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 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
60 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
61 related applications. Previously an "H" was used where available info
62 says that "M" should be, so change to match.
64 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
65 as arguments, so an implementation trying to copy these into a local
66 buffer was taking a taint-enforcement trap. Fix by using dynamically
67 created buffers. Similar fix for radius expansion condition.
69 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
70 reasonable, eg. to count headers. Fix by using dynamically created
71 buffers rather than a local. Do similar fixes for ACL actions "dcc",
72 "log_reject_target", "malware" and "spam"; the arguments are expanded
73 so could be handling tainted values.
75 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
76 broken the (no-op) support for this sendmail command. Restore it
77 to doing nothing, silently, and returning good status.
79 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
80 record path was given (or the default used) without a leading directory
81 path, an error occurred on trying to open it. Use the transport's working
84 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
85 transport. Previously it was the name on the DNS A-record; use instead
86 the head of the CNAME chain leading there (if there is one). This seems
87 to align better with RFC 6125.
89 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
90 smtp_accept_max_per_host allocated resources which were not released
91 when the limit was exceeded. This eventually crashed the daemon. Fix
92 by adding a release action in that path.
94 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
95 expanded; previously using tainted values was rejected. Fix by using
96 dynamically-created buffers.
98 JH/09 Relax restrictions on ACL verify condition needing access to message
99 headers. Previously they were only permitted in data and non-smtp ACLs;
100 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
101 not_blind, header_sender and header_names_ascii verification.
103 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
104 Previously a macro used one argument twice; when called with the
105 argument as an expression having side-effects, incorrect operation
106 resulted. Use an inlineable function.
108 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
109 held open for a verify callout. Previously this wan not accounted for
110 and a corrupt onward SMTP conversation resulted.
112 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
113 Previously they were embedded naively in the constructed address; when
114 needed, strip the quoting and quote the entire local_part.
115 Also make the inbound_srs expansion condition handle quoting.
117 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
118 excluded, not matching the documentation.
120 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
121 was given for the sqlite_dbfile a trap resulted.
123 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
124 "name" argument resulted in a trap. There is no reason to disallow such;
125 this was a coding error.
127 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
128 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
129 spent suspended, ignoring the POSIX definition. Previously we assumed
130 it did and a constant offset from real time could be used as a correction.
131 Change to using the same clock source for the start-of-message and the
132 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
133 exists, just to get a clock slightly more aligned to reality.
135 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
136 RFC says it is optional some validators care. The missing char was not
137 intended but triggered by a line-wrap alignment. Discovery and fix by
138 Guillaume Outters, hacked on by JH.
140 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
141 name being quoted was tainted a trap would be taken. Fix by using
142 dynamically created buffers. The routine could have been called by a
143 rewrite with the "h" flag, by using the "-F" command-line option, or
144 by using a "name=" option on a control=submission ACL modifier.
146 JH/19 SPF: change the Authentication-Results expansion component to give
147 smtp.helo when the sender domain is empty. Previously it gave
150 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
151 not in 127.0.0.0/8 to help in spotting list domains taken over by a
152 domain-parking registrar.
154 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
155 Previously when a whitespace character was specified it was not inserted
156 after removing the newline.
158 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
159 the domain part of the recipient address. This overrides any tls_sni
160 option set, which was previously used.
162 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
165 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
166 is_tainted() had an off-by-one error in the overenthusiastic direction.
167 Find and fix by Gavan. Although NetBSD is not a supported platform for
168 4.94 this bug could affect other platforms.
170 PP/01 Fix default prime selection to be consistent.
171 One path used ike23 still, instead of exim.dev.20160529.3; now both
172 execution flows will use the same DH primes (currently
173 exim.dev.20160529.3).
175 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
176 option in smtp transport, to match the documentation. Previously
177 verification was not being done in some cases where it should have been.
179 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
180 than one server was defined and depending on the platform memory layout
181 details, an internal consistency trap could be hit while walking the list
184 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
185 files. The value is used by the authresults expansion item. Previously
186 if this was used in a router or transport, a crash could result.
188 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
189 have a sufficient incidence of EINTR returns from select that an
190 interaction with other operations done by the main daemon loop exposed
191 a bug in the error-handling. This was benign apart from the log
194 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
195 for consistency with delivered "=>" and failed "**" lines. While we're
196 there, handle PRX and TFO.
198 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
199 applied. This resulted, if any header-line rewrite rules were configured,
200 in a panic-log triggerable by sending a message with a long address in
201 a header. Fix by increasing the arbitrary limit to larger than a single
202 (dewrapped) 5322 header line maximum size.
204 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
205 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
206 RFC 6648 which deprecates X- options in protocols as a general practice.
207 Changeover between the implementations is handled by the mechanisms
210 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
211 a remote which uses callout/hold. Previously the local was queued.
213 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
216 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
217 log line. The asterisk is now consistently appended to the remote IP
218 (and port, if given), and will also be provided on defer and fail log
219 lines. Previously it could be placed on the local IP if that was being
220 logged, and was only provided on delivery lines.
222 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
224 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
225 authenticator client_send option. Previously the next char, after a pair
226 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
227 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
228 documentation. There is still no way to get a leading ^ immediately
229 after a NUL (ie. for the password of a PLAIN method authenticator.
231 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
232 files. For bad sizes read, delete the record and whine to paniclog.
234 JH/38 When logging an AUTH failure, as server, do not include sensitive
235 information. Previously, the credentials would be included if given
236 as part of the AUTH command line and an ACL denied authentication.
238 JH/39 Bug 2691: fix $local_part_data. When the matching list element
239 referred to a file, bad data was returned. This likely also affected
242 JH/40 The gsasl authenticator now supports caching of the salted password
243 generated by the client-side implementation. This required the addition
244 of a new variable: $auth4.
246 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
247 left undeleted; the attempt to re-create it then failed - resulting in
248 the usual "SIGHUP tp have daemon reload configuration" to not work.
249 This affected any platform not supporting "abstract" Unix-domain
250 sockets (i.e. not Linux).
252 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
253 recipients" response to RCPT in a later response, with a 250. The
254 previous coding assumed this would not happen, and under PIPELINING
255 would result in both lost and duplicate recipients for a message.
257 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
258 Previously the weighting was incorrectly applied. Similar fix for socks
259 proxies. Found and fixed by Heiko Schlichting.
261 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
262 not handle sub-lists included using the +namedlist syntax. While
263 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
264 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
267 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
268 Previously the permanent pool was used, so the sore could not be freed.
269 This meant a connection with many messages would use continually-growing
272 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
273 per-pool so as not to waste too much space. Previously a constant size
274 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
275 DOS attacks harder. The cost is wasted memory use in the larger blocks.
277 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
278 same space repeatedly, and a smaller process footprint.
280 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
281 for GnuTLS. Previously the name started "%s" which, while not a bug,
282 looked as if if might be one.
284 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
285 "continued connection") make the $proxy_* variables available. Previously
286 the information was not passed across the exec() call for subsequent
287 transport executions. This also mean that the log lines for the
288 messages can show the proxy information.
290 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
291 receive time. With modern systems the difference is significant.
292 The historical behaviour can be restored by disabling (a new) log_selector
293 "queue_time_exclusive".
295 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
296 content) were specifically tested for. Now, also cover expansions
297 resulting in ACL names and inline ACL content.
299 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
300 making it unusable in complex expressions.
302 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
303 fail with a taint-check on the spoolfile name, and leave the message
306 HS/01 Enforce absolute PID file path name.
308 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
310 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
312 PP/02 Bug 2643: Correct TLS DH constants.
313 A missing NUL termination in our code-generation tool had led to some
314 incorrect Diffie-Hellman constants in the Exim source.
315 Reported by kylon94, code-gen tool fix by Simon Arlott.
317 PP/03 Impose security length checks on various command-line options.
318 Fixes CVE-2020-SPRSS reported by Qualys.
320 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
321 better. Reported by Qualys.
323 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
324 providing a particularly obnoxious sender full name.
327 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
329 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
332 PP/08 Change default for recipients_max from unlimited to 50,000.
334 PP/09 Fix security issue with too many recipients on a message (to remove a
335 known security problem if someone does set recipients_max to unlimited,
336 or if local additions add to the recipient list).
337 Fixes CVE-2020-RCPTL reported by Qualys.
339 PP/10 Fix security issue in SMTP verb option parsing
340 Fixes CVE-2020-EXOPT reported by Qualys.
342 PP/11 Fix security issue in BDAT state confusion.
343 Ensure we reset known-good where we know we need to not be reading BDAT
344 data, as a general case fix, and move the places where we switch to BDAT
345 mode until after various protocol state checks.
346 Fixes CVE-2020-BDATA reported by Qualys.
348 HS/03 Die on "/../" in msglog file names
350 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
351 the Exim runtime user are allowed to create files.
353 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
356 QS/03 When reading the output from interpreted forward files we do not
357 pass the pipe between the parent and the interpreting process to
358 executed child processes (if any).
360 QS/04 Always die if requested from internal logging, even is logging is
363 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
364 the API; compilation noo longer completes with DMARC support included.
365 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
366 been reported on other platforms.
368 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
370 JH/56 Make the majority of info read from config files readonly, for defence-in-
371 depth against exploits. Suggestion by Qualys.
372 Not supported on Solaris 10.
374 JH/57 Fix control=fakreject for a custom message containing tainted data.
375 Previously this resulted in a log complaint, due to a re-expansion present
376 since fakereject was originally introduced.
378 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
379 modifier was given, a loop resulted.
381 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
382 pipelined SMTP command followed the BDAT LAST then it would be
383 incorrectly treated as part of the message body, causing a verification
386 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
387 require looser validation rules than those for 821-level addresses,
388 which only permit IP addresses.
394 JH/01 Avoid costly startup code when not strictly needed. This reduces time
395 for some exim process initialisations. It does mean that the logging
396 of TLS configuration problems is only done for the daemon startup.
398 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
400 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
401 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
404 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
405 DKIM signing is being done. Previously a transport_filter would always
406 disable CHUNKING, falling back to traditional DATA.
408 JH/05 Regard command-line recipients as tainted.
410 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
412 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
413 PAM library frees one of the arguments given to it, despite the
414 documentation. Therefore a plain malloc must be used.
416 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
417 on-stack buffers were used, resulting in a taint trap when DSN information
418 copied from a received message was written into the buffer.
420 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
421 the ordering of its ARC headers. This caused a crash.
423 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
424 a new record was being constructed with information from the peer, a trap
427 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
428 installation would get error messages from DMARC verify, when it hit the
429 nonexistent file indicated by the default. Distros wanting DMARC enabled
430 should both provide the file and set the option.
431 Also enforce no DMARC verification for command-line sourced messages.
433 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
434 could, depending on the platform, hang at the STARTTLS response.
436 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
437 message on a connection. Previously if one message had ARC headers and
438 the following one did not, a crash could result when adding an
439 Authentication-Results: header.
441 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
442 Exim main code and Exim-related utities. The introduction of taint
443 tracking also did many adjustments to string handling. Since then, eximon
444 frequently terminated with an assert failure.
446 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
447 check for 452 responses. This slightly helps the inefficieny of doing
448 a large alias-expansion into a recipient-limited target. The max_rcpt
449 transport option still applies (and at the current default, will override
450 the new feature). The check is done for either cause of synch, and forces
451 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
452 connection. The new facility is not tunable at this time.
454 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
455 library live data was being used, so the results became garbage. Make
456 copies while it is still usable.
458 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
459 on delivery deferred (==) and failed (**) lines (if a delivery was
460 attemtped). Previously it was only on completion (=>) lines.
462 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
463 for the expansion of the server_scram_iter and server_scram_salt options.
465 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
466 are now specifically given a NO_DATA response without hitting the system
467 resolver. The library goes on to do the now-standard TXT lookup.
468 Use of dnsdb lookups is not affected.
470 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
471 only retrieve the errormessage once. Previously two calls to dlerror()
472 were used, and the second one (for mainlog/paniclog) retrieved null
475 JH/20 Taint checking: disallow use of tainted data for
476 - the appendfile transport file and directory options
477 - the pipe transport command
478 - the autoreply transport file, log and once options
479 - file names used by the redirect router (including filter files)
481 - paths used by single-key lookups
482 Previously this was permitted.
484 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
485 adjusted the size of a major service buffer; this failed because the
486 buffer was in use at the time. Change to a compile-time increase in the
487 buffer size, when this authenticator is compiled into exim.
489 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
490 previous fast-mode was untenable in the face of glibs using mmap to
491 support larger malloc requests.
493 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
494 New values supported, if defined on system where compiled:
495 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
496 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
498 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
499 running a limited number of proceses in parallel, a benefit is gained. The
500 amount varies with the platform hardware and load. The use of the option
501 queue_run_in_order means we cannot do this, as ordering becomes
504 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
505 had introduced a string-copy (for ensuring NUL-termination) which was not
506 appropriate for that case, which can include embedded NUL bytes in the
507 block of data. Investigation showed the copy to actually be needless, the
508 data being length-specified.
510 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
511 done during a receiving connection, and both used TLS, global info was
512 used rather than per-connection info for tracking the state of data
513 queued for transmission. This could result in a connection hang.
515 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
516 Previously, when delivering serveral messages down a single connection
517 only the first would provide a SIZE. This was due to the size information
518 not being properly tracked.
520 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
521 TAI (at 37 seconds currently), pretend to be in UTC for time-related
522 expansion and logging. Previously, spurious values such as a future
523 minute could be seen.
525 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
526 it could crash from a null-deref. This could also affect the
527 ${addresses: } operator and ${readsock } item.
529 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
530 message following a mime one, the variable was not reset.
532 JH/30 When an pipelined-connect fails at the first response, assume incorrect
533 cached capability (perhaps the peer reneged?) and immediately retry in
536 JH/31 Fix spurious detection of timeout while writing to transport filter.
538 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
539 an attempt to copy the string was made before checking it.
541 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
542 taint of the lookup key was maintained; we now regard the presence in the
543 filesystem as sufficient validation.
545 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
546 argument is supplied.
548 JH/35 The dsearch lookup now requires that the directory is an absolute path.
549 Previously this was not checked, and nonempty relative paths made an
550 access under Exim's current working directory.
552 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
553 Previously no event was raised.
555 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
556 parameter supplied by the sender MAIL FROM command. Previously it was
557 ignored, and only the check_spool_space option value for the required
560 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
561 the size of the signing public-key. Previously it was instead giving
562 the size of the signature hash.
564 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
565 the default. See the (new) dkim_verify_min_keysizes option.
567 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
568 and an ACL use a lookup for checking either the local_part or domain,
569 stale data could be accessed. Ensure that variable references are
570 dropped between messages.
572 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
573 by the client was not checked as pointing within response data before
574 being used. A malicious client could thus cause an out-of-bounds read and
575 possibly gain authentication. Fix by adding the check.
577 JH/42 Internationalisation: change the default for downconversion in the smtp
578 transport to be "if needed". Previously it was "as previously set" for
579 the message, which usually meant "if needed" for message-submission but
580 "no" for everything else. However, MTAs have been seen using SMTPUTF8
581 even when the envelope addresses did not need it, resulting in forwarding
582 failures to non-supporting MTAs. A downconvert in such cases will be
583 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
584 transport. The change does mean that addresses needing conversion will
585 be converted when previously a delivery failure would occur.
587 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
588 response was received it would be used unchecked in a fail-DSN, violating
589 standards on line-length limits. Truncate if needed.
591 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
592 doesn't work. (Additionally add a new main config option to configure the
599 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
600 side, to decode a TLS 1.3 packet capture.
602 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
603 Previously the default library behaviour applied, sending two, each in
606 JH/03 Debug output for ACL now gives the config file name and line number for
609 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
611 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
613 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
614 buffer overrun for (non-chunking) other transports.
616 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
617 TLS1.3, means that a server rejecting a client certificate is not visible
618 to the client until the first read of encrypted data (typically the
619 response to EHLO). Add detection for that case and treat it as a failed
620 TLS connection attempt, so that the normal retry-in-clear can work (if
621 suitably configured).
623 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
624 and/or domain. Found and fixed by Jason Betts.
626 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
627 configuration). If a CNAME target was not a wellformed name pattern, a
630 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
631 the OS reports them interleaved with other addresses.
633 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
634 used both for input and for a verify callout, both encrypted, SMTP
635 responses being sent by the server could be lost. This resulted in
636 dropped connections and sometimes bounces generated by a peer sending
639 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
640 string. Previously it was possible to cause undefined behaviour in a
641 library routine (usually a crash). Found by "zerons".
643 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
646 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
647 API was removed, so update to use the newer ones.
649 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
650 any timeout set, is taking a long time. Previously we would hang on to a
651 rotated logfile "forever" if the input was arriving with long gaps
652 (a previous attempt to fix addressed lack, for a long time, of initial
655 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
656 shared (NFS) environment. The length of the tempfile name is now
657 4 + 16 ("hdr.$message_exim_id") which might break on file
658 systems which restrict the file name length to lower values.
661 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
662 shared (NFS) environment.
664 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
665 did for all versions <4.90). Notably -M, -m, --invert, -I may be
668 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
669 on some platforms for bit 31.
671 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
672 to changes apparently associated with TLS1.3 handling some of the APIs
673 previously used were either nonfunctional or inappropriate. Strings
674 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
675 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
676 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
677 This affects log line X= elements, the $tls_{in,out}_cipher variables,
678 and the use of specific cipher names in the encrypted= ACL condition.
680 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
682 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
683 verification result was not updated unless hosts_require_ocsp applied.
685 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
686 queue_list_requires_admin set to false, non-admin users were denied the
689 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
690 directory-of-certs mode. Previously they were advertised despite the
693 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
694 A single TCP connection by a client will now hold a TLS connection open
695 for multiple message deliveries, by default. Previously the default was to
698 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
699 default. If built with the facility, DANE will be used. The facility
700 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
702 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
703 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
704 must be defined and you must still, unless you define DISABLE_TLS, manage
705 the the include-dir and library-file requirements that go with that
706 choice. Non-TLS builds are still supported.
708 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
711 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
712 default. If the platform supports and has the facility enabled, it will
713 be requested on all coneections.
715 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
716 controlled by the build-time option SUPPORT_PIPE_CONNECT.
718 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
720 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
721 success-DSN messages. Previously the From: header was always the default
722 one for these; the option was ignored.
724 JH/28 Fix the timeout on smtp response to apply to the whole response.
725 Previously it was reset for every read, so a teergrubing peer sending
726 single bytes within the time limit could extend the connection for a
727 long time. Credit to Qualsys Security Advisory Team for the discovery.
729 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
730 delivery address, which leaked information of the results of local
731 forwarding. Change to the original envelope recipient address, per
734 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
735 requested. Previously not bounce was generated and a log entry of
736 error ignored was made.
738 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
740 JH/32 Introduce a general tainting mechanism for values read from the input
741 channel, and values derived from them. Refuse to expand any tainted
742 values, to catch one form of exploit.
744 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
745 was unused and the unexpanded text used for the test. Found and
746 fixed by Ruben Jenster.
748 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
749 an attempt to use a TLS library read routine dereffed a nul pointer,
752 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
753 them in our smtp response.
755 JH/36 Have the general router option retry_use_local_part default to true when
756 any of the restrictive preconditions are set (to anything). Previously it
757 was only for check_local user. The change removes one item of manual
758 configuration which is required for proper retries when a remote router
759 handles a subset of addresses for a domain.
761 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
762 link count into consideration.
764 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
765 caused the extension of big_buffer, the following lines were ignored.
767 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
768 accordance with RFC 2308. Previously there was no expiry, so a longlived
769 receive process (eg. due to ACL delays) versus a short SOA value could
772 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
774 JH/39 Promote DMARC support to mainline.
776 JH/40 Bug 2452: Add a References: header to DSNs.
778 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
779 parameters. The relevant library call is documented as "Deprecated: This
780 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
781 3.6.0, DH parameters are negotiated following RFC7919."
783 HS/06 Change the default of dnssec_request_domains to "*"
785 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
786 carried on and emitted a BDAT command, even when PIPELINING was not
789 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
790 buffer was used for the filename, resulting in a trap when tainted
791 arguments (eg. $domain) were used.
793 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
794 recommended to avoid a possible server-load attack. The feature can be
795 re-enabled via the openssl_options main cofiguration option.
797 JH/45 local_scan API: documented the current smtp_printf() call. This changed
798 for version 4.90 - adding a "more data" boolean to the arguments.
799 Bumped the ABI version number also, this having been missed previously;
800 release versions 4.90 to 4.92.3 inclusive were effectively broken in
801 respect of usage of smtp_printf() by either local_scan code or libraries
802 accessed via the ${dlfunc } expansion item. Both will need coding
803 adjustment for any calls to smtp_printf() to match the new function
804 signature; a FALSE value for the new argument is always safe.
806 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
807 the file-offset (which the Linux syscall does, and exim expects); this
808 resulted in an indefinite loop.
810 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
811 to do ARC verification. The Authentication-Results: header line added
812 by the configuration then had no ARC item.
814 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
815 an incorrect mode was used for reading the result, resulting in it being
822 JH/01 Remove code calling the customisable local_scan function, unless a new
823 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
825 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
826 non-signal-safe functions being used.
828 JH/03 Bug 2269: When presented with a received message having a stupidly large
829 number of DKIM-Signature headers, disable DKIM verification to avoid
830 a resource-consumption attack. The limit is set at twenty.
832 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
833 report of oldest_pass in ${authres } in consequence, and separate out
834 some descriptions of reasons for verification fail.
836 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
837 files in the spool were present and unlocked. A queue-runner could spot
838 them, resulting in a duplicate delivery. Fix that by doing the unlock
839 after the unlink. Investigation by Tim Stewart. Take the opportunity to
840 add more error-checking on spoolfile handling while that code is being
843 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
844 No known attacks, no CVE, this is defensive hardening.
846 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
847 a queue-runner could start a delivery while other operations were ongoing.
848 Cutthrough delivery was a common victim, resulting in duplicate delivery.
849 Found and investigated by Tim Stewart. Fix by using the open message data
850 file handle rather than opening another, and not locally closing it (which
851 releases a lock) for that case, while creating the temporary .eml format
852 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
854 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
855 $sender_verify_failure/$recipient_verify_failure to "random".
857 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
860 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
861 Previously this would segfault.
863 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
866 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
867 like zero, since the resolver should be doing this for us, But we need one
868 as a CNAME but no MX presence gets the CNAME returned; we need to check
869 that doesn't point to an MX to declare it "no MX returned" rather than
870 "error, loop". A new main option is added so the older capability of
871 following some limited number of chain links is maintained.
873 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
875 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
876 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
877 now-deprecated earlier definitions used only specified the range up to TLS
878 1.2 (in the older-version library docs).
880 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
882 JH/15 Rework TLS client-side context management. Stop using a global, and
883 explicitly pass a context around. This enables future use of TLS for
884 connections to service-daemons (eg. malware scanning) while a client smtp
885 connection is using TLS; with cutthrough connections this is quite likely.
887 JH/16 Fix ARC verification to do AS checks in reverse order.
889 JH/17 Support a "tls" option on the ${readsocket } expansion item.
891 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
892 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
893 Previously the "utf8" would be re-prepended for every additional message.
895 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
896 Previously thery were accepted, resulting in issues when attempting to
897 forward messages to a non-supporting MTA.
899 PP/02 Let -n work with printing macros too, not just options.
901 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
902 one parent address was copied, and bogus data was used at delivery-logging
903 time. Either a crash (after delivery) or bogus log data could result.
904 Discovery and analysis by Tim Stewart.
906 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
907 Previously if the string ended mid-character, we did not insert the
908 promised '?' replacement.
910 PP/04 Documentation: current string operators work on bytes, not codepoints.
912 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
913 should pack well giving a smaller memory footprint so better caching and
914 therefore performance. Group the declarations where this can't be done so
915 that the byte-sized flag variables are not interspersed among pointer
916 variables, giving a better chance of good packing by the compiler.
918 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
919 non-null, to avoid issues with sites running BATV. Previously reports were
920 sent with an empty envelope sender so looked like bounces.
922 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
923 The ignore_error flag wasn't being returned from the filter subprocess so
924 was not set for later routers. Investigation and fix by Matthias Kurz.
926 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
927 and a msg:complete for the whole, when a message is manually removed using
928 -Mrm. Developement by Matthias Kurz, hacked on by JH.
930 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
931 a "Gnu special" function, asprintf() in the DB utility binary builds; I
932 hope that is portable enough.
934 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
935 requiring a known-CA anchor certificate; make it now rely entirely on the
936 TLSA as an anchor. Checking the name on the leaf cert against the name
937 on the A-record for the host is still done for TA (but not for EE mode).
939 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
940 included in delivery lines for non-proxied connections, when compiled with
941 SUPPORT_SOCKS and running with proxy logging enabled.
943 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
944 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
945 move the existing event to fire before the normal logging of message
946 failure so that custom logging is bracketed by normal logging.
948 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
949 msg:fail:internal event. Developement by Matthias Kurz.
951 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
952 far too small for todays use of crypto signatures stored there. Go all
953 the way to the max DNS message size of 64kB, even though this might be
954 overmuch for IOT constrained device use.
956 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
957 copy a string over itself. The library routine is documented as not
958 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
960 JH/32 For main options check_spool_space and check_inode_space, where the
961 platform supports 64b integers, support more than the previous 2^31 kB
962 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
963 the previous G, M, k.
965 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
966 $authenticated_fail_id variable on authentication failure. Previously
969 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
970 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
971 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
972 GNUTLS_SEC_PARAM_MEDIUM.
974 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
975 side. Previously we would continue as if no SNI had been received.
977 JH/36 Harden the handling of string-lists. When a list consisted of a sole
978 "<" character, which should be a list-separator specification, we walked
979 off past the nul-terimation.
981 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
982 causes) even when the retry time is not yet met. Previously they were
983 not, meaning that when (say) an account was over-quota and temp-rejecting,
984 and multiple senders' messages were queued, only one sender would get
985 notified on each configured delay_warning cycle.
987 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
989 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
990 cipher-suites, an error can be left on the stack even for a succeeding
991 accept; this results in impossible error messages when a later operation
994 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
995 return error codes indicating retry. Under TLS1.3 this becomes required.
997 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
998 it only wrote the new authenticators, resulting in a lack of tracking of
999 peer changes of ESMTP extensions until the next cache flush.
1001 JH/41 Fix the loop reading a message header line to check for integer overflow,
1002 and more-often against header_maxsize. Previously a crafted message could
1003 induce a crash of the recive process; now the message is cleanly rejected.
1005 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1006 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1012 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1013 When redis_servers is set to a list of > 1 element, and the Redis servers
1014 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1015 case of MOVED into a DEFER case instead, thus moving the query onto the
1016 next server in the list. For a cluster of N elements, all N servers must
1017 be defined in redis_servers.
1019 GF/02 Catch and remove uninitialized value warning in exiqsumm
1020 Check for existence of @ARGV before looking at $ARGV[0]
1022 JH/01 Replace the store_release() internal interface with store_newblock(),
1023 which internalises the check required to safely use the old one, plus
1024 the allocate and data copy operations duplicated in both (!) of the
1025 extant use locations.
1027 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1028 modifier. This matches the restriction on the commandline.
1030 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1031 Previously only the last row was returned.
1033 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1034 we assumed that tags in the header were well-formed, and parsed the
1035 element content after inspecting only the first char of the tag.
1036 Assumptions at that stage could crash the receive process on malformed
1039 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1040 While running the DKIM ACL we operate on the Permanent memory pool so that
1041 variables created with "set" persist to the DATA ACL. Also (at any time)
1042 DNS lookups that fail create cache records using the Permanent pool. But
1043 expansions release any allocations made on the current pool - so a dnsdb
1044 lookup expansion done in the DKIM ACL releases the memory used for the
1045 DNS negative-cache, and bad things result. Solution is to switch to the
1046 Main pool for expansions.
1047 While we're in that code, add checks on the DNS cache during store_reset,
1048 active in the testsuite.
1049 Problem spotted, and debugging aided, by Wolfgang Breyha.
1051 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1052 When none of the hosts presented to a transport match an already-open
1053 connection, close it and proceed with the list. Previously we would
1054 queue the message. Spotted by Lena with Yahoo, probably involving
1057 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1058 Previously a spurious "250 OK id=" response was appended to the proper
1061 JH/08 The "support for" informational output now, which built with Content
1062 Scanning support, has a line for the malware scanner interfaces compiled
1063 in. Interface can be individually included or not at build time.
1065 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1066 by the template makefile "src/EDITME". The "STREAM" support for an older
1067 ClamAV interface method is removed.
1069 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1070 rows affected is given instead).
1072 JH/11 The runtime Berkeley DB library version is now additionally output by
1073 "exim -d -bV". Previously only the compile-time version was shown.
1075 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1076 SMTP connection. Previously, when one had more recipients than the
1077 first, an abortive onward connection was made. Move to full support for
1078 multiple onward connections in sequence, handling cutthrough connection
1079 for all multi-message initiating connections.
1081 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1082 routers. Previously, a multi-recipient message would fail to match the
1083 onward-connection opened for the first recipient, and cause its closure.
1085 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1086 a timeout on read on a GnuTLS initiating connection, resulting in the
1087 initiating connection being dropped. This mattered most when the callout
1088 was marked defer_ok. Fix to keep the two timeout-detection methods
1091 JH/15 Relax results from ACL control request to enable cutthrough, in
1092 unsupported situations, from error to silently (except under debug)
1093 ignoring. This covers use with PRDR, frozen messages, queue-only and
1096 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1098 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1099 metadata, resulting in a crash in free().
1101 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1102 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1103 Broken also in d185889f4, with init system revamp.
1105 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1106 Previously we abruptly closed the connection after reading a malware-
1107 found indication; now we go on to read the "scan ok" response line,
1110 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1111 ACL. Previously, a crash would result.
1113 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1114 macro text after a replacement (previously it was only once per line) and
1115 by skipping builtin macros when searching for an uppercase lead character.
1117 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1118 for the build is renamed.
1120 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1121 was allocated for every new TLS startup, meaning one per message. Fix
1122 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1124 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1125 reported the original. Fix to report (as far as possible) the ACL
1126 result replacing the original.
1128 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1129 OpenSSL. Certificate information is loaded for every new TLS startup,
1130 and the resources needed to be freed.
1132 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1134 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1137 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1138 DATA response info to the (existing) per-recipient response info for
1139 the "C=" log element. It can have useful tracking info from the
1140 destination system. Patch from Simon Arlott.
1142 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1143 length value. Previously this would segfault.
1145 HS/02 Support Avast multiline protoocol, this allows passing flags to
1146 newer versions of the scanner.
1148 JH/28 Ensure that variables possibly set during message acceptance are marked
1149 dead before release of memory in the daemon loop. This stops complaints
1150 about them when the debug_store option is enabled. Discovered specifically
1151 for sender_rate_period, but applies to a whole set of variables.
1152 Do the same for the queue-runner and queue-list loops, for variables set
1153 from spool message files. Do the same for the SMTP per-message loop, for
1154 certain variables indirectly set in ACL operations.
1156 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1157 as a multi-recipient message from a mailinglist manager). The coding had
1158 an arbitrary cutoff number of characters while checking for more input;
1159 enforced by writing a NUL into the buffer. This corrupted long / fast
1160 input. The problem was exposed more widely when more pipelineing of SMTP
1161 responses was introduced, and one Exim system was feeding another.
1162 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1163 receiving system, and refused recipients seen by the sending system
1164 (propating to people being dropped from mailing lists).
1165 Discovered and pinpointed by David Carter.
1167 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1168 replaced by the ${authresults } expansion.
1170 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1172 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1173 allows proper process termination in container environments.
1175 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1176 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1178 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1179 and "err_perm", deprecated since 4.83 when the RFC-defined words
1180 "temperror" and "permerror" were introduced.
1182 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1183 transport-filters or DKIM-signing. The restriction was lost in the
1184 consolidation of verify-callout and delivery SMTP handling.
1185 Extend the restriction to also cover ARC-signing.
1187 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1188 in defer=pass mode supply a 450 to the initiator. Previously the message
1191 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1192 tls_require_ciphers is used as before.
1194 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1195 "pass_unscanned". Only tmpfails from the scanner are written to
1196 the paniclog, as they may require admin intervention (permission
1197 denied, license issues). Other scanner errors (like decompression
1198 bombs) do not cause a paniclog entry.
1200 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1201 Previously it was possible to log spurious information in receive log
1204 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1205 triggered odd behaviour from Outlook Express clients.
1207 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1210 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1211 since the IETF WG has not yet settled on that versus the original
1212 "bare" representation.
1214 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1215 Previously the millisecond value corrupted the output.
1216 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1217 corrupted the output.
1223 JH/01 Rework error string handling in TLS interface so that the caller in
1224 more cases is responsible for logging. This permits library-sourced
1225 string to be attached to addresses during delivery, and collapses
1226 pairs of long lines into single ones.
1228 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1229 during configuration. Wildcards are allowed and expanded.
1231 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1232 This permits better logging.
1234 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1235 do not close it down and have the child transport start it up again on
1236 the passed-on TCP connection. Instead, proxy the child (and any
1237 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1238 affected: the continued delivery log lines do not have any DNSSEC, TLS
1239 Certificate or OCSP information. TLS cipher information is still logged.
1241 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1242 identical IP addresses on different listening ports. Will also affect
1245 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1246 add noisy ifdef guards to special-case this sillyness.
1247 Patch from Bernd Kuhls.
1249 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1250 than 255 are no longer allowed.
1252 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1253 Disable the session-cache too, which might reduce our load. Since we
1254 currrectly use a new context for every connection, both as server and
1255 client, there is no benefit for these.
1256 GnuTLS appears to not support tickets server-side by default (we don't
1257 call gnutls_session_ticket_enable_server()) but client side is enabled
1258 by default on recent versions (3.1.3 +) unless the PFS priority string
1261 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1262 <https://reproducible-builds.org/specs/source-date-epoch/>.
1264 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1265 the check for any unsuccessful recipients did not notice the limit, and
1266 erroneously found still-pending ones.
1268 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1269 MSG_MORE. Only in-clear (not on TLS connections).
1271 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1272 transport-filter is involved we can buffer the headers in memory for
1273 creating the signature, and read the spool data file once for the
1274 signature and again for transmission.
1276 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1277 4.77 as the kernel support then wasn't solid, having issues in 64bit
1278 mode. Now, it's been long enough. Add support for FreeBSD also.
1280 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1281 case where the routing stage had gathered several addresses to send to
1282 a host before calling the transport for the first, we previously failed
1283 to close down TLS in the old transport process before passing the TCP
1284 connection to the new process. The new one sent a STARTTLS command
1285 which naturally failed, giving a failed delivery and bloating the retry
1286 database. Investigation and fix prototype from Wolfgang Breyha.
1288 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1289 false-negatives in the check that the sender had not preempted a response
1290 or prompt from Exim (running as a server), due to that code's lack of
1291 awareness of the SMTP input buffering.
1293 PP/04 Add commandline_checks_require_admin option.
1294 Exim drops privileges sanely, various checks such as -be aren't a
1295 security problem, as long as you trust local users with access to their
1296 own account. When invoked by services which pass untrusted data to
1297 Exim, this might be an issue. Set this option in main configuration
1298 AND make fixes to the calling application, such as using `--` to stop
1301 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1302 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1303 responses to those, into a single TLS record each way (this usually means
1304 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1307 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1308 affects you only if you're dancing at the edge of the param size limits.
1309 If you are, and this message makes sense to you, then: raise the
1310 configured limit or use OpenSSL 1.1. Nothing we can do for older
1313 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1314 cmdline element to get the documented default one. Previously it was
1317 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1318 are unsuable for non-message-specific reasons - eg. connection timeout,
1319 banner-time rejection.
1321 JH/16 Fix logging of delivery remote port, when specified by router, under
1324 PP/06 Repair manualroute's ability to take options in any order, even if one
1325 is the name of a transport.
1328 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1330 JH/17 Change the list-building routines interface to use the expanding-string
1331 triplet model, for better allocation and copying behaviour.
1333 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1334 Previously it was constructed the first time a possibly-matching string
1335 was met in the configuration file input during startup; now it is done
1338 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1339 compatible one, to avoid the (poorly documented) possibility of a config
1340 file in the working directory redirecting the DB files, possibly correpting
1341 some existing file. CVE-2017-10140 assigned for BDB.
1343 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1344 cache-hot. Previously, although the result was properly cached, the
1345 initial verify call returned a defer.
1347 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1348 the main verify for receipient in uncached-mode.
1350 JH/22 Retire historical build files to an "unsupported" subdir. These are
1351 defined as "ones for which we have no current evidence of testing".
1353 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1354 if present. Previously it was ignored.
1356 JH/24 Start using specified-initialisers in C structure init coding. This is
1357 a C99 feature (it's 2017, so now considered safe).
1359 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1360 if was a fixed-sized field and bitmask ops via macros; it is now more
1363 PP/07 GitHub PR 56: Apply MariaDB build fix.
1364 Patch provided by Jaroslav Škarvada.
1366 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1367 during Coverity cleanups [4.87 JH/47]
1368 Diagnosis and fix provided by Michael Fischer v. Mollard.
1370 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1371 the right size to place the terminating semicolon on its own folded
1372 line, the header hash was calculated to an incorrect value thanks to
1373 the (relaxed) space the fold became.
1375 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1376 and confused the parent.
1378 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1379 which could crash as a result. This could lead to undeliverable messages.
1381 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1384 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1385 matters on fast-turnover and PID-randomising systems, which were getting
1386 out-of-order delivery.
1388 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1389 a possibly-overlapping copy. The symptom was that "Remote host closed
1390 connection in response to HELO" was logged instead of the actual 4xx
1393 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1394 Previously only that bufferd was discarded, resulting in SYMTP command
1397 JH/32 DKIM: when a message has multiple signatures matching an identity given
1398 in dkim_verify_signers, run the dkim acl once for each. Previously only
1399 one run was done. Bug 2189.
1401 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1402 "panic the current process" to "deliberately defer". The panic log is
1403 still written with the problem list name; the mail and reject logs now
1404 get a temp-reject line for the message that was being handled, saying
1405 something like "domains check lookup or other defer". The SMTP 451
1406 message is still "Temporary local problem".
1408 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1409 A crafted sequence of BDAT commands could result in in-use memory beeing
1410 freed. CVE-2017-16943.
1412 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1413 from SMTP input. Previously it was always done; now only done for DATA
1414 and not BDAT commands. CVE-2017-16944.
1416 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1417 to the message (such as an overlong header line). Previously this was
1418 not done and we did not exit BDAT mode. Followon from the previous item
1419 though a different problem.
1425 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1426 than -2003 did; needs libidn2 in addition to libidn.
1428 JH/02 The path option on a pipe transport is now expanded before use.
1430 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1431 Patch provided by "Björn", documentation fix added too.
1433 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1434 missing a wire-to-host endian conversion.
1436 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1437 close after a BDAT command line could be taken as a following command,
1438 giving a synch failure. Fix by only checking for synch immediately
1439 before acknowledging the chunk.
1441 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1442 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1443 macro. Patches provided by Josh Soref.
1445 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1446 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1447 by IANA as a service extension. However, John Klensin suggests that we
1450 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1451 the dkim code may be unix-mode line endings rather than smtp wire-format
1452 CRLF, so prepend a CR to any bare LF.
1454 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1455 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1457 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1458 the first were themselves being wrongly included in the feed into dkim
1459 processing; with most chunk sizes in use this resulted in an incorrect
1460 body hash calculated value.
1462 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1463 DKIM signature block, for verification. Although advised against by
1464 standards it is specifically not ruled illegal.
1466 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1468 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1469 missing a body hash (the bh= tag).
1471 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1472 It seems that HAProxy sends the Proxy Protocol information in clear and
1473 only then does a TLS startup, so do the same.
1475 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1476 TCP connections (such as for Spamd) unless the daemon successfully set
1477 Fast Open mode on its listening sockets. This fixes breakage seen on
1478 too-old kernels or those not configured for Fast Open, at the cost of
1479 requiring both directions being enabled for TFO, and TFO never being used
1480 by non-daemon-related Exim processes.
1482 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1483 endings, at least on the first header line. Try to canonify any that get
1484 past that check, despite the cost.
1486 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1487 now limited to an arbitrary five deep, while parsing addresses with the
1488 strip_excess_angle_brackets option enabled.
1490 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1491 instead leave the unprompted TLS handshake in socket buffer for the
1492 TLS library to consume.
1494 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1496 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1498 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1499 region in one operation, for speed, and this leaves assigned pointers
1500 dangling. Add checks run only under the testsuite which checks all
1501 variables at a store-reset and panics on a dangling pointer; add code
1502 explicitly nulling out all the variables discovered. Fixes one known
1503 bug: a transport crash, where a dangling pointer for $sending_ip_address
1504 originally assigned in a verify callout, is re-used.
1506 PP/06 Drop '.' from @INC in various Perl scripts.
1508 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1510 PP/08 Reduce a number of compilation warnings under clang; building with
1511 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1512 should be warning-free.
1514 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1516 HS/01 Fix portability problems introduced by PP/08 for platforms where
1517 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1519 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1520 chunk. This allows us to accept broken chunked messages. We need a more
1521 general solution here.
1523 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1524 already-broken messages in the queue.
1526 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1528 JH/19 Fix reference counting bug in routing-generated-address tracking.
1534 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1535 supports it and a size is available (ie. the sending peer gave us one).
1537 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1538 years of being deprecated). The replacements are the ACLs
1539 acl_smtp_mime and acl_not_smtp_mime.
1541 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1542 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1543 or even an in-clear connection were permitted. Now, if the host lookup
1544 was dnssec and dane was requested then the host is only used if the
1545 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1546 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1547 if one fails this test.
1548 This means that a poorly-configured remote DNS will make it incommunicado;
1549 but it protects against a DNS-interception attack on it.
1551 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1552 non-noisy when a race steals the message being considered.
1554 JH/05 If main configuration option tls_certificate is unset, generate a
1555 self-signed certificate for inbound TLS connections.
1557 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1558 in rewrites and routers.
1560 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1561 and logged a warning sing 4.83; now they are a configuration file error.
1563 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1564 (lacking @domain). Apply the same qualification processing as RCPT.
1566 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1568 JH/10 Support ${sha256:} applied to a string (as well as the previous
1571 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1572 a cutthrough deliver is pending, as we always want to make a connection.
1573 This also avoids re-routing the message when later placing the cutthrough
1574 connection after a verify cache hit.
1575 Do not update it with the verify result either.
1577 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1578 when routing results in more than one destination address.
1580 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1581 signing (which inhibits the cutthrough capability). Previously only
1582 the presence of an option was tested; now an expansion evaluating as
1583 empty is permissible (obviously it should depend only on data available
1584 when the cutthrough connection is made).
1586 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1587 the relevant preceding SMTP command did not note the pipelining mode.
1589 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1590 Previously they were not counted.
1592 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1593 as one having no matching records. Previously we deferred the message
1594 that needed the lookup.
1596 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1597 distinguished as "(=".
1599 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1600 for missing MX records. Previously it only worked for missing A records.
1602 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1604 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1605 after the data-go-ahead and data-ack. Patch from Jason Betts.
1607 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1608 even for a "none" policy. Patch from Tony Meyer.
1610 JH/22 Fix continued use of a connection for further deliveries. If a port was
1611 specified by a router, it must also match for the delivery to be
1614 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1615 When one of the recipients of a message was unsuitable for the connection
1616 (has no matching addresses), we lost track of needing to mark it
1617 deferred. As a result mail would be lost.
1619 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1621 JH/25 Decoding ACL controls is now done using a binary search; the source code
1622 takes up less space and should be simpler to maintain. Merge the ACL
1623 condition decode tables also, with similar effect.
1625 JH/26 Fix problem with one_time used on a redirect router which returned the
1626 parent address unchanged. A retry would see the parent address marked as
1627 delivered, so not attempt the (identical) child. As a result mail would
1630 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1631 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1632 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1635 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1636 Main config options check_{log,spool}_{inodes,space} are now
1637 100 inodes, 10MB unless set otherwise in the configuration.
1639 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1640 Previously it only applied to the main-section connection policy
1643 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1645 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1646 by me. Added RFC7919 DH primes as an alternative.
1648 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1649 are not in the system include path.
1651 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1652 GnuTLS, when a session startup failed (eg because the client disconnected)
1653 Exim did stdio operations after fclose. This was exposed by a recent
1654 change which nulled out the file handle after the fclose.
1656 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1657 signed directly by the cert-signing cert, rather than an intermediate
1658 OCSP-signing cert. This is the model used by LetsEncrypt.
1660 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1662 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1663 an incoming connection.
1665 HS/02 Bug 1802: Do not half-close the connection after sending a request
1668 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1669 fallback to "prime256v1".
1671 JH/34 SECURITY: Use proper copy of DATA command in error message.
1672 Could leak key material. Remotely exploitable. CVE-2016-9963.
1678 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1679 and 3.4.4 - once the server is enabled to respond to an OCSP request
1680 it does even when not requested, resulting in a stapling non-aware
1681 client dropping the TLS connection.
1683 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1684 support variable-length bit vectors. No functional change.
1686 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1687 The I= interface field on outgoing lines is now after the H= remote
1688 host field, same as incoming lines. There is a separate
1689 outgoing_interface log selector which allows you to disable the
1692 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1693 If not running log_selector +smtp_connection the mainlog would be held
1694 open indefinitely after a "too many connections" event, including to a
1695 deleted file after a log rotate. Leave the per net connection logging
1696 leaving it open for efficiency as that will be quickly detected by the
1697 check on the next write.
1699 HS/01 Bug 1671: Fix post transport crash.
1700 Processing the wait-<transport> messages could crash the delivery
1701 process if the message IDs didn't exist for some reason. When
1702 using 'split_spool_directory=yes' the construction of the spool
1703 file name failed already, exposing the same netto behaviour.
1705 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1706 mime_regex ACL conditions.
1708 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1709 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1710 to HELO, local diagnostic string.
1712 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1713 log line to debug. Even when configured with a tls authenticator many
1714 client connections are expected to not authenticate in this way, so
1715 an authenticate fail is not an error.
1717 HS/02 Add the Exim version string to the process info. This way exiwhat
1718 gives some more detail about the running daemon.
1720 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1721 matter for fast-change records such as DNSBLs.
1723 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1724 retry record, even if constant. There may be multiple transports with
1725 different interface settings and the retry behaviour needs to be kept
1728 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1730 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1732 JH/10 Bug 840: fix log_defer_output option of pipe transport
1734 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1735 hosts_randomize. This matters a lot when combined with mua_wrapper.
1737 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1738 ${quote_pgsql:<string>} operator.
1740 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1742 JH/14 Change header file name for freeradius-client. Relevant if compiling
1743 with Radius support; from the Gentoo tree and checked under Fedora.
1745 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1747 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1748 option transport_filter, meaning no filtering.
1750 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1752 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1753 defaults to "*" (all hosts). The variable is now available when not built
1754 with TLS, default unset, mainly to enable keeping the testsuite sane.
1755 If a server certificate is not supplied (via tls_certificate) an error is
1756 logged, and clients will find TLS connections fail on startup. Presumably
1757 they will retry in-clear.
1758 Packagers of Exim are strongly encouraged to create a server certificate
1759 at installation time.
1761 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1762 with the $config_file variable.
1764 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1765 in transport context, after the attempt, and per-recipient. The latter type
1766 is per host attempted. The event data is the error message, and the errno
1767 information encodes the lookup type (A vs. MX) used for the (first) host,
1768 and the trailing two digits of the smtp 4xx response.
1770 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1771 to write to mainlog (or rejectlog, paniclog) in the window between file
1772 creation and permissions/ownership being changed. Particularly affects
1773 installations where exicyclog is run as root, rather than exim user;
1774 result is that the running daemon panics and dies.
1776 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1778 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1779 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1780 "pri" and "weight". Note that the previous implicit priority given by the
1781 list order is no longer honoured.
1783 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1784 for DKIM processing.
1786 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1787 by defining SUPPORT_SOCKS.
1789 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1790 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1791 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1792 variables are renamed to proxy_{local,external}_{address,port}.
1794 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1795 for a build by defining SUPPORT_I18N
1797 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1798 of the query string, and make ${quote_redis:} do that quoting.
1800 JH/29 Move Events support from Experimental to mainline, enabled by default
1801 and removable for a build by defining DISABLE_EVENT.
1803 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1805 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1806 cached by the daemon.
1808 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1809 by defining LOOKUP_REDIS. The libhiredis library is required.
1811 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1812 keys are given for lookup.
1814 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1815 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1816 only supported when built with TLS support. The PolarSSL SHA routines
1817 are still used when the TLS library is too old for convenient support.
1819 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1820 openssl_options), for security. OpenSSL forces this from version 1.1.0
1821 server-side so match that on older versions.
1823 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1824 allocation for $value could be released as the expansion processing
1825 concluded, but leaving the global pointer active for it.
1827 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1828 and to use the domains and local_parts ACL conditions.
1830 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1831 incorrectly not doubled on cutthrough transmission, hence seen as a
1832 body-termination at the receiving system - resulting in truncated mails.
1833 Commonly the sender saw a TCP-level error, and retransmitted the message
1834 via the normal store-and-forward channel. This could result in duplicates
1835 received - but deduplicating mailstores were liable to retain only the
1836 initial truncated version.
1838 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1840 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1842 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1843 we're in there, support oversigning also; bug 1309.
1845 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1847 HS/04 Add support for keep_environment and add_environment options.
1849 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1850 either intentional arithmetic overflow during PRNG, or testing config-
1853 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1854 delivery resulted in actual delivery. Cancel cutthrough before DATA
1857 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1858 rejects a recipient: pass the reject to the originator.
1860 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1861 Many were false-positives and ignorable, but it's worth fixing the
1864 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1865 for the new environment-manipulation done at startup. Move the routines
1866 from being local to tls.c to being global via the os.c file.
1868 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1869 an extract embedded as result-arg for a map, the first arg for extract
1870 is unavailable so we cannot tell if this is a numbered or keyed
1871 extraction. Accept either.
1877 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1880 JH/02 The smtp transport option "multi_domain" is now expanded.
1882 JH/03 The smtp transport now requests PRDR by default, if the server offers
1885 JH/04 Certificate name checking on server certificates, when exim is a client,
1886 is now done by default. The transport option tls_verify_cert_hostnames
1887 can be used to disable this per-host. The build option
1888 EXPERIMENTAL_CERTNAMES is withdrawn.
1890 JH/05 The value of the tls_verify_certificates smtp transport and main options
1891 default to the word "system" to access the system default CA bundle.
1892 For GnuTLS, only version 3.0.20 or later.
1894 JH/06 Verification of the server certificate for a TLS connection is now tried
1895 (but not required) by default. The verification status is now logged by
1896 default, for both outbound TLS and client-certificate supplying inbound
1899 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1902 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1903 Status Notification (bounce) messages are now MIME format per RFC 3464.
1904 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1905 under the control of the dsn_advertise_hosts option, and routers may
1906 have a dsn_lasthop option.
1908 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1909 default, modifiable by a malware= option. The list separator for
1910 the options can now be changed in the usual way. Bug 68.
1912 JH/10 The smtp_receive_timeout main option is now expanded before use.
1914 JH/11 The incoming_interface log option now also enables logging of the
1915 local interface on delivery outgoing connections.
1917 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1918 if the interface and destination host and port all match.
1920 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1923 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1924 Patch from Andrew Lewis.
1926 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1927 now supports optional time-restrictions, weighting, and priority
1928 modifiers per server. Patch originally by <rommer@active.by>.
1930 JH/16 The spamd_address main option now supports a mixed list of local
1931 and remote servers. Remote servers can be IPv6 addresses, and
1932 specify a port-range.
1934 JH/17 Bug 68: The spamd_address main option now supports an optional
1935 timeout value per server.
1937 JH/18 Bug 1581: Router and transport options headers_add/remove can
1938 now have the list separator specified.
1940 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1943 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1946 JH/21 Support for the A6 type of dns record is withdrawn.
1948 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1949 rather than the verbs used.
1951 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1952 from 255 to 1024 chars.
1954 JH/24 Verification callouts now attempt to use TLS by default.
1956 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1957 are generic router options now. The defaults didn't change.
1959 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1960 Original patch from Alexander Shikoff, worked over by JH.
1962 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1963 files if ZCAT_COMMAND is not executable.
1965 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1967 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1969 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1970 Normally benign, it bites when the pair was led to by a CNAME;
1971 modern usage is to not canonicalize the domain to a CNAME target
1972 (and we were inconsistent anyway for A-only vs AAAA+A).
1974 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1976 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1977 when evaluating $sender_host_dnssec.
1979 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1980 $sender_helo_dnssec variable.
1982 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1984 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1986 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1988 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1989 documented as working, but never had. Support all but $spam_report.
1991 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1992 added for tls authenticator.
1994 HS/03 Add perl_taintmode main config option
2000 TL/01 When running the test suite, the README says that variables such as
2001 no_msglog_check are global and can be placed anywhere in a specific
2002 test's script, however it was observed that placement needed to be near
2003 the beginning for it to behave that way. Changed the runtest perl
2004 script to read through the entire script once to detect and set these
2005 variables, reset to the beginning of the script, and then run through
2006 the script parsing/test process like normal.
2008 TL/02 The BSD's have an arc4random API. One of the functions to induce
2009 adding randomness was arc4random_stir(), but it has been removed in
2010 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2011 function when detected.
2013 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2014 cause callback expansion.
2016 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2017 syntax errors in an expansion can be treated as a string instead of
2018 logging or causing an error, due to the internal use of bool_lax
2019 instead of bool when processing it.
2021 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2022 server certificates when making smtp deliveries.
2024 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2026 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2028 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2030 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2031 Merged patch from Sebastian Wiedenroth.
2033 JH/05 Fix results-pipe from transport process. Several recipients, combined
2034 with certificate use, exposed issues where response data items split
2035 over buffer boundaries were not parsed properly. This eventually
2036 resulted in duplicates being sent. This issue only became common enough
2037 to notice due to the introduction of connection certificate information,
2038 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2040 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2041 size buffer was used, resulting in syntax errors when an expansion
2044 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2045 version 3.3.6 or later.
2047 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2048 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2049 both become "event_action", the variables become $event_name, $event_data
2050 and $event_defer_errno. There is a new variable $verify_mode, usable in
2051 routers, transports and related events. The tls:cert event is now also
2052 raised for inbound connections, if the main configuration event_action
2055 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2056 early OCSP support, but no stapling (appears to be less than 1.0.0).
2058 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2059 server certificate names available under the smtp transport option
2060 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2063 JH/10 Time-related extraction expansions from certificates now use the main
2064 option "timezone" setting for output formatting, and are consistent
2065 between OpenSSL and GnuTLS compilations. Bug 1541.
2067 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2068 encoded parameter in the incoming message. Bug 1558.
2070 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2071 include certificate info, eximon was claiming there were spoolfile
2074 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2076 JH/14 Log delivery-related information more consistently, using the sequence
2077 "H=<name> [<ip>]" wherever possible.
2079 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2080 are problematic for Debian distribution, omit them from the release
2083 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2085 JH/16 Fix string representation of time values on 64bit time_t architectures.
2088 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2089 output list separator was used.
2094 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2095 checkers that were complaining about end of non-void function with no
2098 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2099 This was a regression introduced in 4.83 by another bugfix.
2101 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2103 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2104 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2110 TF/01 Correctly close the server side of TLS when forking for delivery.
2112 When a message was received over SMTP with TLS, Exim failed to clear up
2113 the incoming connection properly after forking off the child process to
2114 deliver the message. In some situations the subsequent outgoing
2115 delivery connection happened to have the same fd number as the incoming
2116 connection previously had. Exim would try to use TLS and fail, logging
2117 a "Bad file descriptor" error.
2119 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2120 utilities have not been installed.
2122 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2123 temporary space as the ACL may create new global variables.
2125 TL/01 LDAP support uses per connection or global context settings, depending
2126 upon the detected version of the libraries at build time.
2128 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2129 to extract and use the src ip:port in logging and expansions as if it
2130 were a direct connection from the outside internet. PPv2 support was
2131 updated based on HAProxy spec change in May 2014.
2133 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2135 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2136 Properly escape header and check for NULL return.
2138 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2141 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2143 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2144 characters in header names, implemented as a verify condition.
2145 Contributed by Michael Fischer v. Mollard.
2147 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2148 results permerror and temperror. Previous values are deprecated but
2149 still accepted. In a future release, err_perm and err_temp will be
2150 completely removed, which will be a backward incompatibility if the
2151 ACL tests for either of these two old results. Patch contributed by
2152 user bes-internal on the mailing list.
2154 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2156 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2157 selectors, in both main and reject logs.
2159 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2162 JH/07 Add malware type "sock" for talking to simple daemon.
2164 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2166 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2167 routers/transports under cutthrough routing.
2169 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2170 numbers. Touch up "bool" conditional to keep the same definition.
2172 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2174 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2176 JH/12 Expand items in router/transport headers_add or headers_remove lists
2177 individually rather than the list as a whole. Bug 1452.
2179 Required for reasonable handling of multiple headers_ options when
2180 they may be empty; requires that headers_remove items with embedded
2181 colons must have them doubled (or the list-separator changed).
2183 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2184 view the policy declared in the DMARC record. Currently, $dmarc_status
2185 is a combined value of both the record presence and the result of the
2188 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2190 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2191 dnslookup router and the smtp transport (applying to the forward
2194 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2195 of ldap servers used for a specific lookup. Patch provided by Heiko
2198 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2199 New variable $lookup_dnssec_authenticated for observability.
2201 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2202 Patch submitted by Lars Timman.
2204 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2206 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2207 Requires trusted mode and valid format message id, aborts otherwise.
2208 Patch contributed by Heiko Schlichting.
2210 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2211 certextract with support for various fields. Bug 1358.
2213 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2214 is requested by default, modifiable by smtp transport option
2217 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2218 operate on certificate variables to give certificate fingerprints
2219 Also new ${sha256:cert_variable}.
2221 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2223 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2224 Christian Aistleitner.
2226 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2228 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2229 file. Patch from Wolfgang Breyha.
2231 JH/25 Expand the coverage of the delivery $host and $host_address to
2232 client authenticators run in verify callout. Bug 1476.
2234 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2235 align with daemon_smtp_ports. Bug 72.
2237 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2238 support and error reporting did not work properly.
2240 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2241 and is readable. Patch from Andrew Colin Kissa.
2243 TL/14 Enhance documentation of ${run expansion and how it parses the
2244 commandline after expansion, particularly in the case when an
2245 unquoted variable expansion results in an empty value.
2247 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2249 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2250 by renaming away from it. Addresses GH issue 15, reported by
2253 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2255 TL/15 SECURITY: prevent double expansion in math comparison functions
2256 (can expand unsanitized data). Not remotely exploitable.
2263 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2265 PP/02 Make -n do something, by making it not do something.
2266 When combined with -bP, the name of an option is not output.
2268 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2271 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2272 $sender_host_name and config options to manage this, and basic check
2275 PP/05 DSCP support for outbound connections and control modifier for inbound.
2277 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2278 (Only plugin which currently uses this is kerberos4, which nobody should
2279 be using, but we should make it available and other future plugins might
2280 conceivably use it, even though it would break NAT; stuff *should* be
2281 using channel bindings instead).
2283 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2284 name; added for Sendmail compatibility; requires admin caller.
2285 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2286 just ignore it); requires trusted caller.
2287 Also parse but ignore: -Ac -Am -X<logfile>
2290 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2292 TL/02 Add +smtp_confirmation as a default logging option.
2294 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2295 Patch by Magnus Holmgren from 2007-02-20.
2297 TL/04 Bugzilla 1281 - Spec typo.
2298 Bugzilla 1283 - Spec typo.
2299 Bugzilla 1290 - Spec grammar fixes.
2301 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2303 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2305 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2306 mailing list by by Dmitry Isaikin.
2308 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2310 JH/02 Support "G" suffix to numbers in ${if comparisons.
2312 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2314 NM/01 Bugzilla 1197 - Spec typo
2315 Bugzilla 1196 - Spec examples corrections
2317 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2319 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2320 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2323 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2324 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2326 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2327 "acl {{name}{arg}...}", and optional args on acl condition
2330 JH/05 Permit multiple router/transport headers_add/remove lines.
2332 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2334 JH/07 Avoid using a waiting database for a single-message-only transport.
2335 Performance patch from Paul Fisher. Bugzilla 1262.
2337 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2340 JH/09 Add $headers_added variable, with content from use of ACL modifier
2341 add_header (but not yet added to the message). Bugzilla 199.
2343 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2344 Pulled from Bugzilla 817 by Wolfgang Breyha.
2346 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2348 (nb: this is the same fix as in Exim 4.80.1)
2350 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2353 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2356 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2358 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2359 advertise SMTP AUTH mechanism to us, instead of a generic
2360 protocol violation error. Also, make Exim more robust to bad
2361 data from the Dovecot auth socket.
2363 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2365 When a queue runner is handling a message, Exim first routes the
2366 recipient addresses, during which it prunes them based on the retry
2367 hints database. After that it attempts to deliver the message to
2368 any remaining recipients. It then updates the hints database using
2371 So if a recipient address works intermittently, it can get repeatedly
2372 deferred at routing time. The retry hints record remains fresh so the
2373 address never reaches the final cutoff time.
2375 This is a fairly common occurrence when a user is bumping up against
2376 their storage quota. Exim had some logic in its local delivery code
2377 to deal with this. However it did not apply to per-recipient defers
2378 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2380 This change adds a proper retry rule check during routing so that the
2381 final cutoff time is checked against the message's age. We only do
2382 this check if there is an address retry record and there is not a
2383 domain retry record; this implies that previous attempts to handle
2384 the address had the retry_use_local_parts option turned on. We use
2385 this as an approximation for the destination being like a local
2386 delivery, as in LMTP.
2388 I suspect this new check makes the old local delivery cutoff check
2389 redundant, but I have not verified this so I left the code in place.
2391 TF/02 Correct gecos expansion when From: is a prefix of the username.
2393 Test 0254 submits a message to Exim with the header
2397 When I ran the test suite under the user fanf2, Exim expanded
2398 the header to contain my full name, whereas it should have added
2399 a Resent-Sender: header. It erroneously treats any prefix of the
2400 username as equal to the username.
2402 This change corrects that bug.
2404 GF/01 DCC debug and logging tidyup
2405 Error conditions log to paniclog rather than rejectlog.
2406 Debug lines prefixed by "DCC: " to remove any ambiguity.
2408 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2410 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2411 Bug spotted by Jeremy Harris; was flawed since initial commit.
2412 Would have resulted in OCSP responses post-SNI triggering an Exim
2413 NULL dereference and crash.
2415 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2417 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2418 Bug detection, analysis and fix by Samuel Thibault.
2419 Bugzilla 1331, Debian bug #698092.
2421 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2423 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2424 Server implementation by Todd Lyons, client by JH.
2425 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2426 config variable "prdr_enable" controls whether the server
2427 advertises the facility. If the client requests PRDR a new
2428 acl_data_smtp_prdr ACL is called once for each recipient, after
2429 the body content is received and before the acl_smtp_data ACL.
2430 The client is controlled by both of: a hosts_try_prdr option
2431 on the smtp transport, and the server advertisement.
2432 Default client logging of deliveries and rejections involving
2433 PRDR are flagged with the string "PRDR".
2435 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2436 fclose(). Diagnosis by Todd Lyons.
2438 PP/17 Update configure.default to handle IPv6 localhost better.
2439 Patch by Alain Williams (plus minor tweaks).
2442 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2443 This is now consistent with GnuTLS, and is now documented: the
2444 previous undocumented portable approach to treating the option as
2445 unset was to force an expansion failure. That still works, and
2446 an empty string is now equivalent.
2448 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2449 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2450 not performing validation itself.
2452 PP/20 Added force_command boolean option to pipe transport.
2453 Patch from Nick Koston, of cPanel Inc.
2455 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2458 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2460 PP/21 Fix eximon continuous updating with timestamped log-files.
2461 Broken in a format-string cleanup in 4.80, missed when I repaired the
2462 other false fix of the same issue.
2463 Report and fix from Heiko Schlichting.
2466 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2467 Report from Prashanth Katuri.
2469 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2470 It's SecureTransport, so affects any MacOS clients which use the
2471 system-integrated TLS libraries, including email clients.
2473 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2474 using a MIME ACL for non-SMTP local injection.
2475 Report and assistance in diagnosis by Warren Baker.
2477 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2479 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2481 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2482 last id that failed so it may be referenced in subsequent ACL's.
2484 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2487 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2488 ldap library initialization, allowing self-signed CA's to be
2489 used. Also properly sets require_cert option later in code by
2490 using NULL (global ldap config) instead of ldap handle (per
2491 session). Bug diagnosis and testing by alxgomz.
2493 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2494 the src/util/ subdirectory.
2496 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2497 renamed to Transport Post Delivery Action by Jeremy Harris, as
2500 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2501 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2502 redis_servers = needs to be configured which will be used by the redis
2503 lookup. Patch from Warren Baker, of The Packet Hub.
2505 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2507 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2508 hostname or reverse DNS when processing a host list. Used suggestions
2509 from multiple comments on this bug.
2511 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2513 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2514 Missed a few lines, added it to make the runtest require no keyboard
2517 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2518 contains upper case chars. Make router use caseful_local_part.
2520 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2521 support when GnuTLS has been built with p11-kit.
2527 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2529 This, or similar/improved, will also be change PP/11 of 4.82.
2535 PP/01 Handle short writes when writing local log-files.
2536 In practice, only affects FreeBSD (8 onwards).
2537 Bugzilla 1053, with thanks to Dmitry Isaikin.
2539 NM/01 Bugzilla 949 - Documentation tweak
2541 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2544 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2546 PP/02 Implemented gsasl authenticator.
2548 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2550 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2551 `pkg-config foo` for cflags/libs.
2553 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2554 with rest of GSASL and with heimdal_gssapi.
2556 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2557 `pkg-config foo` for cflags/libs for the TLS implementation.
2559 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2560 properties get this fed in as external SSF. A number of robustness
2561 and debugging improvements to the cyrus_sasl authenticator.
2563 PP/08 cyrus_sasl server now expands the server_realm option.
2565 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2566 Patch by Jeremy Harris.
2568 PP/10 Added dbmjz lookup type.
2570 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2572 PP/12 MAIL args handles TAB as well as SP, for better interop with
2573 non-compliant senders.
2574 Analysis and variant patch by Todd Lyons.
2576 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2577 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2578 Patch from Dirk Mueller <dmueller@suse.com>
2580 PP/13 tls_peerdn now print-escaped for spool files.
2581 Observed some $tls_peerdn in wild which contained \n, which resulted
2582 in spool file corruption.
2584 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2585 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2586 or write after TLS renegotiation, which otherwise led to messages
2589 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2590 as a tracking header (ie: a signed header comes before the signature).
2591 Patch from Wolfgang Breyha.
2593 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2594 comma-sep list; embedded commas doubled.
2596 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2598 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2600 Report and patch from Dmitry Banschikov.
2602 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2603 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2604 failures. We appear to now support TLS1.1+ with Exim.
2606 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2607 lets Exim select keys and certificates based upon TLS SNI from client.
2608 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2609 before an outbound SMTP session. New log_selector, +tls_sni.
2611 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2612 NULL dereference. Report and patch from Alun Jones.
2614 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2615 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2616 Not seeing resolver debug output on NetBSD, but suspect this is a
2617 resolver implementation change.
2619 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2620 Left warnings. Added "eximon gdb" invocation mode.
2622 PP/21 Defaulting "accept_8bitmime" to true, not false.
2624 PP/22 Added -bw for inetd wait mode support.
2626 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2627 locate the relevant includes and libraries. Made this the default.
2629 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2630 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2632 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2633 This may cause build issues on older platforms.
2635 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2636 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2637 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2638 Added SNI support via GnuTLS too.
2639 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2641 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2643 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2644 Applied second patch from Janne, implementing suggestion to default
2645 multiple-strings-in-record handling to match SPF spec.
2647 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2649 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2650 read-only, out of scope).
2651 Patch from Wolfgang Breyha, report from Stuart Northfield.
2653 PP/29 Fix three issues highlighted by clang analyser static analysis.
2654 Only crash-plausible issue would require the Cambridge-specific
2655 iplookup router and a misconfiguration.
2656 Report from Marcin Mirosław.
2658 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2660 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2661 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2662 As part of this, removing so much warning spew let me fix some minor
2663 real issues in debug logging.
2665 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2666 assignment on my part. Fixed.
2668 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2669 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2670 Janne Snabb (who went above and beyond: thank you).
2672 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2673 string otherwise requires a connection and a bunch more work and it's
2674 relatively easy to get wrong. Should also expose TLS library linkage
2677 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2678 64-bit ${eval} (JH/03).
2680 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2681 GNU libc to support some of the 64-bit stuff, should not lead to
2682 conflicts. Defined before os.h is pulled in, so if a given platform
2683 needs to override this, it can.
2685 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2686 protection layer was required, which is not implemented.
2687 Bugzilla 1254, patch from Wolfgang Breyha.
2689 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2690 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2691 tls_dhparam take prime identifiers. Also unbreak combination of
2692 OpenSSL+DH_params+TLSSNI.
2694 PP/39 Disable SSLv2 by default in OpenSSL support.
2700 PP/01 Solaris build fix for Oracle's LDAP libraries.
2701 Bugzilla 1109, patch from Stephen Usher.
2703 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2705 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2708 TF/02 Fix a couple more cases where we did not log the error message
2709 when unlink() failed. See also change 4.74-TF/03.
2711 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2712 lock up or crash if it happened to be inside a call to libc when it
2713 got a SIGUSR1 from exiwhat.
2715 The SIGUSR1 handler appends the current process status to the process
2716 log which is later printed by exiwhat. It used to use the general
2717 purpose logging code to do this, but several functions it calls are
2718 not safe for signals.
2720 The new output code in the SIGUSR1 handler is specific to the process
2721 log, and simple enough that it's easy to inspect for signal safety.
2722 Removing some special cases also simplifies the general logging code.
2723 Removing the spurious timestamps from the process log simplifies
2726 TF/04 Improved ratelimit ACL condition.
2728 The /noupdate option has been deprecated in favour of /readonly which
2729 has clearer semantics. The /leaky, /strict, and /readonly update modes
2730 are mutually exclusive. The update mode is no longer included in the
2731 database key; it just determines when the database is updated. (This
2732 means that when you upgrade Exim will forget old rate measurements.)
2734 Exim now checks that the per_* options are used with an update mode that
2735 makes sense for the current ACL. For example, when Exim is processing a
2736 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2737 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2738 must specify per_mail/readonly. If you omit the update mode it defaults to
2739 /leaky where that makes sense (as before) or /readonly where required.
2741 The /noupdate option is now undocumented but still supported for
2742 backwards compatibility. It is equivalent to /readonly except that in
2743 ACLs where /readonly is required you may specify /leaky/noupdate or
2744 /strict/noupdate which are treated the same as /readonly.
2746 A useful new feature is the /count= option. This is a generalization
2747 of the per_byte option, so that you can measure the throughput of other
2748 aggregate values. For example, the per_byte option is now equivalent
2749 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2751 The per_rcpt option has been generalized using the /count= mechanism
2752 (though it's more complicated than the per_byte equivalence). When it is
2753 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2754 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2755 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2756 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2757 Note that using per_rcpt with a non-readonly update mode in more than
2758 one ACL will cause the recipients to be double-counted. (The per_mail
2759 and per_byte options don't have this problem.)
2761 The handling of very low rates has changed slightly. If the computed rate
2762 is less than the event's count (usually one) then this event is the first
2763 after a long gap. In this case the rate is set to the same as this event's
2764 count, so that the first message of a spam run is counted properly.
2766 The major new feature is a mechanism for counting the rate of unique
2767 events. The new per_addr option counts the number of different
2768 recipients that someone has sent messages to in the last time period. It
2769 behaves like per_rcpt if all the recipient addresses are different, but
2770 duplicate recipient addresses do not increase the measured rate. Like
2771 the /count= option this is a general mechanism, so the per_addr option
2772 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2773 example, measure the rate that a client uses different sender addresses
2774 with the options per_mail/unique=$sender_address. There are further
2775 details in the main documentation.
2777 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2779 TF/06 Removed a few PCRE remnants.
2781 TF/07 Automatically extract Exim's version number from tags in the git
2782 repository when doing development or release builds.
2784 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2785 Bugzilla 879. Patch from Paul Fisher.
2787 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2788 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2791 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2793 PP/05 Variable $av_failed, true if the AV scanner deferred.
2794 Bugzilla 1078. Patch from John Horne.
2796 PP/06 Stop make process more reliably on build failure.
2797 Bugzilla 1087. Patch from Heiko Schlittermann.
2799 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2800 Bugzilla 1089. Patch from Heiko Schlittermann.
2802 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2803 Bugzilla 1131. Patch from Holger Weiß.
2805 PP/09 Handle IPv6 addresses with SPF.
2806 Bugzilla 860. Patch from Wolfgang Breyha.
2808 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2810 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2813 PP/11 match_* no longer expand right-hand-side by default.
2814 New compile-time build option, EXPAND_LISTMATCH_RHS.
2815 New expansion conditions, "inlist", "inlisti".
2817 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2819 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2821 PP/14 fix log_write() format string regression from TF/03.
2822 Bugzilla 1152. Patch from Dmitry Isaikin.
2828 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2830 PP/02 Harmonised TLS library version reporting; only show if debugging.
2831 Layout now matches that introduced for other libraries in 4.74 PP/03.
2833 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2835 PP/04 New "dns_use_edns0" global option.
2837 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2840 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2841 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2843 TK/01 Updated PolarSSL code to 0.14.2.
2844 Bugzilla 1097. Patch from Andreas Metzler.
2846 PP/07 Catch divide-by-zero in ${eval:...}.
2847 Fixes bugzilla 1102.
2849 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2852 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2853 format-string attack -- SECURITY: remote arbitrary code execution.
2855 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2856 time unintentionally subject to list matching rules, letting the header
2857 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2858 arbitrary string expansion). This allowed for information disclosure.
2860 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2861 INT_MIN/-1 -- value coerced to INT_MAX.
2867 NM/01 Workaround for PCRE version dependency in version reporting
2870 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2871 This fixes portability to compilers other than gcc, notably
2872 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2874 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2875 makefiles for portability to HP-UX and POSIX correctness.
2877 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2878 Also via indented variable definition in the Makefile.
2879 (Debugging by Oliver Heesakkers).
2881 PP/02 Restore caching of spamd results with expanded spamd_address.
2882 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2884 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2885 Improves build reliability. Fix from: Frank Elsner
2887 NM/02 Fix wide character breakage in the rfc2047 coding
2888 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2890 NM/03 Allow underscore in dnslist lookups
2891 Fixes bug 1026. Patch from Graeme Fowler
2893 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2894 Code patches from Adam Ciarcinski of NetBSD.
2896 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2899 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2900 is logged, to avoid truncation. Patch from John Horne.
2902 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2903 Patch from Jakob Hirsch.
2905 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2906 SQL string expansion failure details.
2907 Patch from Andrey Oktyabrski.
2909 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2910 Patch from Simon Arlott.
2912 PP/09 New lookups functionality failed to compile on old gcc which rejects
2913 extern declarations in function scope.
2914 Patch from Oliver Fleischmann
2916 PP/10 Use sig_atomic_t for flags set from signal handlers.
2917 Check getgroups() return and improve debugging.
2918 Fixed developed for diagnosis in bug 927 (which turned out to be
2921 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2922 Patch from Mark Zealey.
2924 PP/12 Bugzilla 1056: Improved spamd server selection.
2925 Patch from Mark Zealey.
2927 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2928 Based on patch from Heiko Schlittermann.
2930 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2931 Patch from Uwe Doering, sign-off by Michael Haardt.
2933 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2934 variable declaration deep within a block. Bug and patch from
2937 PP/15 lookups-Makefile IRIX compatibility coercion.
2939 PP/16 Make DISABLE_DKIM build knob functional.
2941 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2942 Patch by Simon Arlott
2944 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2945 variable argument macros. Our copy now differs from upstream.
2951 TF/01 Failure to get a lock on a hints database can have serious
2952 consequences so log it to the panic log.
2954 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2955 controlled using the smtp_confirmation log selector.
2957 TF/03 Include the error message when we fail to unlink a spool file.
2959 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2960 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2961 for maintaining out-of-tree patches for some time.
2963 PP/01 Bugzilla 139: Documentation and portability issues.
2964 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2965 Handle per-OS dynamic-module compilation flags.
2967 PP/02 Let /dev/null have normal permissions.
2968 The 4.73 fixes were a little too stringent and complained about the
2969 permissions on /dev/null. Exempt it from some checks.
2970 Reported by Andreas M. Kirchwitz.
2972 PP/03 Report version information for many libraries, including
2973 Exim version information for dynamically loaded libraries. Created
2974 version.h, now support a version extension string for distributors
2975 who patch heavily. Dynamic module ABI change.
2977 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2978 privilege escalation vulnerability whereby the Exim run-time user
2979 can cause root to append content of the attacker's choosing to
2982 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2985 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2986 If dropping privileges for untrusted macros, we disabled normal logging
2987 on the basis that it would fail; for the Exim run-time user, this is not
2988 the case, and it resulted in successful deliveries going unlogged.
2989 Fixed. Reported by Andreas Metzler.
2995 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2996 only prepend for the Resent-* case. Fixes regression introduced in
2997 Exim 4.70 by NM/22 for Bugzilla 607.
2999 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3000 increasing numbers of administrators be bitten by this.
3002 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3004 PP/03 Bugzilla 994: added openssl_options main configuration option.
3006 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3008 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3010 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3012 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3013 without a peer certificate, leading to a segfault because of an
3014 assumption that peers always have certificates. Be a little more
3015 paranoid. Problem reported by Martin Tscholak.
3017 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3018 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3019 NB: ClamAV planning to remove STREAM in "middle of 2010".
3020 CL also introduces -bmalware, various -d+acl logging additions and
3021 more caution in buffer sizes.
3023 PP/09 Implemented reverse_ip expansion operator.
3025 PP/10 Bugzilla 937: provide a "debug" ACL control.
3027 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3029 PP/12 Bugzilla 973: Implement --version.
3031 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3033 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3035 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3037 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3038 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3039 ignore trailing whitespace.
3041 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3043 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3046 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3047 Notification from Dr Andrew Aitchison.
3049 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3050 ExtendedDetectionInfo response format.
3051 Notification from John Horne.
3053 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3056 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3057 XSL and documented dependency on system catalogs, with examples of how
3060 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3063 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3064 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3065 configuration file which is writeable by the Exim user or group.
3067 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3068 of configuration files to cover files specified with the -C option if
3069 they are going to be used with root privileges, not just the default
3072 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3073 option (effectively making it always true).
3075 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3076 files to be used while preserving root privileges.
3078 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3079 that rogue child processes cannot use them.
3081 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3082 run-time user, instead of root.
3084 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3085 Exim run-time user without dropping privileges.
3087 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3088 result string, instead of calling string_vformat() twice with the same
3091 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3092 for other users. Others should always drop root privileges if they use
3093 -C on the command line, even for a whitelisted configure file.
3095 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3097 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3103 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3104 $data_path, and $header_path variables; fixed documentation bugs and
3107 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3108 exipick to access non-standard spools, including the "frozen" queue
3111 NM/01 Bugzilla 965: Support mysql stored procedures.
3112 Patch from Alain Williams
3114 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3116 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3117 Patch from Andreas Metzler
3119 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3120 Patch from Kirill Miazine
3122 NM/05 Bugzilla 671: Added umask to procmail example.
3124 JJ/03 installed exipick 20100323.0, fixing doc bug
3126 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3127 directory. Notification and patch from Dan Rosenberg.
3129 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3131 TK/02 Improve log output when DKIM signing operation fails.
3133 MH/01 Treat the transport option dkim_domain as a colon separated
3134 list, not as a single string, and sign the message with each element,
3135 omitting multiple occurences of the same signer.
3137 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3138 Bugzilla 985, 986. Patch by Simon Arlott
3140 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3141 Patch by Simon Arlott
3143 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3144 MBX locking. Notification from Dan Rosenberg.
3150 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3152 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3154 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3156 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3158 NM/04 Bugzilla 915: Flag broken perl installation during build.
3164 TK/01 Added patch by Johannes Berg that expands the main option
3165 "spamd_address" if it starts with a dollar sign.
3167 TK/02 Write list of recipients to X-Envelope-Sender header when building
3168 the mbox-format spool file for content scanning (suggested by Jakob
3171 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3172 (http://www.dcc-servers.net/) support via dccifd. Activated by
3173 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3175 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3176 by Mark Daniel Reidel <mr@df.eu>.
3178 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3179 When building exim an external PCRE library is now needed -
3180 PCRE is a system library on the majority of modern systems.
3181 See entry on PCRE_LIBS in EDITME file.
3183 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3184 conversation. Added nologin parameter to request.
3185 Patch contributed by Kirill Miazine.
3187 TF/01 Do not log submission mode rewrites if they do not change the address.
3189 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3191 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3192 log files in place. Contributed by Roberto Lima.
3194 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3196 TF/03 Bugzilla 615: When checking the local_parts router precondition
3197 after a local_part_suffix or local_part_prefix option, Exim now
3198 does not use the address's named list lookup cache, since this
3199 contains cached lookups for the whole local part.
3201 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3202 Robert Millan. Documentation is in experimental-spec.txt.
3204 TF/04 Bugzilla 668: Fix parallel build (make -j).
3206 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3208 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3209 Patch provided by Jan Srzednicki.
3211 TF/05 Leading white space used to be stripped from $spam_report which
3212 wrecked the formatting. Now it is preserved.
3214 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3215 that they are available at delivery time.
3217 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3219 TF/08 TLS error reporting now respects the incoming_interface and
3220 incoming_port log selectors.
3222 TF/09 Produce a more useful error message if an SMTP transport's hosts
3223 setting expands to an empty string.
3225 NM/06 Bugzilla 744: EXPN did not work under TLS.
3226 Patch provided by Phil Pennock.
3228 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3229 Patch provided by Richard Godbee.
3231 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3232 acl_smtp_notquit, added index entry.
3234 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3235 Patch provided by Eugene Bujak.
3237 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3238 accept(). Patch provided by Maxim Dounin.
3240 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3241 Patch provided by Phil Pennock.
3243 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3245 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3246 Patch provided by Brad "anomie" Jorsch.
3248 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3249 Patch provided by Dean Brooks.
3251 TK/05 Add native DKIM support (does not depend on external libraries).
3253 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3254 Patch provided by Graeme Fowler.
3256 NM/16 Bugzilla 851: Documentation example syntax fix.
3258 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3260 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3263 NM/19 Bugzilla 745: TLS version reporting.
3264 Patch provided by Phil Pennock.
3266 NM/20 Bugzilla 167: bool: condition support.
3267 Patch provided by Phil Pennock.
3269 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3270 clients. Patch provided by Phil Pennock.
3272 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3273 Patch provided by Brad "anomie" Jorsch.
3275 NM/23 Bugzilla 687: Fix misparses in eximstats.
3276 Patch provided by Heiko Schlittermann.
3278 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3279 Patch provided by Heiko Schlittermann.
3281 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3282 plus update to original patch.
3284 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3286 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3287 Patch provided by David Brownlee.
3289 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3291 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3293 NM/30 Bugzilla 888: TLS documentation bugfixes.
3295 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3297 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3298 Unlike the original bugzilla I have changed all shell scripts in src tree.
3300 NM/33 Bugzilla 898: Transport filter timeout fix.
3301 Patch by Todd Rinaldo.
3303 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3304 Patch by Serge Demonchaux.
3306 NM/35 Bugzilla 39: Base64 decode bug fixes.
3307 Patch by Jakob Hirsch.
3309 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3311 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3313 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3315 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3321 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3322 ALT-N's libdkim that I have put here:
3323 http://duncanthrax.net/exim-experimental/
3325 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3326 were called 'true' and it seems that C99 defines that as a reserved
3327 keyword to be used with 'bool' variable types. That means you could
3328 not include C99-style headers which use bools without triggering
3329 build errors in sieve.c.
3331 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3332 as mailq or other aliases. Changed the --help handling significantly
3333 to do whats expected. exim_usage() emits usage/help information.
3335 SC/01 Added the -bylocaldomain option to eximstats.
3337 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3339 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3341 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3347 PH/01 Another patch from the Sieve maintainer.
3349 PH/02 When an IPv6 address is converted to a string for single-key lookup
3350 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3351 dots are used instead of colons so that keys in lsearch files need not
3352 contain colons. This was done some time before quoting was made available
3353 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3354 (notated using the quote facility) so as to distinguish them from IPv4
3355 keys. This meant that lookups for IP addresses in host lists did not work
3356 for iplsearch lookups.
3358 This has been fixed by arranging for IPv6 addresses to be expressed with
3359 colons if the lookup type is iplsearch. This is not incompatible, because
3360 previously such lookups could never work.
3362 The situation is now rather anomalous, since one *can* have colons in
3363 ordinary lsearch keys. However, making the change in all cases is
3364 incompatible and would probably break a number of configurations.
3366 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3369 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3370 conversion specification without a maximum field width, thereby enabling
3371 a rogue spamd server to cause a buffer overflow. While nobody in their
3372 right mind would setup Exim to query an untrusted spamd server, an
3373 attacker that gains access to a server running spamd could potentially
3374 exploit this vulnerability to run arbitrary code as the Exim user.
3376 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3377 $primary_hostname instead of what libspf2 thinks the hosts name is.
3379 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3380 a directory entry by the name of the lookup key. Previously, if a
3381 symlink pointed to a non-existing file or a file in a directory that
3382 Exim lacked permissions to read, a lookup for a key matching that
3383 symlink would fail. Now it is enough that a matching directory entry
3384 exists, symlink or not. (Bugzilla 503.)
3386 PH/03 The body_linecount and body_zerocount variables are now exported in the
3389 PH/04 Added the $dnslist_matched variable.
3391 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3392 This means they are set thereafter only if the connection becomes
3395 PH/06 Added the client_condition to authenticators so that some can be skipped
3396 by clients under certain conditions.
3398 PH/07 The error message for a badly-placed control=no_multiline_responses left
3399 "_responses" off the end of the name.
3401 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3403 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3404 (without spaces) instead of just copying the configuration text.
3406 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3408 PH/11 Added $max_received_linelength.
3410 PH/12 Added +ignore_defer and +include_defer to host lists.
3412 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3413 way in which PCRE > 7.0 is built.
3415 PH/14 Implemented queue_only_load_latch.
3417 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3418 MAIL command. The effect was to mangle the value on 64-bit systems.
3420 PH/16 Another patch from the Sieve maintainer.
3422 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3424 PH/18 If a system quota error occurred while trying to create the file for
3425 a maildir delivery, the message "Mailbox is full" was not appended to the
3426 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3427 only to a quota excession during the actual writing of the file.
3429 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3430 characters?) which causes problems in log lines. The DN values are now
3431 passed through string_printing() before being added to log lines.
3433 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3434 and InterBase are left for another time.)
3436 PH/21 Added message_body_newlines option.
3438 PH/22 Guard against possible overflow in moan_check_errorcopy().
3440 PH/23 POSIX allows open() to be a macro; guard against that.
3442 PH/24 If the recipient of an error message contained an @ in the local part
3443 (suitably quoted, of course), incorrect values were put in $domain and
3444 $local_part during the evaluation of errors_copy.
3450 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3451 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3454 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3455 issue a MAIL command.
3457 PH/02 In an ACL statement such as
3459 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3461 if a client was not listed at all, or was listed with a value other than
3462 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3463 the condition was not true (as it should be), so access was not denied.
3464 The bug was that the ! inversion was incorrectly passed on to the second
3465 item. This has been fixed.
3467 PH/03 Added additional dnslists conditions == and =& which are different from
3468 = and & when the dns lookup returns more than one IP address.
3470 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3471 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3473 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3474 FSYNC, which compiles an option called disable_fsync that allows for
3475 bypassing fsync(). The documentation is heavily laced with warnings.
3477 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3479 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3480 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3481 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3482 including adding "make clean"; (3) Added -fPIC when compiling the test
3483 dynamically loaded module, to get rid of a warning.
3485 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3486 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3487 = 0s. The bug is otherwise harmless.
3489 PH/07 There was a bug in the dovecot authenticator such that the value of
3490 $auth1 could be overwritten, and so not correctly preserved, after a
3491 successful authentication. This usually meant that the value preserved by
3492 the server_setid option was incorrect.
3494 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3496 PH/09 Installed PCRE release 7.0.
3498 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3499 run for batched SMTP input. It is now run at the start of every message
3500 in the batch. While fixing this I discovered that the process information
3501 (output by running exiwhat) was not always getting set for -bs and -bS
3502 input. This is fixed, and it now also says "batched" for BSMTP.
3504 PH/11 Added control=no_pipelining.
3506 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3507 patch, slightly modified), and move the expansion of helo_data till after
3508 the connection is made in the smtp transport (so it can use these
3511 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3513 PH/14 Added log_selector = +pid.
3515 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3517 PH/16 Add ${if forany and ${if forall.
3519 PH/17 Added dsn_from option to vary the From: line in DSNs.
3521 PH/18 Flush SMTP output before performing a callout, unless control =
3522 no_callout_flush is set.
3524 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3525 was true (the default) a successful delivery failed to delete the retry
3526 item, thus causing premature timeout of the address. The bug is now
3529 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3531 PH/21 Long custom messages for fakedefer and fakereject are now split up
3532 into multiline responses in the same way that messages for "deny" and
3533 other ACL rejections are.
3535 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3536 with slight modification.
3538 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3539 draft, changing the syntax and factoring some duplicate code".
3541 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3542 for deliveries of the second and subsequent messages over the same SMTP
3545 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3546 ${reduce, with only minor "tidies".
3548 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3550 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3551 expansion side effects.
3553 PH/27 When a message times out after an over-quota error from an Exim-imposed
3554 quota, the bounce message says "mailbox is full". This message was not
3555 being given when it was a system quota that was exceeded. It now should
3558 MH/03 Made $recipients available in local_scan(). local_scan() already has
3559 better access to the recipient list through recipients_list[], but
3560 $recipients can be useful in postmaster-provided expansion strings.
3562 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3563 in the case of a MAIL command with additional options following the
3564 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3565 were accidentally chopped off.
3567 PH/29 SMTP synchronization checks are implemented when a command is read -
3568 there is a check that no more input is waiting when there shouldn't be
3569 any. However, for some commands, a delay in an ACL can mean that it is
3570 some time before the response is written. In this time, more input might
3571 arrive, invalidly. So now there are extra checks after an ACL has run for
3572 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3573 pipelining has not been advertised.
3575 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3577 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3578 returned in the SMTP response when smtp_return_error_details was false.
3579 This has been fixed.
3581 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3582 instead of the C I/O that was originally supplied, because problems were
3583 reported on Solaris.
3585 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3586 Exim which did not show up earlier: it was assuming that a call to
3587 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3588 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3589 was a macro that became an assignment, so it seemed to work. This has
3590 changed to a proper function call with a void return, hence the compile
3591 error. Exim's code has been fixed.
3593 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3596 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3598 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3600 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3601 because it turns out to be too big or there is a timeout) there is no
3602 "Completed" line in the log. When some messages of this type were
3603 selected by exigrep, they were listed as "not completed". Others were
3604 picked up by some special patterns. I have improved the selection
3605 criteria to be more general.
3607 PH/38 The host_find_failed option in the manualroute router can now be set
3608 to "ignore", to completely ignore a host whose IP address cannot be
3609 found. If all hosts are ignored, the behaviour is controlled by the new
3610 host_all_ignored option.
3612 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3613 homing or because of multiple MX records with /mx) generated more than
3614 one IP address, and the following item turned out to be the local host,
3615 all the secondary addresses of the first item were incorrectly removed
3616 from the list, along with the local host and any following hosts (which
3617 is what is supposed to happen).
3619 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3620 whoever called Exim into the -H file. In the case of the daemon it was
3621 behaving confusingly. When first started, it used values for whoever
3622 started the daemon, but after a SIGHUP it used the Exim user (because it
3623 calls itself on a restart). I have changed the code so that it now always
3626 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3627 message are rejected with the same error (e.g. no authentication or bad
3628 sender address), and a DATA command is nevertheless sent (as can happen
3629 with PIPELINING or a stupid MUA), the error message that was given to the
3630 RCPT commands is included in the rejection of the DATA command. This is
3631 intended to be helpful for MUAs that show only the final error to their
3634 PH/42 Another patch from the Sieve maintainer.
3636 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3637 Eximstats - Fixed some broken HTML links and added missing column headers
3639 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3640 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3642 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3644 PH/43 Yet another patch from the Sieve maintainer.
3646 PH/44 I found a way to check for a TCP/IP connection going away before sending
3647 the response to the final '.' that terminates a message, but only in the
3648 case where the client has not sent further data following the '.'
3649 (unfortunately, this is allowed). However, in many cases there won't be
3650 any further data because there won't be any more messages to send. A call
3651 to select() can be used: if it shows that the input is "ready", there is
3652 either input waiting, or the socket has been closed. An attempt to read
3653 the next input character can distinguish the two cases. Previously, Exim
3654 would have sent an OK response which the client would never have see.
3655 This could lead to message repetition. This fix should cure that, at
3656 least in a lot of common cases.
3658 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3659 advertised in response to EHLO.
3665 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3666 fixed by 4.65/MH/01 (is this a record?) are fixed:
3668 (i) An empty string was always treated as zero by the numeric comparison
3669 operators. This behaviour has been restored.
3671 (ii) It is documented that the numeric comparison operators always treat
3672 their arguments as decimal numbers. This was broken in that numbers
3673 starting with 0 were being interpreted as octal.
3675 While fixing these problems I realized that there was another issue that
3676 hadn't been noticed. Values of message_size_limit (both the global option
3677 and the transport option) were treated as octal if they started with 0.
3678 The documentation was vague. These values are now always treated as
3679 decimal, and I will make that clear in the documentation.
3685 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3686 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3689 MH/01 Don't check that the operands of numeric comparison operators are
3690 integers when their expansion is in "skipping" mode (fixes bug
3691 introduced by 4.64-PH/07).
3693 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3694 child addresses, Exim now panics and dies. Previously, because the count
3695 is held in a short int, deliveries were likely to be lost. As such a
3696 large number of recipients for a single message is ridiculous
3697 (performance will be very, very poor), I have chosen to impose a limit
3698 rather than extend the field.
3704 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3705 leftover -K file (the existence of which was triggered by #402).
3706 While we were at it, introduced process PID as part of the -K
3707 filename. This should rule out race conditions when creating
3710 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3711 processing considerably. Previous code took too long for large mails,
3712 triggering a timeout which in turn triggers #401.
3714 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3715 in the DK code in transports.c. sendfile() is not really portable,
3716 hence the _LINUX specificness.
3718 TF/01 In the add_headers option to the mail command in an Exim filter,
3719 there was a bug that Exim would claim a syntax error in any
3720 header after the first one which had an odd number of characters
3723 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3724 callout verification, Exim cached a "reject" for the entire domain. This
3725 is correct for most verifications, but it is not correct for a recipient
3726 verification with use_sender or use_postmaster set, because in that case
3727 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3728 case of MAIL FROM:<> rejection from other early rejections (e.g.
3729 rejection of HELO). When verifying a recipient using a non-null MAIL
3730 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3731 Whatever the result of the callout, the value of the domain cache is
3732 left unchanged (for any other kind of callout, getting as far as trying
3733 RCPT means that the domain itself is ok).
3735 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3738 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3739 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3740 session, the daemon ignores ECONNECT errors and logs others; it now
3741 ignores EPIPE as well.
3743 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3744 (quoted-printable decoding).
3746 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3747 later the small subsequent patch to fix an introduced bug.
3749 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3751 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3753 PH/08 An error is now given if message_size_limit is specified negative.
3755 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3756 to be given (somewhat) arbitrary names.
3758 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3761 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3762 miscellaneous code fixes
3764 PH/10 Added the log_reject_target ACL modifier to specify where to log
3767 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3768 hostname. This is wrong, because it relates to the incoming message (and
3769 probably the interface on which it is arriving) and not to the outgoing
3770 callout (which could be using a different interface). This has been
3771 changed to use the value of the helo_data option from the smtp transport
3772 instead - this is what is used when a message is actually being sent. If
3773 there is no remote transport (possible with a router that sets up host
3774 addresses), $smtp_active_hostname is used.
3776 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3777 tweaks were necessary in order to get it to work (see also 21 below):
3778 (a) The code assumed that strncpy() returns a negative number on buffer
3779 overflow, which isn't the case. Replaced with Exim's string_format()
3781 (b) There were several signed/unsigned issues. I just did the minimum
3782 hacking in of casts. There is scope for a larger refactoring.
3783 (c) The code used strcasecmp() which is not a standard C function.
3784 Replaced with Exim's strcmpic() function.
3785 (d) The code set only $1; it now sets $auth1 as well.
3786 (e) A simple test gave the error "authentication client didn't specify
3787 service in request". It would seem that Dovecot has changed its
3788 interface. Fortunately there's a specification; I followed it and
3789 changed what the client sends and it appears to be working now.
3791 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3794 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3795 address A is aliased to B and C, where B exists and C does not. Without
3796 -v the output is "A verified" because verification stops after a
3797 successful redirection if more than one address is generated. However,
3798 with -v the child addresses are also verified. Exim was outputting "A
3799 failed to verify" and then showing the successful verification for C,
3800 with its parentage. It now outputs "B failed to verify", showing B's
3801 parentage before showing the successful verification of C.
3803 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3804 look up a TXT record in a specific list after matching in a combined
3807 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3808 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3809 they consult the DNS. I had assumed they would set it the way they
3810 wanted; and indeed my experiments on Linux seem to show that in some
3811 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3812 To be on the safe side, however, I have now made the interface to
3813 host_find_byname() similar to host_find_bydns(), with an argument
3814 containing the DNS resolver options. The host_find_byname() function now
3815 sets these options at its start, just as host_find_bydns() does. The smtp
3816 transport options dns_qualify_single and dns_search_parents are passed to
3817 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3818 of host_find_byname() use the default settings of RES_DEFNAMES
3819 (qualify_single) but not RES_DNSRCH (search_parents).
3821 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3822 spool_read_header() do less string testing, by means of a preliminary
3823 switch on the second character of optional "-foo" lines. (This is
3824 overdue, caused by the large number of possibilities that now exist.
3825 Originally there were few.) While I was there, I also converted the
3826 str(n)cmp tests so they don't re-test the leading "-" and the first
3827 character, in the hope this might squeeze out yet more improvement.
3829 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3830 flag allowing group syntax was set by the header_syntax check but not
3831 turned off, possible causing trouble later; (2) The flag was not being
3832 set at all for the header_verify test, causing "group"-style headers to
3833 be rejected. I have now set it in this case, and also caused header_
3834 verify to ignore an empty address taken from a group. While doing this, I
3835 came across some other cases where the code for allowing group syntax
3836 while scanning a header line wasn't quite right (mostly, not resetting
3837 the flag correctly in the right place). These bugs could have caused
3838 trouble for malformed header lines. I hope it is now all correct.
3840 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3841 with the "reply" argument non-NULL. The code, however (which originally
3842 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3843 but it didn't always do it. This confused somebody who was copying the
3844 code for some other use. I have removed all the tests.
3846 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3847 feature that was used to support insecure browsers during the U.S. crypto
3848 embargo. It requires special client support, and Exim is probably the
3849 only MTA that supported it -- and would never use it because real RSA is
3850 always available. This code has been removed, because it had the bad
3851 effect of slowing Exim down by computing (never used) parameters for the
3852 RSA_EXPORT functionality.
3854 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3855 authenticator to fail if there's a tab character in the incoming data
3856 (there should never be unless someone is messing about, as it's supposed
3857 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3858 if the connection is using TLS or if the remote IP is the same as the
3859 local IP, and the "valid-client-cert option" if a client certificate has
3862 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3863 authenticators. This can be used for authorization after authentication
3864 succeeds. (In the case of plaintext, it servers for both authentication
3867 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3868 if any retry times were supplied.
3870 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3871 connection, where there is no client IP address to check. In this
3872 situation, the verify now always succeeds.
3874 PH/25 Applied John Jetmore's -Mset patch.
3876 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3878 PH/27 In a string expansion for a processed (not raw) header when multiple
3879 headers of the same name were present, leading whitespace was being
3880 removed from all of them, but trailing whitespace was being removed only
3881 from the last one. Now trailing whitespace is removed from each header
3882 before concatenation. Completely empty headers in a concatenation (as
3883 before) are ignored.
3885 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3886 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3888 PH/29 [Removed. This was a change that I later backed out, and forgot to
3889 correct the ChangeLog entry (that I had efficiently created) before
3890 committing the later change.]
3892 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3893 address errors (4xx response to RCPT) over the same connection as other
3894 messages routed to the same hosts. Such deliveries are always "forced",
3895 so retry times are not inspected. This resulted in far too many retries
3896 for the affected addresses. The effect occurred only when there were more
3897 hosts than the hosts_max_try setting in the smtp transport when it had
3898 the 4xx errors. Those hosts that it had tried were not added to the list
3899 of hosts for which the message was waiting, so if all were tried, there
3900 was no problem. Two fixes have been applied:
3902 (i) If there are any address or message errors in an SMTP delivery, none
3903 of the hosts (tried or untried) are now added to the list of hosts
3904 for which the message is waiting, so the message should not be a
3905 candidate for sending over the same connection that was used for a
3906 successful delivery of some other message. This seems entirely
3907 reasonable: after all the message is NOT "waiting for some host".
3908 This is so "obvious" that I'm not sure why it wasn't done
3909 previously. Hope I haven't missed anything, but it can't do any
3910 harm, as the worst effect is to miss an optimization.
3912 (ii) If, despite (i), such a delivery is accidentally attempted, the
3913 routing retry time is respected, so at least it doesn't keep
3914 hammering the server.
3916 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3917 in ${readsocket because some servers need this prod.
3919 PH/32 Added some extra debug output when updating a wait-xxx database.
3921 PH/33 The hint "could be header name not terminated by colon", which has been
3922 given for certain expansion errors for a long time, was not being given
3923 for the ${if def:h_colon_omitted{... case.
3925 PH/34 The spec says: "With one important exception, whenever a domain list is
3926 being scanned, $domain contains the subject domain." There was at least
3927 one case where this was not true.
3929 PH/35 The error "getsockname() failed: connection reset by peer" was being
3930 written to the panic log as well as the main log, but it isn't really
3931 panic-worthy as it just means the connection died rather early on. I have
3932 removed the panic log writing for the ECONNRESET error when getsockname()
3935 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3936 runs only) independently of the message's sender address. This meant
3937 that, if the 4xx error was in fact related to the sender, a different
3938 message to the same recipient with a different sender could confuse
3939 things. In particular, this can happen when sending to a greylisting
3940 server, but other circumstances could also provoke similar problems.
3941 I have changed the default so that the retry time for these errors is now
3942 based a combination of the sender and recipient addresses. This change
3943 can be overridden by setting address_retry_include_sender=false in the
3946 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3947 remote server are returned as part of bounce messages. This was not
3948 happening for LMTP over a pipe (the lmtp transport), but now it is the
3949 same for both kinds of LMTP.
3951 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3952 in header lines that were in fact CNAMEs. This is no longer the case.
3954 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3955 and queue runs started by the daemon processed all messages. This has
3956 been fixed so that -R and -S can now usefully be given with -q<time>.
3958 PH/40 Import PCRE release 6.7 (fixes some bugs).
3960 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3962 PH/42 Give an error if -q is specified more than once.
3964 PH/43 Renamed the variables $interface_address and $interface_port as
3965 $received_ip_address and $received_port, to make it clear that these
3966 values apply to message reception, and not to the outgoing interface when
3967 a message is delivered. (The old names remain recognized, of course.)
3969 PH/44 There was no timeout on the connect() call when using a Unix domain
3970 socket in the ${readsocket expansion. There now is.
3972 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3973 be meaningful with "accept".
3975 SC/01 Eximstats V1.43
3976 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3978 SC/02 Eximstats V1.44
3979 Use a glob alias rather than an array ref in the generated
3980 parser. This improves both readability and performance.
3982 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3983 Collect SpamAssassin and rejection statistics.
3984 Don't display local sender or destination tables unless
3985 there is data to show.
3986 Added average volumes into the top table text output.
3988 SC/04 Eximstats V1.46
3989 Collect data on the number of addresses (recipients)
3990 as well as the number of messages.
3992 SC/05 Eximstats V1.47
3993 Added 'Message too big' to the list of mail rejection
3994 reasons (thanks to Marco Gaiarin).
3996 SC/06 Eximstats V1.48
3997 Mainlog lines which have GMT offsets and are too short to
3998 have a flag are now skipped.
4000 SC/07 Eximstats V1.49 (Alain Williams)
4001 Added the -emptyok flag.
4003 SC/08 Eximstats V1.50
4004 Fixes for obtaining the IP address from reject messages.
4006 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4007 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4008 whitespace changes from 4.64-PH/27
4010 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4013 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4014 are found, allow negative numbers in numeric criteria)
4016 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4018 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4019 to match changes made in 4.64-PH/43
4021 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4023 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4024 "domain" is actually an IP address, and force a failure. This locks out
4025 those revolvers/nameservers that support "A-for-A" lookups, in
4026 contravention of the specifications.
4028 PH/48 When a host name was looked up from an IP address, and the subsequent
4029 forward lookup of the name timed out, the host name was left in
4030 $sender_host_name, contrary to the specification.
4032 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4033 restricted to single-key lookups, Exim was not diagnosing an error if
4034 * or *@ was used with a query-style lookup.
4036 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4038 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4039 long ago, but noone interested enough thought of it. Let's just say that
4040 the "1.1" means that there are some new functions that weren't there at
4041 some point in the past.
4043 PH/51 Error processing for expansion failure of helo_data from an smtp
4044 transport during callout processing was broken.
4046 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4047 tested/used via the -bh/-bhc/-bs options.
4049 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4050 bug, fixed in subsequent PCRE releases).
4052 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4053 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4055 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4061 SC/01 Use a glob alias rather than an array ref in eximstats generated
4062 parser. This improves both readability and performance.
4064 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4065 Don't display local sender or destination tables in eximstats unless
4066 there is data to show.
4067 Added average volumes into the eximstats top table text output.
4069 SC/03 Collect data on the number of addresses (recipients) as well
4070 as the number of messages in eximstats.
4072 TF/01 Correct an error in the documentation for the redirect router. Exim
4073 does (usually) call initgroups() when daemonizing.
4075 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4076 with consistent privilege compared to when running as a daemon.
4078 TF/03 Note in the spec that $authenticated_id is not set for local
4079 submissions from trusted users.
4081 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4082 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4084 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4085 by adding some example configuration directives to the default
4086 configuration file. A little bit of work is required to uncomment the
4087 directives and define how usernames and passwords are checked, but
4088 there is now a framework to start from.
4090 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4091 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4092 without this. I don't know how relevant this is to other LDAP libraries.
4094 PH/02 Add the verb name to the "unknown ACL verb" error.
4096 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4098 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4100 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4101 directory not expanded when it should be if an expanded home directory
4102 was set for the address (which is overridden by the transport).
4104 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4107 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4108 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4109 because it is too late at that time, and has no effect.
4111 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4112 security issue with \' (bugzilla #107). I could not use the
4113 PQescapeStringConn() function, because it needs a PGconn value as one of
4116 PH/08 When testing addresses using -bt, indicate those final addresses that
4117 are duplicates that would not cause an additional delivery. At least one
4118 person was confused, thinking that -bt output corresponded to deliveries.
4119 (Suppressing duplicates isn't a good idea as you lose the information
4120 about possibly different redirections that led to the duplicates.)
4122 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4123 systems where poll() doesn't work, in particular OS X.
4125 PH/10 Added more information to debugging output for retry time not reached.
4127 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4128 operations in malware.c.
4130 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4133 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4134 syslog_duplication set false, log lines that would normally be written
4135 both the the main log and to the reject log were not written to syslog at
4138 PH/14 In the default configuration, change the use of "message" in ACL warn
4139 statements to "add_header".
4141 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4142 not followed by a command (e.g. "seen endif").
4144 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4145 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4148 PH/17 Added extra conditions to the default value of delay_warning_condition
4152 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4153 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4154 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4157 The Auto-Submitted: and various List- headers are standardised, whereas I
4158 don't think Precedence: ever was.
4160 PH/18 Refactored debugging code in route_finduser() to show more information,
4161 in particular, the error code if getpwnam() issues one.
4163 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4164 This is apparently needed in addition to the PH/07 change above to avoid
4165 any possible encoding problems.
4167 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4168 but not after initializing Perl.
4170 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4171 output them only if debugging. By default they are written stderr,
4172 apparently, which is not desirable.
4174 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4177 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4180 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4182 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4183 authenticated or an ident call has been made. Suppress the default
4184 values for $authenticated_id and $authenticated_sender (but permit -oMai
4185 and -oMas) when testing with -bh.
4187 PH/24 Re-jigged the order of the tests in the default configuration so that the
4188 tests for valid domains and recipients precede the DNS black list and CSA
4189 tests, on the grounds that those ones are more expensive.
4191 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4192 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4193 This bug exists in every version of Exim that I still have, right back to
4196 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4197 However, an attempt to turn on case-sensitivity in a regex key by
4198 including (?-i) didn't work because the subject string was already
4199 lowercased, and the effects were non-intuitive. It turns out that a
4200 one-line patch can be used to allow (?-i) to work as expected.
4206 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4207 other effects) broke the use of negated acl sub-conditions.
4209 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4212 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4213 "Deny" causes Exim to reject the incoming connection with a 554 error.
4214 Unfortunately, if there is a major crisis, such as a disk failure,
4215 tcp-wrappers gives "deny", whereas what one would like would be some
4216 kind of temporary error. A kludge has been added to help with this.
4217 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4218 554 error is used if errno is still zero or contains ENOENT (which occurs
4219 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4222 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4224 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4225 errors. Otherwise a message that provokes a temporary error (when other
4226 messages do not) can cause a whole host to time out.
4228 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4229 addresses were routed directly to files or pipes from a redirect router.
4230 File deliveries just didn't batch; pipe deliveries might have suffered
4233 PH/06 A failure to get a lock for a hints database would erroneously always say
4234 "Failed to get write lock", even when it was really a read lock.
4236 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4237 of 0600. This has been changed to use the value of the lockfile_mode
4238 option (which defaults to 0600).
4240 PH/08 Applied small patch from the Sieve maintainer.
4242 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4243 folder from quota calculations, a direct delivery into this folder messed
4244 up the contents of the maildirsize file. This was because the regex was
4245 used only to exclude .Trash (or whatever) when the size of the mailbox
4246 was calculated. There was no check that a delivery was happening into an
4247 excluded directory. This bug has been fixed by ignoring all quota
4248 processing for deliveries into excluded directories.
4250 PH/10 Added the maildirfolder_create_regex option to appendfile.
4256 PH/01 The code for finding all the local interface addresses on a FreeBSD
4257 system running IPv6 was broken. This may well have applied to all BSD
4258 systems, as well as to others that have similar system calls. The broken
4259 code found IPv4 interfaces correctly, but gave incorrect values for the
4260 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4261 that it would not match correctly against @[] and not recognize the IPv6
4264 PH/02 The ipliteral router was not recognizing addresses of the form user@
4265 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4267 PH/03 Added disable_ipv6.
4269 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4270 decoded form, because it is most often used to construct To: headers
4271 lines in autoreplies, and the decoded form may well be syntactically
4272 invalid. However, $reply_address has leading white space removed, and all
4273 newlines turned into spaces so that the autoreply transport does not
4276 PH/05 If group was specified without a user on a router, and no group or user
4277 was specified on a transport, the group from the router was ignored.
4279 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4280 for visible compile-time settings that can be used to change these
4281 numbers, for those that want even more. Backwards compatibility with old
4282 spool files has been maintained. However, going back to a previous Exim
4283 release will lost any variables that are in spool files.
4285 PH/07 Two small changes when running in the test harness: increase delay when
4286 passing a TCP/IP connection to a new process, in case the original
4287 process has to generate a bounce, and remove special handling of
4288 127.0.0.2 (sic), which is no longer necessary.
4290 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4291 be the same on different OS.
4293 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4296 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4297 whether --show-vars was specified or not
4299 JJ/02 exipick: Added support for new ACL variable spool format introduced
4302 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4303 syntactically invalid From: or Reply-to: line, and a filter used this to
4304 generate an autoreply, and therefore failed to obtain an address for the
4305 autoreply, Exim could try to deliver to a non-existent relative file
4306 name, causing unrelated and misleading errors. What now happens is that
4307 it logs this as a hard delivery error, but does not attempt to create a
4310 PH/11 The exinext utility has a -C option for testing purposes, but although
4311 the given file was scanned by exinext itself; it wasn't being passed on
4312 when Exim was called.
4314 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4315 an end-of-file indication when reading a command response.
4317 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4318 compiled. In many other places in Exim, IPv6 addresses are always
4319 recognized, so I have changed this. It also means that IPv4 domain
4320 literals of the form [IPV4:n.n.n.n] are now always recognized.
4322 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4323 used if the router is not running as root, for example, when verifying at
4324 ACL time, or when using -bh. The debugging output from this situation was
4325 non-existent - all you got was a failure to exec. I have made two
4328 (a) Failures to set uid/gid, the current directory, or a process leader
4329 in a subprocess such as that created by queryprogram now generate
4330 suitable debugging output when -d is set.
4332 (b) The queryprogram router detects when it is not running as root,
4333 outputs suitable debugging information if -d is set, and then runs
4334 the subprocess without attempting to change uid/gid.
4336 PH/15 Minor change to Makefile for building test_host (undocumented testing
4339 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4340 additional section of a DNS packet that returns MX or SRV records.
4341 Instead, it always explicitly searches for A/AAAA records. This avoids
4342 major problems that occur when a DNS server includes only records of one
4343 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4344 fixed another bug: if SRV records were looked up and the corresponding
4345 address records were *not* found in the additional section, the port
4346 values from the SRV records were lost.
4348 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4349 using the correct key (the original address) when searching the retry
4350 rules in order to find which one to use for generating the retry hint.
4352 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4353 adding the default one. Similarly, if it contains a Reply-To: header, the
4354 errors_reply_to option, if set, is not used.
4356 PH/19 When calculating a retry time, Exim used to measure the "time since
4357 failure" by looking at the "first failed" field in the retry record. Now
4358 it does not use this if it is later than than the arrival time of the
4359 message. Instead it uses the arrival time. This makes for better
4360 behaviour in cases where some deliveries succeed, thus re-setting the
4361 "first failed" field. An example is a quota failure for a huge message
4362 when small messages continue to be delivered. Without this change, the
4363 "time since failure" will always be short, possible causing more frequent
4364 delivery attempts for the huge message than are intended.
4365 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4367 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4368 $1, $2, $3) because the numerical variables can be reset during some
4369 expansion items (e.g. "match"), thereby losing the authentication data.
4371 PH/21 Make -bV show the size of off_t variables so that the test suite can
4372 decide whether to run tests for quotas > 2G.
4374 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4375 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4376 filecount value is greater than 2G or if a quota value is greater than 2G
4377 on a system where the size of off_t is not greater than 4, a panic error
4380 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4381 never match. The debug and -bh output now contains an explicit error
4382 message indicating a malformed IPv4 address or mask.
4384 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4385 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4386 PH/23 above applies.
4388 PH/25 Do not write to syslog when running in the test harness. The only
4389 occasion when this arises is a failure to open the main or panic logs
4390 (for which there is an explicit test).
4392 PH/26 Added the /no_tell option to "control=freeze".
4394 PH/27 If a host name lookup failed very early in a connection, for example, if
4395 the IP address matched host_lookup and the reverse lookup yielded a name
4396 that did not have a forward lookup, an error message of the form "no IP
4397 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4398 could be logged. Now it outputs the IP address instead of "NULL".
4400 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4401 allows the sender and the authenticated sender to be set when
4402 submitting a message from within Exim. Since child_open_exim() is
4403 documented for local_scan(), the new function should be too.
4405 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4406 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4407 results in an empty string is now treated as unset.
4409 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4411 PH/31 Added sender_verify_fail logging option.
4413 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4414 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4415 tidied the source and removed it altogether.
4417 PH/33 When a queue run was abandoned because the load average was too high, a
4418 log line was always written; now it is written only if the queue_run log
4419 selector is set. In addition, the log line for abandonment now contains
4420 information about the queue run such as the pid. This is always present
4421 in "start" and "stop" lines but was omitted from the "abandon" line.
4423 PH/34 Omit spaces between a header name and the colon in the error message that
4424 is given when verify = headers_syntax fails (if there are lots of them,
4425 the message gets confusing).
4427 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4428 names, as there are now some PTR records that contain slashes. This check
4429 is only to protect against broken name servers that fall over on strange
4430 characters, so the fact that it applies to all lookups doesn't matter.
4432 PH/36 Now that the new test suite is complete, we can remove some of the
4433 special code in Exim that was needed for the old test suite. For example,
4434 sorting DNS records because real resolvers return them in an arbitrary
4435 order. The new test suite's fake resolver always returns records in the
4438 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4439 bounces) except when queue_only is set, to avoid logging races between
4440 the different processes.
4442 PH/38 Panic-die if .include specifies a non-absolute path.
4444 PH/39 A tweak to the "H" retry rule from its user.
4446 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4447 a label. They prevented compilation on older perls.
4449 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4450 a warning to be raised on newish perls.
4452 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4453 on queue. Changes to match documented behaviour of showing count of
4454 messages matching specified criteria.
4456 PH/40 Changed the default ident timeout from 30s to 5s.
4458 PH/41 Added support for the use of login_cap features, on those BSD systems
4459 that have them, for controlling the resources used by pipe deliveries.
4461 PH/42 The content-scanning code uses fopen() to create files in which to put
4462 message data. Previously it was not paying any attention to the mode of
4463 the files. Exim runs with umask(0) because the rest of the code creates
4464 files with open(), and sets the required mode explicitly. Thus, these
4465 files were ending up world-writeable. This was not a big issue, because,
4466 being within the spool directory, they were not world-accessible. I have
4467 created a function called modefopen, which takes an additional mode
4468 argument. It sets umask(777), creates the file, chmods it to the required
4469 mode, then resets the umask. All the relevant calls to fopen() in the
4470 content scanning code have been changed to use this function.
4472 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4473 to 24 hours. This avoids potential overflow problems when processing G
4474 and H retry rules. I suspect nobody ever tinkers with this value.
4476 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4478 PH/45 When the plaintext authenticator is running as a client, the server's
4479 challenges are checked to ensure they are valid base64 strings. By
4480 default, the authentication attempt is cancelled if an invalid string is
4481 received. Setting client_ignore_invalid_base64 true ignores these errors.
4482 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4483 they are received. Thus, the responses can be made to depend on the
4484 challenges. If an invalid string is ignored, an empty string is placed in
4487 PH/46 Messages that are created by the autoreply transport now contains a
4488 References: header, in accordance with RFCs 2822 and 3834.
4490 PH/47 Added authenticated_sender_force to the smtp transport.
4492 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4494 PH/49 Installed latest patch from the Sieve maintainer.
4496 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4497 was also set, the appendfile transport was unnecessarily scanning a
4498 directory of message files (e.g. for maildir delivery) to find the count
4499 of files (along with the size), even though it did not need this
4500 information. It now does the scan only if it needs to find either the
4501 size of the count of files.
4503 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4505 PH/52 Two bugs concerned with error handling when the smtp transport is
4508 (i) Exim was not creating retry information for temporary errors given
4509 for individual recipients after the DATA command when the smtp transport
4510 was used in LMTP mode. This meant that they could be retried too
4511 frequently, and not timed out correctly.
4513 (ii) Exim was setting the flag that allows error details to be returned
4514 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4515 recipients that were returned after the DATA command.
4517 PH/53 This is related to PH/52, but is more general: for any failing address,
4518 when detailed error information was permitted to be returned to the
4519 sender, but the error was temporary, then after the final timeout, only
4520 "retry timeout exceeded" was returned. Now it returns the full error as
4521 well as "retry timeout exceeded".
4523 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4524 do this, and (what is worse) MTAs that accept it.
4526 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4527 will now be deprecated.
4529 PH/56 New os.c-cygwin from the Cygwin maintainer.
4531 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4532 formats (previously only available in exim formats via -bpr, -bpru,
4533 and -bpra. Now also available in native and exiqgrep formats)
4535 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4536 with very large, slow to parse queues
4538 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4540 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4542 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4543 responses to authentication challenges, though it was showing the
4544 challenges; (ii) I've removed the CR characters from the debug output for
4547 PH/58 Allow for the insertion of a newline as well as a space when a string
4548 is turned into more than one encoded-word during RFC 2047 encoding. The
4549 Sieve code now uses this.
4551 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4552 data_4xx, lost_connection, tls_required.
4554 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4555 message was being sent as an SMTP response.
4557 PH/61 Add -l and -k options to exicyclog.
4559 PH/62 When verifying, if an address was redirected to one new address, so that
4560 verification continued, and the new address failed or deferred after
4561 having set something in $address_data, the value of $address_data was not
4562 passed back to the ACL. This was different to the case when no
4563 redirection occurred. The value is now passed back in both cases.
4565 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4566 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4567 use login_cap.h, so on its own it isn't the distinguishing feature. The
4568 new name refers directly to the setclassresources() function.
4570 PH/65 Added configuration files for NetBSD3.
4572 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4574 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4575 is preferred over IPv4.
4577 PH/68 The bounce_return_message and bounce_return_body options were not being
4578 honoured for bounces generated during the reception of non-SMTP messages.
4579 In particular, this applied to messages rejected by the ACL. This bug has
4580 been fixed. However, if bounce_return_message is true and bounce_return_
4581 body is false, the headers that are returned for a non-SMTP message
4582 include only those that have been read before the error was detected.
4583 (In the case of an ACL rejection, they have all been read.)
4585 PH/69 The HTML version of the specification is now built in a directory called
4586 spec_html instead of spec.html, because the latter looks like a path with
4587 a MIME-type, and this confuses some software.
4589 PH/70 Catch two compiler warnings in sieve.c.
4591 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4592 function verify_get_ident() calls ip_connect() to connect a socket, but
4593 if the "connect()" function timed out, ip_connect() used to close the
4594 socket. However, verify_get_ident() also closes the socket later, and in
4595 between Exim writes to the log, which may get opened at this point. When
4596 the socket was closed in ip_connect(), the log could get the same file
4597 descriptor number as the socket. This naturally causes chaos. The fix is
4598 not to close the socket in ip_connect(); the socket should be closed by
4599 the function that creates it. There was only one place in the code where
4600 this was missing, in the iplookup router, which I don't think anybody now
4601 uses, but I've fixed it anyway.
4603 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4604 well as to direct DNS lookups. Otherwise the handling of names in host
4605 lists is inconsistent and therefore confusing.
4611 PH/01 Two changes to the default runtime configuration:
4613 (1) Move the checks for relay_from_hosts and authenticated clients from
4614 after to before the (commented out) DNS black list checks.
4616 (2) Add control=submission to the relay_from_hosts and authenticated
4617 clients checks, on the grounds that messages accepted by these
4618 statements are most likely to be submissions.
4620 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4622 (1) Generate an error if the third argument for the ${prvs expansion is
4625 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4628 (3) Reset the variables that are obtained from the first argument of
4629 ${prvscheck and used in the second argument before leaving the code,
4630 because their memory is reclaimed, so using them afterwards may do
4633 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4634 one (it's much easier than Tom thought :-).
4636 (5) Because of (4), we can now allow for the use of $prvscheck_result
4637 inside the third argument.
4639 PH/03 For some reason, the default setting of PATH when running a command from
4640 a pipe transport was just "/usr/bin". I have changed it to
4643 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4644 anything to be listed in the output from -bV.
4646 PH/05 When a filter generated an autoreply, the entire To: header line was
4647 quoted in the delivery log line, like this:
4649 => >A.N.Other <ano@some.domain> <original@ddress> ...
4651 This has been changed so that it extracts the operative address. There
4652 may be more than one such address. If so, they are comma-separated, like
4655 => >ano@some.domain,ona@other.domain <original@ddress> ...
4657 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4658 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4659 not being looked up in the rDNS to get a host name, Exim was showing the
4660 IP address twice in Received: lines, even though the IP addresses were
4661 identical. For example:
4663 Received: from [1.2.3.4] (helo=[1.2.3.4])
4665 However, if the real host name was known, it was omitting the HELO data
4666 if it matched the actual IP address. This has been tidied up so that it
4667 doesn't show the same IP address twice.
4669 PH/07 When both +timestamp and +memory debugging was on, the value given by
4670 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4671 called by the debug printing, thereby overwriting the timestamp buffer.
4672 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4674 PH/08 When the original message was included in an autoreply transport, it
4675 always said "this is a copy of the message, including all the headers",
4676 even if body_only or headers_only was set. It now gives an appropriate
4679 PH/09 Applied a patch from the Sieve maintainer which:
4681 o fixes some comments
4682 o adds the (disabled) notify extension core
4683 o adds some debug output for the result of if/elsif tests
4684 o points to the current vacation draft in the documentation
4685 and documents the missing references header update
4689 o fixes a bug in processing the envelope test (when testing
4690 multiple envelope elements, the last element determined the
4693 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4694 Electronic Mail") by including:
4696 Auto-submitted: auto-generated
4698 in the messages that it generates (bounce messages and others, such as
4699 warnings). In the case of bounce messages for non-SMTP messages, there was
4700 also a typo: it was using "Auto_submitted" (underscore instead of
4701 hyphen). Since every message generated by Exim is necessarily in response
4702 to another message, thes have all been changed to:
4704 Auto-Submitted: auto-replied
4706 in accordance with these statements in the RFC:
4708 The auto-replied keyword:
4710 - SHOULD be used on messages sent in direct response to another
4711 message by an automatic process,
4713 - MUST NOT be used on manually-generated messages,
4715 - MAY be used on Delivery Status Notifications (DSNs) and Message
4716 Disposition Notifications (MDNs),
4718 - MUST NOT be used on messages generated by automatic or periodic
4719 processes, except for messages which are automatic responses to
4722 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4723 to the default Received: header definition.
4725 PH/12 Added log selector acl_warn_skipped (default on).
4727 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4728 variables because (a) they are in the wrong storage pool and (b) even if
4729 they were copied, it wouldn't work properly because of the caching.
4731 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4732 checking when decoding. Apparently there are clients that generate
4733 overlong encoded strings. Why am I not surprised?
4735 PH/15 If the first argument of "${if match_address" was not empty, but did not
4736 contain an "@" character, Exim crashed. Now it writes a panic log message
4737 and treats the condition as false.
4739 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4741 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4742 extension "envelope-auth". The code is finished and in agreement with
4743 other implementations, but there is no documentation so far and in fact,
4744 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4745 not changing the active code.
4747 Print executed "if" and "elsif" statements when debugging is used. This
4748 helps a great deal to understand what a filter does.
4750 Document more things not specified clearly in RFC3028. I had all this
4751 sorted out, when out of a sudden new issues came to my mind. Oops."
4753 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4756 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4757 canonical form (as documented). However, after a host name lookup from
4758 the IP address, check_host() was doing a simple string comparison with
4759 addresses acquired from the DNS when checking that the found name did
4760 have the original IP as one of its addresses. Since any found IPv6
4761 addresses are likely to be in abbreviated form, the comparison could
4762 fail. Luckily, there already exists a function for doing the comparison
4763 by converting both addresses to binary, so now that is used instead of
4764 the text comparison.
4766 PH/20 There was another similar case to PH/19, when a complete host name was
4767 given in a host list; looking up its IP address could give an abbreviated
4768 form, whereas the current host's name might or might not be abbreviated.
4769 The same fix has been applied.
4775 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4776 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4779 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4780 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4782 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4784 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4785 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4786 Also, for "Auto-Submitted", treat anything other than "no" as
4787 non-personal, in accordance with RFC 3834. (Previously it treated
4788 anything starting "auto-" as non-personal.)
4790 TF/01 The control=submission/name=... option had a problem with syntax
4791 errors if the name included a slash character. The /name= option
4792 now slurps the rest of the string, so it can include any characters
4793 but it must come last in the list of options (after /sender_retain
4796 PH/05 Some modifications to the interface to the fake nameserver for the new
4804 TK/01 Added the "success_on_redirect" address verification option. See
4805 NewStuff for rationale and an example.
4807 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4809 PH/02 Patch to exigrep to allow it to work on syslog lines.
4811 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4812 fread() to skip over the body file's header line, because in Cygwin the
4813 header line is locked and is inaccessible.
4815 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4816 co-exist for some time) to make it clear that it is the Exim ID that is
4817 referenced, not the Message-ID: header line.
4819 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4820 string_format() function, because snprintf() does not exist on all
4823 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4824 ${stat: expansion item.
4826 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4827 protocol synchronization error", to keep the pedants happy.
4829 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4830 well as for IRIX systems, when gcc is being used. See the host.c source
4833 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4835 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4838 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4839 transport and to the smtp transport in LMTP mode.
4841 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4843 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4844 run a filter in a subprocess. This could lead to confusion in subsequent
4845 lookups in the parent process. There should also be a search_tidyup() at
4846 the end of the subprocess.
4848 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4849 only if the host matched helo_try_verify_hosts, which caused the
4850 verification to occur when the EHLO/HELO command was issued. The ACL just
4851 tested the remembered result. Now, if a previous verification attempt has
4852 not happened, "verify = helo" does it there and then.
4854 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4856 TK/03 Fix log output including CR from clamd.
4858 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4859 exist provoked a memory error which could cause a segfault.
4861 PH/15 Installed PCRE 6.2
4863 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4865 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4866 of the problem. Specifically, suggested +O2 rather than +O1 for the
4869 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4871 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4872 record, the port was not correctly logged when the outgoing_port log
4873 selector was set (it logged the transort's default port).
4875 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4876 fallback_hosts, and "hosts" in the smtp transport.
4878 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4879 host errors such as "Connection refused".
4881 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4882 authentication with radiusclient 0.4.9:
4884 - Error returned from rc_read_config was caught wrongly
4885 - Username/password not passed on to radius server due to wrong length.
4887 The presumption is that some radiusclient API changes for 4.51/PH/17
4888 were not taken care of correctly. The code is still untested by me (my
4889 Linux distribution still has 0.3.2 of radiusclient), but it was
4890 contributed by a Radius user.
4892 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4893 expanding the "port" option of the smtp transport.
4895 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4896 while reading a MIME header. Thanks to Tom Hughes for a patch.
4898 PH/24 Include config.h inside local_scan.h so that configuration settings are
4901 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4902 that in an ACL for RCPT (for example), you can examine exactly what was
4905 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4906 commands, but it was not correctly comparing the address with the actual
4907 client host address. Thus, it would show the EHLO address in Received:
4908 header lines when this was not necessary.
4910 PH/27 Added the % operator to ${eval:}.
4912 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4913 it should be ignoring failures (because with -C, for example, it has lost
4914 privilege). It wasn't ignoring creation failures other than "already
4917 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4920 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4921 because an input line was too long, either on its own, or by virtue of
4922 too many continuations, the temporary file was not being removed, and the
4923 return code was incorrect.
4925 PH/31 Missing "BOOL" in function definition in filtertest.c.
4927 PH/32 Applied Sieve patches from the maintainer.
4929 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4931 PH/33 Added "verify = not_blind".
4933 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4934 Local/Makefile (with some defaults set). These are used in built scripts
4935 such as exicyclog, but they have never been used in the exim_install
4936 script (though there are many overriding facilities there). I have
4937 arranged that the exim_install script now takes note of these two
4940 PH/35 Installed configuration files for Dragonfly.
4942 PH/36 When a locally submitted message by a trusted user did not contain a
4943 From: header, and the sender address was obtained from -f or from an SMTP
4944 MAIL command, and the trusted user did not use -F to supply a sender
4945 name, $originator_name was incorrectly used when constructing a From:
4946 header. Furthermore, $originator_name was used for submission mode
4947 messages from external hosts without From: headers in a similar way,
4948 which is clearly wrong.
4950 PH/37 Added control=suppress_local_fixups.
4952 PH/38 When log_selector = +received_sender was set, and the addition of the
4953 sender made the log line's construction buffer exactly full, or one byte
4954 less than full, an overflow happened when the terminating "\n" was
4957 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4958 when the result of a list match is failure because a DNS lookup failed.
4960 PH/40 RM_COMMAND is now used in the building process.
4962 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4963 the "build-* directories that it finds.
4965 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4966 address in a domain literal was a prefix of an interface address.
4968 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4969 when verifying a sender address, unless rewrite_headers is false.
4971 PH/44 Wrote a long comment about why errors_to addresses are verified as
4972 recipients, not senders.
4974 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4975 the ratelimit ACL was added.
4977 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4979 PH/46 Added extra information about PostgreSQL errors to the error string.
4981 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4982 avoiding the need to install special zones in a real server. This is
4983 backwards compatible; if it can't find the fake resolver, it drops back.
4984 Thus, both old and new test suites can be run.
4986 TF/02 Added util/ratelimit.pl
4988 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4989 clock is set back in time.
4991 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4992 Candler <B.Candler@pobox.com>.
4994 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4995 OK for addresses that are the result of redirections.
4997 PH/48 A number of further additions for the benefit of the new test suite,
4998 including a fake gethostbyname() that interfaces to the fake DNS resolver
5001 TF/06 The fix for widen_domains has also been applied to qualify_single and
5002 search_parents which are the other dnslookup options that can cause
5005 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5008 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5010 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5011 service" if it encountered an MX record with an empty target hostname.
5012 The message is now "an MX or SRV record indicated no SMTP service".
5014 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5015 if the DNS of the sending site is misconfigured. This is quite a
5016 common situation. This change restores the behaviour of treating a
5017 helo verification defer as a failure.
5019 PH/51 If self=fail was set on a router, the bounce message did not include the
5020 actual error message.
5026 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5028 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5029 command itself ended in error, the underlying message about the transport
5030 filter timeout was being overwritten with the pipe command error. Now the
5031 underlying error message should be appended to the second error message.
5033 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5035 PH/02 Reduce the amount of output that "make" produces by default. Full output
5036 can still be requested.
5038 PH/03 The warning log line about a condition test deferring for a "warn" verb
5039 was being output only once per connection, rather than after each
5040 occurrence (because it was using the same function as for successful
5041 "warn" verbs). This seems wrong, so I have changed it.
5043 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5044 it should not have, which might have caused a crash in the right
5045 circumstances, but probably never did.
5047 PH/04 Installed a modified version of Tony Finch's patch to make submission
5048 mode fix the return path as well as the Sender: header line, and to
5049 add a /name= option so that you can make the user's friendly name appear
5052 TF/03 Added the control = fakedefer ACL modifier.
5054 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5055 Mark Lowes for thorough testing.
5057 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5059 TK/03 Merged latest SRS patch from Miles Wilton.
5061 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5062 with the definition in sysexits.h (which is #included earlier).
5063 Fortunately, Exim does not actually use EX_OK. The code used to try to
5064 preserve the sysexits.h value, by assuming that macro definitions were
5065 scanned for macro replacements. I have been disabused of this notion,
5066 so now the code just undefines EX_OK before #including unistd.h.
5068 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5069 in the smtp transport. When a block could not be written in a single
5070 write() function, the timeout was being re-applied to each part-write.
5071 This seems wrong - if the receiver was accepting one byte at a time it
5072 would take for ever. The timeout is now adjusted when this happens. It
5073 doesn't have to be particularly precise.
5075 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5076 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5078 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5079 without a domain if the check to <postmaster@domain> fails.
5081 SC/01 Eximstats: added -xls and the ability to specify output files
5082 (patch written by Frank Heydlauf).
5084 SC/02 Eximstats: use FileHandles for outputting results.
5086 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5088 SC/04 Eximstats: fixed display of large numbers with -nvr option
5090 SC/05 Eximstats: fixed merging of reports with empty tables.
5092 SC/06 Eximstats: added the -include_original_destination flag
5094 SC/07 Eximstats: removed tabs and trailing whitespace.
5096 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5098 TK/06 MBOX spool code: Add real "From " MBOX separator line
5099 so the .eml file is really in mbox format (even though
5100 most programs do not really care). Patch from Alex Miller.
5102 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5103 The latter is generated from $received_to and is only set if the
5104 message has one envelope recipient. SA can use these headers,
5105 obviously out-of-the-box. Patch from Alex Miller.
5107 PH/08 The ${def test on a variable was returning false if the variable's
5108 value was "0", contrary to what the specification has always said!
5109 The result should be true unless the variable is empty.
5111 PH/09 The syntax error of a character other than { following "${if
5112 def:variable_name" (after optional whitespace) was not being diagnosed.
5113 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5114 accidental colon was present, for example, could give incorrect results.
5116 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5117 result is used (not including appendfile, where other changes are about
5120 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5121 This involved changing a lot of size variables from int to off_t. It
5122 should work with maildirs and everything.
5124 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5125 spamd dying while we are connected to it.
5127 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5130 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5131 function for BDB 4.3.
5133 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5135 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5136 the sources for PCRE have been re-arranged and more files are now
5139 PH/15 The code I had for printing potentially long long variables in PH/11
5140 above was not the best (it lost precision). The length of off_t variables
5141 is now inspected at build time, and an appropriate printing format (%ld
5142 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5143 to be "long long int" or "long int". This is needed for the internal
5144 formatting function string_vformat().
5146 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5147 the configuration file to be ":syslog", then the script "guesses" where
5148 the logs files are, rather than using the compiled in default. In our
5149 case the guess is not the same as the compiled default, so the script
5150 suddenly stopped working when I started to use syslog. The patch checks
5151 to see if log_file_path is "". If so, it attempts to read it from exim
5152 with no configuration file to get the compiled in version, before it
5153 falls back to the previous guessing code."
5155 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5156 implementing BATV in an Exim configuration. See NewStuff for the gory
5159 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5160 Makefile that are specific to HP-UX.
5162 PH/18 If the "use_postmaster" option was set for a recipient callout together
5163 with the "random" option, the postmaster address was used as the MAIL
5164 FROM address for the random test, but not for the subsequent recipient
5165 test. It is now used for both.
5167 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5168 patch removes a few documentation additions to RFC 3028, because the
5169 latest draft now contains them. It adds the new en;ascii-case comparator
5170 and a new error check for 8bit text in MIME parts. Comparator and
5171 require names are now matched exactly. I enabled the subaddress
5172 extension, but it is not well tested yet (read: it works for me)."
5174 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5175 rework some of the code of TK/09 above to avoid the hardwired use of
5176 "%lld" and "long long". Replaced the call to snprintf() with a call to
5179 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5180 records point to non-existent hosts", "retry timeout exceeded", and
5181 "retry time not reached for any host after a long failure period".
5183 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5184 experimental DomainKeys support:
5186 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5187 (2) On an error such as an illegally used "control", the wrong name for
5188 the control was given.
5190 These problems did NOT occur unless DomainKeys support was compiled.
5192 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5194 PH/24 Added ${if match_ip condition.
5196 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5197 databases so that it will be absolutely obvious if a crash occurs in the
5198 DB library. This is a regular occurrence (often caused by mis-matched
5201 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5202 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5203 picked up on a user's system that detects such things. There doesn't seem
5204 to be a gcc warning option for this - only an attribute that has to be
5205 put on the function's prototype. It seems that in Fedora Core 4 they have
5206 set this on a number of new functions. No doubt there will be more in due
5209 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5210 specify a transport. However, if an address that was verified by such a
5211 router was the subject of a callout, Exim crashed because it tried to
5212 read the rcpt_include_affixes from the non-existent transport. Now it
5213 just assumes that the setting of that option is false. This bug was
5214 introduced by 4.51/PH/31.
5216 PH/28 Changed -d+all to exclude +memory, because that information is very
5217 rarely of interest, but it makes the output a lot bigger. People tend to
5218 do -d+all out of habit.
5220 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5221 code in os-type was giving problems when libc.so lives in lib64, like on
5224 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5225 aren't the modern standard, and it seems that some systems' include files
5226 don't always have them. Exim was already checking for some of the newer
5227 ones like T_AAAA, and defining it itself. I've added checks for all the
5228 record types that Exim uses.
5230 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5231 not automatically generating a new one, as it is supposed to. This
5232 prevented TLS from working. If the file did exist, but contained invalid
5233 data, a new version was generated, as expected. It was only the case of a
5234 non-existent file that was broken.
5236 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5237 with a change in libdomainkeys > 0.64.
5239 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5240 from DNS. If the selector record carries the flag, it now has
5241 precedence over the domain-wide flag.
5243 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5245 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5246 the use of a port name that isn't defined in /etc/services) occurred, the
5247 message was deferred as in a normal delivery, and thus remained on the
5248 spool, instead of being failed because of the mua_wrapper setting. This
5249 is now fixed, and I tidied up some of the mua_wrapper messages at the
5252 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5253 the messages in a hash of arrays rather than using individual hashes.
5254 This is a bit cleaner and results in dramatic memory savings, albeit
5255 at a slight CPU cost.
5257 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5258 as requested by Marc Sherman.
5260 SC/10 Eximstats: added histograms for user specified patterns as requested
5263 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5265 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5266 fopen() in the content-scanning modules that did not already have it.
5272 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5273 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5275 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5277 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5279 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5280 to test Sieve filters that use "vacation".
5282 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5283 that changes the way the GnuTLS parameters are stored in the cache file.
5284 The new format can be generated externally. For backward compatibility,
5285 if the data in the cache doesn't make sense, Exim assumes it has read an
5286 old-format file, and it generates new data and writes a new file. This
5287 means that you can't go back to an older release without removing the
5290 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5291 work if there are any delivery delays because "one_time" forces the
5292 parent to be marked "delivered", so its unseen clone is never tried
5293 again. For this reason, Exim now forbids the simultaneous setting of
5296 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5297 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5298 entry if you want to know the details. The fix, however, neglected to
5299 consider the case when local delivery batching is involved. The test for
5300 "previously delivered" was not happening when checking to see if an
5301 address could be batched with a previous (undelivered) one; under
5302 certain circumstances this could lead to multiple deliveries to the same
5305 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5306 in its include files, and this causes problems building Exim.
5308 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5309 header_syntax) but Exim was just ignoring anything given after a slash.
5310 In particular, this caused confusion with an attempt to use "verify =
5311 reverse_host_lookup/defer_ok". An error is now given when options are
5312 supplied for verify items that do not have them. (Maybe reverse_host_
5313 lookup should have a defer_ok option, but that's a different point.)
5315 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5316 defined by RFC 821) to 2048, because there were problems with some AUTH
5317 commands, and RFC 1869 says the size should be increased for extended
5318 SMTP commands that take arguments.
5320 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5323 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5324 "unknown" error; now it says that the functionality isn't in the binary.
5326 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5327 an address' error message when a string expansion fails (syntax or
5328 whatever). Otherwise the password may appear in the log. Following change
5329 PH/42 below, there is no longer a chance of it appearing in a bounce
5332 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5334 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5335 of its addresses was ever tried. (Bugzilla bug #2.)
5337 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5338 the result incorrectly in the debug output. (It correctly added a newline
5339 to what was transported.)
5341 TF/01 Added $received_time.
5343 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5344 commented out examples of how to interface to a virus scanner and to
5345 SpamAssassin. Also added commented examples of av_scanner and
5346 spamd_address settings.
5348 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5349 and controls are allowed in which ACLs. There were a couple of minor
5350 errors. Some of the entries in the conditions table (which is a table of
5351 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5352 negation of where the condition IS allowed.
5354 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5356 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5357 header file does not have a version number, so I've had to invent a new
5358 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5359 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5360 radiusclient), but it was contributed by a Radius user.
5362 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5363 files or directories, for OpenSSL.
5365 PH/19 When an Exim process that is running as root has to create an Exim log
5366 file, it does so in a subprocess that runs as exim:exim so as to get the
5367 ownership right at creation (otherwise, other Exim processes might see
5368 the file with the wrong ownership). There was no test for failure of this
5369 fork() call, which would lead to the process getting stuck as it waited
5370 for a non-existent subprocess. Forks do occasionally fail when resources
5371 run out. I reviewed all the other calls to fork(); they all seem to check
5374 PH/20 When checking for unexpected SMTP input at connect time (before writing
5375 the banner), Exim was not dealing correctly with a non-positive return
5376 from the read() function. If the client had disconnected by this time,
5377 the result was a log entry for a synchronization error with an empty
5378 string after "input=" when read() returned zero. If read() returned -1
5379 (an event I could not check), uninitialized data bytes were printed.
5380 There were reports of junk text (parts of files, etc) appearing after
5383 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5385 PH/22 Added support for macro redefinition, and (re)definition in between
5386 driver and ACL definitions.
5388 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5389 forgetting to use the resulting value; it was using the unexpanded value.
5391 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5392 hadn't been configured. The fix is from Juergen Kreileder, who
5393 understands it better than I do:
5395 "Here's what I see happening with three configured cyrus_sasl
5396 authenticators configured (plain, login, cram-md5):
5398 On startup auth_cyrus_sasl_init() gets called for each of these.
5399 This means three calls to sasl_listmech() without a specified mech_list.
5400 => SASL tests which mechs of all available mechs actually work
5401 => three warnings about OTP not working
5402 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5404 With the patch, sasl_listmech() also gets called three times. But now
5405 SASL's mech_list option is set to the server_mech specified in the the
5406 authenticator. Or in other words, the answer from sasl_listmech()
5407 gets limited to just the mech you're testing for (which is different
5409 => the return list contains just 'plain' or 'login', 'cram-md5' or
5410 nothing depending on the value of ob->server_mech.
5412 I've just tested the patch: Authentication still works fine,
5413 unavailable mechs specified in the exim configuration are still
5414 caught, and the auth.log warnings about OTP are gone."
5416 PH/25 When debugging is enabled, the contents of the command line are added
5417 to the debugging output, even when log_selector=+arguments is not
5420 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5421 answer is "GNU", and only if the return is "GNU/something" is the answer
5424 PH/27 $acl_verify_message is now set immediately after the failure of a
5425 verification in an ACL, and so is available in subsequent modifiers. In
5426 particular, the message can be preserved by coding like this:
5428 warn !verify = sender
5429 set acl_m0 = $acl_verify_message
5431 Previously, $acl_verify_message was set only while expanding "message"
5432 and "log_message" when a very denied access.
5434 PH/28 Modified OS/os.c-Linux with
5436 -#ifndef OS_LOAD_AVERAGE
5437 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5439 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5440 nomenclature these days.)
5442 PH/29 Installed patch from the Sieve maintainer that adds the options
5443 sieve_useraddress and sieve_subaddress to the redirect router.
5445 PH/30 In these circumstances:
5446 . Two addresses routed to the same list of hosts;
5447 . First host does not offer TLS;
5448 . First host accepts first address;
5449 . First host gives temporary error to second address;
5450 . Second host offers TLS and a TLS session is established;
5451 . Second host accepts second address.
5452 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5453 and peerdn, if requested) that were in fact used only for the second
5456 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5457 attention to any local part prefix or suffix that was matched by the
5458 router that accepted the address. It now behaves in the same way as it
5459 does for delivery: the affixes are removed from the local part unless
5460 rcpt_include_affixes is set on the transport.
5462 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5463 timeout during the DATA phase of an incoming message.
5465 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5466 applied a patch sanctioned by the Sieve maintainer.
5468 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5469 the uid or gid is negative. A case of a negative gid caused this to be
5470 noticed. The fix allows for either to be negative.
5472 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5473 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5476 PH/36 The change PH/12 above was broken. Fixed it.
5478 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5479 the grounds that routing the same address twice would always produce the
5480 same answer. This might have been true once, but it is certainly no
5481 longer true now. Routing a child address may depend on the previous
5482 routing that produced that child. Some complicated redirection strategies
5483 went wrong when messages had multiple recipients, and made Exim's
5484 behaviour dependent on the order in which the addresses were given.
5486 I have moved the duplicate checking until after the routing is complete.
5487 Exim scans the addresses that are assigned to local and remote
5488 transports, and removes any duplicates. This means that more work will be
5489 done, as duplicates will always all be routed, but duplicates are
5490 presumably rare, so I don't expect this is of any significance.
5492 For deliveries to pipes, files, and autoreplies, the duplicate checking
5493 still happens during the routing process, since they are not going to be
5496 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5497 It corrects a timeout issue with spamd. This is Ian's comment: "The
5498 background is that sometimes spamd either never reads data from a
5499 connection it has accepted, or it never writes response data. The exiscan
5500 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5501 blindly assumes that writes won't block so it may never time out."
5503 PH/39 Allow G after quota size as well as K and M.
5505 PH/40 The value set for $authenticated_id in an authenticator may not contain
5506 binary zeroes or newlines because the value is written to log lines and
5507 to spool files. There was no check on this. Now the value is run through
5508 the string_printing() function so that such characters are converted to
5509 printable escape sequences.
5511 PH/41 $message_linecount is a new variable that contains the total number of
5512 lines in the message. Compare $body_linecount, which is the count for the
5515 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5516 bounce and delay warning messages, except in certain special cases, which
5519 (a) An SMTP error message from a remote host;
5520 (b) A message specified in a :fail: redirection;
5521 (c) A message specified in a "fail" command in a system filter;
5522 (d) A message specified in a FAIL return from the queryprogram router;
5523 (e) A message specified by the cannot_route_message router option.
5525 In these cases only, Exim does include the error details in bounce and
5526 warning messages. There are also a few cases where bland messages such
5527 as "unrouteable address" or "local delivery error" are given.
5529 PH/43 $value is now also set for the "else" part of a ${run expansion.
5531 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5532 being worked on, but at least Exim now implements the latest version to
5535 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5536 process to complete was treated as a delivery failure, a timeout while
5537 writing the message to the pipe was logged, but erroneously treated as a
5538 successful delivery. Such timeouts include transport filter timeouts. For
5539 consistency with the overall process timeout, these timeouts are now
5540 treated as errors, giving rise to delivery failures by default. However,
5541 there is now a new Boolean option for the pipe transport called
5542 timeout_defer, which, if set TRUE, converts the failures into defers for
5543 both kinds of timeout. A transport filter timeout is now identified in
5546 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5547 systems where "make" and "gmake" are different, calling "gmake" at top
5548 level broke things. I've arranged for the value of $(MAKE) to be passed
5549 from the Makefile to this script so that it can call the same version of
5553 A note about Exim versions 4.44 and 4.50
5554 ----------------------------------------
5556 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5557 changes of various kinds. As a consequence, a big documentation update was
5558 needed. This delayed the release for rather longer than seemed good, especially
5559 in the light of a couple of (minor) security issues. Therefore, the changes
5560 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5561 release. So 4.44 and 4.50 are in effect two different branches that both start
5564 I have left the 4.50 change log unchanged; it contains all the changes since
5565 4.43. The change log for 4.44 is below; many of its items are identical to
5566 those for 4.50. This seems to be the most sensible way to preserve the
5567 historical information.
5573 1. Minor wording change to the doc/README.SIEVE file.
5575 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5576 computation of the current number of files was incorrect.
5578 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5579 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5582 4. Give more explanation in the error message when the command for a transport
5583 filter fails to execute.
5585 5. There are several places where Exim runs a non-Exim command in a
5586 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5587 was being done only for the command run by the queryprogram router. It is
5588 now done for all such subprocesses. The other cases are: ${run, transport
5589 filters, and the commands run by the lmtp and pipe transports.
5591 6. Added CONFIGURE_GROUP build-time option.
5593 7. Some older OS have a limit of 256 on the maximum number of file
5594 descriptors. Exim was using setrlimit() to set 1000 as a large value
5595 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5596 systems. I've change it so that if it can't get 1000, it tries for 256.
5598 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5599 was an oversight, and furthermore, ever since the addition of extra
5600 controls (e.g. 4.43/32), the checks on when to allow different forms of
5601 "control" were broken. There should now be diagnostics for all cases when a
5602 control that does not make sense is encountered.
5604 9. Added the /retain_sender option to "control=submission".
5606 10. $recipients is now available in the predata ACL (oversight).
5608 11. Tidy the search cache before the fork to do a delivery from a message
5609 received from the command line. Otherwise the child will trigger a lookup
5610 failure and thereby defer the delivery if it tries to use (for example) a
5611 cached ldap connection that the parent has called unbind on.
5613 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5614 of $address_data from the recipient verification was clobbered by the
5615 sender verification.
5617 13. The value of address_data from a sender verification is now available in
5618 $sender_address_data in subsequent conditions in the ACL statement.
5620 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5622 15. Added a new option "connect=<time>" to callout options, to set a different
5625 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5626 was its contents. (It was OK if the option was not defined at all.)
5628 17. A "Completed" log line is now written for messages that are removed from
5629 the spool by the -Mrm option.
5631 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5632 information about exactly what failed.
5634 19. Added -dd to debug only the daemon process.
5636 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5637 handles timeouts, both on the server side and network timeouts. Renamed the
5638 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5640 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5641 It is now set to "smtps".
5643 22. $host_address is now set to the target address during the checking of
5644 ignore_target_hosts.
5646 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5647 being passed; this would have caused $sender_host_name to have been used if
5648 matching the list had actually called for a host name (not very likely,
5649 since this list is usually IP addresses). A host name is now passed as
5652 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5653 code that specifies a non-restarting handler (typically sigaction() in
5654 modern systems) in an attempt to fix a rare and obscure crash bug.
5656 25. Narrowed the window for a race in the daemon that could cause it to ignore
5657 SIGCHLD signals. This is not a major problem, because they are used only to
5658 wake it up if nothing else does.
5660 26. A malformed maildirsize file could cause Exim to calculate negative values
5661 for the mailbox size or file count. Odd effects could occur as a result.
5662 The maildirsize information is now recalculated if the size or filecount
5665 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5666 support for a long time. Removed HAVE_SYS_VFS_H.
5668 28. Installed the latest version of exipick from John Jetmore.
5670 29. In an address list, if the pattern was not a regular expression, an empty
5671 subject address (from a bounce message) matched only if the pattern was an
5672 empty string. Non-empty patterns were not even tested. This was the wrong
5673 because it is perfectly reasonable to use an empty address as part of a
5674 database query. An empty address is now tested by patterns that are
5675 lookups. However, all the other forms of pattern expect the subject to
5676 contain a local part and a domain, and therefore, for them, an empty
5677 address still always fails if the pattern is not itself empty.
5679 30. Exim went into a mad DNS loop when attempting to do a callout where the
5680 host was specified on an smtp transport, and looking it up yielded more
5681 than one IP address.
5683 31. Re-factored the code for checking spool and log partition space into a
5684 function that finds that data and another that does the check. The former
5685 is then used to implement four new variables: $spool_space, $log_space,
5686 $spool_inodes, and $log_inodes.
5688 32. The RFC2047 encoding function was originally intended for short strings
5689 such as real names; it was not keeping to the 75-character limit for
5690 encoded words that the RFC imposes. It now respects the limit, and
5691 generates multiple encoded words if necessary. To be on the safe side, I
5692 have increased the buffer size for the ${rfc2047: expansion operator from
5695 33. It is now permitted to omit both strings after an "if" condition; if the
5696 condition is true, the result is "true". As before, when the second string
5697 is omitted, a false condition yields an empty string. This makes it less
5698 cumbersome to write custom ACL and router conditions.
5700 34. Failure to deliver a bounce message always caused it to be frozen, even if
5701 there was an errors_to setting on the router. The errors_to setting is now
5704 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5705 canonical form (fully expanded) before being placed in
5706 $sender_host_address.
5708 36. The table in the code that translates DNS record types into text (T_A to
5709 "A" for instance) was missing entries for NS and CNAME. It is just possible
5710 that this could have caused confusion if both these types were looked up
5711 for the same domain, because the text type is used as part of Exim's
5712 per-process caching. But the chance of anyone hitting this buglet seems
5715 37. The dnsdb lookup has been extended in a number of ways.
5717 (1) There is a new type, "zns", which walks up the domain tree until it
5718 finds some nameserver records. It should be used with care.
5720 (2) There is a new type, "mxh", which is like "mx" except that it returns
5721 just the host names, not the priorities.
5723 (3) It is now possible to give a list of domains (or IP addresses) to be
5724 looked up. The behaviour when one of the lookups defers can be
5725 controlled by a keyword.
5727 (4) It is now possible to specify the separator character for use when
5728 multiple records are returned.
5730 38. The dnslists ACL condition has been extended: it is now possible to supply
5731 a list of IP addresses and/or domains to be looked up in a particular DNS
5734 39. Added log_selector=+queue_time_overall.
5736 40. When running the queue in the test harness, wait just a tad after forking a
5737 delivery process, to get repeatability of debugging output.
5739 41. Include certificate and key file names in error message when GnuTLS fails
5740 to set them up, because the GnuTLS error message doesn't include the name
5741 of the failing file when there is a problem reading it.
5743 42. Allow both -bf and -bF in the same test run.
5745 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5747 44. The "Exiscan patch" is now merged into the mainline Exim source.
5749 45. Sometimes the final signoff response after QUIT could fail to get
5750 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5751 < 0 before doing a fflush(). This bug looks as though it goes back to the
5752 introduction of TLS in release 3.20, but "sometimes" must have been rare
5753 because the tests only now provoked it.
5755 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5756 (this can affect the format of dates).
5758 47. exim_tidydb, when checking for the continued existence of a message for
5759 which it has found a message-specific retry record, was not finding
5760 messages that were in split spool directories. Consequently, it was
5761 deleting retry records that should have stayed in existence.
5763 48. Steve fixed some bugs in eximstats.
5765 49. The SPA authentication driver was not abandoning authentication and moving
5766 on to the next authenticator when an expansion was forced to fail,
5767 contradicting the general specification for all authenticators. Instead it
5768 was generating a temporary error. It now behaves as specified.
5770 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5771 (the order specifies the preference for clients). The order is now AES256,
5772 AES128, 3DES, ARCFOUR128.
5774 51. Small patch to Sieve code - explicitly set From: when generating an
5777 52. Exim crashed if a remote delivery caused a very long error message to be
5778 recorded - for instance if somebody sent an entire SpamAssassin report back
5779 as a large number of 550 error lines. This bug was coincidentally fixed by
5780 increasing the size of one of Exim's internal buffers (big_buffer) that
5781 happened as part of the Exiscan merge. However, to be on the safe side, I
5782 have made the code more robust (and fixed the comments that describe what
5785 53. Now that there can be additional text after "Completed" in log lines (if
5786 the queue_time_overall log selector is set), a one-byte patch to exigrep
5787 was needed to allow it to recognize "Completed" as not the last thing in
5790 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5791 patch that reportedly fixes this has been added. I am not expert enough to
5792 create a test for it. This is what the patch creator wrote:
5794 "I found a little strange behaviour of ldap code when working with
5795 Windows 2003 AD Domain, where users was placed in more than one
5796 Organization Units. When I tried to give exim partial DN, the exit code
5797 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5798 But simultaneously result of request was absolutely normal ldap result,
5799 so I produce this patch..."
5801 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5802 REFERENCE, so I have modified the code to exclude the patch when that macro
5805 55. Some experimental protocols are using DNS PTR records for new purposes. The
5806 keys for these records are domain names, not reversed IP addresses. The
5807 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5808 leaves it alone. Component reversal etc. now happens only for IP addresses.
5811 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5813 57. Double the size of the debug message buffer (to 2048) so that more of very
5814 long debug lines gets shown.
5816 58. The exicyclog utility now does better if the number of log files to keep
5817 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5819 59. Two changes related to the smtp_active_hostname option:
5821 (1) $smtp_active_hostname is now available as a variable.
5822 (2) The default for smtp_banner uses $smtp_active_hostname instead
5823 of $primary_hostname.
5825 60. The host_aton() function is supposed to be passed a string that is known
5826 to be a valid IP address. However, in the case of IPv6 addresses, it was
5827 not checking this. This is a hostage to fortune. Exim now panics and dies
5828 if the condition is not met. A case was found where this could be provoked
5829 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5830 components; fortuitously, this particular loophole had already been fixed
5831 by change 4.50/55 above.
5833 If there are any other similar loopholes, the new check in host_aton()
5834 itself should stop them being exploited. The report I received stated that
5835 data on the command line could provoke the exploit when Exim was running as
5836 exim, but did not say which command line option was involved. All I could
5837 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5838 running as the user.
5841 61. There was a buffer overflow vulnerability in the SPA authentication code
5842 (which came originally from the Samba project). I have added a test to the
5843 spa_base64_to_bits() function which I hope fixes it.
5846 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5847 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5849 63. The daemon start-up calls getloadavg() while still root for those OS that
5850 need the first call to be done as root, but it missed one case: when
5851 deliver_queue_load_max is set with deliver_drop_privilege. This is
5852 necessary for the benefit of the queue runner, because there is no re-exec
5853 when deliver_drop_privilege is set.
5855 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5856 This has been fixed.
5858 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5859 was in use, was not putting the data itself into the right store pool;
5860 consequently, it could be overwritten for a subsequent message in the same
5861 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5864 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5866 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5867 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5868 the function were treating the return as a boolean value, which happened to
5869 work because 0=false and not-0=true, but is not correct code.
5871 68. The host_aton() function was not handling scoped IPv6 addresses (those
5872 with, for example, "%eth0" on the end) correctly.
5874 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5875 negated items (that is, ~something) in unsigned ints. Some compilers
5876 apparently mutter when there is no cast.
5878 70. If an address verification called from an ACL failed, and did not produce a
5879 user-specific message (i.e. there was only a "system" message), nothing was
5880 put in $acl_verify_message. In this situation, it now puts the system
5883 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5884 session; change 4.31/43 added the unwanted input to the log line - except
5885 that it did not do this in the start of session case. It now does.
5887 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5888 This is wrong and can cause the other end to generate a synchronization
5889 error if it is another Exim or anything else that does the synchronization
5890 check. A QUIT command is no longer sent after a timeout.
5892 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5893 during host lookups.
5895 74. The defer_ok option of callout verification was not working if it was used
5896 when verifying addresses in header lines, that is, for this case:
5898 verify = header_sender/callout=defer_ok
5900 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5901 those file descriptors could be used for SMTP connections. If anything
5902 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5903 could be sent to the SMTP client, causing chaos. The daemon now opens
5904 stdin, stdout, and stderr to /dev/null when it puts itself into the
5907 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5908 log by default. The user can override this with suitable Perl magic.
5910 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5911 the log message when discard triggers, was not working for the DATA ACL or
5912 for the non-SMTP ACL.
5914 78. Error message wording change in sieve.c.
5916 79. If smtp_accept_max_per_host was set, the number of connections could be
5917 restricted to fewer than expected, because the daemon was trying to set up
5918 a new connection before checking whether the processes handling previous
5919 connections had finished. The check for completed processes is now done
5920 earlier. On busy systems, this bug wouldn't be noticed because something
5921 else would have woken the daemon, and it would have reaped the completed
5924 80. If a message was submitted locally by a user whose login name contained one
5925 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5926 It caused a spool format error. I have fixed the spool reading code. A
5927 related problem was that the "from" clause in the Received: line became
5928 illegal because of the space(s). It is now covered by ${quote_local_part.
5930 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5933 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5935 83. Patch from Sieve maintainer to fix unterminated string problem in
5936 "vacation" handling.
5938 84. Some minor changes to the Linux configuration files to help with other
5939 OS variants using glibc.
5941 85. One more patch for Sieve to update vacation handling to latest spec.
5944 ----------------------------------------------------
5945 See the note above about the 4.44 and 4.50 releases.
5946 ----------------------------------------------------
5952 1. Change 4.43/35 introduced a bug that caused file counts to be
5953 incorrectly computed when quota_filecount was set in an appendfile
5956 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5957 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5960 3. Give more explanation in the error message when the command for a transport
5961 filter fails to execute.
5963 4. There are several places where Exim runs a non-Exim command in a
5964 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5965 was being done only for the command run by the queryprogram router. It is
5966 now done for all such subprocesses. The other cases are: ${run, transport
5967 filters, and the commands run by the lmtp and pipe transports.
5969 5. Some older OS have a limit of 256 on the maximum number of file
5970 descriptors. Exim was using setrlimit() to set 1000 as a large value
5971 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5972 systems. I've change it so that if it can't get 1000, it tries for 256.
5974 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5975 was an oversight, and furthermore, ever since the addition of extra
5976 controls (e.g. 4.43/32), the checks on when to allow different forms of
5977 "control" were broken. There should now be diagnostics for all cases when a
5978 control that does not make sense is encountered.
5980 7. $recipients is now available in the predata ACL (oversight).
5982 8. Tidy the search cache before the fork to do a delivery from a message
5983 received from the command line. Otherwise the child will trigger a lookup
5984 failure and thereby defer the delivery if it tries to use (for example) a
5985 cached ldap connection that the parent has called unbind on.
5987 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5988 of $address_data from the recipient verification was clobbered by the
5989 sender verification.
5991 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5992 was its contents. (It was OK if the option was not defined at all.)
5994 11. A "Completed" log line is now written for messages that are removed from
5995 the spool by the -Mrm option.
5997 12. $host_address is now set to the target address during the checking of
5998 ignore_target_hosts.
6000 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6001 being passed; this would have caused $sender_host_name to have been used if
6002 matching the list had actually called for a host name (not very likely,
6003 since this list is usually IP addresses). A host name is now passed as
6006 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6007 code that specifies a non-restarting handler (typically sigaction() in
6008 modern systems) in an attempt to fix a rare and obscure crash bug.
6010 15. Narrowed the window for a race in the daemon that could cause it to ignore
6011 SIGCHLD signals. This is not a major problem, because they are used only to
6012 wake it up if nothing else does.
6014 16. A malformed maildirsize file could cause Exim to calculate negative values
6015 for the mailbox size or file count. Odd effects could occur as a result.
6016 The maildirsize information is now recalculated if the size or filecount
6019 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6020 support for a long time. Removed HAVE_SYS_VFS_H.
6022 18. Updated exipick to current release from John Jetmore.
6024 19. Allow an empty sender to be matched against a lookup in an address list.
6025 Previously the only cases considered were a regular expression, or an
6028 20. Exim went into a mad DNS lookup loop when doing a callout where the
6029 host was specified on the transport, if the DNS lookup yielded more than
6032 21. The RFC2047 encoding function was originally intended for short strings
6033 such as real names; it was not keeping to the 75-character limit for
6034 encoded words that the RFC imposes. It now respects the limit, and
6035 generates multiple encoded words if necessary. To be on the safe side, I
6036 have increased the buffer size for the ${rfc2047: expansion operator from
6039 22. Failure to deliver a bounce message always caused it to be frozen, even if
6040 there was an errors_to setting on the router. The errors_to setting is now
6043 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6044 canonical form (fully expanded) before being placed in
6045 $sender_host_address.
6047 24. Updated eximstats to version 1.33
6049 25. Include certificate and key file names in error message when GnuTLS fails
6050 to set them up, because the GnuTLS error message doesn't include the name
6051 of the failing file when there is a problem reading it.
6053 26. Expand error message when OpenSSL has problems setting up cert/key files.
6056 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6057 (this can affect the format of dates).
6059 28. exim_tidydb, when checking for the continued existence of a message for
6060 which it has found a message-specific retry record, was not finding
6061 messages that were in split spool directories. Consequently, it was
6062 deleting retry records that should have stayed in existence.
6064 29. eximstats updated to version 1.35
6065 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6066 1.35 - bugfix such that pie charts by volume are generated correctly
6068 30. The SPA authentication driver was not abandoning authentication and moving
6069 on to the next authenticator when an expansion was forced to fail,
6070 contradicting the general specification for all authenticators. Instead it
6071 was generating a temporary error. It now behaves as specified.
6073 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6074 (the order specifies the preference for clients). The order is now AES256,
6075 AES128, 3DES, ARCFOUR128.
6077 31. Small patch to Sieve code - explicitly set From: when generating an
6080 32. Exim crashed if a remote delivery caused a very long error message to be
6081 recorded - for instance if somebody sent an entire SpamAssassin report back
6082 as a large number of 550 error lines. This bug was coincidentally fixed by
6083 increasing the size of one of Exim's internal buffers (big_buffer) that
6084 happened as part of the Exiscan merge. However, to be on the safe side, I
6085 have made the code more robust (and fixed the comments that describe what
6088 33. Some experimental protocols are using DNS PTR records for new purposes. The
6089 keys for these records are domain names, not reversed IP addresses. The
6090 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6091 leaves it alone. Component reversal etc. now happens only for IP addresses.
6094 34. The host_aton() function is supposed to be passed a string that is known
6095 to be a valid IP address. However, in the case of IPv6 addresses, it was
6096 not checking this. This is a hostage to fortune. Exim now panics and dies
6097 if the condition is not met. A case was found where this could be provoked
6098 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6099 components; fortuitously, this particular loophole had already been fixed
6100 by change 4.50/55 or 4.44/33 above.
6102 If there are any other similar loopholes, the new check in host_aton()
6103 itself should stop them being exploited. The report I received stated that
6104 data on the command line could provoke the exploit when Exim was running as
6105 exim, but did not say which command line option was involved. All I could
6106 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6107 running as the user.
6110 35. There was a buffer overflow vulnerability in the SPA authentication code
6111 (which came originally from the Samba project). I have added a test to the
6112 spa_base64_to_bits() function which I hope fixes it.
6115 36. The daemon start-up calls getloadavg() while still root for those OS that
6116 need the first call to be done as root, but it missed one case: when
6117 deliver_queue_load_max is set with deliver_drop_privilege. This is
6118 necessary for the benefit of the queue runner, because there is no re-exec
6119 when deliver_drop_privilege is set.
6121 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6122 was in use, was not putting the data itself into the right store pool;
6123 consequently, it could be overwritten for a subsequent message in the same
6124 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6127 38. Sometimes the final signoff response after QUIT could fail to get
6128 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6129 < 0 before doing a fflush(). This bug looks as though it goes back to the
6130 introduction of TLS in release 3.20, but "sometimes" must have been rare
6131 because the tests only now provoked it.
6137 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6138 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6139 Now it returns an integer. A number of calls were still expecting a T/F
6140 return. Fortuitously, in all cases, the tests worked in OK situations,
6141 which is the norm. However, things would have gone wrong on any write
6142 failures on the smtp file descriptor. This function is used when sending
6143 messages over SMTP and also when doing verify callouts.
6145 2. When Exim is called to do synchronous delivery of a locally submitted
6146 message (the -odf or -odi options), it no longer closes stderr before doing
6149 3. Implemented the mua_wrapper option.
6151 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6153 5. Implemented the functions header_remove(), header_testname(),
6154 header_add_at_position(), and receive_remove_recipient(), and exported them
6157 6. If an ACL "warn" statement specified the addition of headers, Exim already
6158 inserted X-ACL-Warn: at the start if there was no header name. However, it
6159 was not making this test for the second and subsequent header lines if
6160 there were newlines in the string. This meant that an invalid header could
6161 be inserted if Exim was badly configured.
6163 7. Allow an ACL "warn" statement to add header lines at the start or after all
6164 the Received: headers, as well as at the end.
6166 8. Added the rcpt_4xx retry error code.
6168 9. Added postmaster_mailfrom=xxx to callout verification option.
6170 10. Added mailfrom=xxxx to the callout verification option, for verify=
6173 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6174 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6176 12. Inserted some casts to stop certain compilers warning when using pointer
6177 differences as field lengths or precisions in printf-type calls (mostly
6178 affecting debugging statements).
6180 13. Added optional readline() support for -be (dynamically loaded).
6182 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6183 same clock tick as a message's arrival, so that its received time was the
6184 same as the "first fail" time on the retry record, and that message
6185 remained on the queue past the ultimate address timeout, every queue runner
6186 would try a delivery (because it was past the ultimate address timeout) but
6187 after another failure, the ultimate address timeout, which should have then
6188 bounced the address, did not kick in. This was a "< instead of <=" error;
6189 in most cases the first failure would have been in the next clock tick
6190 after the received time, and all would be well.
6192 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6193 being recognized when the domain list was tested by the match_domain
6194 condition in an expansion string.
6196 16. Added the ${str2b64: operator.
6198 17. Exim was always calling setrlimit() to set a large limit for the number of
6199 processes, without checking whether the existing limit was already
6200 adequate. (It did check for the limit on file descriptors.) Furthermore,
6201 errors from getrlimit() and setrlimit() were being ignored. Now they are
6202 logged to the main and panic logs, but Exim does carry on, to try to do its
6203 job under whatever limits there are.
6205 18. Imported PCRE 5.0.
6207 19. Trivial typo in log message " temporarily refused connection" (the leading
6210 20. If the log selector return_path_on_delivery was set and an address was
6211 redirected to /dev/null, the delivery process crashed because it assumed
6212 that a return path would always be set for a "successful" delivery. In this
6213 case, the whole delivery is bypassed as an optimization, and therefore no
6216 21. Internal re-arrangement: the function for sending a challenge and reading
6217 a response while authentication was assuming a zero-terminated challenge
6218 string. It's now changed to take a pointer and a length, to allow for
6219 binary data in such strings.
6221 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6223 23. Exim was not respecting finduser_retries when seeking the login of the
6224 uid under which it was called; it was always trying 10 times. (The default
6225 setting of finduser_retries is zero.) Also, it was sleeping after the final
6226 failure, which is pointless.
6228 24. Implemented tls_on_connect_ports.
6230 25. Implemented acl_smtp_predata.
6232 26. If the domain in control=submission is set empty, Exim assumes that the
6233 authenticated id is a complete email address when it generates From: or
6234 Sender: header lines.
6236 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6237 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6238 chown and chgrp in /bin and hostname in /usr/bin.
6240 28. Exim was keeping the "process log" file open after each use, just as it
6241 does for the main log. This opens the possibility of it remaining open for
6242 long periods when the USR1 signal hits a daemon. Occasional processlog
6243 errors were reported, that could have been caused by this. Anyway, it seems
6244 much more sensible not to leave this file open at all, so that is what now
6247 29. The long-running daemon process does not normally write to the log once it
6248 has entered its main loop, and it closes the log before doing so. This is
6249 so that log files can straightforwardly be renamed and moved. However,
6250 there are a couple of unusual error situations where the daemon does write
6251 log entries, and I had neglected to close the log afterwards.
6253 30. The text of an SMTP error response that was received during a remote
6254 delivery was being truncated at 512 bytes. This is too short for some of
6255 the long messages that one sometimes sees. I've increased the limit to
6258 31. It is now possible to make retry rules that apply only when a message has a
6259 specific sender, in particular, an empty sender.
6261 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6262 it possible to be selective about when SMTP synchronization is enforced.
6264 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6266 32. Implemented hosts_connection_nolog.
6268 33. Added an ACL for QUIT.
6270 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6273 35. Added mailbox_size and mailbox_filecount to appendfile.
6275 36. Added control = no_multiline_responses to ACLs.
6277 37. There was a bug in the logic of the code that waits for the clock to tick
6278 in the case where the clock went backwards by a substantial amount such
6279 that the microsecond fraction of "now" was more than the microsecond
6280 fraction of "then" (but the whole seconds number was less).
6282 38. Added support for the libradius Radius client library this is found on
6283 FreeBSD (previously only the radiusclient library was supported).
6289 1. When certain lookups returned multiple values in the form name=value, the
6290 quoting of the values was not always being done properly. Specifically:
6291 (a) If the value started with a double quote, but contained no whitespace,
6293 (b) If the value contained whitespace other than a space character (i.e.
6294 tabs or newlines or carriage returns) it was not quoted.
6295 This fix has been applied to the mysql and pgsql lookups by writing a
6296 separate quoting function and calling it from the lookup code. The fix
6297 should probably also be applied to nisplus, ibase and oracle lookups, but
6298 since I cannot test any of those, I have not disturbed their existing code.
6300 2. A hit in the callout cache for a specific address caused a log line with no
6301 reason for rejecting RCPT. Now it says "Previous (cached) callout
6302 verification failure".
6304 3. There was an off-by-one bug in the queryprogram router. An over-long
6305 return line was truncated at 256 instead of 255 characters, thereby
6306 overflowing its buffer with the terminating zero. As well as fixing this, I
6307 have increased the buffer size to 1024 (and made a note to document this).
6309 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6310 when Exim is waiting for an SMTP response from a remote server, Exim
6311 restarts its select() call on the socket, thereby resetting its timeout.
6312 This is not a problem when such interrupts are rare. Somebody set up a cron
6313 job to run exiwhat every 2 minutes, which is less than the normal select()
6314 timeout (5 or 10 minutes). This meant that the select() timeout never
6315 kicked in because it was always reset. I have fixed this by comparing the
6316 time when an interrupt arrives with the time at the start of the first call
6317 to select(). If more time than the timeout has elapsed, the interrupt is
6318 treated as a timeout.
6320 5. Some internal re-factoring in preparation for the addition of Sieve
6321 extensions (by MH). In particular, the "personal" test is moved to a
6322 separate function, and given an option for scanning Cc: and Bcc: (which is
6323 not set for Exim filters).
6325 6. When Exim created an email address using the login of the caller as the
6326 local part (e.g. when creating a From: or Sender: header line), it was not
6327 quoting the local part when it contained special characters such as @.
6329 7. Installed new OpenBSD configuration files.
6331 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6332 try to make them clearer.
6334 9. Callout options, other than the timeout value, were being ignored when
6335 verifying sender addresses in header lines. For example, when using
6337 verify = header_sender/callout=no_cache
6339 the cache was (incorrectly) being used.
6341 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6342 only the Cygwin environment.
6344 11. When return_path_on_delivery was set as a log selector, if different remote
6345 addresses in the same message used different return paths and parallel
6346 remote delivery occurred, the wrong values would sometimes be logged.
6347 (Whenever a remote delivery process finished, the return path value from
6348 the most recently started remote delivery process was logged.)
6350 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6351 lines when AUTH and/or TLS are in use. This is the "received protocol"
6352 field. Exim used to use "asmtp" for authenticated SMTP, without any
6353 indication (in the protocol name) for TLS use. Now it follows the RFC and
6354 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6355 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6356 names appear in log lines as well as in Received: header lines.
6358 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6359 extensions, and comparison tests, and to fix some bugs.
6361 14. Changes to the "personal" filter test:
6363 (1) The test was buggy in that it was just doing the equivalent of
6364 "contains" tests on header lines. For example, if a user's address was
6365 anne@some.where, the "personal" test would incorrectly be true for
6367 To: susanne@some.where
6369 This test is now done by extracting each address from the header in turn,
6370 and checking the entire address. Other tests that are part of "personal"
6371 are now done using regular expressions (for example, to check local parts
6372 of addresses in From: header lines).
6374 (2) The list of non-personal local parts in From: addresses has been
6375 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6376 taken from the Sieve specification recommendations.
6378 (3) If the message contains any header line starting with "List-" it is
6379 treated as non-personal.
6381 (4) The test for "circular" in the Subject: header line has been removed
6382 because it now seems ill-conceived.
6384 15. Minor typos in src/EDITME comments corrected.
6386 16. Installed latest exipick from John Jetmore.
6388 17. If headers_add on a router specified a text string that was too long for
6389 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6390 of string_sprintf() is now avoided.
6392 18. $message_body_size was not set (it was always zero) when running the DATA
6393 ACL and the local_scan() function.
6395 19. For the "mail" command in an Exim filter, no default was being set for
6396 the once_repeat time, causing a random time value to be used if "once" was
6397 specified. (If the value happened to be <= 0, no repeat happened.) The
6398 default is now 0s, meaning "never repeat". The "vacation" command was OK
6399 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6400 (I found it when inspecting the code).
6402 20. There is now an overall timeout for performing a callout verification. It
6403 defaults to 4 times the callout timeout, which applies to individual SMTP
6404 commands during the callout. The overall timeout applies when there is more
6405 than one host that can be tried. The timeout is checked before trying the
6406 next host. This prevents very long delays if there are a large number of
6407 hosts and all are timing out (e.g. when the network connections are timing
6408 out). The value of the overall timeout can be changed by specifying an
6409 additional sub-option for "callout", called "maxwait". For example:
6411 verify = sender/callout=5s,maxwait=20s
6413 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6414 to the end before writing, but this should make it even safer).
6416 22. Exim was forgetting that it had advertised PIPELINING for the second and
6417 subsequent messages on an SMTP connection. It was also not resetting its
6418 memory on STARTTLS and an internal HELO.
6420 23. When Exim logs an SMTP synchronization error within a session, it now
6421 records whether PIPELINING has been advertised or not.
6423 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6424 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6425 rather than long int.
6427 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6429 26. Added the never_mail option to autoreply.
6435 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6436 crash if the getsockname() call failed; this can happen if a connection is
6437 closed very soon after it is established. The problem was simply in the
6438 order in which certain operations were done, causing Exim to try to write
6439 to the SMTP stream before it had set up the file descriptor. The bug has
6440 been fixed by making things happen in the correct order.
6446 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6447 before the connection was closed, thus losing the rejection response.
6449 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6450 some early Solaris releases, but causes trouble in current releases where
6451 socklen_t is defined.
6453 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6456 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6459 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6460 that releasing the top of it at the end releases what was used for sub-
6461 expansions (unless the block got too big). However, discard this block if
6462 the first thing is a variable or header, so that we can use its block when
6463 it is dynamic (useful for very large $message_headers, for example).
6465 6. Lookups now cache *every* query, not just the most recent. A new, separate
6466 store pool is used for this. It can be recovered when all lookup caches are
6467 flushed. Lookups now release memory at the end of their result strings.
6468 This has involved some general refactoring of the lookup sources.
6470 7. Some code has been added to the store_xxx() functions to reduce the amount
6471 of flapping under certain conditions.
6473 8. log_incoming_interface used to affect only the <= reception log lines. Now
6474 it causes the local interface and port to be added to several more SMTP log
6475 lines, for example "SMTP connection from", and rejection lines.
6477 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6479 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6481 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6482 could be overwritten at the end of the current message (or the start of a
6483 new message if it was set in a HELO ACL). The value is now preserved for
6484 the duration of the SMTP connection.
6486 12. If a transport had a headers_rewrite setting, and a matching header line
6487 contained an unqualified address, that address was qualified, even if it
6488 did not match any rewriting rules. The underlying bug was that the values
6489 of the flags that permit the existence of unqualified sender and recipient
6490 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6491 non-local messages, and by -bnq for local messages) were not being
6492 preserved with the message after it was received.
6494 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6495 "next input=" as part of the text comprising the host identity instead of
6496 the correct text. The code was using the same buffer for two different
6497 strings. However, depending on which order the printing function evaluated
6498 its arguments, the bug did not always show up. Under Linux, for example, my
6499 test suite worked just fine.
6501 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6502 doesn't work with some older versions of Perl. It has been changed to "my",
6503 which in any case is probably the better facility to use.
6505 15. A really picky compiler found some instances of statements for creating
6506 error messages that either had too many or two few arguments for the format
6509 16. The size of the buffer for calls to the DNS resolver has been increased
6510 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6511 for addresses that have a lot of PTR records. This alleviates a problem; it
6512 does not fully solve it.
6514 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6515 buffer now truncates the list and logs the incident, which is the same
6516 action as happens when Exim is looking up a host name and its aliases.
6517 Previously in this situation something unpredictable would happen;
6518 sometimes it was "internal error: store_reset failed".
6520 18. If a server dropped the connection unexpectedly when an Exim client was
6521 using GnuTLS and trying to read a response, the client delivery process
6522 crashed while trying to generate an error log message.
6524 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6525 string, for example:
6527 warn message = H1: something\nH2: something
6529 the text was added as a single header line from Exim's point of view
6530 though it ended up OK in the delivered message. However, searching for the
6531 second and subsequent header lines using $h_h2: did not work. This has been
6532 fixed. Similarly, if a system filter added multiple headers in this way,
6533 the routers could not see them.
6535 20. Expanded the error message when iplsearch is called with an invalid key to
6536 suggest using net-iplsearch in a host list.
6538 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6539 ACLs are no longer actually imposed (and a message to that effect is
6542 22. If a "gecos" field in a passwd entry contained escaped characters, in
6543 particular, if it contained a \" sequence, Exim got it wrong when building
6544 a From: or a Sender: header from that name. A second bug also caused
6545 incorrect handling when an unquoted " was present following a character
6546 that needed quoting.
6548 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6549 was not being matched caselessly.
6551 24. Arranged for all hyphens in the exim.8 source to be escaped with
6554 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6555 properly. Recipient callout cache records were still being keyed to include
6556 the sender, even when use_sender was set false. This led to far more
6557 callouts that were necessary. The sender is no longer included in the key
6558 when use_sender is false.
6560 26. Added "control = submission" modifier to ACLs.
6562 27. Added the ${base62d: operator to decode base 62 numbers.
6564 28. dnsdb lookups can now access SRV records.
6566 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6567 the configuration file.
6569 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6570 (-v) mode. This makes the output for a verbose queue run more intelligible.
6572 31. Added a use_postmaster feature to recipient callouts.
6574 32. Added the $body_zerocount variable, containing the number of binary zero
6575 bytes in the message body.
6577 33. The time of last modification of the "new" subdirectory is now used as the
6578 "mailbox time last read" when there is a quota error for a maildir
6581 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6583 35. Added +ignore_unknown as a special item in host lists.
6585 36. Code for decoding IPv6 addresses in host lists is now included, even if
6586 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6587 address was recognized as an IP address, but was then not correctly decoded
6588 into binary, causing unexpected and incorrect effects when compared with
6595 1. Very minor rewording of debugging text in manualroute to say "list of
6596 hosts" instead of "hostlist".
6598 2. If verify=header_syntax was set, and a header line with an unqualified
6599 address (no domain) and a large number of spaces between the end of the
6600 name and the colon was received, the reception process suffered a buffer
6601 overflow, and (when I tested it) crashed. This was caused by some obsolete
6602 code that should have been removed. The fix is to remove it!
6604 3. When running in the test harness, delay a bit after writing a bounce
6605 message to get a bit more predictability in the log output.
6607 4. Added a call to search_tidyup() just before forking a reception process. In
6608 theory, someone could use a lookup in the expansion of smtp_accept_max_
6609 per_host which, without the tidyup, could leave open a database connection.
6611 5. Added the variables $recipient_data and $sender_data which get set from a
6612 lookup success in an ACL "recipients" or "senders" condition, or a router
6613 "senders" option, similar to $domain_data and $local_part_data.
6615 6. Moved the writing of debug_print from before to after the "senders" test
6618 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6619 problems for message scanning, either using a data ACL, or using
6620 local_scan() because the Received: header was not generated till after they
6621 were called (in order to set the time as the time of reception completion).
6622 I have revised the way this works. The header is now generated after the
6623 body is received, but before the ACL or local_scan() are called. After they
6624 are run, the timestamp in the header is updated.
6630 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6631 before starting a queue runner without re-exec. This happened only when
6632 deliver_drop_privilege was set or when the Exim user was set to root. The
6633 effect of the bug was that timeouts during subsequent deliveries caused
6634 crashes instead of being properly handled. The handler is now left at its
6635 default (and expected) setting.
6637 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6638 message, again when deliver_drop_privilege is set or Exim is run as root.
6639 The bug described in (1) was not present in this case, but the tidying up
6640 of the other signals was missing. I have made the two cases consistent.
6642 3. The ignore_target_hosts setting on a manualroute router was being ignored
6643 for hosts that were looked up using the /MX notation.
6645 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6648 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6649 operated on the sender address. After changing the $sender_address to <>
6650 for the sender address verify, Exim was re-instated it as the original
6651 (before rewriting) address, but remembering that it had rewritten it, so it
6652 wasn't rewriting it again. This bug also had the effect of breaking the
6653 sender address verification caching when the sender address was rewritten.
6655 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6656 This has been changed so that if the ip literal address matches
6657 ignore_target_hosts, the router declines.
6659 7. Added expansion conditions match_domain, match_address, and match_local_
6660 part (NOT match_host).
6662 8. The placeholder for the Received: header didn't have a length field set.
6664 9. Added code to Exim itself and to exim_lock to test for a specific race
6665 condition that could lead to file corruption when using MBX delivery. The
6666 issue is with the lockfile that is created in /tmp. If this file is removed
6667 after a process has opened it but before that process has acquired a lock,
6668 there is the potential for a second process to recreate the file and also
6669 acquire a lock. This could lead to two Exim processes writing to the file
6670 at the same time. The added code performs the same test as UW imapd; it
6671 checks after acquiring the lock that its file descriptor still refers to
6672 the same named file.
6674 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6675 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6678 11. Added the smtp_active_hostname option. If used, this will typically be made
6679 to depend on the incoming interface address. Because $interface_address is
6680 not set up until the daemon has forked a reception process, error responses
6681 that can happen earlier (such as "too many connections") no longer contain
6684 12. If an expansion in a condition on a "warn" statement fails because a lookup
6685 defers, the "warn" statement is abandoned, and the next ACL statement is
6686 processed. Previously this caused the whole ACL to be aborted.
6688 13. Added the iplsearch lookup type.
6690 14. Added ident_timeout as a log selector.
6692 15. Added tls_certificate_verified as a log selector.
6694 16. Added a global option tls_require_ciphers (compare the smtp transport
6695 option of the same name). This controls incoming TLS connections.
6697 17. I finally figured out how to make tls_require_ciphers do a similar thing
6698 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6699 before starting the TLS session.
6701 18. Tabs are now shown as \t in -bP output.
6703 19. If the log selector return_path_on_delivery was set, Exim crashed when
6704 bouncing a message because it had too many Received: header lines.
6706 20. If two routers both had headers_remove settings, and the first one included
6707 a superfluous trailing colon, the final name in the first list and the
6708 first name in the second list were incorrectly joined into one item (with a
6709 colon in the middle).
6715 1. Added -C and -D options to the exinext utility, mainly to make it easier
6716 to include in the automated testing, but these could be helpful when
6717 multiple configurations are in use.
6719 2. The exinext utility was not formatting the output nicely when there was
6720 an alternate port involved in the retry record key, nor when there was a
6721 message id as well (for retries that were specific to a specific message
6722 and a specific host). It was also confused by IPv6 addresses, because of
6723 the additional colons they contain. I have fixed the IPv4 problem, and
6724 patched it up to do a reasonable job for IPv6.
6726 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6727 delivery, the log line now contains "pipelined" if PIPELINING was used.
6729 4. An SMTP transport process used to panic and die if the bind() call to set
6730 an explicit outgoing interface failed. This has been changed; it is now
6731 treated in the same way as a connect() failure.
6733 5. A reference to $sender_host_name in the part of a conditional expansion
6734 that was being skipped was still causing a DNS lookup. This no longer
6737 6. The def: expansion condition was not recognizing references to header lines
6738 that used bh_ and bheader_.
6740 7. Added the _cache feature to named lists.
6742 8. The code for checking quota_filecount in the appendfile transport was
6743 allowing one more file than it should have been.
6745 9. For compatibility with Sendmail, the command line option
6753 and sets the incoming protocol and host name (for trusted callers). The
6754 host name and its colon can be omitted when only the protocol is to be set.
6755 Note the Exim already has two private options, -pd and -ps, that refer to
6756 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6757 "s", but I don't think that's a major issue.
6759 10. A number of refactoring changes to the code, none of which should affect
6762 (a) The number of logging options was getting close to filling up the
6763 32-bit word that was used as a bit map. I have split them into two classes:
6764 those that are passed in the argument to log_write(), and those that are
6765 only ever tested independently outside of that function. These are now in
6766 separate 32-bit words, so there is plenty of room for expansion again.
6767 There is no change in the user interface or the logging behaviour.
6769 (b) When building, for example, log lines, the code previously used a
6770 macro that called string_cat() twice, in order to add two strings. This is
6771 not really sufficiently general. Furthermore, there was one instance where
6772 it was actually wrong because one of the argument was used twice, and in
6773 one call a function was used. (As it happened, calling the function twice
6774 did not affect the overall behaviour.) The macro has been replaced by a
6775 function that can join an arbitrary number of extra strings onto a growing
6778 (c) The code for expansion conditions now uses a table and a binary chop
6779 instead of a serial search (which was left over from when there were very
6780 few conditions). Also, it now recognizes conditions like "pam" even when
6781 the relevant support is not compiled in: a suitably worded error message is
6782 given if an attempt is made to use such a condition.
6784 11. Added ${time_interval:xxxxx}.
6786 12. A bug was causing one of the ddress fields not to be passed back correctly
6787 from remote delivery subprocesses. The field in question was not being
6788 subsequently used, so this caused to problems in practice.
6790 13. Added new log selectors queue_time and deliver_time.
6792 14. Might have fixed a bug in maildirsizefile handling that threw up
6793 "unexpected character" debug warnings, and recalculated the data
6794 unnecessarily. In any case, I expanded the warning message to give more
6797 15. Added the message "Restricted characters in address" to the statements in
6798 the default ACL that block characters like @ and % in local parts.
6800 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6801 Three changes have been made:
6803 (a) There was a serious bug; a negative response to MAIL caused the whole
6804 recipient domain to be cached as invalid, thereby blocking all messages
6805 to all local parts at the same domain, from all senders. This bug has
6806 been fixed. The domain is no longer cached after a negative response to
6807 MAIL if the sender used is not empty.
6809 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6812 (c) A new callout option, "use_sender" has been added for people who want
6813 the modified behaviour.
6819 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6822 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6823 indeed breaks things for older releases.
6825 3. Added additional logging to the case where there is a problem reading data
6826 from a filter that is running in a subprocess using a pipe, in order to
6827 try to track down a specific problem.
6829 4. Testing facility fudge: when running in the test harness and attempting
6830 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6831 getting "No route to host". Convert this to a timeout.
6833 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6836 6. Some OS don't have socklen_t but use size_t instead. This affects the
6837 fifth argument of getsockopt() amongst other things. This is now
6838 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6839 can be set for individual OS. I have set it for SunOS5, OSF1, and
6840 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6841 some earlier ones do not.
6843 7. Change 4.30/15 was not doing the test caselessly.
6845 8. The standard form for an IPv6 address literal was being rejected by address
6846 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6847 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6848 this, as well as the form without the "IPv6" on the front (but only when
6849 address literals are enabled, of course).
6851 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6853 10. Exim crashed if a message with an empty sender address specified by -f
6854 encountered a router with an errors_to setting. This could be provoked only
6855 by a command such as
6859 where an empty string was supplied; "<>" did not hit this bug.
6861 11. Installed PCRE release 4.5.
6863 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6864 remained set. It is now erased.
6866 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6867 times from message ids (which are base 36 rather than the normal 62).
6869 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6870 were being counted as actual protocol errors, and logged if the log
6871 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6872 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6873 and DATA following a set of rejected RCPTs do not count as protocol errors.
6874 In other words, Exim assumes they were pipelined, though this may not
6875 actually be the case. Of course, in all cases the client gets an
6876 appropriate error code.
6878 15. If a lookup fails in an ACL condition, a message about the failure may
6879 be available; it is used if testing the ACL cannot continue, because most
6880 such messages specify what the cause of the deferral is. However, some
6881 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6882 that caused an old message to be retained and used if a later statement
6883 caused a defer, replacing the real cause of the deferral.
6885 16. If an IP address had so many PTR records that the DNS lookup buffer
6886 was not large enough to hold them, Exim could crash while trying to process
6887 the truncated data. It now detects and logs this case.
6889 17. Further to 4.21/58, another change has been made: if (and only if) the
6890 first line of a message (the first header line) ends with CRLF, a bare LF
6891 in a subsequent header line has a space inserted after it, so as not to
6892 terminate the header.
6894 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6895 unnecessarily, used atoi() instead of strtol(), and didn't check the
6896 termination when getting file sizes from file names by regex.
6898 19. Completely re-implemented the support for maildirsize files, in the light
6899 of a number of problems with the previous contributed implementation
6900 (4.30/29). In particular:
6902 . If the quota is zero, the maildirsize file is maintained, but no quota is
6905 . If the maildir directory does not exist, it is created before any attempt
6906 to write a maildirsize file.
6908 . The quota value in the file is just a cache; if the quota is changed in
6909 the transport, the new value overrides.
6911 . A regular expression is available for excluding directories from the
6914 20. The autoreply transport checks the characters in options that define the
6915 message's headers; it allows continued headers, but it was checking with
6916 isspace() after an embedded newline instead of explicitly looking for a
6919 21. If all the "regular" hosts to which an address was routed had passed their
6920 expiry times, and had not reached their retry times, the address was
6921 bounced, even if fallback hosts were defined. Now Exim should go on to try
6924 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6925 equivalent code in the SMTP transport. Some hosts send humungous responses
6926 to HELO/EHLO, more than 1024 it seems.
6928 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6929 gives compiler warnings in some environments. I've now done it "properly",
6932 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6933 (because of problems with the built-in one) was declared to return uschar *
6934 instead of char *, causing compiler failure.
6936 25. Fixed a file descriptor leak when processing alias/forward files.
6938 26. Fixed a minor format string issue in dbfn.c.
6940 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6942 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6943 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6946 29. When a sender address is verified, it is cached, to save repeating the test
6947 when there is more than one recipient in a message. However, when the
6948 verification involves a callout, it is possible for different callout
6949 options to be set for different recipients. It is too complicated to keep
6950 track of this in the cache, so now Exim always runs a verification when a
6951 callout is required, relying on the callout cache for the optimization.
6952 The overhead is duplication of the address routing, but this should not be
6955 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6956 to be verified with callout=postmaster, and the main callout worked but the
6957 postmaster check failed, the verification correctly failed. However, if a
6958 subsequent RCPT command asked for sender verification *without* the
6959 postmaster check, incorrect caching caused this verification also to fail,
6962 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6963 it was not caching the DNS options (qualify_single, search_parents) that
6964 were used when the lookup failed. A subsequent lookup with different
6965 options therefore always gave the same answer, though there were cases
6966 where it should not have. (Example: a "domains = !$mx_any" option on a
6967 dnslookup router: the "domains" option is always processed without any
6968 widening, but the router might have qualify_single set.) Now Exim uses the
6969 cached value only when the same options are set.
6971 32. Added John Jetmore's "exipick" utility to the distribution.
6973 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6974 than a timeout (e.g. a certificate is required, and is not provided), an
6975 Exim server now closes the connection immediately. Previously it waited for
6976 the client to close - but if the client is SSL, it seems that they each
6977 wait for each other, leading to a delay before one of them times out.
6979 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6980 maintained 0.8.x compatibility because I don't think many are using it, and
6981 it is clearly obsolete.
6983 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6986 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6987 client certificate was expired. A simple patch fixes this, though I don't
6988 understand the full logic of why the verify callback is called multiple
6991 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6992 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6993 which causes problems with some clients (such as the Certicom SSL Plus
6994 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6995 disables the coutermeasure allowing Eudora to connect."
6997 38. Exim was not checking that a write() to a log file succeeded. This could
6998 lead to Bad Things if a log got too big, in particular if it hit a file
6999 size limit. Exim now panics and dies if it cannot write to a log file, just
7000 as it does if it cannot open a log file.
7002 39. Modified OS/Makefile-Linux so that it now contains
7004 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7006 The two -D definitions ensure that Exim is compiled with large file
7007 support, which makes it possible to handle log files that are bigger than
7010 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7011 instance) a domain was checked against a named list that involved a lookup,
7012 causing $domain_data to be set, then another domain was checked against the
7013 same list, then the first domain was re-checked, the value of $domain_data
7014 after the final check could be wrong. In particular, if the second check
7015 failed, it could be set empty. This bug probably also applied to
7018 41. The strip_trailing_dot option was not being applied to the address given
7019 with the -f command-line option.
7021 42. The code for reading a message's header from the spool was incrementing
7022 $received_count, but never initializing it. This meant that the value was
7023 incorrect (doubled) while delivering a message in the same process in which
7024 it was received. In the most common configuration of Exim, this never
7025 happens - a fresh exec is done - but it can happen when
7026 deliver_drop_privilege is set.
7028 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7029 it now includes up to 150 characters of the unexpected data in the log
7032 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7033 and building data strings. The size of both of these buffers was 10 000
7034 bytes - far larger than anybody would *ever* want, thought I. Needless to
7035 say, somebody hit the limit. I have increased the maximum line length to
7036 20 000 and the maximum data length of concatenated lines to 100 000. I have
7037 also fixed two bugs, because there was no checking on these buffers. Tsk,
7038 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7039 buffer is too small.
7041 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7042 lsearch lookups. Now it does.
7044 46. When parsing a route_list item in a manualroute router, a fixed-length
7045 buffer was used for the list of hosts. I made this 1024 bytes long,
7046 thinking that nobody would ever have a list of hosts that long. Wrong.
7047 Somebody had a whole pile of complicated expansion conditions, and the
7048 string was silently truncated, leading to an expansion error. It turns out
7049 that it is easier to change to an unlimited length (owing to other changes
7050 that have happened since this code was originally written) than to build
7051 structure for giving a limitation error. The length of the item that
7052 expands into the list of hosts is now unlimited.
7054 47. The lsearch lookup could not handle data where the length of text line was
7055 more than 4095 characters. Such lines were truncated, leading to shortened
7056 data being returned. It should now handle lines of any length.
7058 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7059 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7062 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7063 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7064 no longer happens in comment lines. A list of replacements is now placed
7065 at the head of all of the source files, except those whose only change is
7066 to replace PERL_COMMAND in the very first #! line.
7068 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7069 messages on the queue, with a bottom-up merge sort, using code contributed
7070 by Michael Haardt. This should make operations like -bp somewhat faster on
7071 large queues. It won't affect queue runners, except when queue_run_in_order
7074 51. Installed eximstats 1.31 in the distribution.
7076 52. Added support for SRV lookups to the dnslookup router.
7078 53. If an ACL referred to $message_body or $message_body_end, the value was not
7079 reset for any messages that followed in the same SMTP session.
7081 54. The store-handling optimization for building very long strings was not
7082 differentiating between the different store pools. I don't think this
7083 actually made any difference in practice, but I've tidied it.
7085 55. While running the routers to verify a sender address, $sender_address
7086 was still set to the sender address. This is wrong, because when routing to
7087 send a bounce to the sender, it would be empty. Therefore, I have changed
7088 it so that, while verifying a sender address, $sender_address is set to <>.
7089 (There is no change to what happens when verifying a recipient address.)
7091 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7092 target A or AAAA records (if not already returned) without resetting the
7093 qualify_single or search_parents options of the DNS resolver. These are
7094 inappropriate in this case because the targets of MX and SRV records must
7095 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7096 target that, when qualified, matched something in the local domain. These
7097 two options are now turned off when doing these lookups.
7099 57. It seems that at least some releases of Reiserfs (which does not have the
7100 concept of a fixed number of inodes) returns zero and not -1 for the
7101 number of available inodes. This interacted badly with check_spool_inodes,
7102 which assumed that -1 was the "no such thing" setting. What I have done is
7103 to check that the total number of inodes is greater than zero before doing
7104 the test of how many are available.
7106 58. When a "warn" ACL statement has a log_message modifier, the message is
7107 remembered, and not repeated. This is to avoid a lot of repetition when a
7108 message has many recipients that cause the same warning to be written.
7109 However, Exim was preserving the list of already written lines for an
7110 entire SMTP session, which doesn't seem right. The memory is now reset if a
7111 new message is started.
7113 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7114 debugging output unless log_rewrite was also set.
7116 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7117 of (int)(handle) when we know that handle contains (void *)(-1).
7119 61. The Exim daemon panic-logs an error return when it closes the incoming
7120 connection. However "connection reset by peer" seems to be common, and
7121 isn't really an error worthy of noting specially, so that particular error
7124 62. When Exim is trying to find all the local interfaces, it used to panic and
7125 die if the ioctl to get the interface flags failed. However, it seems that
7126 on at least one OS (Solaris 9) it is possible to have an interface that is
7127 included in the list of interfaces, but for which you get a failure error
7128 for this call. This happens when the interface is not "plumbed" into a
7129 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7130 failure of the "get flags" call assumes that the interface is down.
7132 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7133 makes life easier for people who are doing arithmetic on fields extracted
7134 from dates, where you often get leading zeros that should not be
7135 interpreted as octal.
7137 64. Added qualify_domain to the redirect router, to override the global
7140 65. If a pathologically long header line contained very many addresses (the
7141 report of this problem mentioned 10 000) and each of them was rewritten,
7142 Exim could use up a very large amount of memory. (It kept on making new
7143 copies of the header line as it rewrote, and never released the old ones.)
7144 At the expense of a bit more processing, the header rewriting function has
7145 been changed so that it no longer eats memory in this way.
7147 66. The generation of the Received: header has been moved from the time that a
7148 message starts to be received, to the time that it finishes. The timestamp
7149 in the Received: header should now be very close to that of the <= log
7150 line. There are two side-effects of this change:
7152 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7153 logged header lines no longer include the local Received: line, because
7154 it has not yet been created. The same applies to a copy of the message
7155 that is returned to a non-SMTP sender when a message is rejected.
7157 (b) When a filter file is tested using -bf, no additional Received: header
7158 is added to the test message. After some thought, I decided that this
7161 This change does not affect the value of $received_for. It is still set
7162 after address rewriting, but before local_scan() is called.
7164 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7166 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7167 gave an unhelpful panic error message, and a defer error. I have managed to
7168 change this behaviour so that it now rejects any supplied certificate,
7169 which seems right, as the list of acceptable certificates is empty.
7171 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7172 gave an unhelpful defer error. I have not managed to make this reject any
7173 supplied certificates, but the error message it gives is "no certificate
7174 supplied", which is not helpful.
7176 70. exigrep's output now also includes lines that are not associated with any
7177 message, but which match the given pattern. Implemented by a patch from
7178 Martin Sluka, which also tidied up the Perl a bit.
7180 71. Recipient callout verification, like sender verification, was using <> in
7181 the MAIL FROM command. This isn't really the right thing, since the actual
7182 sender may affect whether the remote host accepts the recipient or not. I
7183 have changed it to use the actual sender in the callout; this means that
7184 the cache record is now keyed on a recipient/sender pair, not just the
7185 recipient address. There doesn't seem to be a real danger of callout loops,
7186 since a callout by the remote host to check the sender would use <>.
7187 [SEE ABOVE: changed after hitting problems.]
7189 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7190 temporary errors. However, in the case of such a code being given after
7191 the end of a data transmission (i.e. after ".") Exim was failing to write
7192 a retry record for the message. (Yes, there was some broken host that was
7193 actually sending 8xx at this point.)
7195 73. An unknown lookup type in a host list could cause Exim to panic-die when
7196 the list was checked. (An example that provoked this was putting <; in the
7197 middle of a list instead of at the start.) If this happened during a DATA
7198 ACL check, a -D file could be left lying around. This kind of configuration
7199 error no longer causes Exim to die; instead it causes a defer error. The
7200 incident is still logged to the main and panic logs.
7202 74. Buglet left over from Exim 3 conversion. The message "too many messages
7203 in one connection" was written to the rejectlog but not the mainlog, except
7204 when address rewriting (yes!) was being logged.
7206 75. Added write_rejectlog option.
7208 76. When a system filter was run not as root (that is, when system_filter_user
7209 was set), the values of the $n variables were not being returned to the
7210 main process; thus, they were not subsequently available in the $sn
7213 77. Added +return_path_on_delivery log selector.
7215 78. A connection timeout was being treated differently from recipients deferred
7216 when testing hosts_max_try with a message that was older than the host's
7217 retry timeout. (The host should not be counted, thus allowing all hosts to
7218 be tried at least once before bouncing.) This may have been the cause of an
7219 occasionally reported bug whereby a message would remain on the queue
7220 longer than the retry timeout, but would be bounced if a delivery was
7221 forced. I say "may" because I never totally pinned down the problem;
7222 setting up timeout/retry tests is difficult. See also the next item.
7224 79. The ultimate address timeout was not being applied to errors that involved
7225 a combination of host plus message (for example, a timeout on a MAIL
7226 command). When an address resolved to a number of possible hosts, and they
7227 were not all tried for each delivery (e.g. because of hosts_max_try), a
7228 message could remain on the queue longer than the retry timeout.
7230 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7233 81. Fixed an obscure SMTP outgoing bug which required at least the following
7234 conditions: (a) there was another message waiting for the same server;
7235 (b) the server returned 5xx to all RCPT commands in the first message so
7236 that the message was not completed; (c) the server dropped the connection
7237 or gave a negative response to the RSET that Exim sends to abort the
7238 transaction. The observed case was a dropped connection after DATA that had
7239 been sent in pipelining mode. That is, the server had advertised PIPELINING
7240 but was not implementing it correctly. The effect of the bug was incorrect
7241 behaviour, such as trying another host, and this could lead to a crash.
7247 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7248 and daemon.c were passed as pointers to ints; they should have been
7249 pointers to socklen_t variables (which are typically unsigned ints).
7251 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7254 3. Fixed a really odd bug that affected only the testing scheme; patching a
7255 certain fixed string in the binary changed the value of another string that
7256 happened to be identical to the end of the original first string.
7258 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7259 name", it returns that address as the IP address. On some operating
7260 systems (e.g. Solaris), it also passes back the IP address string as the
7261 "host name". However, on others (e.g. Linux), it passes back an empty
7262 string. Exim wasn't checking for this, and was changing the host name to an
7263 empty string, assuming it had been canonicalized.
7265 5. Although rare, it is permitted to have more than one PTR record for a given
7266 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7267 all the names associated with an address, because they do in Solaris.
7268 However, it seems that they do not in Linux for data that comes from the
7269 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7270 I found this out when I moved to a new Linux workstation and tried to run
7271 the Exim test suite.
7273 To get round this problem I have changed the code so that it now does its
7274 own call to the DNS to look up PTR records when searching for a host name.
7275 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7276 addresses that are only in /etc/hosts are still found.
7278 This behaviour is, however, controlled by an option called host_lookup_
7279 order, which defaults to "bydns:byaddr". If people want to use the other
7280 order, or indeed, just use one or the other means of lookup, they can
7281 specify it in this variable.
7283 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7284 some operating systems, this comes back from gethostbyaddr() as an empty
7285 string, and this is what Exim used to test for. However, it seems that in
7286 other systems, "." is yielded. Exim now tests for this case too.
7288 7. The values of check_spool_space and check_log_space are now held internally
7289 as a number of kilobytes instead of an absolute number of bytes. If a
7290 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7291 kilobyte. This means that much larger values can be stored.
7293 8. Exim monitor: an attempt to get the action menu when not actually pointing
7294 at a message produces an empty menu entitled "No message selected". This
7295 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7296 no entries in it ("Shell widget menu has zero width and/or height"). So I
7297 have added a single, blank menu entry in this case.
7299 9. Added ${quote_local_part.
7301 10. MIME decoding is now applied to the contents of Subject: header lines when
7304 11. Now that a reference to $sender_host_address automatically causes a reverse
7305 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7306 host lookup before query-style lookups in lists that might use this
7307 variable. This has therefore been abolished, and the "net-" prefix is no
7308 longer necessary for query-style lookups.
7310 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7311 have been a typo for LFLAGS, so it has been changed.
7313 13. The install script calls Exim with "-C /dev/null" in order to find the
7314 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7315 to be output. However, since Exim outputs its version number before the
7316 error, it didn't break the script. It just looked ugly. I fixed this by
7317 always allowing "-C /dev/null" if the caller is root.
7319 14. Ignore overlarge ACL variable number when reading spool file - insurance
7320 against a later release with more variables having written the file.
7322 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7323 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7324 this, as well as the form without the "IPv6" on the front.
7326 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7327 OS/Makefile-Darwin file.
7329 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7330 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7332 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7333 variable. However, some OS use ints for inodes. I've added cast to long int
7334 to get rid of the compiler warning.
7336 19. I had forgotten to lock out "/../" in configuration file names when
7337 ALT_CONFIG_PREFIX was set.
7339 20. Routers used for verification do not need to specify transports. However,
7340 if such a router generated a host list, and callout was configured, Exim
7341 crashed, because it could not find a port number from the (non-existent)
7342 transport. It now assumes port 25 in this circumstance.
7344 21. Added the -t option to exigrep.
7346 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7347 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7348 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7349 them is set, LOOKUP_LSEARCH is forced.
7351 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7353 24. Added sender and host information to the "rejected by local_scan()" log
7354 line; previously there was no indication of these.
7356 25. Added .include_if_exists.
7358 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7359 Linux. It turns out that not all file systems support this. Apparently some
7360 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7361 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7362 not supported on the file descriptor, is now ignored when Exim is trying to
7363 sync a directory. This applies only to Linux.
7365 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7367 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7368 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7371 29. Added support for maildirsize files from supplied patch (modified a bit).
7373 30. The use of :fail: followed by an empty string could lead Exim to respond to
7374 sender verification failures with (e.g.):
7376 550 Verification failed for <xxx>
7377 550 Sender verify failed
7379 where the first response line was missing the '-' that indicates it is not
7380 the final line of the response.
7382 31. The loop for finding the name of the user that called Exim had a hardwired
7383 limit of 10; it now uses the value of finduser_retries, which is used for
7384 all other user lookups.
7386 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7389 33. Exim was neglecting to zero errno before one call of strtol() when
7390 expanding a string and expecting an integer value. On some systems this
7391 resulted in spurious "integer overflow" errors. Also, it was casting the
7392 result into an int without checking.
7394 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7395 did not work. The code looks as if it has *never* worked, though it appears
7396 to have been documented since at least release 1.62. I have made it work.
7398 35. The "timeout_DNS" error in retry rules, also documented since at least
7399 1.62, also never worked. As it isn't clear exactly what this means, and
7400 clearly it isn't a major issue, I have abolished the feature by treating it
7401 as "timeout", and writing a warning to the main and panic logs.
7403 36. The display of retry rules for -brt wasn't always showing the error code
7406 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7407 timeout_connect_A, timeout_connect_MX.
7409 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7410 to the empty sender.
7412 39. The daemon was not analysing the content of -oX till after it had closed
7413 stderr and disconnected from the controlling terminal. This meant that any
7414 syntax errors were only noted on the panic log, and the return code from
7415 the command was 0. By re-arranging the code a little, I've made the
7416 decoding happen first, so such errors now appear on stderr, and the return
7417 code is 1. However, the actual setting up of the sockets still happens in
7418 the disconnected process, so errors there are still only recorded on the
7421 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7422 connections (as happens on some IP stacks) was logged at start up time as
7423 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7424 it from "IPv6 and IPv4", which means that two separate sockets are being
7427 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7428 says whether AF_INET or AF_INET6 was passed as an argument.
7430 42. Exiwhat output was messed up when time zones were included in log
7433 43. Exiwhat now gives more information about the daemon's listening ports,
7434 and whether -tls-on-connect was used.
7436 44. The "port" option of the smtp transport is now expanded.
7438 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7439 silently ignored. Now an error message is written to the main and panic
7442 46. There's a new ACL modifier called "logwrite" which writes to a log file
7443 as soon as it is encountered.
7445 47. Added $local_user_uid and $local_user_gid at routing time.
7447 48. Exim crashed when trying to verify a sender address that was being
7450 49. Exim was recognizing only a space character after ".include". It now also
7451 recognizes a tab character.
7453 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7454 extracting the relevant information from the specification. The man page no
7455 longer contains scrambled data for the -d option, and I've added a section
7456 at the front about calling Exim under different names.
7458 51. Added "extra_headers" argument to the "mail" command in filter files.
7460 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7463 53. Installed eximstats 1.29.
7465 54. Added transport_filter_timeout as a generic transport option.
7467 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7468 Cc: header lines. This was required by RFC 822, but it not required by RFC
7471 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7472 incoming messages that did not have them. Now it does so only if the
7473 message originates locally, that is, if there is no associated remote host
7474 address. When Resent- header lines are present, this applies to the Resent-
7475 lines rather than the non-Resent- lines.
7477 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7478 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7480 58. Messages for configuration errors now include the name of the main
7481 configuration file - useful now that there may be more than one file in a
7482 list (.included file names were always shown).
7484 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7485 for those rare installations that do not start the daemon as root or run it
7486 setuid root. I've cut out the call to initgroups() if the daemon is not
7489 60. The Exim user and group can now be bound into the binary as text strings
7490 that are looked up at the start of Exim's processing.
7492 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7494 62. Added $mailstore_basename variable.
7496 63. Installed patch to sieve.c from Michael Haardt.
7498 64. When Exim failed to open the panic log after failing to open the main log,
7499 the original message it was trying to log was written to stderr and debug
7500 output, but if they were not available (the usual case in production), it
7501 was lost. Now it is written to syslog before the two lines that record the
7502 failures to open the logs.
7504 65. Users' Exim filters run in subprocesses under the user's uid. It is
7505 possible for a "deliver" command or an alias in a "personal" command to
7506 provoke an address rewrite. If logging of address rewriting is configured,
7507 this fails because the process is not running as root or exim. There may be
7508 a better way of dealing with this, but for the moment (because 4.30 needs
7509 to be released), I have disabled address rewrite logging when running a
7510 filter in a non-root, non-exim process.
7516 1. The buildconfig auxiliary program wasn't quoting the value set for
7517 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7518 not defined. This bug was masked in 4.22 by the effect that was fixed in
7521 2. Some messages that were rejected after a message id was allocated were
7522 shown as "incomplete" by exigrep. It no longer does this for messages that
7523 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7525 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7526 have allow_domain_literals set, the ID did not get logged in the <= line.
7527 Domain literals are now always recognized in Message-ID: header lines.
7529 4. The first argument for a ${extract expansion item is the key name or field
7530 number. Leading and trailing spaces in this item were not being ignored,
7531 causing some misleading effects.
7533 5. When deliver_drop_privilege was set, single queue runner processes started
7534 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7535 same command in the process it spins off) were not dropping privilege.
7537 6. When the daemon running as "exim" started a queue runner, it always
7538 re-executed Exim in the spun-off process. This is a waste of effort when
7539 deliver_drop_privilege is set. The new process now just calls the
7540 queue-runner function directly.
7546 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7549 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7550 ignored. Though the use of -oP was forcing the writing of a pid file, it
7551 was always written to the default place.
7553 3. If the message "no IP address found for host xxxx" is generated during
7554 incoming verification, it is now followed by identification of the incoming
7555 connection (so you can more easily find what provoked it).
7557 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7559 5. Added some features to "harden" Exim a bit more against certain attacks:
7561 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7562 be put in Local/Makefile. This is like the never_users runtime option,
7563 but it cannot be overridden. The default setting is "root".
7565 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7566 prefix string with which any file named in a -C command line option
7569 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7570 is retained for -C and -D only if the caller of Exim is root. Without
7571 it, the exim user may also use -C and -D and retain privilege.
7573 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7574 command line option is disabled.
7576 6. Macro names set by the -D option must start with an upper case letter, just
7577 like macro names defined in the configuration file.
7579 7. Added "dereference=" facility to LDAP.
7581 8. Two instances of the typo "uknown" in the source files are fixed.
7583 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7584 the Configure-Makefile script screwed up while processing it.
7586 10. Incorporated PCRE 4.4.
7588 11. The SMTP synchronization check was not operating right at the start of an
7589 SMTP session. For example, it could not catch a HELO sent before the client
7590 waited for the greeting. There is now a check for outstanding input at the
7591 point when the greeting is written. Because of the duplex, asynchronous
7592 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7593 way, but not yet received, when the check is performed.
7595 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7596 on TCP/IP sockets, because this apparently causes some broken clients to
7599 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7600 unchanged) from the Cygwin maintainer.
7602 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7603 was supported instead of testing for mbx. Effectively a typo.
7605 15. The spa authenticator server code was not checking that the input it
7606 received was valid base64.
7608 16. The debug output line for the "set" modifier in ACLs was not showing the
7609 name of the variable that was being set.
7611 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7613 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7614 lookup on its own. Something else was needed to trigger the lookup. For
7615 example, a match in host_lookup or the need for a host name in a host list.
7616 Now, if $sender_host_name is referenced and the host name has not yet been
7617 looked up, a lookup is performed. If the lookup fails, the variable remains
7618 empty, and $host_lookup_failed is set to "1".
7620 19. Added "eqi" as a case-independent comparison operator.
7622 20. The saslauthd authentication condition could segfault if neither service
7623 nor realm was specified.
7625 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7626 error message that was logged was misleading, and incoming SMTP
7627 connections were dropped. The message is now more accurate, and temporary
7628 errors are given to SMTP connections.
7630 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7631 (except RSET) with a 421 error, until QUIT is received. However, it was
7632 failing to send a response to QUIT.
7634 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7635 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7636 ACL is now run after the helo_try_verify_hosts code.
7638 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7639 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7640 case-independent, so other case variants are also recognized). Apparently
7641 some systems use these upper case variants.
7643 25. If more than two messages were waiting for the same host, and a transport
7644 filter was specified for the transport, Exim sent two messages over the
7645 same TCP/IP connection, and then failed with "socket operation on non-
7646 socket" when it tried to send the third.
7648 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7650 27. The extern definition of crypt16() in expand.c was not being excluded when
7651 the OS had its own crypt16() function.
7653 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7654 as a preferred synonym for return_size_limit, both as an option and as an
7657 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7659 30. Changed the default configuration ACL to relax the local part checking rule
7660 for addresses that are not in any local domains. For these addresses,
7661 slashes and pipe symbols are allowed within local parts, but the sequence
7662 /../ is explicitly forbidden.
7664 31. SPA server authentication was not clearing the challenge buffer before
7667 32. log_message in a "warn" ACL statement was writing to the reject log as
7668 well as to the main log, which contradicts the documentation and doesn't
7669 seem right (because no rejection is happening). So I have stopped it.
7671 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7672 However, I am unable to do any testing of this.
7674 34. Fixed an infelicity in the appendfile transport. When checking directories
7675 for a mailbox, to see if any needed to be created, it was accidentally
7676 using path names with one or more superfluous leading slashes; tracing
7677 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7679 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7680 contents are added to the log line that is written for every discarded
7681 recipient. (Previously a log_message setting was ignored.)
7683 36. The ${quote: operator now quotes the string if it is empty.
7685 37. The install script runs exim in order to find its version number. If for
7686 some reason other than non-existence or emptiness, which it checks, it
7687 could not run './exim', it was installing it with an empty version number,
7688 i.e. as "exim-". This error state is now caught, and the installation is
7691 38. An argument was missing from the function that creates an error message
7692 when Exim fails to connect to the socket for saslauthd authentication.
7693 This could cause Exim to crash, or give a corrupted message.
7695 39. Added isip, isip4, and isip6 to ${if conditions.
7697 40. The ACL variables $acl_xx are now saved with the message, and can be
7698 accessed later in routers, transports, and filters.
7700 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7701 strings in the file are not string-expanded.
7703 42. If a MAIL command specified a SIZE value that was too large to fit into an
7704 int variable, the check against message_size_limit failed. Such values are
7705 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7706 day this will have to be increased, but I don't think I want to be around
7707 when emails are that large.
7714 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7715 iconv() is not standard in FreeBSD.
7717 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7718 IPv6 enabled. The observed symptom was a segmentation fault on return from
7719 the function os_common_find_running_interfaces() in src/os.c.
7721 3. In the check_special_case() function in daemon.c I had used "errno" as an
7722 argument name, which causes warnings on some systems. This was basically a
7723 typo, since it was named "eno" in the comments!
7725 4. The code that waits for the clock to tick (at a resolution of some fraction
7726 of a second) so as to ensure message-id uniqueness was always waiting for
7727 at least one whole tick, when it could have waited for less. [This is
7728 almost certainly not relevant at current processor speeds, where it is
7729 unlikely to ever wait at all. But we try to future-proof.]
7731 5. The function that sleeps for a time interval that includes fractions of a
7732 second contained a race. It did not block SIGALRM between setting the
7733 timer, and suspending (a couple of lines later). If the interval was short
7734 and the sigsuspend() was delayed until after it had expired, the suspension
7735 never ended. On busy systems this could lead to processes getting stuck for
7738 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7739 process, before it forks any delivery processes. The open lookup caching
7740 mechanism meant that the open file or database connection was passed into
7741 the delivery process. The problem was that delivery processes always tidy
7742 up cached lookup data. This could cause a problem for the next delivery
7743 process started by the queue runner, because the external queue runner
7744 process does not know about the closure. So the next delivery process
7745 still has data in the lookup cache. In the case of a file lookup, there was
7746 no problem because closing a file descriptor in a subprocess doesn't affect
7747 the parent. However, if the lookup was caching a connection to a database,
7748 the connection was closed, and the second delivery process was likely to
7749 see errors such as "PGSQL: query failed: server closed the connection
7750 unexpectedly". The problem has been fixed by closing all cached lookups
7751 in a queue runner before running a delivery process.
7753 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7754 seem to have the "const" qualifier which it has on other OS. I've
7757 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7758 *of the same type* (client or server) with the same public name that an
7759 error should be diagnosed.
7761 9. When Exim looked up a host name for an IP address, but failed to find the
7762 original IP address when looking up the host name (a safety check), it
7763 output the message "<ip address> does not match any IP for NULL", which was
7764 confusing, to say the least. The bug was that the host name should have
7765 appeared instead of "NULL".
7767 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7768 user that is built into the binary, and the -C or -D options is used, root
7769 privilege is dropped before the configuration file is read. In addition,
7770 logging is switched to stderr instead of the normal log files. If the
7771 configuration then re-defines the Exim user, the unprivileged environment
7772 is probably not what is expected, so Exim logs a panic warning message (but
7775 However, if deliver_drop_privilege is set, the unprivileged state may well
7776 be exactly what is intended, so the warning has been cut out in that case,
7777 and Exim is allowed to try to write to its normal log files.
7783 1. smtp_return_error_details was not giving details for temporary sender
7784 or receiver verification errors.
7786 2. Diagnose a configuration error if two authenticators have the same public
7789 3. Exim used not to create the message log file for a message until the first
7790 delivery attempt. This could be confusing when incoming messages were held
7791 for policy or load reasons. The message log file is now created at the time
7792 the message is received, and an initial "Received" line is written to it.
7794 4. The automatically generated man page for command line options had a minor
7795 bug that caused no ill effects; however, a more serious problem was that
7796 the procedure for building the man page automatically didn't always
7797 operate. Consequently, release 4.20 contains an out-of-date version. This
7798 shouldn't happen again.
7800 5. When building Exim with embedded Perl support, the script that builds the
7801 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7802 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7804 6. The freeze_tell option was not being used for messages that were frozen on
7805 arrival, either by an ACL or by local_scan().
7807 7. Added the smtp_incomplete_transaction log selector.
7809 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7810 was accepting AUTH without a new EHLO.
7812 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7813 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7816 10. Logging of TCP/IP connections (when configured) now happens in the main
7817 daemon process instead of the child process, so that the TCP/IP connection
7818 count is more accurate (but it can never be perfect).
7820 11. The use of "drop" in a nested ACL was not being handled correctly in the
7821 outer ACL. Now, if condition failure induced by the nested "drop" causes
7822 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7823 or "require"), the connection is dropped.
7825 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7826 that yield "discard" can now be used with an "accept" or a "discard" verb,
7827 but an error is generated for any others (because I can't see a useful way
7828 to define what should happen).
7830 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7831 are now processed in the same way as lines in the Exim configuration file.
7832 In particular, continuation lines are supported.
7834 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7836 15. Added -ti meaning -t -i.
7838 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7839 domains, and warn by logging if others are found.
7841 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7842 returned from GIFCONF when Exim is trying to find the list of interfaces on
7843 a host. The code in os.c has been modified to copy each ifreq to an aligned
7844 structure in all cases.
7846 Also, in some cases, the returned ifreq's were being copied to a 'struct
7847 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7848 means the last couple of bytes of an IPv6 address could be chopped if the
7849 ifreq contained only a normal sockaddr (14 bytes storage).
7851 18. Named domain lists were not supported in the hosts_treat_as_local option.
7852 An entry such as +xxxx was not recognized, and was treated as a literal
7855 19. Ensure that header lines added by a DATA ACL are included in the reject log
7856 if the ACL subsequently rejects the message.
7858 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7859 MD5 (which is deprecated).
7861 21. When testing a filter file using -bf, Exim was writing a message when it
7862 took the sender from a "From " line in the message, but it was not doing so
7863 when it took $return_path from a Return-Path: header line. It now does.
7865 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7866 with a valid header line field name (a series of printing characters
7867 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7869 23. Changed "disc" in the source to "disk" to conform to the documentation and
7870 the book and for uniformity.
7872 24. Ignore Sendmail's -Ooption=value command line item.
7874 25. When execve() failed while trying to run a command in a pipe transport,
7875 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7876 could be confused with a return value of 69 from the command itself. This
7877 has been changed to 127, the value the shell returns if it is asked to run
7878 a non-existent command. The wording for the related log line suggests a
7879 non-existent command as the problem.
7881 26. If received_header_text expands to an empty string, do not add a Received:
7882 header line to the message. (Well, it adds a token one on the spool, but
7883 marks it "old" so that it doesn't get used or transmitted.)
7885 27. Installed eximstats 1.28 (addition of -nt option).
7887 28. There was no check for failure on the call to getsockname() in the daemon
7888 code. This can fail if there is a shortage of resources on the system, with
7889 ENOMEM, for example. A temporary error is now given on failure.
7891 29. Contrary to the C standard, it seems that in some environments, the
7892 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7893 program. Exim now does this explicitly; it affects the formatting of
7894 timestamps using strftime().
7896 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7897 complaints. I've now initialized all the variables, to avoid this.
7899 32. Header lines added by a system filter were not being "seen" during
7900 transport-time rewrites.
7902 33. The info_callback() function passed to OpenSSL is set up with type void
7903 (*)(SSL *, int, int), as described somewhere. However, when calling the
7904 function (actually a macro) that sets it up, the type void(*)() is
7905 expected. I've put in a cast to prevent warnings from picky compilers.
7907 34. If a DNS black list lookup found a CNAME record, but there were no A
7908 records associated with the domain it pointed at, Exim crashed.
7910 35. If a DNS black list lookup returned more than one A record, Exim ignored
7911 all but the first. It now scans all returned addresses if a particular IP
7912 value is being sought. In this situation, the contents of the
7913 $dnslist_value variable are a list of all the addresses, separated by a
7916 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7917 to accept a host name and all its aliases if the forward lookup for any of
7918 them yielded the IP address of the incoming connection. Now it accepts only
7919 those names whose forward lookup yields the correct IP address. Any other
7920 names are discarded. This closes a loophole whereby a rogue DNS
7921 administrator could create reverse DNS records to break through a
7922 wildcarded host restriction in an ACL.
7924 37. If a user filter or a system filter that ran in a subprocess used any of
7925 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7926 the wrong values were passed to the pipe command ($thisaddress had the
7927 value of $0, $0 had the value of $1, etc). This bug was introduced by
7928 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7930 38. Improved the line breaking for long SMTP error messages from ACLs.
7931 Previously, if there was no break point between 40 and 75 characters, Exim
7932 left the rest of the message alone. Two changes have been made: (a) I've
7933 reduced the minimum length to 35 characters; (b) if it can't find a break
7934 point between 35 and 75 characters, it looks ahead and uses the first one
7935 that it finds. This may give the occasional overlong line, but at least the
7936 remaining text gets split now.
7938 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7939 file descriptors might be low, and that setting 1000 would always raise it.
7940 It turns out that in some environments, the limit is already over 1000 and
7941 that lowering it causes trouble. So now Exim takes care not to decrease it.
7943 40. When delivering a message, the value of $return_path is set to $sender_
7944 address at the start of routing (routers may change the value). By an
7945 oversight, this default was not being set up when an address was tested by
7946 -bt or -bv, which affected the outcome if any router or filter referred to
7949 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7950 log only when "message" or "log_message" are set. However, if one of the
7951 conditions was an address verification, or a call to a nested ACL, the
7952 messages generated by the underlying test were being passed through. This
7953 no longer happens. The underlying message is available in $acl_verify_
7954 message for both "message" and "log_message" expansions, so it can be
7955 passed through if needed.
7957 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7958 new expansion $bh_ to give the encoded byte string without charset
7959 translation. Translation happens only if iconv() is available; HAVE_ICONV
7960 indicates this at build time. HEADERS_CHARSET gives the charset to
7961 translate to; headers_charset can change it in the configuration, and
7962 "headers charset" can change it in an individual filter file.
7964 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7965 that creates RFC 2047 encoded "words" labels them as that charset instead
7966 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7967 expansion operator; (ii) when Exim creates a From: line for a local
7968 message; (iii) when a header line is rewritten to include a "phrase" part.
7970 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7971 buggy, causing it to skip the first lines of messages whose message ID
7972 ended in 'D'. This would not have bitten before Exim release 4.14, because
7973 message IDs were unlikely to end in 'D' before then. The effect was to have
7974 incorrect size information for certain domains.
7976 45. #include "config.h" was missing at the start of the crypt16.c module. This
7977 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7980 46. If there was a timeout during a "random" callout check, Exim treated it as
7981 a failure of the random address, and carried on sending RSET and the real
7982 address. If the delay was just some slowness somewhere, the response to the
7983 original RCPT would be taken as a response to RSET and so on, causing
7984 mayhem of various kinds.
7986 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7987 when I implemented it. It didn't allow for the fact that some option values
7988 may legitimately be negative (e.g. size_addition), and it didn't even do
7989 the right test for positive values.
7991 48. Domain names in DNS records are case-independent. Exim always looks them up
7992 in lower case. Some resolvers return domain names in exactly the case they
7993 appear in the zone file, that is, they may contain uppercase letters. Not
7994 all resolvers do this - some return always lower case. Exim was treating a
7995 change of case by a resolver as a change of domain, similar to a widening
7996 of a domain abbreviation. This triggered its re-routing code and so it was
7997 trying to route what was effectively the same domain again. This normally
7998 caused routing to fail (because the router wouldn't handle the domain
7999 twice). Now Exim checks for this case specially, and just changes the
8000 casing of the domain that it ultimately uses when it transmits the message
8003 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8006 50. If a filter generated a file delivery with a non-absolute name (possible if
8007 no home directory exists for the router), the forbid_file option was not
8010 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8011 the existing equality matching.
8013 52. Exim was using ints instead of ino_t variables in some places where it was
8014 dealing with inode numbers.
8016 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8017 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8018 called TMPDIR, and if it finds it is different, it changes its value.
8020 54. The smtp_printf() function is now made available to local_scan() so
8021 additional output lines can be written before returning. There is also an
8022 smtp_fflush() function to enable the detection of a dropped connection.
8023 The variables smtp_input and smtp_batched_input are exported to
8026 55. Changed the default runtime configuration: the message "Unknown user"
8027 has been removed from the ACL, and instead placed on the localuser router,
8028 using the cannot_route_message feature. This means that any verification
8029 failures that generate their own messages won't get overridden. Similarly,
8030 the "Unrouteable address" message that was in the ACL for unverifiable
8031 relay addresses has also been removed.
8033 56. Added hosts_avoid_esmtp to the smtp transport.
8035 57. The exicyclog script was not checking for the esoteric option
8036 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8037 will work only if exicyclog is run under the appropriate euid.
8039 58. Following a discussion on the list, the rules by which Exim recognises line
8040 endings on incoming messages have been changed. The -dropcr and drop_cr
8041 options are now no-ops, retained only for backwards compatibility. The
8042 following line terminators are recognized: LF CRLF CR. However, special
8043 processing applies to CR:
8045 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8046 nor a local message in the state where . is a terminator.
8048 (ii) If a bare CR is encountered in a header line, an extra space is added
8049 after the line terminator so as not to end the header. The reasoning
8050 behind this is that bare CRs in header lines are most likely either
8051 to be mistakes, or people trying to play silly games.
8053 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8054 was being incorrectly given as 18 bytes larger than it should have been.
8055 This is a VOB (very old bug).
8057 60. This may never have affected anything current, but just in case it has:
8058 When the local host is found other than at the start of a list of hosts,
8059 the local host, those with the same MX, and any that follow, are discarded.
8060 When the list in question was part of a longer list of hosts, the following
8061 hosts (not currently being processed) were also being discarded. This no
8062 longer happens. I'm not sure if this situation could ever has previously
8065 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8068 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8069 header. This is something that is recommended in a new Internet Draft, and
8070 is something that is documented as being done by Sendmail. There are two
8071 possible values. For messages generated by the autoreply transport, Exim
8074 Auto-Submitted: auto-replied
8076 whereas for all other generated messages (e.g. bounces) it adds
8078 Auto-Submitted: auto-generated
8080 63. The "personal" condition in filters now includes a test for the
8081 Auto-Submitted: header. If it contains the string "auto-" the message it
8082 not considered personal.
8084 64. Added rcpt_include_affixes as a generic transport option.
8086 65. Added queue_only_override (default true).
8088 66. Added the syslog_duplication option.
8090 67. If what should have been the first header line of a message consisted of
8091 a space followed by a colon, Exim was mis-interpreting it as a header line.
8092 It isn't of course - it is syntactically invalid and should therefore be
8093 treated as the start of the message body. The misbehaviour could have
8094 caused a number of strange effects, including loss of data in subsequent
8095 header lines, and spool format errors.
8097 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8098 client host had authenticated. This control can now be exercised by an ACL
8099 for more flexibility.
8101 69. By default, callouts do not happen when testing with -bh. There is now a
8102 variant, -bhc, which does actually run the callout code, including
8103 consulting and updating the callout cache.
8105 70. Added support for saslauthd authentication, courtesy of Alexander
8108 71. If statvfs() failed on the spool or log directories while checking their
8109 size for availability, Exim confusingly gave the error "space shortage".
8110 Furthermore, in debugging mode it crashed with a floating point exception.
8111 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8112 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8113 serious problem, Exim now writes to the main and panic logs when this
8114 happens, with details of the failure. It then refuses to accept the
8115 incoming message, giving the message "spool directory problem" or "log
8116 directory problem" with a 421 code for SMTP messages.
8118 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8119 0, 1, and 2 exist, because some OS complain for execs without them (see
8120 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8121 descriptors. However, the code omitted to check that the open succeeded,
8122 causing mysterious errors if for some reason the permissions on /dev/null
8123 got screwed. Now Exim writes a message to the main and panic logs, and
8124 bombs out if it can't open /dev/null.
8126 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8127 interact so that it is all more flexible. It is supposed to remain
8128 backwards compatible. Also added extra_local_interfaces.
8130 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8131 to bomb out with an assertion failure - to the client this appears as a
8132 connection drop. This problem occurs in the part of the code that was taken
8133 from the Samba project. Fortunately, the assertion is in a very simple
8134 function, so I have fixed this by reproducing the function inline in the
8135 one place where it is called, and arranging for authentication to fail
8136 instead of killing the process with assert().
8138 75. The SPA client code was not working when the server requested OEM rather
8139 than Unicode encoding.
8141 76. Added code to make require_files with a specific uid setting more usable in
8142 the case where statting the file as root fails - usually a non-root-mounted
8143 NFS file system. When this happens and the failure is EACCES, Exim now
8144 forks a subprocess and does the per-uid checking as the relevant uid.
8146 77. Added process_log_path.
8148 78. If log_file_path was not explicitly set, a setting of check_log_space or
8149 check_log_inodes was ignored.
8151 79. If a space check for the spool or log partitions fails, the incident is now
8152 logged. Of course, in the latter case the data may get lost...
8154 80. Added the %p formatting code to string_format() so that it can be used to
8155 print addresses in debug_print(). Adjusted all the address printing in the
8156 debugging in store.c to use %p rather than %d.
8158 81. There was a concern that a line of code in smtp_in.c could overflow a
8159 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8160 initially expressed, the concern was not well-founded, because trailing
8161 spaces are removed early. However, if the trailing spaces were followed by
8162 a NULL, they did not get removed, so the overflow was possible. Two fixes
8165 (a) I re-wrote the offending code in a cleaner fashion.
8166 (b) If an incoming SMTP command contains a NULL character, it is rejected
8169 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8170 runs initgroups(), so that if the Exim user is in any additional groups,
8171 they will be used during message reception.
8177 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob