58, another change has been made: if (and only if) the
7245 first line of a message (the first header line) ends with CRLF, a bare LF
7246 in a subsequent header line has a space inserted after it, so as not to
7247 terminate the header.
7249 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7250 unnecessarily, used atoi() instead of strtol(), and didn't check the
7251 termination when getting file sizes from file names by regex.
7253 19. Completely re-implemented the support for maildirsize files, in the light
7254 of a number of problems with the previous contributed implementation
7255 (4.30/29). In particular:
7257 . If the quota is zero, the maildirsize file is maintained, but no quota is
7260 . If the maildir directory does not exist, it is created before any attempt
7261 to write a maildirsize file.
7263 . The quota value in the file is just a cache; if the quota is changed in
7264 the transport, the new value overrides.
7266 . A regular expression is available for excluding directories from the
7269 20. The autoreply transport checks the characters in options that define the
7270 message's headers; it allows continued headers, but it was checking with
7271 isspace() after an embedded newline instead of explicitly looking for a
7274 21. If all the "regular" hosts to which an address was routed had passed their
7275 expiry times, and had not reached their retry times, the address was
7276 bounced, even if fallback hosts were defined. Now Exim should go on to try
7279 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7280 equivalent code in the SMTP transport. Some hosts send humungous responses
7281 to HELO/EHLO, more than 1024 it seems.
7283 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7284 gives compiler warnings in some environments. I've now done it "properly",
7287 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7288 (because of problems with the built-in one) was declared to return uschar *
7289 instead of char *, causing compiler failure.
7291 25. Fixed a file descriptor leak when processing alias/forward files.
7293 26. Fixed a minor format string issue in dbfn.c.
7295 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7297 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7298 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7301 29. When a sender address is verified, it is cached, to save repeating the test
7302 when there is more than one recipient in a message. However, when the
7303 verification involves a callout, it is possible for different callout
7304 options to be set for different recipients. It is too complicated to keep
7305 track of this in the cache, so now Exim always runs a verification when a
7306 callout is required, relying on the callout cache for the optimization.
7307 The overhead is duplication of the address routing, but this should not be
7310 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7311 to be verified with callout=postmaster, and the main callout worked but the
7312 postmaster check failed, the verification correctly failed. However, if a
7313 subsequent RCPT command asked for sender verification *without* the
7314 postmaster check, incorrect caching caused this verification also to fail,
7317 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7318 it was not caching the DNS options (qualify_single, search_parents) that
7319 were used when the lookup failed. A subsequent lookup with different
7320 options therefore always gave the same answer, though there were cases
7321 where it should not have. (Example: a "domains = !$mx_any" option on a
7322 dnslookup router: the "domains" option is always processed without any
7323 widening, but the router might have qualify_single set.) Now Exim uses the
7324 cached value only when the same options are set.
7326 32. Added John Jetmore's "exipick" utility to the distribution.
7328 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7329 than a timeout (e.g. a certificate is required, and is not provided), an
7330 Exim server now closes the connection immediately. Previously it waited for
7331 the client to close - but if the client is SSL, it seems that they each
7332 wait for each other, leading to a delay before one of them times out.
7334 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7335 maintained 0.8.x compatibility because I don't think many are using it, and
7336 it is clearly obsolete.
7338 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7341 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7342 client certificate was expired. A simple patch fixes this, though I don't
7343 understand the full logic of why the verify callback is called multiple
7346 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7347 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7348 which causes problems with some clients (such as the Certicom SSL Plus
7349 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7350 disables the coutermeasure allowing Eudora to connect."
7352 38. Exim was not checking that a write() to a log file succeeded. This could
7353 lead to Bad Things if a log got too big, in particular if it hit a file
7354 size limit. Exim now panics and dies if it cannot write to a log file, just
7355 as it does if it cannot open a log file.
7357 39. Modified OS/Makefile-Linux so that it now contains
7359 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7361 The two -D definitions ensure that Exim is compiled with large file
7362 support, which makes it possible to handle log files that are bigger than
7365 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7366 instance) a domain was checked against a named list that involved a lookup,
7367 causing $domain_data to be set, then another domain was checked against the
7368 same list, then the first domain was re-checked, the value of $domain_data
7369 after the final check could be wrong. In particular, if the second check
7370 failed, it could be set empty. This bug probably also applied to
7373 41. The strip_trailing_dot option was not being applied to the address given
7374 with the -f command-line option.
7376 42. The code for reading a message's header from the spool was incrementing
7377 $received_count, but never initializing it. This meant that the value was
7378 incorrect (doubled) while delivering a message in the same process in which
7379 it was received. In the most common configuration of Exim, this never
7380 happens - a fresh exec is done - but it can happen when
7381 deliver_drop_privilege is set.
7383 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7384 it now includes up to 150 characters of the unexpected data in the log
7387 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7388 and building data strings. The size of both of these buffers was 10 000
7389 bytes - far larger than anybody would *ever* want, thought I. Needless to
7390 say, somebody hit the limit. I have increased the maximum line length to
7391 20 000 and the maximum data length of concatenated lines to 100 000. I have
7392 also fixed two bugs, because there was no checking on these buffers. Tsk,
7393 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7394 buffer is too small.
7396 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7397 lsearch lookups. Now it does.
7399 46. When parsing a route_list item in a manualroute router, a fixed-length
7400 buffer was used for the list of hosts. I made this 1024 bytes long,
7401 thinking that nobody would ever have a list of hosts that long. Wrong.
7402 Somebody had a whole pile of complicated expansion conditions, and the
7403 string was silently truncated, leading to an expansion error. It turns out
7404 that it is easier to change to an unlimited length (owing to other changes
7405 that have happened since this code was originally written) than to build
7406 structure for giving a limitation error. The length of the item that
7407 expands into the list of hosts is now unlimited.
7409 47. The lsearch lookup could not handle data where the length of text line was
7410 more than 4095 characters. Such lines were truncated, leading to shortened
7411 data being returned. It should now handle lines of any length.
7413 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7414 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7417 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7418 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7419 no longer happens in comment lines. A list of replacements is now placed
7420 at the head of all of the source files, except those whose only change is
7421 to replace PERL_COMMAND in the very first #! line.
7423 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7424 messages on the queue, with a bottom-up merge sort, using code contributed
7425 by Michael Haardt. This should make operations like -bp somewhat faster on
7426 large queues. It won't affect queue runners, except when queue_run_in_order
7429 51. Installed eximstats 1.31 in the distribution.
7431 52. Added support for SRV lookups to the dnslookup router.
7433 53. If an ACL referred to $message_body or $message_body_end, the value was not
7434 reset for any messages that followed in the same SMTP session.
7436 54. The store-handling optimization for building very long strings was not
7437 differentiating between the different store pools. I don't think this
7438 actually made any difference in practice, but I've tidied it.
7440 55. While running the routers to verify a sender address, $sender_address
7441 was still set to the sender address. This is wrong, because when routing to
7442 send a bounce to the sender, it would be empty. Therefore, I have changed
7443 it so that, while verifying a sender address, $sender_address is set to <>.
7444 (There is no change to what happens when verifying a recipient address.)
7446 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7447 target A or AAAA records (if not already returned) without resetting the
7448 qualify_single or search_parents options of the DNS resolver. These are
7449 inappropriate in this case because the targets of MX and SRV records must
7450 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7451 target that, when qualified, matched something in the local domain. These
7452 two options are now turned off when doing these lookups.
7454 57. It seems that at least some releases of Reiserfs (which does not have the
7455 concept of a fixed number of inodes) returns zero and not -1 for the
7456 number of available inodes. This interacted badly with check_spool_inodes,
7457 which assumed that -1 was the "no such thing" setting. What I have done is
7458 to check that the total number of inodes is greater than zero before doing
7459 the test of how many are available.
7461 58. When a "warn" ACL statement has a log_message modifier, the message is
7462 remembered, and not repeated. This is to avoid a lot of repetition when a
7463 message has many recipients that cause the same warning to be written.
7464 However, Exim was preserving the list of already written lines for an
7465 entire SMTP session, which doesn't seem right. The memory is now reset if a
7466 new message is started.
7468 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7469 debugging output unless log_rewrite was also set.
7471 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7472 of (int)(handle) when we know that handle contains (void *)(-1).
7474 61. The Exim daemon panic-logs an error return when it closes the incoming
7475 connection. However "connection reset by peer" seems to be common, and
7476 isn't really an error worthy of noting specially, so that particular error
7479 62. When Exim is trying to find all the local interfaces, it used to panic and
7480 die if the ioctl to get the interface flags failed. However, it seems that
7481 on at least one OS (Solaris 9) it is possible to have an interface that is
7482 included in the list of interfaces, but for which you get a failure error
7483 for this call. This happens when the interface is not "plumbed" into a
7484 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7485 failure of the "get flags" call assumes that the interface is down.
7487 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7488 makes life easier for people who are doing arithmetic on fields extracted
7489 from dates, where you often get leading zeros that should not be
7490 interpreted as octal.
7492 64. Added qualify_domain to the redirect router, to override the global
7495 65. If a pathologically long header line contained very many addresses (the
7496 report of this problem mentioned 10 000) and each of them was rewritten,
7497 Exim could use up a very large amount of memory. (It kept on making new
7498 copies of the header line as it rewrote, and never released the old ones.)
7499 At the expense of a bit more processing, the header rewriting function has
7500 been changed so that it no longer eats memory in this way.
7502 66. The generation of the Received: header has been moved from the time that a
7503 message starts to be received, to the time that it finishes. The timestamp
7504 in the Received: header should now be very close to that of the <= log
7505 line. There are two side-effects of this change:
7507 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7508 logged header lines no longer include the local Received: line, because
7509 it has not yet been created. The same applies to a copy of the message
7510 that is returned to a non-SMTP sender when a message is rejected.
7512 (b) When a filter file is tested using -bf, no additional Received: header
7513 is added to the test message. After some thought, I decided that this
7516 This change does not affect the value of $received_for. It is still set
7517 after address rewriting, but before local_scan() is called.
7519 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7521 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7522 gave an unhelpful panic error message, and a defer error. I have managed to
7523 change this behaviour so that it now rejects any supplied certificate,
7524 which seems right, as the list of acceptable certificates is empty.
7526 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7527 gave an unhelpful defer error. I have not managed to make this reject any
7528 supplied certificates, but the error message it gives is "no certificate
7529 supplied", which is not helpful.
7531 70. exigrep's output now also includes lines that are not associated with any
7532 message, but which match the given pattern. Implemented by a patch from
7533 Martin Sluka, which also tidied up the Perl a bit.
7535 71. Recipient callout verification, like sender verification, was using <> in
7536 the MAIL FROM command. This isn't really the right thing, since the actual
7537 sender may affect whether the remote host accepts the recipient or not. I
7538 have changed it to use the actual sender in the callout; this means that
7539 the cache record is now keyed on a recipient/sender pair, not just the
7540 recipient address. There doesn't seem to be a real danger of callout loops,
7541 since a callout by the remote host to check the sender would use <>.
7542 [SEE ABOVE: changed after hitting problems.]
7544 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7545 temporary errors. However, in the case of such a code being given after
7546 the end of a data transmission (i.e. after ".") Exim was failing to write
7547 a retry record for the message. (Yes, there was some broken host that was
7548 actually sending 8xx at this point.)
7550 73. An unknown lookup type in a host list could cause Exim to panic-die when
7551 the list was checked. (An example that provoked this was putting <; in the
7552 middle of a list instead of at the start.) If this happened during a DATA
7553 ACL check, a -D file could be left lying around. This kind of configuration
7554 error no longer causes Exim to die; instead it causes a defer error. The
7555 incident is still logged to the main and panic logs.
7557 74. Buglet left over from Exim 3 conversion. The message "too many messages
7558 in one connection" was written to the rejectlog but not the mainlog, except
7559 when address rewriting (yes!) was being logged.
7561 75. Added write_rejectlog option.
7563 76. When a system filter was run not as root (that is, when system_filter_user
7564 was set), the values of the $n variables were not being returned to the
7565 main process; thus, they were not subsequently available in the $sn
7568 77. Added +return_path_on_delivery log selector.
7570 78. A connection timeout was being treated differently from recipients deferred
7571 when testing hosts_max_try with a message that was older than the host's
7572 retry timeout. (The host should not be counted, thus allowing all hosts to
7573 be tried at least once before bouncing.) This may have been the cause of an
7574 occasionally reported bug whereby a message would remain on the queue
7575 longer than the retry timeout, but would be bounced if a delivery was
7576 forced. I say "may" because I never totally pinned down the problem;
7577 setting up timeout/retry tests is difficult. See also the next item.
7579 79. The ultimate address timeout was not being applied to errors that involved
7580 a combination of host plus message (for example, a timeout on a MAIL
7581 command). When an address resolved to a number of possible hosts, and they
7582 were not all tried for each delivery (e.g. because of hosts_max_try), a
7583 message could remain on the queue longer than the retry timeout.
7585 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7588 81. Fixed an obscure SMTP outgoing bug which required at least the following
7589 conditions: (a) there was another message waiting for the same server;
7590 (b) the server returned 5xx to all RCPT commands in the first message so
7591 that the message was not completed; (c) the server dropped the connection
7592 or gave a negative response to the RSET that Exim sends to abort the
7593 transaction. The observed case was a dropped connection after DATA that had
7594 been sent in pipelining mode. That is, the server had advertised PIPELINING
7595 but was not implementing it correctly. The effect of the bug was incorrect
7596 behaviour, such as trying another host, and this could lead to a crash.
7602 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7603 and daemon.c were passed as pointers to ints; they should have been
7604 pointers to socklen_t variables (which are typically unsigned ints).
7606 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7609 3. Fixed a really odd bug that affected only the testing scheme; patching a
7610 certain fixed string in the binary changed the value of another string that
7611 happened to be identical to the end of the original first string.
7613 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7614 name", it returns that address as the IP address. On some operating
7615 systems (e.g. Solaris), it also passes back the IP address string as the
7616 "host name". However, on others (e.g. Linux), it passes back an empty
7617 string. Exim wasn't checking for this, and was changing the host name to an
7618 empty string, assuming it had been canonicalized.
7620 5. Although rare, it is permitted to have more than one PTR record for a given
7621 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7622 all the names associated with an address, because they do in Solaris.
7623 However, it seems that they do not in Linux for data that comes from the
7624 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7625 I found this out when I moved to a new Linux workstation and tried to run
7626 the Exim test suite.
7628 To get round this problem I have changed the code so that it now does its
7629 own call to the DNS to look up PTR records when searching for a host name.
7630 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7631 addresses that are only in /etc/hosts are still found.
7633 This behaviour is, however, controlled by an option called host_lookup_
7634 order, which defaults to "bydns:byaddr". If people want to use the other
7635 order, or indeed, just use one or the other means of lookup, they can
7636 specify it in this variable.
7638 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7639 some operating systems, this comes back from gethostbyaddr() as an empty
7640 string, and this is what Exim used to test for. However, it seems that in
7641 other systems, "." is yielded. Exim now tests for this case too.
7643 7. The values of check_spool_space and check_log_space are now held internally
7644 as a number of kilobytes instead of an absolute number of bytes. If a
7645 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7646 kilobyte. This means that much larger values can be stored.
7648 8. Exim monitor: an attempt to get the action menu when not actually pointing
7649 at a message produces an empty menu entitled "No message selected". This
7650 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7651 no entries in it ("Shell widget menu has zero width and/or height"). So I
7652 have added a single, blank menu entry in this case.
7654 9. Added ${quote_local_part.
7656 10. MIME decoding is now applied to the contents of Subject: header lines when
7659 11. Now that a reference to $sender_host_address automatically causes a reverse
7660 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7661 host lookup before query-style lookups in lists that might use this
7662 variable. This has therefore been abolished, and the "net-" prefix is no
7663 longer necessary for query-style lookups.
7665 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7666 have been a typo for LFLAGS, so it has been changed.
7668 13. The install script calls Exim with "-C /dev/null" in order to find the
7669 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7670 to be output. However, since Exim outputs its version number before the
7671 error, it didn't break the script. It just looked ugly. I fixed this by
7672 always allowing "-C /dev/null" if the caller is root.
7674 14. Ignore overlarge ACL variable number when reading spool file - insurance
7675 against a later release with more variables having written the file.
7677 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7678 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7679 this, as well as the form without the "IPv6" on the front.
7681 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7682 OS/Makefile-Darwin file.
7684 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7685 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7687 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7688 variable. However, some OS use ints for inodes. I've added cast to long int
7689 to get rid of the compiler warning.
7691 19. I had forgotten to lock out "/../" in configuration file names when
7692 ALT_CONFIG_PREFIX was set.
7694 20. Routers used for verification do not need to specify transports. However,
7695 if such a router generated a host list, and callout was configured, Exim
7696 crashed, because it could not find a port number from the (non-existent)
7697 transport. It now assumes port 25 in this circumstance.
7699 21. Added the -t option to exigrep.
7701 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7702 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7703 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7704 them is set, LOOKUP_LSEARCH is forced.
7706 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7708 24. Added sender and host information to the "rejected by local_scan()" log
7709 line; previously there was no indication of these.
7711 25. Added .include_if_exists.
7713 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7714 Linux. It turns out that not all file systems support this. Apparently some
7715 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7716 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7717 not supported on the file descriptor, is now ignored when Exim is trying to
7718 sync a directory. This applies only to Linux.
7720 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7722 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7723 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7726 29. Added support for maildirsize files from supplied patch (modified a bit).
7728 30. The use of :fail: followed by an empty string could lead Exim to respond to
7729 sender verification failures with (e.g.):
7731 550 Verification failed for <xxx>
7732 550 Sender verify failed
7734 where the first response line was missing the '-' that indicates it is not
7735 the final line of the response.
7737 31. The loop for finding the name of the user that called Exim had a hardwired
7738 limit of 10; it now uses the value of finduser_retries, which is used for
7739 all other user lookups.
7741 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7744 33. Exim was neglecting to zero errno before one call of strtol() when
7745 expanding a string and expecting an integer value. On some systems this
7746 resulted in spurious "integer overflow" errors. Also, it was casting the
7747 result into an int without checking.
7749 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7750 did not work. The code looks as if it has *never* worked, though it appears
7751 to have been documented since at least release 1.62. I have made it work.
7753 35. The "timeout_DNS" error in retry rules, also documented since at least
7754 1.62, also never worked. As it isn't clear exactly what this means, and
7755 clearly it isn't a major issue, I have abolished the feature by treating it
7756 as "timeout", and writing a warning to the main and panic logs.
7758 36. The display of retry rules for -brt wasn't always showing the error code
7761 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7762 timeout_connect_A, timeout_connect_MX.
7764 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7765 to the empty sender.
7767 39. The daemon was not analysing the content of -oX till after it had closed
7768 stderr and disconnected from the controlling terminal. This meant that any
7769 syntax errors were only noted on the panic log, and the return code from
7770 the command was 0. By re-arranging the code a little, I've made the
7771 decoding happen first, so such errors now appear on stderr, and the return
7772 code is 1. However, the actual setting up of the sockets still happens in
7773 the disconnected process, so errors there are still only recorded on the
7776 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7777 connections (as happens on some IP stacks) was logged at start up time as
7778 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7779 it from "IPv6 and IPv4", which means that two separate sockets are being
7782 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7783 says whether AF_INET or AF_INET6 was passed as an argument.
7785 42. Exiwhat output was messed up when time zones were included in log
7788 43. Exiwhat now gives more information about the daemon's listening ports,
7789 and whether -tls-on-connect was used.
7791 44. The "port" option of the smtp transport is now expanded.
7793 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7794 silently ignored. Now an error message is written to the main and panic
7797 46. There's a new ACL modifier called "logwrite" which writes to a log file
7798 as soon as it is encountered.
7800 47. Added $local_user_uid and $local_user_gid at routing time.
7802 48. Exim crashed when trying to verify a sender address that was being
7805 49. Exim was recognizing only a space character after ".include". It now also
7806 recognizes a tab character.
7808 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7809 extracting the relevant information from the specification. The man page no
7810 longer contains scrambled data for the -d option, and I've added a section
7811 at the front about calling Exim under different names.
7813 51. Added "extra_headers" argument to the "mail" command in filter files.
7815 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7818 53. Installed eximstats 1.29.
7820 54. Added transport_filter_timeout as a generic transport option.
7822 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7823 Cc: header lines. This was required by RFC 822, but it not required by RFC
7826 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7827 incoming messages that did not have them. Now it does so only if the
7828 message originates locally, that is, if there is no associated remote host
7829 address. When Resent- header lines are present, this applies to the Resent-
7830 lines rather than the non-Resent- lines.
7832 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7833 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7835 58. Messages for configuration errors now include the name of the main
7836 configuration file - useful now that there may be more than one file in a
7837 list (.included file names were always shown).
7839 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7840 for those rare installations that do not start the daemon as root or run it
7841 setuid root. I've cut out the call to initgroups() if the daemon is not
7844 60. The Exim user and group can now be bound into the binary as text strings
7845 that are looked up at the start of Exim's processing.
7847 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7849 62. Added $mailstore_basename variable.
7851 63. Installed patch to sieve.c from Michael Haardt.
7853 64. When Exim failed to open the panic log after failing to open the main log,
7854 the original message it was trying to log was written to stderr and debug
7855 output, but if they were not available (the usual case in production), it
7856 was lost. Now it is written to syslog before the two lines that record the
7857 failures to open the logs.
7859 65. Users' Exim filters run in subprocesses under the user's uid. It is
7860 possible for a "deliver" command or an alias in a "personal" command to
7861 provoke an address rewrite. If logging of address rewriting is configured,
7862 this fails because the process is not running as root or exim. There may be
7863 a better way of dealing with this, but for the moment (because 4.30 needs
7864 to be released), I have disabled address rewrite logging when running a
7865 filter in a non-root, non-exim process.
7871 1. The buildconfig auxiliary program wasn't quoting the value set for
7872 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7873 not defined. This bug was masked in 4.22 by the effect that was fixed in
7876 2. Some messages that were rejected after a message id was allocated were
7877 shown as "incomplete" by exigrep. It no longer does this for messages that
7878 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7880 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7881 have allow_domain_literals set, the ID did not get logged in the <= line.
7882 Domain literals are now always recognized in Message-ID: header lines.
7884 4. The first argument for a ${extract expansion item is the key name or field
7885 number. Leading and trailing spaces in this item were not being ignored,
7886 causing some misleading effects.
7888 5. When deliver_drop_privilege was set, single queue runner processes started
7889 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7890 same command in the process it spins off) were not dropping privilege.
7892 6. When the daemon running as "exim" started a queue runner, it always
7893 re-executed Exim in the spun-off process. This is a waste of effort when
7894 deliver_drop_privilege is set. The new process now just calls the
7895 queue-runner function directly.
7901 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7904 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7905 ignored. Though the use of -oP was forcing the writing of a pid file, it
7906 was always written to the default place.
7908 3. If the message "no IP address found for host xxxx" is generated during
7909 incoming verification, it is now followed by identification of the incoming
7910 connection (so you can more easily find what provoked it).
7912 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7914 5. Added some features to "harden" Exim a bit more against certain attacks:
7916 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7917 be put in Local/Makefile. This is like the never_users runtime option,
7918 but it cannot be overridden. The default setting is "root".
7920 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7921 prefix string with which any file named in a -C command line option
7924 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7925 is retained for -C and -D only if the caller of Exim is root. Without
7926 it, the exim user may also use -C and -D and retain privilege.
7928 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7929 command line option is disabled.
7931 6. Macro names set by the -D option must start with an upper case letter, just
7932 like macro names defined in the configuration file.
7934 7. Added "dereference=" facility to LDAP.
7936 8. Two instances of the typo "uknown" in the source files are fixed.
7938 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7939 the Configure-Makefile script screwed up while processing it.
7941 10. Incorporated PCRE 4.4.
7943 11. The SMTP synchronization check was not operating right at the start of an
7944 SMTP session. For example, it could not catch a HELO sent before the client
7945 waited for the greeting. There is now a check for outstanding input at the
7946 point when the greeting is written. Because of the duplex, asynchronous
7947 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7948 way, but not yet received, when the check is performed.
7950 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7951 on TCP/IP sockets, because this apparently causes some broken clients to
7954 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7955 unchanged) from the Cygwin maintainer.
7957 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7958 was supported instead of testing for mbx. Effectively a typo.
7960 15. The spa authenticator server code was not checking that the input it
7961 received was valid base64.
7963 16. The debug output line for the "set" modifier in ACLs was not showing the
7964 name of the variable that was being set.
7966 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7968 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7969 lookup on its own. Something else was needed to trigger the lookup. For
7970 example, a match in host_lookup or the need for a host name in a host list.
7971 Now, if $sender_host_name is referenced and the host name has not yet been
7972 looked up, a lookup is performed. If the lookup fails, the variable remains
7973 empty, and $host_lookup_failed is set to "1".
7975 19. Added "eqi" as a case-independent comparison operator.
7977 20. The saslauthd authentication condition could segfault if neither service
7978 nor realm was specified.
7980 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7981 error message that was logged was misleading, and incoming SMTP
7982 connections were dropped. The message is now more accurate, and temporary
7983 errors are given to SMTP connections.
7985 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7986 (except RSET) with a 421 error, until QUIT is received. However, it was
7987 failing to send a response to QUIT.
7989 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7990 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7991 ACL is now run after the helo_try_verify_hosts code.
7993 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7994 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7995 case-independent, so other case variants are also recognized). Apparently
7996 some systems use these upper case variants.
7998 25. If more than two messages were waiting for the same host, and a transport
7999 filter was specified for the transport, Exim sent two messages over the
8000 same TCP/IP connection, and then failed with "socket operation on non-
8001 socket" when it tried to send the third.
8003 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8005 27. The extern definition of crypt16() in expand.c was not being excluded when
8006 the OS had its own crypt16() function.
8008 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8009 as a preferred synonym for return_size_limit, both as an option and as an
8012 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8014 30. Changed the default configuration ACL to relax the local part checking rule
8015 for addresses that are not in any local domains. For these addresses,
8016 slashes and pipe symbols are allowed within local parts, but the sequence
8017 /../ is explicitly forbidden.
8019 31. SPA server authentication was not clearing the challenge buffer before
8022 32. log_message in a "warn" ACL statement was writing to the reject log as
8023 well as to the main log, which contradicts the documentation and doesn't
8024 seem right (because no rejection is happening). So I have stopped it.
8026 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8027 However, I am unable to do any testing of this.
8029 34. Fixed an infelicity in the appendfile transport. When checking directories
8030 for a mailbox, to see if any needed to be created, it was accidentally
8031 using path names with one or more superfluous leading slashes; tracing
8032 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8034 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8035 contents are added to the log line that is written for every discarded
8036 recipient. (Previously a log_message setting was ignored.)
8038 36. The ${quote: operator now quotes the string if it is empty.
8040 37. The install script runs exim in order to find its version number. If for
8041 some reason other than non-existence or emptiness, which it checks, it
8042 could not run './exim', it was installing it with an empty version number,
8043 i.e. as "exim-". This error state is now caught, and the installation is
8046 38. An argument was missing from the function that creates an error message
8047 when Exim fails to connect to the socket for saslauthd authentication.
8048 This could cause Exim to crash, or give a corrupted message.
8050 39. Added isip, isip4, and isip6 to ${if conditions.
8052 40. The ACL variables $acl_xx are now saved with the message, and can be
8053 accessed later in routers, transports, and filters.
8055 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8056 strings in the file are not string-expanded.
8058 42. If a MAIL command specified a SIZE value that was too large to fit into an
8059 int variable, the check against message_size_limit failed. Such values are
8060 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8061 day this will have to be increased, but I don't think I want to be around
8062 when emails are that large.
8069 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8070 iconv() is not standard in FreeBSD.
8072 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8073 IPv6 enabled. The observed symptom was a segmentation fault on return from
8074 the function os_common_find_running_interfaces() in src/os.c.
8076 3. In the check_special_case() function in daemon.c I had used "errno" as an
8077 argument name, which causes warnings on some systems. This was basically a
8078 typo, since it was named "eno" in the comments!
8080 4. The code that waits for the clock to tick (at a resolution of some fraction
8081 of a second) so as to ensure message-id uniqueness was always waiting for
8082 at least one whole tick, when it could have waited for less. [This is
8083 almost certainly not relevant at current processor speeds, where it is
8084 unlikely to ever wait at all. But we try to future-proof.]
8086 5. The function that sleeps for a time interval that includes fractions of a
8087 second contained a race. It did not block SIGALRM between setting the
8088 timer, and suspending (a couple of lines later). If the interval was short
8089 and the sigsuspend() was delayed until after it had expired, the suspension
8090 never ended. On busy systems this could lead to processes getting stuck for
8093 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8094 process, before it forks any delivery processes. The open lookup caching
8095 mechanism meant that the open file or database connection was passed into
8096 the delivery process. The problem was that delivery processes always tidy
8097 up cached lookup data. This could cause a problem for the next delivery
8098 process started by the queue runner, because the external queue runner
8099 process does not know about the closure. So the next delivery process
8100 still has data in the lookup cache. In the case of a file lookup, there was
8101 no problem because closing a file descriptor in a subprocess doesn't affect
8102 the parent. However, if the lookup was caching a connection to a database,
8103 the connection was closed, and the second delivery process was likely to
8104 see errors such as "PGSQL: query failed: server closed the connection
8105 unexpectedly". The problem has been fixed by closing all cached lookups
8106 in a queue runner before running a delivery process.
8108 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8109 seem to have the "const" qualifier which it has on other OS. I've
8112 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8113 *of the same type* (client or server) with the same public name that an
8114 error should be diagnosed.
8116 9. When Exim looked up a host name for an IP address, but failed to find the
8117 original IP address when looking up the host name (a safety check), it
8118 output the message "<ip address> does not match any IP for NULL", which was
8119 confusing, to say the least. The bug was that the host name should have
8120 appeared instead of "NULL".
8122 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8123 user that is built into the binary, and the -C or -D options is used, root
8124 privilege is dropped before the configuration file is read. In addition,
8125 logging is switched to stderr instead of the normal log files. If the
8126 configuration then re-defines the Exim user, the unprivileged environment
8127 is probably not what is expected, so Exim logs a panic warning message (but
8130 However, if deliver_drop_privilege is set, the unprivileged state may well
8131 be exactly what is intended, so the warning has been cut out in that case,
8132 and Exim is allowed to try to write to its normal log files.
8138 1. smtp_return_error_details was not giving details for temporary sender
8139 or receiver verification errors.
8141 2. Diagnose a configuration error if two authenticators have the same public
8144 3. Exim used not to create the message log file for a message until the first
8145 delivery attempt. This could be confusing when incoming messages were held
8146 for policy or load reasons. The message log file is now created at the time
8147 the message is received, and an initial "Received" line is written to it.
8149 4. The automatically generated man page for command line options had a minor
8150 bug that caused no ill effects; however, a more serious problem was that
8151 the procedure for building the man page automatically didn't always
8152 operate. Consequently, release 4.20 contains an out-of-date version. This
8153 shouldn't happen again.
8155 5. When building Exim with embedded Perl support, the script that builds the
8156 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8157 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8159 6. The freeze_tell option was not being used for messages that were frozen on
8160 arrival, either by an ACL or by local_scan().
8162 7. Added the smtp_incomplete_transaction log selector.
8164 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8165 was accepting AUTH without a new EHLO.
8167 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8168 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8171 10. Logging of TCP/IP connections (when configured) now happens in the main
8172 daemon process instead of the child process, so that the TCP/IP connection
8173 count is more accurate (but it can never be perfect).
8175 11. The use of "drop" in a nested ACL was not being handled correctly in the
8176 outer ACL. Now, if condition failure induced by the nested "drop" causes
8177 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8178 or "require"), the connection is dropped.
8180 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8181 that yield "discard" can now be used with an "accept" or a "discard" verb,
8182 but an error is generated for any others (because I can't see a useful way
8183 to define what should happen).
8185 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8186 are now processed in the same way as lines in the Exim configuration file.
8187 In particular, continuation lines are supported.
8189 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8191 15. Added -ti meaning -t -i.
8193 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8194 domains, and warn by logging if others are found.
8196 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8197 returned from GIFCONF when Exim is trying to find the list of interfaces on
8198 a host. The code in os.c has been modified to copy each ifreq to an aligned
8199 structure in all cases.
8201 Also, in some cases, the returned ifreq's were being copied to a 'struct
8202 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8203 means the last couple of bytes of an IPv6 address could be chopped if the
8204 ifreq contained only a normal sockaddr (14 bytes storage).
8206 18. Named domain lists were not supported in the hosts_treat_as_local option.
8207 An entry such as +xxxx was not recognized, and was treated as a literal
8210 19. Ensure that header lines added by a DATA ACL are included in the reject log
8211 if the ACL subsequently rejects the message.
8213 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8214 MD5 (which is deprecated).
8216 21. When testing a filter file using -bf, Exim was writing a message when it
8217 took the sender from a "From " line in the message, but it was not doing so
8218 when it took $return_path from a Return-Path: header line. It now does.
8220 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8221 with a valid header line field name (a series of printing characters
8222 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8224 23. Changed "disc" in the source to "disk" to conform to the documentation and
8225 the book and for uniformity.
8227 24. Ignore Sendmail's -Ooption=value command line item.
8229 25. When execve() failed while trying to run a command in a pipe transport,
8230 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8231 could be confused with a return value of 69 from the command itself. This
8232 has been changed to 127, the value the shell returns if it is asked to run
8233 a non-existent command. The wording for the related log line suggests a
8234 non-existent command as the problem.
8236 26. If received_header_text expands to an empty string, do not add a Received:
8237 header line to the message. (Well, it adds a token one on the spool, but
8238 marks it "old" so that it doesn't get used or transmitted.)
8240 27. Installed eximstats 1.28 (addition of -nt option).
8242 28. There was no check for failure on the call to getsockname() in the daemon
8243 code. This can fail if there is a shortage of resources on the system, with
8244 ENOMEM, for example. A temporary error is now given on failure.
8246 29. Contrary to the C standard, it seems that in some environments, the
8247 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8248 program. Exim now does this explicitly; it affects the formatting of
8249 timestamps using strftime().
8251 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8252 complaints. I've now initialized all the variables, to avoid this.
8254 32. Header lines added by a system filter were not being "seen" during
8255 transport-time rewrites.
8257 33. The info_callback() function passed to OpenSSL is set up with type void
8258 (*)(SSL *, int, int), as described somewhere. However, when calling the
8259 function (actually a macro) that sets it up, the type void(*)() is
8260 expected. I've put in a cast to prevent warnings from picky compilers.
8262 34. If a DNS black list lookup found a CNAME record, but there were no A
8263 records associated with the domain it pointed at, Exim crashed.
8265 35. If a DNS black list lookup returned more than one A record, Exim ignored
8266 all but the first. It now scans all returned addresses if a particular IP
8267 value is being sought. In this situation, the contents of the
8268 $dnslist_value variable are a list of all the addresses, separated by a
8271 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8272 to accept a host name and all its aliases if the forward lookup for any of
8273 them yielded the IP address of the incoming connection. Now it accepts only
8274 those names whose forward lookup yields the correct IP address. Any other
8275 names are discarded. This closes a loophole whereby a rogue DNS
8276 administrator could create reverse DNS records to break through a
8277 wildcarded host restriction in an ACL.
8279 37. If a user filter or a system filter that ran in a subprocess used any of
8280 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8281 the wrong values were passed to the pipe command ($thisaddress had the
8282 value of $0, $0 had the value of $1, etc). This bug was introduced by
8283 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8285 38. Improved the line breaking for long SMTP error messages from ACLs.
8286 Previously, if there was no break point between 40 and 75 characters, Exim
8287 left the rest of the message alone. Two changes have been made: (a) I've
8288 reduced the minimum length to 35 characters; (b) if it can't find a break
8289 point between 35 and 75 characters, it looks ahead and uses the first one
8290 that it finds. This may give the occasional overlong line, but at least the
8291 remaining text gets split now.
8293 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8294 file descriptors might be low, and that setting 1000 would always raise it.
8295 It turns out that in some environments, the limit is already over 1000 and
8296 that lowering it causes trouble. So now Exim takes care not to decrease it.
8298 40. When delivering a message, the value of $return_path is set to $sender_
8299 address at the start of routing (routers may change the value). By an
8300 oversight, this default was not being set up when an address was tested by
8301 -bt or -bv, which affected the outcome if any router or filter referred to
8304 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8305 log only when "message" or "log_message" are set. However, if one of the
8306 conditions was an address verification, or a call to a nested ACL, the
8307 messages generated by the underlying test were being passed through. This
8308 no longer happens. The underlying message is available in $acl_verify_
8309 message for both "message" and "log_message" expansions, so it can be
8310 passed through if needed.
8312 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8313 new expansion $bh_ to give the encoded byte string without charset
8314 translation. Translation happens only if iconv() is available; HAVE_ICONV
8315 indicates this at build time. HEADERS_CHARSET gives the charset to
8316 translate to; headers_charset can change it in the configuration, and
8317 "headers charset" can change it in an individual filter file.
8319 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8320 that creates RFC 2047 encoded "words" labels them as that charset instead
8321 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8322 expansion operator; (ii) when Exim creates a From: line for a local
8323 message; (iii) when a header line is rewritten to include a "phrase" part.
8325 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8326 buggy, causing it to skip the first lines of messages whose message ID
8327 ended in 'D'. This would not have bitten before Exim release 4.14, because
8328 message IDs were unlikely to end in 'D' before then. The effect was to have
8329 incorrect size information for certain domains.
8331 45. #include "config.h" was missing at the start of the crypt16.c module. This
8332 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8335 46. If there was a timeout during a "random" callout check, Exim treated it as
8336 a failure of the random address, and carried on sending RSET and the real
8337 address. If the delay was just some slowness somewhere, the response to the
8338 original RCPT would be taken as a response to RSET and so on, causing
8339 mayhem of various kinds.
8341 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8342 when I implemented it. It didn't allow for the fact that some option values
8343 may legitimately be negative (e.g. size_addition), and it didn't even do
8344 the right test for positive values.
8346 48. Domain names in DNS records are case-independent. Exim always looks them up
8347 in lower case. Some resolvers return domain names in exactly the case they
8348 appear in the zone file, that is, they may contain uppercase letters. Not
8349 all resolvers do this - some return always lower case. Exim was treating a
8350 change of case by a resolver as a change of domain, similar to a widening
8351 of a domain abbreviation. This triggered its re-routing code and so it was
8352 trying to route what was effectively the same domain again. This normally
8353 caused routing to fail (because the router wouldn't handle the domain
8354 twice). Now Exim checks for this case specially, and just changes the
8355 casing of the domain that it ultimately uses when it transmits the message
8358 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8361 50. If a filter generated a file delivery with a non-absolute name (possible if
8362 no home directory exists for the router), the forbid_file option was not
8365 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8366 the existing equality matching.
8368 52. Exim was using ints instead of ino_t variables in some places where it was
8369 dealing with inode numbers.
8371 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8372 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8373 called TMPDIR, and if it finds it is different, it changes its value.
8375 54. The smtp_printf() function is now made available to local_scan() so
8376 additional output lines can be written before returning. There is also an
8377 smtp_fflush() function to enable the detection of a dropped connection.
8378 The variables smtp_input and smtp_batched_input are exported to
8381 55. Changed the default runtime configuration: the message "Unknown user"
8382 has been removed from the ACL, and instead placed on the localuser router,
8383 using the cannot_route_message feature. This means that any verification
8384 failures that generate their own messages won't get overridden. Similarly,
8385 the "Unrouteable address" message that was in the ACL for unverifiable
8386 relay addresses has also been removed.
8388 56. Added hosts_avoid_esmtp to the smtp transport.
8390 57. The exicyclog script was not checking for the esoteric option
8391 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8392 will work only if exicyclog is run under the appropriate euid.
8394 58. Following a discussion on the list, the rules by which Exim recognises line
8395 endings on incoming messages have been changed. The -dropcr and drop_cr
8396 options are now no-ops, retained only for backwards compatibility. The
8397 following line terminators are recognized: LF CRLF CR. However, special
8398 processing applies to CR:
8400 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8401 nor a local message in the state where . is a terminator.
8403 (ii) If a bare CR is encountered in a header line, an extra space is added
8404 after the line terminator so as not to end the header. The reasoning
8405 behind this is that bare CRs in header lines are most likely either
8406 to be mistakes, or people trying to play silly games.
8408 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8409 was being incorrectly given as 18 bytes larger than it should have been.
8410 This is a VOB (very old bug).
8412 60. This may never have affected anything current, but just in case it has:
8413 When the local host is found other than at the start of a list of hosts,
8414 the local host, those with the same MX, and any that follow, are discarded.
8415 When the list in question was part of a longer list of hosts, the following
8416 hosts (not currently being processed) were also being discarded. This no
8417 longer happens. I'm not sure if this situation could ever has previously
8420 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8423 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8424 header. This is something that is recommended in a new Internet Draft, and
8425 is something that is documented as being done by Sendmail. There are two
8426 possible values. For messages generated by the autoreply transport, Exim
8429 Auto-Submitted: auto-replied
8431 whereas for all other generated messages (e.g. bounces) it adds
8433 Auto-Submitted: auto-generated
8435 63. The "personal" condition in filters now includes a test for the
8436 Auto-Submitted: header. If it contains the string "auto-" the message it
8437 not considered personal.
8439 64. Added rcpt_include_affixes as a generic transport option.
8441 65. Added queue_only_override (default true).
8443 66. Added the syslog_duplication option.
8445 67. If what should have been the first header line of a message consisted of
8446 a space followed by a colon, Exim was mis-interpreting it as a header line.
8447 It isn't of course - it is syntactically invalid and should therefore be
8448 treated as the start of the message body. The misbehaviour could have
8449 caused a number of strange effects, including loss of data in subsequent
8450 header lines, and spool format errors.
8452 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8453 client host had authenticated. This control can now be exercised by an ACL
8454 for more flexibility.
8456 69. By default, callouts do not happen when testing with -bh. There is now a
8457 variant, -bhc, which does actually run the callout code, including
8458 consulting and updating the callout cache.
8460 70. Added support for saslauthd authentication, courtesy of Alexander
8463 71. If statvfs() failed on the spool or log directories while checking their
8464 size for availability, Exim confusingly gave the error "space shortage".
8465 Furthermore, in debugging mode it crashed with a floating point exception.
8466 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8467 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8468 serious problem, Exim now writes to the main and panic logs when this
8469 happens, with details of the failure. It then refuses to accept the
8470 incoming message, giving the message "spool directory problem" or "log
8471 directory problem" with a 421 code for SMTP messages.
8473 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8474 0, 1, and 2 exist, because some OS complain for execs without them (see
8475 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8476 descriptors. However, the code omitted to check that the open succeeded,
8477 causing mysterious errors if for some reason the permissions on /dev/null
8478 got screwed. Now Exim writes a message to the main and panic logs, and
8479 bombs out if it can't open /dev/null.
8481 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8482 interact so that it is all more flexible. It is supposed to remain
8483 backwards compatible. Also added extra_local_interfaces.
8485 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8486 to bomb out with an assertion failure - to the client this appears as a
8487 connection drop. This problem occurs in the part of the code that was taken
8488 from the Samba project. Fortunately, the assertion is in a very simple
8489 function, so I have fixed this by reproducing the function inline in the
8490 one place where it is called, and arranging for authentication to fail
8491 instead of killing the process with assert().
8493 75. The SPA client code was not working when the server requested OEM rather
8494 than Unicode encoding.
8496 76. Added code to make require_files with a specific uid setting more usable in
8497 the case where statting the file as root fails - usually a non-root-mounted
8498 NFS file system. When this happens and the failure is EACCES, Exim now
8499 forks a subprocess and does the per-uid checking as the relevant uid.
8501 77. Added process_log_path.
8503 78. If log_file_path was not explicitly set, a setting of check_log_space or
8504 check_log_inodes was ignored.
8506 79. If a space check for the spool or log partitions fails, the incident is now
8507 logged. Of course, in the latter case the data may get lost...
8509 80. Added the %p formatting code to string_format() so that it can be used to
8510 print addresses in debug_print(). Adjusted all the address printing in the
8511 debugging in store.c to use %p rather than %d.
8513 81. There was a concern that a line of code in smtp_in.c could overflow a
8514 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8515 initially expressed, the concern was not well-founded, because trailing
8516 spaces are removed early. However, if the trailing spaces were followed by
8517 a NULL, they did not get removed, so the overflow was possible. Two fixes
8520 (a) I re-wrote the offending code in a cleaner fashion.
8521 (b) If an incoming SMTP command contains a NULL character, it is rejected
8524 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8525 runs initgroups(), so that if the Exim user is in any additional groups,
8526 they will be used during message reception.
8532 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob