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
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes, when delivery is immediate. Previously debugging
74 stopped any time Exim re-execs.
80 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
81 related applications. Previously an "H" was used where available info
82 says that "M" should be, so change to match.
84 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
85 as arguments, so an implementation trying to copy these into a local
86 buffer was taking a taint-enforcement trap. Fix by using dynamically
87 created buffers. Similar fix for radius expansion condition.
89 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
90 reasonable, eg. to count headers. Fix by using dynamically created
91 buffers rather than a local. Do similar fixes for ACL actions "dcc",
92 "log_reject_target", "malware" and "spam"; the arguments are expanded
93 so could be handling tainted values.
95 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
96 broken the (no-op) support for this sendmail command. Restore it
97 to doing nothing, silently, and returning good status.
99 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
100 record path was given (or the default used) without a leading directory
101 path, an error occurred on trying to open it. Use the transport's working
104 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
105 transport. Previously it was the name on the DNS A-record; use instead
106 the head of the CNAME chain leading there (if there is one). This seems
107 to align better with RFC 6125.
109 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
110 smtp_accept_max_per_host allocated resources which were not released
111 when the limit was exceeded. This eventually crashed the daemon. Fix
112 by adding a release action in that path.
114 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
115 expanded; previously using tainted values was rejected. Fix by using
116 dynamically-created buffers.
118 JH/09 Relax restrictions on ACL verify condition needing access to message
119 headers. Previously they were only permitted in data and non-smtp ACLs;
120 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
121 not_blind, header_sender and header_names_ascii verification.
123 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
124 Previously a macro used one argument twice; when called with the
125 argument as an expression having side-effects, incorrect operation
126 resulted. Use an inlineable function.
128 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
129 held open for a verify callout. Previously this wan not accounted for
130 and a corrupt onward SMTP conversation resulted.
132 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
133 Previously they were embedded naively in the constructed address; when
134 needed, strip the quoting and quote the entire local_part.
135 Also make the inbound_srs expansion condition handle quoting.
137 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
138 excluded, not matching the documentation.
140 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
141 was given for the sqlite_dbfile a trap resulted.
143 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
144 "name" argument resulted in a trap. There is no reason to disallow such;
145 this was a coding error.
147 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
148 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
149 spent suspended, ignoring the POSIX definition. Previously we assumed
150 it did and a constant offset from real time could be used as a correction.
151 Change to using the same clock source for the start-of-message and the
152 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
153 exists, just to get a clock slightly more aligned to reality.
155 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
156 RFC says it is optional some validators care. The missing char was not
157 intended but triggered by a line-wrap alignment. Discovery and fix by
158 Guillaume Outters, hacked on by JH.
160 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
161 name being quoted was tainted a trap would be taken. Fix by using
162 dynamically created buffers. The routine could have been called by a
163 rewrite with the "h" flag, by using the "-F" command-line option, or
164 by using a "name=" option on a control=submission ACL modifier.
166 JH/19 SPF: change the Authentication-Results expansion component to give
167 smtp.helo when the sender domain is empty. Previously it gave
170 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
171 not in 127.0.0.0/8 to help in spotting list domains taken over by a
172 domain-parking registrar.
174 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
175 Previously when a whitespace character was specified it was not inserted
176 after removing the newline.
178 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
179 the domain part of the recipient address. This overrides any tls_sni
180 option set, which was previously used.
182 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
185 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
186 is_tainted() had an off-by-one error in the overenthusiastic direction.
187 Find and fix by Gavan. Although NetBSD is not a supported platform for
188 4.94 this bug could affect other platforms.
190 PP/01 Fix default prime selection to be consistent.
191 One path used ike23 still, instead of exim.dev.20160529.3; now both
192 execution flows will use the same DH primes (currently
193 exim.dev.20160529.3).
195 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
196 option in smtp transport, to match the documentation. Previously
197 verification was not being done in some cases where it should have been.
199 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
200 than one server was defined and depending on the platform memory layout
201 details, an internal consistency trap could be hit while walking the list
204 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
205 files. The value is used by the authresults expansion item. Previously
206 if this was used in a router or transport, a crash could result.
208 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
209 have a sufficient incidence of EINTR returns from select that an
210 interaction with other operations done by the main daemon loop exposed
211 a bug in the error-handling. This was benign apart from the log
214 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
215 for consistency with delivered "=>" and failed "**" lines. While we're
216 there, handle PRX and TFO.
218 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
219 applied. This resulted, if any header-line rewrite rules were configured,
220 in a panic-log triggerable by sending a message with a long address in
221 a header. Fix by increasing the arbitrary limit to larger than a single
222 (dewrapped) 5322 header line maximum size.
224 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
225 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
226 RFC 6648 which deprecates X- options in protocols as a general practice.
227 Changeover between the implementations is handled by the mechanisms
230 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
231 a remote which uses callout/hold. Previously the local was queued.
233 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
236 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
237 log line. The asterisk is now consistently appended to the remote IP
238 (and port, if given), and will also be provided on defer and fail log
239 lines. Previously it could be placed on the local IP if that was being
240 logged, and was only provided on delivery lines.
242 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
244 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
245 authenticator client_send option. Previously the next char, after a pair
246 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
247 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
248 documentation. There is still no way to get a leading ^ immediately
249 after a NUL (ie. for the password of a PLAIN method authenticator.
251 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
252 files. For bad sizes read, delete the record and whine to paniclog.
254 JH/38 When logging an AUTH failure, as server, do not include sensitive
255 information. Previously, the credentials would be included if given
256 as part of the AUTH command line and an ACL denied authentication.
258 JH/39 Bug 2691: fix $local_part_data. When the matching list element
259 referred to a file, bad data was returned. This likely also affected
262 JH/40 The gsasl authenticator now supports caching of the salted password
263 generated by the client-side implementation. This required the addition
264 of a new variable: $auth4.
266 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
267 left undeleted; the attempt to re-create it then failed - resulting in
268 the usual "SIGHUP tp have daemon reload configuration" to not work.
269 This affected any platform not supporting "abstract" Unix-domain
270 sockets (i.e. not Linux).
272 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
273 recipients" response to RCPT in a later response, with a 250. The
274 previous coding assumed this would not happen, and under PIPELINING
275 would result in both lost and duplicate recipients for a message.
277 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
278 Previously the weighting was incorrectly applied. Similar fix for socks
279 proxies. Found and fixed by Heiko Schlichting.
281 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
282 not handle sub-lists included using the +namedlist syntax. While
283 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
284 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
287 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
288 Previously the permanent pool was used, so the sore could not be freed.
289 This meant a connection with many messages would use continually-growing
292 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
293 per-pool so as not to waste too much space. Previously a constant size
294 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
295 DOS attacks harder. The cost is wasted memory use in the larger blocks.
297 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
298 same space repeatedly, and a smaller process footprint.
300 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
301 for GnuTLS. Previously the name started "%s" which, while not a bug,
302 looked as if if might be one.
304 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
305 "continued connection") make the $proxy_* variables available. Previously
306 the information was not passed across the exec() call for subsequent
307 transport executions. This also mean that the log lines for the
308 messages can show the proxy information.
310 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
311 receive time. With modern systems the difference is significant.
312 The historical behaviour can be restored by disabling (a new) log_selector
313 "queue_time_exclusive".
315 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
316 content) were specifically tested for. Now, also cover expansions
317 resulting in ACL names and inline ACL content.
319 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
320 making it unusable in complex expressions.
322 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
323 fail with a taint-check on the spoolfile name, and leave the message
326 HS/01 Enforce absolute PID file path name.
328 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
330 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
332 PP/02 Bug 2643: Correct TLS DH constants.
333 A missing NUL termination in our code-generation tool had led to some
334 incorrect Diffie-Hellman constants in the Exim source.
335 Reported by kylon94, code-gen tool fix by Simon Arlott.
337 PP/03 Impose security length checks on various command-line options.
338 Fixes CVE-2020-SPRSS reported by Qualys.
340 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
341 better. Reported by Qualys.
343 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
344 providing a particularly obnoxious sender full name.
347 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
349 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
352 PP/08 Change default for recipients_max from unlimited to 50,000.
354 PP/09 Fix security issue with too many recipients on a message (to remove a
355 known security problem if someone does set recipients_max to unlimited,
356 or if local additions add to the recipient list).
357 Fixes CVE-2020-RCPTL reported by Qualys.
359 PP/10 Fix security issue in SMTP verb option parsing
360 Fixes CVE-2020-EXOPT reported by Qualys.
362 PP/11 Fix security issue in BDAT state confusion.
363 Ensure we reset known-good where we know we need to not be reading BDAT
364 data, as a general case fix, and move the places where we switch to BDAT
365 mode until after various protocol state checks.
366 Fixes CVE-2020-BDATA reported by Qualys.
368 HS/03 Die on "/../" in msglog file names
370 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
371 the Exim runtime user are allowed to create files.
373 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
376 QS/03 When reading the output from interpreted forward files we do not
377 pass the pipe between the parent and the interpreting process to
378 executed child processes (if any).
380 QS/04 Always die if requested from internal logging, even is logging is
383 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
384 the API; compilation noo longer completes with DMARC support included.
385 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
386 been reported on other platforms.
388 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
390 JH/56 Make the majority of info read from config files readonly, for defence-in-
391 depth against exploits. Suggestion by Qualys.
392 Not supported on Solaris 10.
394 JH/57 Fix control=fakreject for a custom message containing tainted data.
395 Previously this resulted in a log complaint, due to a re-expansion present
396 since fakereject was originally introduced.
398 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
399 modifier was given, a loop resulted.
401 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
402 pipelined SMTP command followed the BDAT LAST then it would be
403 incorrectly treated as part of the message body, causing a verification
406 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
407 require looser validation rules than those for 821-level addresses,
408 which only permit IP addresses.
414 JH/01 Avoid costly startup code when not strictly needed. This reduces time
415 for some exim process initialisations. It does mean that the logging
416 of TLS configuration problems is only done for the daemon startup.
418 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
420 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
421 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
424 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
425 DKIM signing is being done. Previously a transport_filter would always
426 disable CHUNKING, falling back to traditional DATA.
428 JH/05 Regard command-line recipients as tainted.
430 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
432 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
433 PAM library frees one of the arguments given to it, despite the
434 documentation. Therefore a plain malloc must be used.
436 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
437 on-stack buffers were used, resulting in a taint trap when DSN information
438 copied from a received message was written into the buffer.
440 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
441 the ordering of its ARC headers. This caused a crash.
443 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
444 a new record was being constructed with information from the peer, a trap
447 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
448 installation would get error messages from DMARC verify, when it hit the
449 nonexistent file indicated by the default. Distros wanting DMARC enabled
450 should both provide the file and set the option.
451 Also enforce no DMARC verification for command-line sourced messages.
453 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
454 could, depending on the platform, hang at the STARTTLS response.
456 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
457 message on a connection. Previously if one message had ARC headers and
458 the following one did not, a crash could result when adding an
459 Authentication-Results: header.
461 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
462 Exim main code and Exim-related utities. The introduction of taint
463 tracking also did many adjustments to string handling. Since then, eximon
464 frequently terminated with an assert failure.
466 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
467 check for 452 responses. This slightly helps the inefficieny of doing
468 a large alias-expansion into a recipient-limited target. The max_rcpt
469 transport option still applies (and at the current default, will override
470 the new feature). The check is done for either cause of synch, and forces
471 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
472 connection. The new facility is not tunable at this time.
474 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
475 library live data was being used, so the results became garbage. Make
476 copies while it is still usable.
478 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
479 on delivery deferred (==) and failed (**) lines (if a delivery was
480 attemtped). Previously it was only on completion (=>) lines.
482 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
483 for the expansion of the server_scram_iter and server_scram_salt options.
485 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
486 are now specifically given a NO_DATA response without hitting the system
487 resolver. The library goes on to do the now-standard TXT lookup.
488 Use of dnsdb lookups is not affected.
490 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
491 only retrieve the errormessage once. Previously two calls to dlerror()
492 were used, and the second one (for mainlog/paniclog) retrieved null
495 JH/20 Taint checking: disallow use of tainted data for
496 - the appendfile transport file and directory options
497 - the pipe transport command
498 - the autoreply transport file, log and once options
499 - file names used by the redirect router (including filter files)
501 - paths used by single-key lookups
502 Previously this was permitted.
504 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
505 adjusted the size of a major service buffer; this failed because the
506 buffer was in use at the time. Change to a compile-time increase in the
507 buffer size, when this authenticator is compiled into exim.
509 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
510 previous fast-mode was untenable in the face of glibs using mmap to
511 support larger malloc requests.
513 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
514 New values supported, if defined on system where compiled:
515 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
516 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
518 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
519 running a limited number of proceses in parallel, a benefit is gained. The
520 amount varies with the platform hardware and load. The use of the option
521 queue_run_in_order means we cannot do this, as ordering becomes
524 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
525 had introduced a string-copy (for ensuring NUL-termination) which was not
526 appropriate for that case, which can include embedded NUL bytes in the
527 block of data. Investigation showed the copy to actually be needless, the
528 data being length-specified.
530 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
531 done during a receiving connection, and both used TLS, global info was
532 used rather than per-connection info for tracking the state of data
533 queued for transmission. This could result in a connection hang.
535 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
536 Previously, when delivering serveral messages down a single connection
537 only the first would provide a SIZE. This was due to the size information
538 not being properly tracked.
540 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
541 TAI (at 37 seconds currently), pretend to be in UTC for time-related
542 expansion and logging. Previously, spurious values such as a future
543 minute could be seen.
545 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
546 it could crash from a null-deref. This could also affect the
547 ${addresses: } operator and ${readsock } item.
549 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
550 message following a mime one, the variable was not reset.
552 JH/30 When an pipelined-connect fails at the first response, assume incorrect
553 cached capability (perhaps the peer reneged?) and immediately retry in
556 JH/31 Fix spurious detection of timeout while writing to transport filter.
558 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
559 an attempt to copy the string was made before checking it.
561 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
562 taint of the lookup key was maintained; we now regard the presence in the
563 filesystem as sufficient validation.
565 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
566 argument is supplied.
568 JH/35 The dsearch lookup now requires that the directory is an absolute path.
569 Previously this was not checked, and nonempty relative paths made an
570 access under Exim's current working directory.
572 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
573 Previously no event was raised.
575 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
576 parameter supplied by the sender MAIL FROM command. Previously it was
577 ignored, and only the check_spool_space option value for the required
580 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
581 the size of the signing public-key. Previously it was instead giving
582 the size of the signature hash.
584 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
585 the default. See the (new) dkim_verify_min_keysizes option.
587 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
588 and an ACL use a lookup for checking either the local_part or domain,
589 stale data could be accessed. Ensure that variable references are
590 dropped between messages.
592 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
593 by the client was not checked as pointing within response data before
594 being used. A malicious client could thus cause an out-of-bounds read and
595 possibly gain authentication. Fix by adding the check.
597 JH/42 Internationalisation: change the default for downconversion in the smtp
598 transport to be "if needed". Previously it was "as previously set" for
599 the message, which usually meant "if needed" for message-submission but
600 "no" for everything else. However, MTAs have been seen using SMTPUTF8
601 even when the envelope addresses did not need it, resulting in forwarding
602 failures to non-supporting MTAs. A downconvert in such cases will be
603 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
604 transport. The change does mean that addresses needing conversion will
605 be converted when previously a delivery failure would occur.
607 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
608 response was received it would be used unchecked in a fail-DSN, violating
609 standards on line-length limits. Truncate if needed.
611 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
612 doesn't work. (Additionally add a new main config option to configure the
619 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
620 side, to decode a TLS 1.3 packet capture.
622 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
623 Previously the default library behaviour applied, sending two, each in
626 JH/03 Debug output for ACL now gives the config file name and line number for
629 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
631 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
633 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
634 buffer overrun for (non-chunking) other transports.
636 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
637 TLS1.3, means that a server rejecting a client certificate is not visible
638 to the client until the first read of encrypted data (typically the
639 response to EHLO). Add detection for that case and treat it as a failed
640 TLS connection attempt, so that the normal retry-in-clear can work (if
641 suitably configured).
643 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
644 and/or domain. Found and fixed by Jason Betts.
646 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
647 configuration). If a CNAME target was not a wellformed name pattern, a
650 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
651 the OS reports them interleaved with other addresses.
653 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
654 used both for input and for a verify callout, both encrypted, SMTP
655 responses being sent by the server could be lost. This resulted in
656 dropped connections and sometimes bounces generated by a peer sending
659 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
660 string. Previously it was possible to cause undefined behaviour in a
661 library routine (usually a crash). Found by "zerons".
663 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
666 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
667 API was removed, so update to use the newer ones.
669 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
670 any timeout set, is taking a long time. Previously we would hang on to a
671 rotated logfile "forever" if the input was arriving with long gaps
672 (a previous attempt to fix addressed lack, for a long time, of initial
675 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
676 shared (NFS) environment. The length of the tempfile name is now
677 4 + 16 ("hdr.$message_exim_id") which might break on file
678 systems which restrict the file name length to lower values.
681 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
682 shared (NFS) environment.
684 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
685 did for all versions <4.90). Notably -M, -m, --invert, -I may be
688 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
689 on some platforms for bit 31.
691 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
692 to changes apparently associated with TLS1.3 handling some of the APIs
693 previously used were either nonfunctional or inappropriate. Strings
694 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
695 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
696 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
697 This affects log line X= elements, the $tls_{in,out}_cipher variables,
698 and the use of specific cipher names in the encrypted= ACL condition.
700 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
702 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
703 verification result was not updated unless hosts_require_ocsp applied.
705 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
706 queue_list_requires_admin set to false, non-admin users were denied the
709 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
710 directory-of-certs mode. Previously they were advertised despite the
713 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
714 A single TCP connection by a client will now hold a TLS connection open
715 for multiple message deliveries, by default. Previously the default was to
718 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
719 default. If built with the facility, DANE will be used. The facility
720 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
722 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
723 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
724 must be defined and you must still, unless you define DISABLE_TLS, manage
725 the the include-dir and library-file requirements that go with that
726 choice. Non-TLS builds are still supported.
728 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
731 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
732 default. If the platform supports and has the facility enabled, it will
733 be requested on all coneections.
735 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
736 controlled by the build-time option SUPPORT_PIPE_CONNECT.
738 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
740 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
741 success-DSN messages. Previously the From: header was always the default
742 one for these; the option was ignored.
744 JH/28 Fix the timeout on smtp response to apply to the whole response.
745 Previously it was reset for every read, so a teergrubing peer sending
746 single bytes within the time limit could extend the connection for a
747 long time. Credit to Qualsys Security Advisory Team for the discovery.
749 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
750 delivery address, which leaked information of the results of local
751 forwarding. Change to the original envelope recipient address, per
754 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
755 requested. Previously not bounce was generated and a log entry of
756 error ignored was made.
758 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
760 JH/32 Introduce a general tainting mechanism for values read from the input
761 channel, and values derived from them. Refuse to expand any tainted
762 values, to catch one form of exploit.
764 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
765 was unused and the unexpanded text used for the test. Found and
766 fixed by Ruben Jenster.
768 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
769 an attempt to use a TLS library read routine dereffed a nul pointer,
772 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
773 them in our smtp response.
775 JH/36 Have the general router option retry_use_local_part default to true when
776 any of the restrictive preconditions are set (to anything). Previously it
777 was only for check_local user. The change removes one item of manual
778 configuration which is required for proper retries when a remote router
779 handles a subset of addresses for a domain.
781 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
782 link count into consideration.
784 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
785 caused the extension of big_buffer, the following lines were ignored.
787 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
788 accordance with RFC 2308. Previously there was no expiry, so a longlived
789 receive process (eg. due to ACL delays) versus a short SOA value could
792 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
794 JH/39 Promote DMARC support to mainline.
796 JH/40 Bug 2452: Add a References: header to DSNs.
798 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
799 parameters. The relevant library call is documented as "Deprecated: This
800 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
801 3.6.0, DH parameters are negotiated following RFC7919."
803 HS/06 Change the default of dnssec_request_domains to "*"
805 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
806 carried on and emitted a BDAT command, even when PIPELINING was not
809 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
810 buffer was used for the filename, resulting in a trap when tainted
811 arguments (eg. $domain) were used.
813 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
814 recommended to avoid a possible server-load attack. The feature can be
815 re-enabled via the openssl_options main cofiguration option.
817 JH/45 local_scan API: documented the current smtp_printf() call. This changed
818 for version 4.90 - adding a "more data" boolean to the arguments.
819 Bumped the ABI version number also, this having been missed previously;
820 release versions 4.90 to 4.92.3 inclusive were effectively broken in
821 respect of usage of smtp_printf() by either local_scan code or libraries
822 accessed via the ${dlfunc } expansion item. Both will need coding
823 adjustment for any calls to smtp_printf() to match the new function
824 signature; a FALSE value for the new argument is always safe.
826 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
827 the file-offset (which the Linux syscall does, and exim expects); this
828 resulted in an indefinite loop.
830 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
831 to do ARC verification. The Authentication-Results: header line added
832 by the configuration then had no ARC item.
834 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
835 an incorrect mode was used for reading the result, resulting in it being
842 JH/01 Remove code calling the customisable local_scan function, unless a new
843 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
845 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
846 non-signal-safe functions being used.
848 JH/03 Bug 2269: When presented with a received message having a stupidly large
849 number of DKIM-Signature headers, disable DKIM verification to avoid
850 a resource-consumption attack. The limit is set at twenty.
852 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
853 report of oldest_pass in ${authres } in consequence, and separate out
854 some descriptions of reasons for verification fail.
856 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
857 files in the spool were present and unlocked. A queue-runner could spot
858 them, resulting in a duplicate delivery. Fix that by doing the unlock
859 after the unlink. Investigation by Tim Stewart. Take the opportunity to
860 add more error-checking on spoolfile handling while that code is being
863 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
864 No known attacks, no CVE, this is defensive hardening.
866 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
867 a queue-runner could start a delivery while other operations were ongoing.
868 Cutthrough delivery was a common victim, resulting in duplicate delivery.
869 Found and investigated by Tim Stewart. Fix by using the open message data
870 file handle rather than opening another, and not locally closing it (which
871 releases a lock) for that case, while creating the temporary .eml format
872 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
874 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
875 $sender_verify_failure/$recipient_verify_failure to "random".
877 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
880 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
881 Previously this would segfault.
883 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
886 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
887 like zero, since the resolver should be doing this for us, But we need one
888 as a CNAME but no MX presence gets the CNAME returned; we need to check
889 that doesn't point to an MX to declare it "no MX returned" rather than
890 "error, loop". A new main option is added so the older capability of
891 following some limited number of chain links is maintained.
893 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
895 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
896 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
897 now-deprecated earlier definitions used only specified the range up to TLS
898 1.2 (in the older-version library docs).
900 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
902 JH/15 Rework TLS client-side context management. Stop using a global, and
903 explicitly pass a context around. This enables future use of TLS for
904 connections to service-daemons (eg. malware scanning) while a client smtp
905 connection is using TLS; with cutthrough connections this is quite likely.
907 JH/16 Fix ARC verification to do AS checks in reverse order.
909 JH/17 Support a "tls" option on the ${readsocket } expansion item.
911 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
912 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
913 Previously the "utf8" would be re-prepended for every additional message.
915 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
916 Previously thery were accepted, resulting in issues when attempting to
917 forward messages to a non-supporting MTA.
919 PP/02 Let -n work with printing macros too, not just options.
921 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
922 one parent address was copied, and bogus data was used at delivery-logging
923 time. Either a crash (after delivery) or bogus log data could result.
924 Discovery and analysis by Tim Stewart.
926 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
927 Previously if the string ended mid-character, we did not insert the
928 promised '?' replacement.
930 PP/04 Documentation: current string operators work on bytes, not codepoints.
932 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
933 should pack well giving a smaller memory footprint so better caching and
934 therefore performance. Group the declarations where this can't be done so
935 that the byte-sized flag variables are not interspersed among pointer
936 variables, giving a better chance of good packing by the compiler.
938 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
939 non-null, to avoid issues with sites running BATV. Previously reports were
940 sent with an empty envelope sender so looked like bounces.
942 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
943 The ignore_error flag wasn't being returned from the filter subprocess so
944 was not set for later routers. Investigation and fix by Matthias Kurz.
946 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
947 and a msg:complete for the whole, when a message is manually removed using
948 -Mrm. Developement by Matthias Kurz, hacked on by JH.
950 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
951 a "Gnu special" function, asprintf() in the DB utility binary builds; I
952 hope that is portable enough.
954 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
955 requiring a known-CA anchor certificate; make it now rely entirely on the
956 TLSA as an anchor. Checking the name on the leaf cert against the name
957 on the A-record for the host is still done for TA (but not for EE mode).
959 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
960 included in delivery lines for non-proxied connections, when compiled with
961 SUPPORT_SOCKS and running with proxy logging enabled.
963 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
964 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
965 move the existing event to fire before the normal logging of message
966 failure so that custom logging is bracketed by normal logging.
968 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
969 msg:fail:internal event. Developement by Matthias Kurz.
971 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
972 far too small for todays use of crypto signatures stored there. Go all
973 the way to the max DNS message size of 64kB, even though this might be
974 overmuch for IOT constrained device use.
976 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
977 copy a string over itself. The library routine is documented as not
978 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
980 JH/32 For main options check_spool_space and check_inode_space, where the
981 platform supports 64b integers, support more than the previous 2^31 kB
982 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
983 the previous G, M, k.
985 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
986 $authenticated_fail_id variable on authentication failure. Previously
989 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
990 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
991 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
992 GNUTLS_SEC_PARAM_MEDIUM.
994 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
995 side. Previously we would continue as if no SNI had been received.
997 JH/36 Harden the handling of string-lists. When a list consisted of a sole
998 "<" character, which should be a list-separator specification, we walked
999 off past the nul-terimation.
1001 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1002 causes) even when the retry time is not yet met. Previously they were
1003 not, meaning that when (say) an account was over-quota and temp-rejecting,
1004 and multiple senders' messages were queued, only one sender would get
1005 notified on each configured delay_warning cycle.
1007 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1009 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1010 cipher-suites, an error can be left on the stack even for a succeeding
1011 accept; this results in impossible error messages when a later operation
1014 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1015 return error codes indicating retry. Under TLS1.3 this becomes required.
1017 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1018 it only wrote the new authenticators, resulting in a lack of tracking of
1019 peer changes of ESMTP extensions until the next cache flush.
1021 JH/41 Fix the loop reading a message header line to check for integer overflow,
1022 and more-often against header_maxsize. Previously a crafted message could
1023 induce a crash of the recive process; now the message is cleanly rejected.
1025 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1026 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1032 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1033 When redis_servers is set to a list of > 1 element, and the Redis servers
1034 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1035 case of MOVED into a DEFER case instead, thus moving the query onto the
1036 next server in the list. For a cluster of N elements, all N servers must
1037 be defined in redis_servers.
1039 GF/02 Catch and remove uninitialized value warning in exiqsumm
1040 Check for existence of @ARGV before looking at $ARGV[0]
1042 JH/01 Replace the store_release() internal interface with store_newblock(),
1043 which internalises the check required to safely use the old one, plus
1044 the allocate and data copy operations duplicated in both (!) of the
1045 extant use locations.
1047 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1048 modifier. This matches the restriction on the commandline.
1050 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1051 Previously only the last row was returned.
1053 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1054 we assumed that tags in the header were well-formed, and parsed the
1055 element content after inspecting only the first char of the tag.
1056 Assumptions at that stage could crash the receive process on malformed
1059 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1060 While running the DKIM ACL we operate on the Permanent memory pool so that
1061 variables created with "set" persist to the DATA ACL. Also (at any time)
1062 DNS lookups that fail create cache records using the Permanent pool. But
1063 expansions release any allocations made on the current pool - so a dnsdb
1064 lookup expansion done in the DKIM ACL releases the memory used for the
1065 DNS negative-cache, and bad things result. Solution is to switch to the
1066 Main pool for expansions.
1067 While we're in that code, add checks on the DNS cache during store_reset,
1068 active in the testsuite.
1069 Problem spotted, and debugging aided, by Wolfgang Breyha.
1071 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1072 When none of the hosts presented to a transport match an already-open
1073 connection, close it and proceed with the list. Previously we would
1074 queue the message. Spotted by Lena with Yahoo, probably involving
1077 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1078 Previously a spurious "250 OK id=" response was appended to the proper
1081 JH/08 The "support for" informational output now, which built with Content
1082 Scanning support, has a line for the malware scanner interfaces compiled
1083 in. Interface can be individually included or not at build time.
1085 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1086 by the template makefile "src/EDITME". The "STREAM" support for an older
1087 ClamAV interface method is removed.
1089 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1090 rows affected is given instead).
1092 JH/11 The runtime Berkeley DB library version is now additionally output by
1093 "exim -d -bV". Previously only the compile-time version was shown.
1095 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1096 SMTP connection. Previously, when one had more recipients than the
1097 first, an abortive onward connection was made. Move to full support for
1098 multiple onward connections in sequence, handling cutthrough connection
1099 for all multi-message initiating connections.
1101 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1102 routers. Previously, a multi-recipient message would fail to match the
1103 onward-connection opened for the first recipient, and cause its closure.
1105 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1106 a timeout on read on a GnuTLS initiating connection, resulting in the
1107 initiating connection being dropped. This mattered most when the callout
1108 was marked defer_ok. Fix to keep the two timeout-detection methods
1111 JH/15 Relax results from ACL control request to enable cutthrough, in
1112 unsupported situations, from error to silently (except under debug)
1113 ignoring. This covers use with PRDR, frozen messages, queue-only and
1116 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1118 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1119 metadata, resulting in a crash in free().
1121 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1122 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1123 Broken also in d185889f4, with init system revamp.
1125 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1126 Previously we abruptly closed the connection after reading a malware-
1127 found indication; now we go on to read the "scan ok" response line,
1130 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1131 ACL. Previously, a crash would result.
1133 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1134 macro text after a replacement (previously it was only once per line) and
1135 by skipping builtin macros when searching for an uppercase lead character.
1137 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1138 for the build is renamed.
1140 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1141 was allocated for every new TLS startup, meaning one per message. Fix
1142 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1144 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1145 reported the original. Fix to report (as far as possible) the ACL
1146 result replacing the original.
1148 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1149 OpenSSL. Certificate information is loaded for every new TLS startup,
1150 and the resources needed to be freed.
1152 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1154 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1157 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1158 DATA response info to the (existing) per-recipient response info for
1159 the "C=" log element. It can have useful tracking info from the
1160 destination system. Patch from Simon Arlott.
1162 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1163 length value. Previously this would segfault.
1165 HS/02 Support Avast multiline protoocol, this allows passing flags to
1166 newer versions of the scanner.
1168 JH/28 Ensure that variables possibly set during message acceptance are marked
1169 dead before release of memory in the daemon loop. This stops complaints
1170 about them when the debug_store option is enabled. Discovered specifically
1171 for sender_rate_period, but applies to a whole set of variables.
1172 Do the same for the queue-runner and queue-list loops, for variables set
1173 from spool message files. Do the same for the SMTP per-message loop, for
1174 certain variables indirectly set in ACL operations.
1176 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1177 as a multi-recipient message from a mailinglist manager). The coding had
1178 an arbitrary cutoff number of characters while checking for more input;
1179 enforced by writing a NUL into the buffer. This corrupted long / fast
1180 input. The problem was exposed more widely when more pipelineing of SMTP
1181 responses was introduced, and one Exim system was feeding another.
1182 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1183 receiving system, and refused recipients seen by the sending system
1184 (propating to people being dropped from mailing lists).
1185 Discovered and pinpointed by David Carter.
1187 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1188 replaced by the ${authresults } expansion.
1190 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1192 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1193 allows proper process termination in container environments.
1195 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1196 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1198 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1199 and "err_perm", deprecated since 4.83 when the RFC-defined words
1200 "temperror" and "permerror" were introduced.
1202 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1203 transport-filters or DKIM-signing. The restriction was lost in the
1204 consolidation of verify-callout and delivery SMTP handling.
1205 Extend the restriction to also cover ARC-signing.
1207 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1208 in defer=pass mode supply a 450 to the initiator. Previously the message
1211 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1212 tls_require_ciphers is used as before.
1214 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1215 "pass_unscanned". Only tmpfails from the scanner are written to
1216 the paniclog, as they may require admin intervention (permission
1217 denied, license issues). Other scanner errors (like decompression
1218 bombs) do not cause a paniclog entry.
1220 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1221 Previously it was possible to log spurious information in receive log
1224 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1225 triggered odd behaviour from Outlook Express clients.
1227 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1230 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1231 since the IETF WG has not yet settled on that versus the original
1232 "bare" representation.
1234 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1235 Previously the millisecond value corrupted the output.
1236 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1237 corrupted the output.
1243 JH/01 Rework error string handling in TLS interface so that the caller in
1244 more cases is responsible for logging. This permits library-sourced
1245 string to be attached to addresses during delivery, and collapses
1246 pairs of long lines into single ones.
1248 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1249 during configuration. Wildcards are allowed and expanded.
1251 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1252 This permits better logging.
1254 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1255 do not close it down and have the child transport start it up again on
1256 the passed-on TCP connection. Instead, proxy the child (and any
1257 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1258 affected: the continued delivery log lines do not have any DNSSEC, TLS
1259 Certificate or OCSP information. TLS cipher information is still logged.
1261 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1262 identical IP addresses on different listening ports. Will also affect
1265 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1266 add noisy ifdef guards to special-case this sillyness.
1267 Patch from Bernd Kuhls.
1269 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1270 than 255 are no longer allowed.
1272 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1273 Disable the session-cache too, which might reduce our load. Since we
1274 currrectly use a new context for every connection, both as server and
1275 client, there is no benefit for these.
1276 GnuTLS appears to not support tickets server-side by default (we don't
1277 call gnutls_session_ticket_enable_server()) but client side is enabled
1278 by default on recent versions (3.1.3 +) unless the PFS priority string
1281 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1282 <https://reproducible-builds.org/specs/source-date-epoch/>.
1284 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1285 the check for any unsuccessful recipients did not notice the limit, and
1286 erroneously found still-pending ones.
1288 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1289 MSG_MORE. Only in-clear (not on TLS connections).
1291 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1292 transport-filter is involved we can buffer the headers in memory for
1293 creating the signature, and read the spool data file once for the
1294 signature and again for transmission.
1296 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1297 4.77 as the kernel support then wasn't solid, having issues in 64bit
1298 mode. Now, it's been long enough. Add support for FreeBSD also.
1300 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1301 case where the routing stage had gathered several addresses to send to
1302 a host before calling the transport for the first, we previously failed
1303 to close down TLS in the old transport process before passing the TCP
1304 connection to the new process. The new one sent a STARTTLS command
1305 which naturally failed, giving a failed delivery and bloating the retry
1306 database. Investigation and fix prototype from Wolfgang Breyha.
1308 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1309 false-negatives in the check that the sender had not preempted a response
1310 or prompt from Exim (running as a server), due to that code's lack of
1311 awareness of the SMTP input buffering.
1313 PP/04 Add commandline_checks_require_admin option.
1314 Exim drops privileges sanely, various checks such as -be aren't a
1315 security problem, as long as you trust local users with access to their
1316 own account. When invoked by services which pass untrusted data to
1317 Exim, this might be an issue. Set this option in main configuration
1318 AND make fixes to the calling application, such as using `--` to stop
1321 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1322 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1323 responses to those, into a single TLS record each way (this usually means
1324 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1327 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1328 affects you only if you're dancing at the edge of the param size limits.
1329 If you are, and this message makes sense to you, then: raise the
1330 configured limit or use OpenSSL 1.1. Nothing we can do for older
1333 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1334 cmdline element to get the documented default one. Previously it was
1337 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1338 are unsuable for non-message-specific reasons - eg. connection timeout,
1339 banner-time rejection.
1341 JH/16 Fix logging of delivery remote port, when specified by router, under
1344 PP/06 Repair manualroute's ability to take options in any order, even if one
1345 is the name of a transport.
1348 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1350 JH/17 Change the list-building routines interface to use the expanding-string
1351 triplet model, for better allocation and copying behaviour.
1353 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1354 Previously it was constructed the first time a possibly-matching string
1355 was met in the configuration file input during startup; now it is done
1358 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1359 compatible one, to avoid the (poorly documented) possibility of a config
1360 file in the working directory redirecting the DB files, possibly correpting
1361 some existing file. CVE-2017-10140 assigned for BDB.
1363 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1364 cache-hot. Previously, although the result was properly cached, the
1365 initial verify call returned a defer.
1367 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1368 the main verify for receipient in uncached-mode.
1370 JH/22 Retire historical build files to an "unsupported" subdir. These are
1371 defined as "ones for which we have no current evidence of testing".
1373 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1374 if present. Previously it was ignored.
1376 JH/24 Start using specified-initialisers in C structure init coding. This is
1377 a C99 feature (it's 2017, so now considered safe).
1379 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1380 if was a fixed-sized field and bitmask ops via macros; it is now more
1383 PP/07 GitHub PR 56: Apply MariaDB build fix.
1384 Patch provided by Jaroslav Škarvada.
1386 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1387 during Coverity cleanups [4.87 JH/47]
1388 Diagnosis and fix provided by Michael Fischer v. Mollard.
1390 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1391 the right size to place the terminating semicolon on its own folded
1392 line, the header hash was calculated to an incorrect value thanks to
1393 the (relaxed) space the fold became.
1395 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1396 and confused the parent.
1398 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1399 which could crash as a result. This could lead to undeliverable messages.
1401 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1404 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1405 matters on fast-turnover and PID-randomising systems, which were getting
1406 out-of-order delivery.
1408 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1409 a possibly-overlapping copy. The symptom was that "Remote host closed
1410 connection in response to HELO" was logged instead of the actual 4xx
1413 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1414 Previously only that bufferd was discarded, resulting in SYMTP command
1417 JH/32 DKIM: when a message has multiple signatures matching an identity given
1418 in dkim_verify_signers, run the dkim acl once for each. Previously only
1419 one run was done. Bug 2189.
1421 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1422 "panic the current process" to "deliberately defer". The panic log is
1423 still written with the problem list name; the mail and reject logs now
1424 get a temp-reject line for the message that was being handled, saying
1425 something like "domains check lookup or other defer". The SMTP 451
1426 message is still "Temporary local problem".
1428 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1429 A crafted sequence of BDAT commands could result in in-use memory beeing
1430 freed. CVE-2017-16943.
1432 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1433 from SMTP input. Previously it was always done; now only done for DATA
1434 and not BDAT commands. CVE-2017-16944.
1436 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1437 to the message (such as an overlong header line). Previously this was
1438 not done and we did not exit BDAT mode. Followon from the previous item
1439 though a different problem.
1445 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1446 than -2003 did; needs libidn2 in addition to libidn.
1448 JH/02 The path option on a pipe transport is now expanded before use.
1450 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1451 Patch provided by "Björn", documentation fix added too.
1453 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1454 missing a wire-to-host endian conversion.
1456 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1457 close after a BDAT command line could be taken as a following command,
1458 giving a synch failure. Fix by only checking for synch immediately
1459 before acknowledging the chunk.
1461 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1462 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1463 macro. Patches provided by Josh Soref.
1465 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1466 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1467 by IANA as a service extension. However, John Klensin suggests that we
1470 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1471 the dkim code may be unix-mode line endings rather than smtp wire-format
1472 CRLF, so prepend a CR to any bare LF.
1474 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1475 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1477 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1478 the first were themselves being wrongly included in the feed into dkim
1479 processing; with most chunk sizes in use this resulted in an incorrect
1480 body hash calculated value.
1482 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1483 DKIM signature block, for verification. Although advised against by
1484 standards it is specifically not ruled illegal.
1486 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1488 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1489 missing a body hash (the bh= tag).
1491 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1492 It seems that HAProxy sends the Proxy Protocol information in clear and
1493 only then does a TLS startup, so do the same.
1495 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1496 TCP connections (such as for Spamd) unless the daemon successfully set
1497 Fast Open mode on its listening sockets. This fixes breakage seen on
1498 too-old kernels or those not configured for Fast Open, at the cost of
1499 requiring both directions being enabled for TFO, and TFO never being used
1500 by non-daemon-related Exim processes.
1502 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1503 endings, at least on the first header line. Try to canonify any that get
1504 past that check, despite the cost.
1506 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1507 now limited to an arbitrary five deep, while parsing addresses with the
1508 strip_excess_angle_brackets option enabled.
1510 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1511 instead leave the unprompted TLS handshake in socket buffer for the
1512 TLS library to consume.
1514 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1516 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1518 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1519 region in one operation, for speed, and this leaves assigned pointers
1520 dangling. Add checks run only under the testsuite which checks all
1521 variables at a store-reset and panics on a dangling pointer; add code
1522 explicitly nulling out all the variables discovered. Fixes one known
1523 bug: a transport crash, where a dangling pointer for $sending_ip_address
1524 originally assigned in a verify callout, is re-used.
1526 PP/06 Drop '.' from @INC in various Perl scripts.
1528 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1530 PP/08 Reduce a number of compilation warnings under clang; building with
1531 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1532 should be warning-free.
1534 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1536 HS/01 Fix portability problems introduced by PP/08 for platforms where
1537 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1539 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1540 chunk. This allows us to accept broken chunked messages. We need a more
1541 general solution here.
1543 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1544 already-broken messages in the queue.
1546 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1548 JH/19 Fix reference counting bug in routing-generated-address tracking.
1554 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1555 supports it and a size is available (ie. the sending peer gave us one).
1557 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1558 years of being deprecated). The replacements are the ACLs
1559 acl_smtp_mime and acl_not_smtp_mime.
1561 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1562 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1563 or even an in-clear connection were permitted. Now, if the host lookup
1564 was dnssec and dane was requested then the host is only used if the
1565 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1566 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1567 if one fails this test.
1568 This means that a poorly-configured remote DNS will make it incommunicado;
1569 but it protects against a DNS-interception attack on it.
1571 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1572 non-noisy when a race steals the message being considered.
1574 JH/05 If main configuration option tls_certificate is unset, generate a
1575 self-signed certificate for inbound TLS connections.
1577 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1578 in rewrites and routers.
1580 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1581 and logged a warning sing 4.83; now they are a configuration file error.
1583 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1584 (lacking @domain). Apply the same qualification processing as RCPT.
1586 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1588 JH/10 Support ${sha256:} applied to a string (as well as the previous
1591 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1592 a cutthrough deliver is pending, as we always want to make a connection.
1593 This also avoids re-routing the message when later placing the cutthrough
1594 connection after a verify cache hit.
1595 Do not update it with the verify result either.
1597 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1598 when routing results in more than one destination address.
1600 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1601 signing (which inhibits the cutthrough capability). Previously only
1602 the presence of an option was tested; now an expansion evaluating as
1603 empty is permissible (obviously it should depend only on data available
1604 when the cutthrough connection is made).
1606 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1607 the relevant preceding SMTP command did not note the pipelining mode.
1609 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1610 Previously they were not counted.
1612 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1613 as one having no matching records. Previously we deferred the message
1614 that needed the lookup.
1616 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1617 distinguished as "(=".
1619 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1620 for missing MX records. Previously it only worked for missing A records.
1622 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1624 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1625 after the data-go-ahead and data-ack. Patch from Jason Betts.
1627 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1628 even for a "none" policy. Patch from Tony Meyer.
1630 JH/22 Fix continued use of a connection for further deliveries. If a port was
1631 specified by a router, it must also match for the delivery to be
1634 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1635 When one of the recipients of a message was unsuitable for the connection
1636 (has no matching addresses), we lost track of needing to mark it
1637 deferred. As a result mail would be lost.
1639 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1641 JH/25 Decoding ACL controls is now done using a binary search; the source code
1642 takes up less space and should be simpler to maintain. Merge the ACL
1643 condition decode tables also, with similar effect.
1645 JH/26 Fix problem with one_time used on a redirect router which returned the
1646 parent address unchanged. A retry would see the parent address marked as
1647 delivered, so not attempt the (identical) child. As a result mail would
1650 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1651 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1652 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1655 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1656 Main config options check_{log,spool}_{inodes,space} are now
1657 100 inodes, 10MB unless set otherwise in the configuration.
1659 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1660 Previously it only applied to the main-section connection policy
1663 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1665 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1666 by me. Added RFC7919 DH primes as an alternative.
1668 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1669 are not in the system include path.
1671 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1672 GnuTLS, when a session startup failed (eg because the client disconnected)
1673 Exim did stdio operations after fclose. This was exposed by a recent
1674 change which nulled out the file handle after the fclose.
1676 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1677 signed directly by the cert-signing cert, rather than an intermediate
1678 OCSP-signing cert. This is the model used by LetsEncrypt.
1680 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1682 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1683 an incoming connection.
1685 HS/02 Bug 1802: Do not half-close the connection after sending a request
1688 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1689 fallback to "prime256v1".
1691 JH/34 SECURITY: Use proper copy of DATA command in error message.
1692 Could leak key material. Remotely exploitable. CVE-2016-9963.
1698 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1699 and 3.4.4 - once the server is enabled to respond to an OCSP request
1700 it does even when not requested, resulting in a stapling non-aware
1701 client dropping the TLS connection.
1703 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1704 support variable-length bit vectors. No functional change.
1706 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1707 The I= interface field on outgoing lines is now after the H= remote
1708 host field, same as incoming lines. There is a separate
1709 outgoing_interface log selector which allows you to disable the
1712 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1713 If not running log_selector +smtp_connection the mainlog would be held
1714 open indefinitely after a "too many connections" event, including to a
1715 deleted file after a log rotate. Leave the per net connection logging
1716 leaving it open for efficiency as that will be quickly detected by the
1717 check on the next write.
1719 HS/01 Bug 1671: Fix post transport crash.
1720 Processing the wait-<transport> messages could crash the delivery
1721 process if the message IDs didn't exist for some reason. When
1722 using 'split_spool_directory=yes' the construction of the spool
1723 file name failed already, exposing the same netto behaviour.
1725 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1726 mime_regex ACL conditions.
1728 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1729 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1730 to HELO, local diagnostic string.
1732 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1733 log line to debug. Even when configured with a tls authenticator many
1734 client connections are expected to not authenticate in this way, so
1735 an authenticate fail is not an error.
1737 HS/02 Add the Exim version string to the process info. This way exiwhat
1738 gives some more detail about the running daemon.
1740 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1741 matter for fast-change records such as DNSBLs.
1743 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1744 retry record, even if constant. There may be multiple transports with
1745 different interface settings and the retry behaviour needs to be kept
1748 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1750 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1752 JH/10 Bug 840: fix log_defer_output option of pipe transport
1754 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1755 hosts_randomize. This matters a lot when combined with mua_wrapper.
1757 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1758 ${quote_pgsql:<string>} operator.
1760 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1762 JH/14 Change header file name for freeradius-client. Relevant if compiling
1763 with Radius support; from the Gentoo tree and checked under Fedora.
1765 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1767 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1768 option transport_filter, meaning no filtering.
1770 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1772 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1773 defaults to "*" (all hosts). The variable is now available when not built
1774 with TLS, default unset, mainly to enable keeping the testsuite sane.
1775 If a server certificate is not supplied (via tls_certificate) an error is
1776 logged, and clients will find TLS connections fail on startup. Presumably
1777 they will retry in-clear.
1778 Packagers of Exim are strongly encouraged to create a server certificate
1779 at installation time.
1781 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1782 with the $config_file variable.
1784 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1785 in transport context, after the attempt, and per-recipient. The latter type
1786 is per host attempted. The event data is the error message, and the errno
1787 information encodes the lookup type (A vs. MX) used for the (first) host,
1788 and the trailing two digits of the smtp 4xx response.
1790 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1791 to write to mainlog (or rejectlog, paniclog) in the window between file
1792 creation and permissions/ownership being changed. Particularly affects
1793 installations where exicyclog is run as root, rather than exim user;
1794 result is that the running daemon panics and dies.
1796 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1798 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1799 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1800 "pri" and "weight". Note that the previous implicit priority given by the
1801 list order is no longer honoured.
1803 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1804 for DKIM processing.
1806 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1807 by defining SUPPORT_SOCKS.
1809 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1810 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1811 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1812 variables are renamed to proxy_{local,external}_{address,port}.
1814 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1815 for a build by defining SUPPORT_I18N
1817 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1818 of the query string, and make ${quote_redis:} do that quoting.
1820 JH/29 Move Events support from Experimental to mainline, enabled by default
1821 and removable for a build by defining DISABLE_EVENT.
1823 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1825 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1826 cached by the daemon.
1828 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1829 by defining LOOKUP_REDIS. The libhiredis library is required.
1831 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1832 keys are given for lookup.
1834 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1835 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1836 only supported when built with TLS support. The PolarSSL SHA routines
1837 are still used when the TLS library is too old for convenient support.
1839 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1840 openssl_options), for security. OpenSSL forces this from version 1.1.0
1841 server-side so match that on older versions.
1843 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1844 allocation for $value could be released as the expansion processing
1845 concluded, but leaving the global pointer active for it.
1847 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1848 and to use the domains and local_parts ACL conditions.
1850 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1851 incorrectly not doubled on cutthrough transmission, hence seen as a
1852 body-termination at the receiving system - resulting in truncated mails.
1853 Commonly the sender saw a TCP-level error, and retransmitted the message
1854 via the normal store-and-forward channel. This could result in duplicates
1855 received - but deduplicating mailstores were liable to retain only the
1856 initial truncated version.
1858 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1860 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1862 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1863 we're in there, support oversigning also; bug 1309.
1865 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1867 HS/04 Add support for keep_environment and add_environment options.
1869 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1870 either intentional arithmetic overflow during PRNG, or testing config-
1873 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1874 delivery resulted in actual delivery. Cancel cutthrough before DATA
1877 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1878 rejects a recipient: pass the reject to the originator.
1880 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1881 Many were false-positives and ignorable, but it's worth fixing the
1884 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1885 for the new environment-manipulation done at startup. Move the routines
1886 from being local to tls.c to being global via the os.c file.
1888 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1889 an extract embedded as result-arg for a map, the first arg for extract
1890 is unavailable so we cannot tell if this is a numbered or keyed
1891 extraction. Accept either.
1897 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1900 JH/02 The smtp transport option "multi_domain" is now expanded.
1902 JH/03 The smtp transport now requests PRDR by default, if the server offers
1905 JH/04 Certificate name checking on server certificates, when exim is a client,
1906 is now done by default. The transport option tls_verify_cert_hostnames
1907 can be used to disable this per-host. The build option
1908 EXPERIMENTAL_CERTNAMES is withdrawn.
1910 JH/05 The value of the tls_verify_certificates smtp transport and main options
1911 default to the word "system" to access the system default CA bundle.
1912 For GnuTLS, only version 3.0.20 or later.
1914 JH/06 Verification of the server certificate for a TLS connection is now tried
1915 (but not required) by default. The verification status is now logged by
1916 default, for both outbound TLS and client-certificate supplying inbound
1919 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1922 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1923 Status Notification (bounce) messages are now MIME format per RFC 3464.
1924 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1925 under the control of the dsn_advertise_hosts option, and routers may
1926 have a dsn_lasthop option.
1928 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1929 default, modifiable by a malware= option. The list separator for
1930 the options can now be changed in the usual way. Bug 68.
1932 JH/10 The smtp_receive_timeout main option is now expanded before use.
1934 JH/11 The incoming_interface log option now also enables logging of the
1935 local interface on delivery outgoing connections.
1937 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1938 if the interface and destination host and port all match.
1940 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1943 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1944 Patch from Andrew Lewis.
1946 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1947 now supports optional time-restrictions, weighting, and priority
1948 modifiers per server. Patch originally by <rommer@active.by>.
1950 JH/16 The spamd_address main option now supports a mixed list of local
1951 and remote servers. Remote servers can be IPv6 addresses, and
1952 specify a port-range.
1954 JH/17 Bug 68: The spamd_address main option now supports an optional
1955 timeout value per server.
1957 JH/18 Bug 1581: Router and transport options headers_add/remove can
1958 now have the list separator specified.
1960 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1963 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1966 JH/21 Support for the A6 type of dns record is withdrawn.
1968 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1969 rather than the verbs used.
1971 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1972 from 255 to 1024 chars.
1974 JH/24 Verification callouts now attempt to use TLS by default.
1976 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1977 are generic router options now. The defaults didn't change.
1979 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1980 Original patch from Alexander Shikoff, worked over by JH.
1982 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1983 files if ZCAT_COMMAND is not executable.
1985 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1987 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1989 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1990 Normally benign, it bites when the pair was led to by a CNAME;
1991 modern usage is to not canonicalize the domain to a CNAME target
1992 (and we were inconsistent anyway for A-only vs AAAA+A).
1994 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1996 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1997 when evaluating $sender_host_dnssec.
1999 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2000 $sender_helo_dnssec variable.
2002 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2004 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2006 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2008 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2009 documented as working, but never had. Support all but $spam_report.
2011 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2012 added for tls authenticator.
2014 HS/03 Add perl_taintmode main config option
2020 TL/01 When running the test suite, the README says that variables such as
2021 no_msglog_check are global and can be placed anywhere in a specific
2022 test's script, however it was observed that placement needed to be near
2023 the beginning for it to behave that way. Changed the runtest perl
2024 script to read through the entire script once to detect and set these
2025 variables, reset to the beginning of the script, and then run through
2026 the script parsing/test process like normal.
2028 TL/02 The BSD's have an arc4random API. One of the functions to induce
2029 adding randomness was arc4random_stir(), but it has been removed in
2030 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2031 function when detected.
2033 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2034 cause callback expansion.
2036 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2037 syntax errors in an expansion can be treated as a string instead of
2038 logging or causing an error, due to the internal use of bool_lax
2039 instead of bool when processing it.
2041 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2042 server certificates when making smtp deliveries.
2044 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2046 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2048 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2050 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2051 Merged patch from Sebastian Wiedenroth.
2053 JH/05 Fix results-pipe from transport process. Several recipients, combined
2054 with certificate use, exposed issues where response data items split
2055 over buffer boundaries were not parsed properly. This eventually
2056 resulted in duplicates being sent. This issue only became common enough
2057 to notice due to the introduction of connection certificate information,
2058 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2060 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2061 size buffer was used, resulting in syntax errors when an expansion
2064 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2065 version 3.3.6 or later.
2067 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2068 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2069 both become "event_action", the variables become $event_name, $event_data
2070 and $event_defer_errno. There is a new variable $verify_mode, usable in
2071 routers, transports and related events. The tls:cert event is now also
2072 raised for inbound connections, if the main configuration event_action
2075 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2076 early OCSP support, but no stapling (appears to be less than 1.0.0).
2078 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2079 server certificate names available under the smtp transport option
2080 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2083 JH/10 Time-related extraction expansions from certificates now use the main
2084 option "timezone" setting for output formatting, and are consistent
2085 between OpenSSL and GnuTLS compilations. Bug 1541.
2087 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2088 encoded parameter in the incoming message. Bug 1558.
2090 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2091 include certificate info, eximon was claiming there were spoolfile
2094 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2096 JH/14 Log delivery-related information more consistently, using the sequence
2097 "H=<name> [<ip>]" wherever possible.
2099 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2100 are problematic for Debian distribution, omit them from the release
2103 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2105 JH/16 Fix string representation of time values on 64bit time_t architectures.
2108 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2109 output list separator was used.
2114 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2115 checkers that were complaining about end of non-void function with no
2118 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2119 This was a regression introduced in 4.83 by another bugfix.
2121 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2123 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2124 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2130 TF/01 Correctly close the server side of TLS when forking for delivery.
2132 When a message was received over SMTP with TLS, Exim failed to clear up
2133 the incoming connection properly after forking off the child process to
2134 deliver the message. In some situations the subsequent outgoing
2135 delivery connection happened to have the same fd number as the incoming
2136 connection previously had. Exim would try to use TLS and fail, logging
2137 a "Bad file descriptor" error.
2139 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2140 utilities have not been installed.
2142 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2143 temporary space as the ACL may create new global variables.
2145 TL/01 LDAP support uses per connection or global context settings, depending
2146 upon the detected version of the libraries at build time.
2148 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2149 to extract and use the src ip:port in logging and expansions as if it
2150 were a direct connection from the outside internet. PPv2 support was
2151 updated based on HAProxy spec change in May 2014.
2153 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2155 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2156 Properly escape header and check for NULL return.
2158 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2161 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2163 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2164 characters in header names, implemented as a verify condition.
2165 Contributed by Michael Fischer v. Mollard.
2167 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2168 results permerror and temperror. Previous values are deprecated but
2169 still accepted. In a future release, err_perm and err_temp will be
2170 completely removed, which will be a backward incompatibility if the
2171 ACL tests for either of these two old results. Patch contributed by
2172 user bes-internal on the mailing list.
2174 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2176 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2177 selectors, in both main and reject logs.
2179 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2182 JH/07 Add malware type "sock" for talking to simple daemon.
2184 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2186 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2187 routers/transports under cutthrough routing.
2189 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2190 numbers. Touch up "bool" conditional to keep the same definition.
2192 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2194 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2196 JH/12 Expand items in router/transport headers_add or headers_remove lists
2197 individually rather than the list as a whole. Bug 1452.
2199 Required for reasonable handling of multiple headers_ options when
2200 they may be empty; requires that headers_remove items with embedded
2201 colons must have them doubled (or the list-separator changed).
2203 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2204 view the policy declared in the DMARC record. Currently, $dmarc_status
2205 is a combined value of both the record presence and the result of the
2208 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2210 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2211 dnslookup router and the smtp transport (applying to the forward
2214 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2215 of ldap servers used for a specific lookup. Patch provided by Heiko
2218 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2219 New variable $lookup_dnssec_authenticated for observability.
2221 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2222 Patch submitted by Lars Timman.
2224 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2226 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2227 Requires trusted mode and valid format message id, aborts otherwise.
2228 Patch contributed by Heiko Schlichting.
2230 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2231 certextract with support for various fields. Bug 1358.
2233 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2234 is requested by default, modifiable by smtp transport option
2237 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2238 operate on certificate variables to give certificate fingerprints
2239 Also new ${sha256:cert_variable}.
2241 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2243 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2244 Christian Aistleitner.
2246 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2248 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2249 file. Patch from Wolfgang Breyha.
2251 JH/25 Expand the coverage of the delivery $host and $host_address to
2252 client authenticators run in verify callout. Bug 1476.
2254 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2255 align with daemon_smtp_ports. Bug 72.
2257 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2258 support and error reporting did not work properly.
2260 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2261 and is readable. Patch from Andrew Colin Kissa.
2263 TL/14 Enhance documentation of ${run expansion and how it parses the
2264 commandline after expansion, particularly in the case when an
2265 unquoted variable expansion results in an empty value.
2267 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2269 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2270 by renaming away from it. Addresses GH issue 15, reported by
2273 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2275 TL/15 SECURITY: prevent double expansion in math comparison functions
2276 (can expand unsanitized data). Not remotely exploitable.
2283 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2285 PP/02 Make -n do something, by making it not do something.
2286 When combined with -bP, the name of an option is not output.
2288 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2291 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2292 $sender_host_name and config options to manage this, and basic check
2295 PP/05 DSCP support for outbound connections and control modifier for inbound.
2297 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2298 (Only plugin which currently uses this is kerberos4, which nobody should
2299 be using, but we should make it available and other future plugins might
2300 conceivably use it, even though it would break NAT; stuff *should* be
2301 using channel bindings instead).
2303 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2304 name; added for Sendmail compatibility; requires admin caller.
2305 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2306 just ignore it); requires trusted caller.
2307 Also parse but ignore: -Ac -Am -X<logfile>
2310 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2312 TL/02 Add +smtp_confirmation as a default logging option.
2314 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2315 Patch by Magnus Holmgren from 2007-02-20.
2317 TL/04 Bugzilla 1281 - Spec typo.
2318 Bugzilla 1283 - Spec typo.
2319 Bugzilla 1290 - Spec grammar fixes.
2321 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2323 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2325 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2326 mailing list by by Dmitry Isaikin.
2328 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2330 JH/02 Support "G" suffix to numbers in ${if comparisons.
2332 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2334 NM/01 Bugzilla 1197 - Spec typo
2335 Bugzilla 1196 - Spec examples corrections
2337 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2339 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2340 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2343 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2344 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2346 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2347 "acl {{name}{arg}...}", and optional args on acl condition
2350 JH/05 Permit multiple router/transport headers_add/remove lines.
2352 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2354 JH/07 Avoid using a waiting database for a single-message-only transport.
2355 Performance patch from Paul Fisher. Bugzilla 1262.
2357 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2360 JH/09 Add $headers_added variable, with content from use of ACL modifier
2361 add_header (but not yet added to the message). Bugzilla 199.
2363 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2364 Pulled from Bugzilla 817 by Wolfgang Breyha.
2366 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2368 (nb: this is the same fix as in Exim 4.80.1)
2370 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2373 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2376 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2378 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2379 advertise SMTP AUTH mechanism to us, instead of a generic
2380 protocol violation error. Also, make Exim more robust to bad
2381 data from the Dovecot auth socket.
2383 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2385 When a queue runner is handling a message, Exim first routes the
2386 recipient addresses, during which it prunes them based on the retry
2387 hints database. After that it attempts to deliver the message to
2388 any remaining recipients. It then updates the hints database using
2391 So if a recipient address works intermittently, it can get repeatedly
2392 deferred at routing time. The retry hints record remains fresh so the
2393 address never reaches the final cutoff time.
2395 This is a fairly common occurrence when a user is bumping up against
2396 their storage quota. Exim had some logic in its local delivery code
2397 to deal with this. However it did not apply to per-recipient defers
2398 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2400 This change adds a proper retry rule check during routing so that the
2401 final cutoff time is checked against the message's age. We only do
2402 this check if there is an address retry record and there is not a
2403 domain retry record; this implies that previous attempts to handle
2404 the address had the retry_use_local_parts option turned on. We use
2405 this as an approximation for the destination being like a local
2406 delivery, as in LMTP.
2408 I suspect this new check makes the old local delivery cutoff check
2409 redundant, but I have not verified this so I left the code in place.
2411 TF/02 Correct gecos expansion when From: is a prefix of the username.
2413 Test 0254 submits a message to Exim with the header
2417 When I ran the test suite under the user fanf2, Exim expanded
2418 the header to contain my full name, whereas it should have added
2419 a Resent-Sender: header. It erroneously treats any prefix of the
2420 username as equal to the username.
2422 This change corrects that bug.
2424 GF/01 DCC debug and logging tidyup
2425 Error conditions log to paniclog rather than rejectlog.
2426 Debug lines prefixed by "DCC: " to remove any ambiguity.
2428 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2430 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2431 Bug spotted by Jeremy Harris; was flawed since initial commit.
2432 Would have resulted in OCSP responses post-SNI triggering an Exim
2433 NULL dereference and crash.
2435 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2437 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2438 Bug detection, analysis and fix by Samuel Thibault.
2439 Bugzilla 1331, Debian bug #698092.
2441 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2443 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2444 Server implementation by Todd Lyons, client by JH.
2445 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2446 config variable "prdr_enable" controls whether the server
2447 advertises the facility. If the client requests PRDR a new
2448 acl_data_smtp_prdr ACL is called once for each recipient, after
2449 the body content is received and before the acl_smtp_data ACL.
2450 The client is controlled by both of: a hosts_try_prdr option
2451 on the smtp transport, and the server advertisement.
2452 Default client logging of deliveries and rejections involving
2453 PRDR are flagged with the string "PRDR".
2455 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2456 fclose(). Diagnosis by Todd Lyons.
2458 PP/17 Update configure.default to handle IPv6 localhost better.
2459 Patch by Alain Williams (plus minor tweaks).
2462 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2463 This is now consistent with GnuTLS, and is now documented: the
2464 previous undocumented portable approach to treating the option as
2465 unset was to force an expansion failure. That still works, and
2466 an empty string is now equivalent.
2468 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2469 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2470 not performing validation itself.
2472 PP/20 Added force_command boolean option to pipe transport.
2473 Patch from Nick Koston, of cPanel Inc.
2475 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2478 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2480 PP/21 Fix eximon continuous updating with timestamped log-files.
2481 Broken in a format-string cleanup in 4.80, missed when I repaired the
2482 other false fix of the same issue.
2483 Report and fix from Heiko Schlichting.
2486 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2487 Report from Prashanth Katuri.
2489 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2490 It's SecureTransport, so affects any MacOS clients which use the
2491 system-integrated TLS libraries, including email clients.
2493 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2494 using a MIME ACL for non-SMTP local injection.
2495 Report and assistance in diagnosis by Warren Baker.
2497 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2499 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2501 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2502 last id that failed so it may be referenced in subsequent ACL's.
2504 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2507 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2508 ldap library initialization, allowing self-signed CA's to be
2509 used. Also properly sets require_cert option later in code by
2510 using NULL (global ldap config) instead of ldap handle (per
2511 session). Bug diagnosis and testing by alxgomz.
2513 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2514 the src/util/ subdirectory.
2516 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2517 renamed to Transport Post Delivery Action by Jeremy Harris, as
2520 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2521 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2522 redis_servers = needs to be configured which will be used by the redis
2523 lookup. Patch from Warren Baker, of The Packet Hub.
2525 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2527 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2528 hostname or reverse DNS when processing a host list. Used suggestions
2529 from multiple comments on this bug.
2531 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2533 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2534 Missed a few lines, added it to make the runtest require no keyboard
2537 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2538 contains upper case chars. Make router use caseful_local_part.
2540 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2541 support when GnuTLS has been built with p11-kit.
2547 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2549 This, or similar/improved, will also be change PP/11 of 4.82.
2555 PP/01 Handle short writes when writing local log-files.
2556 In practice, only affects FreeBSD (8 onwards).
2557 Bugzilla 1053, with thanks to Dmitry Isaikin.
2559 NM/01 Bugzilla 949 - Documentation tweak
2561 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2564 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2566 PP/02 Implemented gsasl authenticator.
2568 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2570 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2571 `pkg-config foo` for cflags/libs.
2573 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2574 with rest of GSASL and with heimdal_gssapi.
2576 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2577 `pkg-config foo` for cflags/libs for the TLS implementation.
2579 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2580 properties get this fed in as external SSF. A number of robustness
2581 and debugging improvements to the cyrus_sasl authenticator.
2583 PP/08 cyrus_sasl server now expands the server_realm option.
2585 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2586 Patch by Jeremy Harris.
2588 PP/10 Added dbmjz lookup type.
2590 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2592 PP/12 MAIL args handles TAB as well as SP, for better interop with
2593 non-compliant senders.
2594 Analysis and variant patch by Todd Lyons.
2596 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2597 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2598 Patch from Dirk Mueller <dmueller@suse.com>
2600 PP/13 tls_peerdn now print-escaped for spool files.
2601 Observed some $tls_peerdn in wild which contained \n, which resulted
2602 in spool file corruption.
2604 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2605 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2606 or write after TLS renegotiation, which otherwise led to messages
2609 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2610 as a tracking header (ie: a signed header comes before the signature).
2611 Patch from Wolfgang Breyha.
2613 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2614 comma-sep list; embedded commas doubled.
2616 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2618 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2620 Report and patch from Dmitry Banschikov.
2622 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2623 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2624 failures. We appear to now support TLS1.1+ with Exim.
2626 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2627 lets Exim select keys and certificates based upon TLS SNI from client.
2628 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2629 before an outbound SMTP session. New log_selector, +tls_sni.
2631 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2632 NULL dereference. Report and patch from Alun Jones.
2634 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2635 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2636 Not seeing resolver debug output on NetBSD, but suspect this is a
2637 resolver implementation change.
2639 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2640 Left warnings. Added "eximon gdb" invocation mode.
2642 PP/21 Defaulting "accept_8bitmime" to true, not false.
2644 PP/22 Added -bw for inetd wait mode support.
2646 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2647 locate the relevant includes and libraries. Made this the default.
2649 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2650 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2652 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2653 This may cause build issues on older platforms.
2655 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2656 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2657 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2658 Added SNI support via GnuTLS too.
2659 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2661 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2663 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2664 Applied second patch from Janne, implementing suggestion to default
2665 multiple-strings-in-record handling to match SPF spec.
2667 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2669 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2670 read-only, out of scope).
2671 Patch from Wolfgang Breyha, report from Stuart Northfield.
2673 PP/29 Fix three issues highlighted by clang analyser static analysis.
2674 Only crash-plausible issue would require the Cambridge-specific
2675 iplookup router and a misconfiguration.
2676 Report from Marcin Mirosław.
2678 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2680 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2681 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2682 As part of this, removing so much warning spew let me fix some minor
2683 real issues in debug logging.
2685 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2686 assignment on my part. Fixed.
2688 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2689 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2690 Janne Snabb (who went above and beyond: thank you).
2692 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2693 string otherwise requires a connection and a bunch more work and it's
2694 relatively easy to get wrong. Should also expose TLS library linkage
2697 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2698 64-bit ${eval} (JH/03).
2700 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2701 GNU libc to support some of the 64-bit stuff, should not lead to
2702 conflicts. Defined before os.h is pulled in, so if a given platform
2703 needs to override this, it can.
2705 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2706 protection layer was required, which is not implemented.
2707 Bugzilla 1254, patch from Wolfgang Breyha.
2709 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2710 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2711 tls_dhparam take prime identifiers. Also unbreak combination of
2712 OpenSSL+DH_params+TLSSNI.
2714 PP/39 Disable SSLv2 by default in OpenSSL support.
2720 PP/01 Solaris build fix for Oracle's LDAP libraries.
2721 Bugzilla 1109, patch from Stephen Usher.
2723 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2725 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2728 TF/02 Fix a couple more cases where we did not log the error message
2729 when unlink() failed. See also change 4.74-TF/03.
2731 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2732 lock up or crash if it happened to be inside a call to libc when it
2733 got a SIGUSR1 from exiwhat.
2735 The SIGUSR1 handler appends the current process status to the process
2736 log which is later printed by exiwhat. It used to use the general
2737 purpose logging code to do this, but several functions it calls are
2738 not safe for signals.
2740 The new output code in the SIGUSR1 handler is specific to the process
2741 log, and simple enough that it's easy to inspect for signal safety.
2742 Removing some special cases also simplifies the general logging code.
2743 Removing the spurious timestamps from the process log simplifies
2746 TF/04 Improved ratelimit ACL condition.
2748 The /noupdate option has been deprecated in favour of /readonly which
2749 has clearer semantics. The /leaky, /strict, and /readonly update modes
2750 are mutually exclusive. The update mode is no longer included in the
2751 database key; it just determines when the database is updated. (This
2752 means that when you upgrade Exim will forget old rate measurements.)
2754 Exim now checks that the per_* options are used with an update mode that
2755 makes sense for the current ACL. For example, when Exim is processing a
2756 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2757 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2758 must specify per_mail/readonly. If you omit the update mode it defaults to
2759 /leaky where that makes sense (as before) or /readonly where required.
2761 The /noupdate option is now undocumented but still supported for
2762 backwards compatibility. It is equivalent to /readonly except that in
2763 ACLs where /readonly is required you may specify /leaky/noupdate or
2764 /strict/noupdate which are treated the same as /readonly.
2766 A useful new feature is the /count= option. This is a generalization
2767 of the per_byte option, so that you can measure the throughput of other
2768 aggregate values. For example, the per_byte option is now equivalent
2769 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2771 The per_rcpt option has been generalized using the /count= mechanism
2772 (though it's more complicated than the per_byte equivalence). When it is
2773 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2774 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2775 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2776 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2777 Note that using per_rcpt with a non-readonly update mode in more than
2778 one ACL will cause the recipients to be double-counted. (The per_mail
2779 and per_byte options don't have this problem.)
2781 The handling of very low rates has changed slightly. If the computed rate
2782 is less than the event's count (usually one) then this event is the first
2783 after a long gap. In this case the rate is set to the same as this event's
2784 count, so that the first message of a spam run is counted properly.
2786 The major new feature is a mechanism for counting the rate of unique
2787 events. The new per_addr option counts the number of different
2788 recipients that someone has sent messages to in the last time period. It
2789 behaves like per_rcpt if all the recipient addresses are different, but
2790 duplicate recipient addresses do not increase the measured rate. Like
2791 the /count= option this is a general mechanism, so the per_addr option
2792 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2793 example, measure the rate that a client uses different sender addresses
2794 with the options per_mail/unique=$sender_address. There are further
2795 details in the main documentation.
2797 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2799 TF/06 Removed a few PCRE remnants.
2801 TF/07 Automatically extract Exim's version number from tags in the git
2802 repository when doing development or release builds.
2804 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2805 Bugzilla 879. Patch from Paul Fisher.
2807 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2808 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2811 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2813 PP/05 Variable $av_failed, true if the AV scanner deferred.
2814 Bugzilla 1078. Patch from John Horne.
2816 PP/06 Stop make process more reliably on build failure.
2817 Bugzilla 1087. Patch from Heiko Schlittermann.
2819 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2820 Bugzilla 1089. Patch from Heiko Schlittermann.
2822 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2823 Bugzilla 1131. Patch from Holger Weiß.
2825 PP/09 Handle IPv6 addresses with SPF.
2826 Bugzilla 860. Patch from Wolfgang Breyha.
2828 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2830 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2833 PP/11 match_* no longer expand right-hand-side by default.
2834 New compile-time build option, EXPAND_LISTMATCH_RHS.
2835 New expansion conditions, "inlist", "inlisti".
2837 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2839 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2841 PP/14 fix log_write() format string regression from TF/03.
2842 Bugzilla 1152. Patch from Dmitry Isaikin.
2848 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2850 PP/02 Harmonised TLS library version reporting; only show if debugging.
2851 Layout now matches that introduced for other libraries in 4.74 PP/03.
2853 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2855 PP/04 New "dns_use_edns0" global option.
2857 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2860 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2861 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2863 TK/01 Updated PolarSSL code to 0.14.2.
2864 Bugzilla 1097. Patch from Andreas Metzler.
2866 PP/07 Catch divide-by-zero in ${eval:...}.
2867 Fixes bugzilla 1102.
2869 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2872 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2873 format-string attack -- SECURITY: remote arbitrary code execution.
2875 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2876 time unintentionally subject to list matching rules, letting the header
2877 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2878 arbitrary string expansion). This allowed for information disclosure.
2880 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2881 INT_MIN/-1 -- value coerced to INT_MAX.
2887 NM/01 Workaround for PCRE version dependency in version reporting
2890 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2891 This fixes portability to compilers other than gcc, notably
2892 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2894 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2895 makefiles for portability to HP-UX and POSIX correctness.
2897 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2898 Also via indented variable definition in the Makefile.
2899 (Debugging by Oliver Heesakkers).
2901 PP/02 Restore caching of spamd results with expanded spamd_address.
2902 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2904 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2905 Improves build reliability. Fix from: Frank Elsner
2907 NM/02 Fix wide character breakage in the rfc2047 coding
2908 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2910 NM/03 Allow underscore in dnslist lookups
2911 Fixes bug 1026. Patch from Graeme Fowler
2913 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2914 Code patches from Adam Ciarcinski of NetBSD.
2916 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2919 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2920 is logged, to avoid truncation. Patch from John Horne.
2922 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2923 Patch from Jakob Hirsch.
2925 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2926 SQL string expansion failure details.
2927 Patch from Andrey Oktyabrski.
2929 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2930 Patch from Simon Arlott.
2932 PP/09 New lookups functionality failed to compile on old gcc which rejects
2933 extern declarations in function scope.
2934 Patch from Oliver Fleischmann
2936 PP/10 Use sig_atomic_t for flags set from signal handlers.
2937 Check getgroups() return and improve debugging.
2938 Fixed developed for diagnosis in bug 927 (which turned out to be
2941 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2942 Patch from Mark Zealey.
2944 PP/12 Bugzilla 1056: Improved spamd server selection.
2945 Patch from Mark Zealey.
2947 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2948 Based on patch from Heiko Schlittermann.
2950 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2951 Patch from Uwe Doering, sign-off by Michael Haardt.
2953 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2954 variable declaration deep within a block. Bug and patch from
2957 PP/15 lookups-Makefile IRIX compatibility coercion.
2959 PP/16 Make DISABLE_DKIM build knob functional.
2961 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2962 Patch by Simon Arlott
2964 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2965 variable argument macros. Our copy now differs from upstream.
2971 TF/01 Failure to get a lock on a hints database can have serious
2972 consequences so log it to the panic log.
2974 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2975 controlled using the smtp_confirmation log selector.
2977 TF/03 Include the error message when we fail to unlink a spool file.
2979 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2980 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2981 for maintaining out-of-tree patches for some time.
2983 PP/01 Bugzilla 139: Documentation and portability issues.
2984 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2985 Handle per-OS dynamic-module compilation flags.
2987 PP/02 Let /dev/null have normal permissions.
2988 The 4.73 fixes were a little too stringent and complained about the
2989 permissions on /dev/null. Exempt it from some checks.
2990 Reported by Andreas M. Kirchwitz.
2992 PP/03 Report version information for many libraries, including
2993 Exim version information for dynamically loaded libraries. Created
2994 version.h, now support a version extension string for distributors
2995 who patch heavily. Dynamic module ABI change.
2997 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2998 privilege escalation vulnerability whereby the Exim run-time user
2999 can cause root to append content of the attacker's choosing to
3002 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3005 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3006 If dropping privileges for untrusted macros, we disabled normal logging
3007 on the basis that it would fail; for the Exim run-time user, this is not
3008 the case, and it resulted in successful deliveries going unlogged.
3009 Fixed. Reported by Andreas Metzler.
3015 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3016 only prepend for the Resent-* case. Fixes regression introduced in
3017 Exim 4.70 by NM/22 for Bugzilla 607.
3019 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3020 increasing numbers of administrators be bitten by this.
3022 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3024 PP/03 Bugzilla 994: added openssl_options main configuration option.
3026 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3028 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3030 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3032 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3033 without a peer certificate, leading to a segfault because of an
3034 assumption that peers always have certificates. Be a little more
3035 paranoid. Problem reported by Martin Tscholak.
3037 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3038 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3039 NB: ClamAV planning to remove STREAM in "middle of 2010".
3040 CL also introduces -bmalware, various -d+acl logging additions and
3041 more caution in buffer sizes.
3043 PP/09 Implemented reverse_ip expansion operator.
3045 PP/10 Bugzilla 937: provide a "debug" ACL control.
3047 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3049 PP/12 Bugzilla 973: Implement --version.
3051 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3053 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3055 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3057 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3058 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3059 ignore trailing whitespace.
3061 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3063 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3066 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3067 Notification from Dr Andrew Aitchison.
3069 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3070 ExtendedDetectionInfo response format.
3071 Notification from John Horne.
3073 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3076 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3077 XSL and documented dependency on system catalogs, with examples of how
3080 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3083 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3084 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3085 configuration file which is writeable by the Exim user or group.
3087 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3088 of configuration files to cover files specified with the -C option if
3089 they are going to be used with root privileges, not just the default
3092 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3093 option (effectively making it always true).
3095 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3096 files to be used while preserving root privileges.
3098 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3099 that rogue child processes cannot use them.
3101 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3102 run-time user, instead of root.
3104 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3105 Exim run-time user without dropping privileges.
3107 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3108 result string, instead of calling string_vformat() twice with the same
3111 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3112 for other users. Others should always drop root privileges if they use
3113 -C on the command line, even for a whitelisted configure file.
3115 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3117 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3123 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3124 $data_path, and $header_path variables; fixed documentation bugs and
3127 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3128 exipick to access non-standard spools, including the "frozen" queue
3131 NM/01 Bugzilla 965: Support mysql stored procedures.
3132 Patch from Alain Williams
3134 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3136 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3137 Patch from Andreas Metzler
3139 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3140 Patch from Kirill Miazine
3142 NM/05 Bugzilla 671: Added umask to procmail example.
3144 JJ/03 installed exipick 20100323.0, fixing doc bug
3146 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3147 directory. Notification and patch from Dan Rosenberg.
3149 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3151 TK/02 Improve log output when DKIM signing operation fails.
3153 MH/01 Treat the transport option dkim_domain as a colon separated
3154 list, not as a single string, and sign the message with each element,
3155 omitting multiple occurences of the same signer.
3157 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3158 Bugzilla 985, 986. Patch by Simon Arlott
3160 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3161 Patch by Simon Arlott
3163 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3164 MBX locking. Notification from Dan Rosenberg.
3170 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3172 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3174 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3176 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3178 NM/04 Bugzilla 915: Flag broken perl installation during build.
3184 TK/01 Added patch by Johannes Berg that expands the main option
3185 "spamd_address" if it starts with a dollar sign.
3187 TK/02 Write list of recipients to X-Envelope-Sender header when building
3188 the mbox-format spool file for content scanning (suggested by Jakob
3191 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3192 (http://www.dcc-servers.net/) support via dccifd. Activated by
3193 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3195 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3196 by Mark Daniel Reidel <mr@df.eu>.
3198 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3199 When building exim an external PCRE library is now needed -
3200 PCRE is a system library on the majority of modern systems.
3201 See entry on PCRE_LIBS in EDITME file.
3203 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3204 conversation. Added nologin parameter to request.
3205 Patch contributed by Kirill Miazine.
3207 TF/01 Do not log submission mode rewrites if they do not change the address.
3209 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3211 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3212 log files in place. Contributed by Roberto Lima.
3214 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3216 TF/03 Bugzilla 615: When checking the local_parts router precondition
3217 after a local_part_suffix or local_part_prefix option, Exim now
3218 does not use the address's named list lookup cache, since this
3219 contains cached lookups for the whole local part.
3221 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3222 Robert Millan. Documentation is in experimental-spec.txt.
3224 TF/04 Bugzilla 668: Fix parallel build (make -j).
3226 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3228 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3229 Patch provided by Jan Srzednicki.
3231 TF/05 Leading white space used to be stripped from $spam_report which
3232 wrecked the formatting. Now it is preserved.
3234 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3235 that they are available at delivery time.
3237 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3239 TF/08 TLS error reporting now respects the incoming_interface and
3240 incoming_port log selectors.
3242 TF/09 Produce a more useful error message if an SMTP transport's hosts
3243 setting expands to an empty string.
3245 NM/06 Bugzilla 744: EXPN did not work under TLS.
3246 Patch provided by Phil Pennock.
3248 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3249 Patch provided by Richard Godbee.
3251 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3252 acl_smtp_notquit, added index entry.
3254 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3255 Patch provided by Eugene Bujak.
3257 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3258 accept(). Patch provided by Maxim Dounin.
3260 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3261 Patch provided by Phil Pennock.
3263 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3265 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3266 Patch provided by Brad "anomie" Jorsch.
3268 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3269 Patch provided by Dean Brooks.
3271 TK/05 Add native DKIM support (does not depend on external libraries).
3273 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3274 Patch provided by Graeme Fowler.
3276 NM/16 Bugzilla 851: Documentation example syntax fix.
3278 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3280 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3283 NM/19 Bugzilla 745: TLS version reporting.
3284 Patch provided by Phil Pennock.
3286 NM/20 Bugzilla 167: bool: condition support.
3287 Patch provided by Phil Pennock.
3289 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3290 clients. Patch provided by Phil Pennock.
3292 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3293 Patch provided by Brad "anomie" Jorsch.
3295 NM/23 Bugzilla 687: Fix misparses in eximstats.
3296 Patch provided by Heiko Schlittermann.
3298 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3299 Patch provided by Heiko Schlittermann.
3301 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3302 plus update to original patch.
3304 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3306 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3307 Patch provided by David Brownlee.
3309 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3311 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3313 NM/30 Bugzilla 888: TLS documentation bugfixes.
3315 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3317 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3318 Unlike the original bugzilla I have changed all shell scripts in src tree.
3320 NM/33 Bugzilla 898: Transport filter timeout fix.
3321 Patch by Todd Rinaldo.
3323 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3324 Patch by Serge Demonchaux.
3326 NM/35 Bugzilla 39: Base64 decode bug fixes.
3327 Patch by Jakob Hirsch.
3329 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3331 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3333 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3335 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3341 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3342 ALT-N's libdkim that I have put here:
3343 http://duncanthrax.net/exim-experimental/
3345 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3346 were called 'true' and it seems that C99 defines that as a reserved
3347 keyword to be used with 'bool' variable types. That means you could
3348 not include C99-style headers which use bools without triggering
3349 build errors in sieve.c.
3351 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3352 as mailq or other aliases. Changed the --help handling significantly
3353 to do whats expected. exim_usage() emits usage/help information.
3355 SC/01 Added the -bylocaldomain option to eximstats.
3357 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3359 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3361 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3367 PH/01 Another patch from the Sieve maintainer.
3369 PH/02 When an IPv6 address is converted to a string for single-key lookup
3370 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3371 dots are used instead of colons so that keys in lsearch files need not
3372 contain colons. This was done some time before quoting was made available
3373 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3374 (notated using the quote facility) so as to distinguish them from IPv4
3375 keys. This meant that lookups for IP addresses in host lists did not work
3376 for iplsearch lookups.
3378 This has been fixed by arranging for IPv6 addresses to be expressed with
3379 colons if the lookup type is iplsearch. This is not incompatible, because
3380 previously such lookups could never work.
3382 The situation is now rather anomalous, since one *can* have colons in
3383 ordinary lsearch keys. However, making the change in all cases is
3384 incompatible and would probably break a number of configurations.
3386 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3389 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3390 conversion specification without a maximum field width, thereby enabling
3391 a rogue spamd server to cause a buffer overflow. While nobody in their
3392 right mind would setup Exim to query an untrusted spamd server, an
3393 attacker that gains access to a server running spamd could potentially
3394 exploit this vulnerability to run arbitrary code as the Exim user.
3396 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3397 $primary_hostname instead of what libspf2 thinks the hosts name is.
3399 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3400 a directory entry by the name of the lookup key. Previously, if a
3401 symlink pointed to a non-existing file or a file in a directory that
3402 Exim lacked permissions to read, a lookup for a key matching that
3403 symlink would fail. Now it is enough that a matching directory entry
3404 exists, symlink or not. (Bugzilla 503.)
3406 PH/03 The body_linecount and body_zerocount variables are now exported in the
3409 PH/04 Added the $dnslist_matched variable.
3411 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3412 This means they are set thereafter only if the connection becomes
3415 PH/06 Added the client_condition to authenticators so that some can be skipped
3416 by clients under certain conditions.
3418 PH/07 The error message for a badly-placed control=no_multiline_responses left
3419 "_responses" off the end of the name.
3421 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3423 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3424 (without spaces) instead of just copying the configuration text.
3426 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3428 PH/11 Added $max_received_linelength.
3430 PH/12 Added +ignore_defer and +include_defer to host lists.
3432 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3433 way in which PCRE > 7.0 is built.
3435 PH/14 Implemented queue_only_load_latch.
3437 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3438 MAIL command. The effect was to mangle the value on 64-bit systems.
3440 PH/16 Another patch from the Sieve maintainer.
3442 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3444 PH/18 If a system quota error occurred while trying to create the file for
3445 a maildir delivery, the message "Mailbox is full" was not appended to the
3446 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3447 only to a quota excession during the actual writing of the file.
3449 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3450 characters?) which causes problems in log lines. The DN values are now
3451 passed through string_printing() before being added to log lines.
3453 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3454 and InterBase are left for another time.)
3456 PH/21 Added message_body_newlines option.
3458 PH/22 Guard against possible overflow in moan_check_errorcopy().
3460 PH/23 POSIX allows open() to be a macro; guard against that.
3462 PH/24 If the recipient of an error message contained an @ in the local part
3463 (suitably quoted, of course), incorrect values were put in $domain and
3464 $local_part during the evaluation of errors_copy.
3470 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3471 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3474 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3475 issue a MAIL command.
3477 PH/02 In an ACL statement such as
3479 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3481 if a client was not listed at all, or was listed with a value other than
3482 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3483 the condition was not true (as it should be), so access was not denied.
3484 The bug was that the ! inversion was incorrectly passed on to the second
3485 item. This has been fixed.
3487 PH/03 Added additional dnslists conditions == and =& which are different from
3488 = and & when the dns lookup returns more than one IP address.
3490 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3491 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3493 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3494 FSYNC, which compiles an option called disable_fsync that allows for
3495 bypassing fsync(). The documentation is heavily laced with warnings.
3497 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3499 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3500 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3501 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3502 including adding "make clean"; (3) Added -fPIC when compiling the test
3503 dynamically loaded module, to get rid of a warning.
3505 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3506 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3507 = 0s. The bug is otherwise harmless.
3509 PH/07 There was a bug in the dovecot authenticator such that the value of
3510 $auth1 could be overwritten, and so not correctly preserved, after a
3511 successful authentication. This usually meant that the value preserved by
3512 the server_setid option was incorrect.
3514 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3516 PH/09 Installed PCRE release 7.0.
3518 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3519 run for batched SMTP input. It is now run at the start of every message
3520 in the batch. While fixing this I discovered that the process information
3521 (output by running exiwhat) was not always getting set for -bs and -bS
3522 input. This is fixed, and it now also says "batched" for BSMTP.
3524 PH/11 Added control=no_pipelining.
3526 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3527 patch, slightly modified), and move the expansion of helo_data till after
3528 the connection is made in the smtp transport (so it can use these
3531 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3533 PH/14 Added log_selector = +pid.
3535 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3537 PH/16 Add ${if forany and ${if forall.
3539 PH/17 Added dsn_from option to vary the From: line in DSNs.
3541 PH/18 Flush SMTP output before performing a callout, unless control =
3542 no_callout_flush is set.
3544 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3545 was true (the default) a successful delivery failed to delete the retry
3546 item, thus causing premature timeout of the address. The bug is now
3549 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3551 PH/21 Long custom messages for fakedefer and fakereject are now split up
3552 into multiline responses in the same way that messages for "deny" and
3553 other ACL rejections are.
3555 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3556 with slight modification.
3558 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3559 draft, changing the syntax and factoring some duplicate code".
3561 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3562 for deliveries of the second and subsequent messages over the same SMTP
3565 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3566 ${reduce, with only minor "tidies".
3568 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3570 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3571 expansion side effects.
3573 PH/27 When a message times out after an over-quota error from an Exim-imposed
3574 quota, the bounce message says "mailbox is full". This message was not
3575 being given when it was a system quota that was exceeded. It now should
3578 MH/03 Made $recipients available in local_scan(). local_scan() already has
3579 better access to the recipient list through recipients_list[], but
3580 $recipients can be useful in postmaster-provided expansion strings.
3582 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3583 in the case of a MAIL command with additional options following the
3584 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3585 were accidentally chopped off.
3587 PH/29 SMTP synchronization checks are implemented when a command is read -
3588 there is a check that no more input is waiting when there shouldn't be
3589 any. However, for some commands, a delay in an ACL can mean that it is
3590 some time before the response is written. In this time, more input might
3591 arrive, invalidly. So now there are extra checks after an ACL has run for
3592 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3593 pipelining has not been advertised.
3595 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3597 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3598 returned in the SMTP response when smtp_return_error_details was false.
3599 This has been fixed.
3601 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3602 instead of the C I/O that was originally supplied, because problems were
3603 reported on Solaris.
3605 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3606 Exim which did not show up earlier: it was assuming that a call to
3607 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3608 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3609 was a macro that became an assignment, so it seemed to work. This has
3610 changed to a proper function call with a void return, hence the compile
3611 error. Exim's code has been fixed.
3613 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3616 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3618 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3620 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3621 because it turns out to be too big or there is a timeout) there is no
3622 "Completed" line in the log. When some messages of this type were
3623 selected by exigrep, they were listed as "not completed". Others were
3624 picked up by some special patterns. I have improved the selection
3625 criteria to be more general.
3627 PH/38 The host_find_failed option in the manualroute router can now be set
3628 to "ignore", to completely ignore a host whose IP address cannot be
3629 found. If all hosts are ignored, the behaviour is controlled by the new
3630 host_all_ignored option.
3632 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3633 homing or because of multiple MX records with /mx) generated more than
3634 one IP address, and the following item turned out to be the local host,
3635 all the secondary addresses of the first item were incorrectly removed
3636 from the list, along with the local host and any following hosts (which
3637 is what is supposed to happen).
3639 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3640 whoever called Exim into the -H file. In the case of the daemon it was
3641 behaving confusingly. When first started, it used values for whoever
3642 started the daemon, but after a SIGHUP it used the Exim user (because it
3643 calls itself on a restart). I have changed the code so that it now always
3646 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3647 message are rejected with the same error (e.g. no authentication or bad
3648 sender address), and a DATA command is nevertheless sent (as can happen
3649 with PIPELINING or a stupid MUA), the error message that was given to the
3650 RCPT commands is included in the rejection of the DATA command. This is
3651 intended to be helpful for MUAs that show only the final error to their
3654 PH/42 Another patch from the Sieve maintainer.
3656 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3657 Eximstats - Fixed some broken HTML links and added missing column headers
3659 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3660 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3662 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3664 PH/43 Yet another patch from the Sieve maintainer.
3666 PH/44 I found a way to check for a TCP/IP connection going away before sending
3667 the response to the final '.' that terminates a message, but only in the
3668 case where the client has not sent further data following the '.'
3669 (unfortunately, this is allowed). However, in many cases there won't be
3670 any further data because there won't be any more messages to send. A call
3671 to select() can be used: if it shows that the input is "ready", there is
3672 either input waiting, or the socket has been closed. An attempt to read
3673 the next input character can distinguish the two cases. Previously, Exim
3674 would have sent an OK response which the client would never have see.
3675 This could lead to message repetition. This fix should cure that, at
3676 least in a lot of common cases.
3678 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3679 advertised in response to EHLO.
3685 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3686 fixed by 4.65/MH/01 (is this a record?) are fixed:
3688 (i) An empty string was always treated as zero by the numeric comparison
3689 operators. This behaviour has been restored.
3691 (ii) It is documented that the numeric comparison operators always treat
3692 their arguments as decimal numbers. This was broken in that numbers
3693 starting with 0 were being interpreted as octal.
3695 While fixing these problems I realized that there was another issue that
3696 hadn't been noticed. Values of message_size_limit (both the global option
3697 and the transport option) were treated as octal if they started with 0.
3698 The documentation was vague. These values are now always treated as
3699 decimal, and I will make that clear in the documentation.
3705 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3706 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3709 MH/01 Don't check that the operands of numeric comparison operators are
3710 integers when their expansion is in "skipping" mode (fixes bug
3711 introduced by 4.64-PH/07).
3713 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3714 child addresses, Exim now panics and dies. Previously, because the count
3715 is held in a short int, deliveries were likely to be lost. As such a
3716 large number of recipients for a single message is ridiculous
3717 (performance will be very, very poor), I have chosen to impose a limit
3718 rather than extend the field.
3724 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3725 leftover -K file (the existence of which was triggered by #402).
3726 While we were at it, introduced process PID as part of the -K
3727 filename. This should rule out race conditions when creating
3730 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3731 processing considerably. Previous code took too long for large mails,
3732 triggering a timeout which in turn triggers #401.
3734 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3735 in the DK code in transports.c. sendfile() is not really portable,
3736 hence the _LINUX specificness.
3738 TF/01 In the add_headers option to the mail command in an Exim filter,
3739 there was a bug that Exim would claim a syntax error in any
3740 header after the first one which had an odd number of characters
3743 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3744 callout verification, Exim cached a "reject" for the entire domain. This
3745 is correct for most verifications, but it is not correct for a recipient
3746 verification with use_sender or use_postmaster set, because in that case
3747 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3748 case of MAIL FROM:<> rejection from other early rejections (e.g.
3749 rejection of HELO). When verifying a recipient using a non-null MAIL
3750 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3751 Whatever the result of the callout, the value of the domain cache is
3752 left unchanged (for any other kind of callout, getting as far as trying
3753 RCPT means that the domain itself is ok).
3755 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3758 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3759 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3760 session, the daemon ignores ECONNECT errors and logs others; it now
3761 ignores EPIPE as well.
3763 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3764 (quoted-printable decoding).
3766 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3767 later the small subsequent patch to fix an introduced bug.
3769 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3771 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3773 PH/08 An error is now given if message_size_limit is specified negative.
3775 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3776 to be given (somewhat) arbitrary names.
3778 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3781 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3782 miscellaneous code fixes
3784 PH/10 Added the log_reject_target ACL modifier to specify where to log
3787 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3788 hostname. This is wrong, because it relates to the incoming message (and
3789 probably the interface on which it is arriving) and not to the outgoing
3790 callout (which could be using a different interface). This has been
3791 changed to use the value of the helo_data option from the smtp transport
3792 instead - this is what is used when a message is actually being sent. If
3793 there is no remote transport (possible with a router that sets up host
3794 addresses), $smtp_active_hostname is used.
3796 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3797 tweaks were necessary in order to get it to work (see also 21 below):
3798 (a) The code assumed that strncpy() returns a negative number on buffer
3799 overflow, which isn't the case. Replaced with Exim's string_format()
3801 (b) There were several signed/unsigned issues. I just did the minimum
3802 hacking in of casts. There is scope for a larger refactoring.
3803 (c) The code used strcasecmp() which is not a standard C function.
3804 Replaced with Exim's strcmpic() function.
3805 (d) The code set only $1; it now sets $auth1 as well.
3806 (e) A simple test gave the error "authentication client didn't specify
3807 service in request". It would seem that Dovecot has changed its
3808 interface. Fortunately there's a specification; I followed it and
3809 changed what the client sends and it appears to be working now.
3811 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3814 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3815 address A is aliased to B and C, where B exists and C does not. Without
3816 -v the output is "A verified" because verification stops after a
3817 successful redirection if more than one address is generated. However,
3818 with -v the child addresses are also verified. Exim was outputting "A
3819 failed to verify" and then showing the successful verification for C,
3820 with its parentage. It now outputs "B failed to verify", showing B's
3821 parentage before showing the successful verification of C.
3823 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3824 look up a TXT record in a specific list after matching in a combined
3827 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3828 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3829 they consult the DNS. I had assumed they would set it the way they
3830 wanted; and indeed my experiments on Linux seem to show that in some
3831 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3832 To be on the safe side, however, I have now made the interface to
3833 host_find_byname() similar to host_find_bydns(), with an argument
3834 containing the DNS resolver options. The host_find_byname() function now
3835 sets these options at its start, just as host_find_bydns() does. The smtp
3836 transport options dns_qualify_single and dns_search_parents are passed to
3837 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3838 of host_find_byname() use the default settings of RES_DEFNAMES
3839 (qualify_single) but not RES_DNSRCH (search_parents).
3841 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3842 spool_read_header() do less string testing, by means of a preliminary
3843 switch on the second character of optional "-foo" lines. (This is
3844 overdue, caused by the large number of possibilities that now exist.
3845 Originally there were few.) While I was there, I also converted the
3846 str(n)cmp tests so they don't re-test the leading "-" and the first
3847 character, in the hope this might squeeze out yet more improvement.
3849 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3850 flag allowing group syntax was set by the header_syntax check but not
3851 turned off, possible causing trouble later; (2) The flag was not being
3852 set at all for the header_verify test, causing "group"-style headers to
3853 be rejected. I have now set it in this case, and also caused header_
3854 verify to ignore an empty address taken from a group. While doing this, I
3855 came across some other cases where the code for allowing group syntax
3856 while scanning a header line wasn't quite right (mostly, not resetting
3857 the flag correctly in the right place). These bugs could have caused
3858 trouble for malformed header lines. I hope it is now all correct.
3860 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3861 with the "reply" argument non-NULL. The code, however (which originally
3862 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3863 but it didn't always do it. This confused somebody who was copying the
3864 code for some other use. I have removed all the tests.
3866 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3867 feature that was used to support insecure browsers during the U.S. crypto
3868 embargo. It requires special client support, and Exim is probably the
3869 only MTA that supported it -- and would never use it because real RSA is
3870 always available. This code has been removed, because it had the bad
3871 effect of slowing Exim down by computing (never used) parameters for the
3872 RSA_EXPORT functionality.
3874 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3875 authenticator to fail if there's a tab character in the incoming data
3876 (there should never be unless someone is messing about, as it's supposed
3877 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3878 if the connection is using TLS or if the remote IP is the same as the
3879 local IP, and the "valid-client-cert option" if a client certificate has
3882 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3883 authenticators. This can be used for authorization after authentication
3884 succeeds. (In the case of plaintext, it servers for both authentication
3887 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3888 if any retry times were supplied.
3890 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3891 connection, where there is no client IP address to check. In this
3892 situation, the verify now always succeeds.
3894 PH/25 Applied John Jetmore's -Mset patch.
3896 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3898 PH/27 In a string expansion for a processed (not raw) header when multiple
3899 headers of the same name were present, leading whitespace was being
3900 removed from all of them, but trailing whitespace was being removed only
3901 from the last one. Now trailing whitespace is removed from each header
3902 before concatenation. Completely empty headers in a concatenation (as
3903 before) are ignored.
3905 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3906 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3908 PH/29 [Removed. This was a change that I later backed out, and forgot to
3909 correct the ChangeLog entry (that I had efficiently created) before
3910 committing the later change.]
3912 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3913 address errors (4xx response to RCPT) over the same connection as other
3914 messages routed to the same hosts. Such deliveries are always "forced",
3915 so retry times are not inspected. This resulted in far too many retries
3916 for the affected addresses. The effect occurred only when there were more
3917 hosts than the hosts_max_try setting in the smtp transport when it had
3918 the 4xx errors. Those hosts that it had tried were not added to the list
3919 of hosts for which the message was waiting, so if all were tried, there
3920 was no problem. Two fixes have been applied:
3922 (i) If there are any address or message errors in an SMTP delivery, none
3923 of the hosts (tried or untried) are now added to the list of hosts
3924 for which the message is waiting, so the message should not be a
3925 candidate for sending over the same connection that was used for a
3926 successful delivery of some other message. This seems entirely
3927 reasonable: after all the message is NOT "waiting for some host".
3928 This is so "obvious" that I'm not sure why it wasn't done
3929 previously. Hope I haven't missed anything, but it can't do any
3930 harm, as the worst effect is to miss an optimization.
3932 (ii) If, despite (i), such a delivery is accidentally attempted, the
3933 routing retry time is respected, so at least it doesn't keep
3934 hammering the server.
3936 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3937 in ${readsocket because some servers need this prod.
3939 PH/32 Added some extra debug output when updating a wait-xxx database.
3941 PH/33 The hint "could be header name not terminated by colon", which has been
3942 given for certain expansion errors for a long time, was not being given
3943 for the ${if def:h_colon_omitted{... case.
3945 PH/34 The spec says: "With one important exception, whenever a domain list is
3946 being scanned, $domain contains the subject domain." There was at least
3947 one case where this was not true.
3949 PH/35 The error "getsockname() failed: connection reset by peer" was being
3950 written to the panic log as well as the main log, but it isn't really
3951 panic-worthy as it just means the connection died rather early on. I have
3952 removed the panic log writing for the ECONNRESET error when getsockname()
3955 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3956 runs only) independently of the message's sender address. This meant
3957 that, if the 4xx error was in fact related to the sender, a different
3958 message to the same recipient with a different sender could confuse
3959 things. In particular, this can happen when sending to a greylisting
3960 server, but other circumstances could also provoke similar problems.
3961 I have changed the default so that the retry time for these errors is now
3962 based a combination of the sender and recipient addresses. This change
3963 can be overridden by setting address_retry_include_sender=false in the
3966 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3967 remote server are returned as part of bounce messages. This was not
3968 happening for LMTP over a pipe (the lmtp transport), but now it is the
3969 same for both kinds of LMTP.
3971 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3972 in header lines that were in fact CNAMEs. This is no longer the case.
3974 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3975 and queue runs started by the daemon processed all messages. This has
3976 been fixed so that -R and -S can now usefully be given with -q<time>.
3978 PH/40 Import PCRE release 6.7 (fixes some bugs).
3980 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3982 PH/42 Give an error if -q is specified more than once.
3984 PH/43 Renamed the variables $interface_address and $interface_port as
3985 $received_ip_address and $received_port, to make it clear that these
3986 values apply to message reception, and not to the outgoing interface when
3987 a message is delivered. (The old names remain recognized, of course.)
3989 PH/44 There was no timeout on the connect() call when using a Unix domain
3990 socket in the ${readsocket expansion. There now is.
3992 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3993 be meaningful with "accept".
3995 SC/01 Eximstats V1.43
3996 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3998 SC/02 Eximstats V1.44
3999 Use a glob alias rather than an array ref in the generated
4000 parser. This improves both readability and performance.
4002 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4003 Collect SpamAssassin and rejection statistics.
4004 Don't display local sender or destination tables unless
4005 there is data to show.
4006 Added average volumes into the top table text output.
4008 SC/04 Eximstats V1.46
4009 Collect data on the number of addresses (recipients)
4010 as well as the number of messages.
4012 SC/05 Eximstats V1.47
4013 Added 'Message too big' to the list of mail rejection
4014 reasons (thanks to Marco Gaiarin).
4016 SC/06 Eximstats V1.48
4017 Mainlog lines which have GMT offsets and are too short to
4018 have a flag are now skipped.
4020 SC/07 Eximstats V1.49 (Alain Williams)
4021 Added the -emptyok flag.
4023 SC/08 Eximstats V1.50
4024 Fixes for obtaining the IP address from reject messages.
4026 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4027 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4028 whitespace changes from 4.64-PH/27
4030 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4033 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4034 are found, allow negative numbers in numeric criteria)
4036 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4038 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4039 to match changes made in 4.64-PH/43
4041 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4043 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4044 "domain" is actually an IP address, and force a failure. This locks out
4045 those revolvers/nameservers that support "A-for-A" lookups, in
4046 contravention of the specifications.
4048 PH/48 When a host name was looked up from an IP address, and the subsequent
4049 forward lookup of the name timed out, the host name was left in
4050 $sender_host_name, contrary to the specification.
4052 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4053 restricted to single-key lookups, Exim was not diagnosing an error if
4054 * or *@ was used with a query-style lookup.
4056 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4058 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4059 long ago, but noone interested enough thought of it. Let's just say that
4060 the "1.1" means that there are some new functions that weren't there at
4061 some point in the past.
4063 PH/51 Error processing for expansion failure of helo_data from an smtp
4064 transport during callout processing was broken.
4066 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4067 tested/used via the -bh/-bhc/-bs options.
4069 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4070 bug, fixed in subsequent PCRE releases).
4072 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4073 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4075 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4081 SC/01 Use a glob alias rather than an array ref in eximstats generated
4082 parser. This improves both readability and performance.
4084 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4085 Don't display local sender or destination tables in eximstats unless
4086 there is data to show.
4087 Added average volumes into the eximstats top table text output.
4089 SC/03 Collect data on the number of addresses (recipients) as well
4090 as the number of messages in eximstats.
4092 TF/01 Correct an error in the documentation for the redirect router. Exim
4093 does (usually) call initgroups() when daemonizing.
4095 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4096 with consistent privilege compared to when running as a daemon.
4098 TF/03 Note in the spec that $authenticated_id is not set for local
4099 submissions from trusted users.
4101 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4102 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4104 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4105 by adding some example configuration directives to the default
4106 configuration file. A little bit of work is required to uncomment the
4107 directives and define how usernames and passwords are checked, but
4108 there is now a framework to start from.
4110 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4111 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4112 without this. I don't know how relevant this is to other LDAP libraries.
4114 PH/02 Add the verb name to the "unknown ACL verb" error.
4116 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4118 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4120 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4121 directory not expanded when it should be if an expanded home directory
4122 was set for the address (which is overridden by the transport).
4124 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4127 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4128 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4129 because it is too late at that time, and has no effect.
4131 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4132 security issue with \' (bugzilla #107). I could not use the
4133 PQescapeStringConn() function, because it needs a PGconn value as one of
4136 PH/08 When testing addresses using -bt, indicate those final addresses that
4137 are duplicates that would not cause an additional delivery. At least one
4138 person was confused, thinking that -bt output corresponded to deliveries.
4139 (Suppressing duplicates isn't a good idea as you lose the information
4140 about possibly different redirections that led to the duplicates.)
4142 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4143 systems where poll() doesn't work, in particular OS X.
4145 PH/10 Added more information to debugging output for retry time not reached.
4147 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4148 operations in malware.c.
4150 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4153 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4154 syslog_duplication set false, log lines that would normally be written
4155 both the the main log and to the reject log were not written to syslog at
4158 PH/14 In the default configuration, change the use of "message" in ACL warn
4159 statements to "add_header".
4161 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4162 not followed by a command (e.g. "seen endif").
4164 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4165 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4168 PH/17 Added extra conditions to the default value of delay_warning_condition
4172 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4173 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4174 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4177 The Auto-Submitted: and various List- headers are standardised, whereas I
4178 don't think Precedence: ever was.
4180 PH/18 Refactored debugging code in route_finduser() to show more information,
4181 in particular, the error code if getpwnam() issues one.
4183 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4184 This is apparently needed in addition to the PH/07 change above to avoid
4185 any possible encoding problems.
4187 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4188 but not after initializing Perl.
4190 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4191 output them only if debugging. By default they are written stderr,
4192 apparently, which is not desirable.
4194 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4197 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4200 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4202 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4203 authenticated or an ident call has been made. Suppress the default
4204 values for $authenticated_id and $authenticated_sender (but permit -oMai
4205 and -oMas) when testing with -bh.
4207 PH/24 Re-jigged the order of the tests in the default configuration so that the
4208 tests for valid domains and recipients precede the DNS black list and CSA
4209 tests, on the grounds that those ones are more expensive.
4211 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4212 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4213 This bug exists in every version of Exim that I still have, right back to
4216 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4217 However, an attempt to turn on case-sensitivity in a regex key by
4218 including (?-i) didn't work because the subject string was already
4219 lowercased, and the effects were non-intuitive. It turns out that a
4220 one-line patch can be used to allow (?-i) to work as expected.
4226 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4227 other effects) broke the use of negated acl sub-conditions.
4229 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4232 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4233 "Deny" causes Exim to reject the incoming connection with a 554 error.
4234 Unfortunately, if there is a major crisis, such as a disk failure,
4235 tcp-wrappers gives "deny", whereas what one would like would be some
4236 kind of temporary error. A kludge has been added to help with this.
4237 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4238 554 error is used if errno is still zero or contains ENOENT (which occurs
4239 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4242 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4244 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4245 errors. Otherwise a message that provokes a temporary error (when other
4246 messages do not) can cause a whole host to time out.
4248 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4249 addresses were routed directly to files or pipes from a redirect router.
4250 File deliveries just didn't batch; pipe deliveries might have suffered
4253 PH/06 A failure to get a lock for a hints database would erroneously always say
4254 "Failed to get write lock", even when it was really a read lock.
4256 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4257 of 0600. This has been changed to use the value of the lockfile_mode
4258 option (which defaults to 0600).
4260 PH/08 Applied small patch from the Sieve maintainer.
4262 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4263 folder from quota calculations, a direct delivery into this folder messed
4264 up the contents of the maildirsize file. This was because the regex was
4265 used only to exclude .Trash (or whatever) when the size of the mailbox
4266 was calculated. There was no check that a delivery was happening into an
4267 excluded directory. This bug has been fixed by ignoring all quota
4268 processing for deliveries into excluded directories.
4270 PH/10 Added the maildirfolder_create_regex option to appendfile.
4276 PH/01 The code for finding all the local interface addresses on a FreeBSD
4277 system running IPv6 was broken. This may well have applied to all BSD
4278 systems, as well as to others that have similar system calls. The broken
4279 code found IPv4 interfaces correctly, but gave incorrect values for the
4280 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4281 that it would not match correctly against @[] and not recognize the IPv6
4284 PH/02 The ipliteral router was not recognizing addresses of the form user@
4285 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4287 PH/03 Added disable_ipv6.
4289 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4290 decoded form, because it is most often used to construct To: headers
4291 lines in autoreplies, and the decoded form may well be syntactically
4292 invalid. However, $reply_address has leading white space removed, and all
4293 newlines turned into spaces so that the autoreply transport does not
4296 PH/05 If group was specified without a user on a router, and no group or user
4297 was specified on a transport, the group from the router was ignored.
4299 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4300 for visible compile-time settings that can be used to change these
4301 numbers, for those that want even more. Backwards compatibility with old
4302 spool files has been maintained. However, going back to a previous Exim
4303 release will lost any variables that are in spool files.
4305 PH/07 Two small changes when running in the test harness: increase delay when
4306 passing a TCP/IP connection to a new process, in case the original
4307 process has to generate a bounce, and remove special handling of
4308 127.0.0.2 (sic), which is no longer necessary.
4310 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4311 be the same on different OS.
4313 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4316 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4317 whether --show-vars was specified or not
4319 JJ/02 exipick: Added support for new ACL variable spool format introduced
4322 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4323 syntactically invalid From: or Reply-to: line, and a filter used this to
4324 generate an autoreply, and therefore failed to obtain an address for the
4325 autoreply, Exim could try to deliver to a non-existent relative file
4326 name, causing unrelated and misleading errors. What now happens is that
4327 it logs this as a hard delivery error, but does not attempt to create a
4330 PH/11 The exinext utility has a -C option for testing purposes, but although
4331 the given file was scanned by exinext itself; it wasn't being passed on
4332 when Exim was called.
4334 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4335 an end-of-file indication when reading a command response.
4337 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4338 compiled. In many other places in Exim, IPv6 addresses are always
4339 recognized, so I have changed this. It also means that IPv4 domain
4340 literals of the form [IPV4:n.n.n.n] are now always recognized.
4342 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4343 used if the router is not running as root, for example, when verifying at
4344 ACL time, or when using -bh. The debugging output from this situation was
4345 non-existent - all you got was a failure to exec. I have made two
4348 (a) Failures to set uid/gid, the current directory, or a process leader
4349 in a subprocess such as that created by queryprogram now generate
4350 suitable debugging output when -d is set.
4352 (b) The queryprogram router detects when it is not running as root,
4353 outputs suitable debugging information if -d is set, and then runs
4354 the subprocess without attempting to change uid/gid.
4356 PH/15 Minor change to Makefile for building test_host (undocumented testing
4359 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4360 additional section of a DNS packet that returns MX or SRV records.
4361 Instead, it always explicitly searches for A/AAAA records. This avoids
4362 major problems that occur when a DNS server includes only records of one
4363 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4364 fixed another bug: if SRV records were looked up and the corresponding
4365 address records were *not* found in the additional section, the port
4366 values from the SRV records were lost.
4368 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4369 using the correct key (the original address) when searching the retry
4370 rules in order to find which one to use for generating the retry hint.
4372 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4373 adding the default one. Similarly, if it contains a Reply-To: header, the
4374 errors_reply_to option, if set, is not used.
4376 PH/19 When calculating a retry time, Exim used to measure the "time since
4377 failure" by looking at the "first failed" field in the retry record. Now
4378 it does not use this if it is later than than the arrival time of the
4379 message. Instead it uses the arrival time. This makes for better
4380 behaviour in cases where some deliveries succeed, thus re-setting the
4381 "first failed" field. An example is a quota failure for a huge message
4382 when small messages continue to be delivered. Without this change, the
4383 "time since failure" will always be short, possible causing more frequent
4384 delivery attempts for the huge message than are intended.
4385 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4387 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4388 $1, $2, $3) because the numerical variables can be reset during some
4389 expansion items (e.g. "match"), thereby losing the authentication data.
4391 PH/21 Make -bV show the size of off_t variables so that the test suite can
4392 decide whether to run tests for quotas > 2G.
4394 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4395 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4396 filecount value is greater than 2G or if a quota value is greater than 2G
4397 on a system where the size of off_t is not greater than 4, a panic error
4400 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4401 never match. The debug and -bh output now contains an explicit error
4402 message indicating a malformed IPv4 address or mask.
4404 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4405 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4406 PH/23 above applies.
4408 PH/25 Do not write to syslog when running in the test harness. The only
4409 occasion when this arises is a failure to open the main or panic logs
4410 (for which there is an explicit test).
4412 PH/26 Added the /no_tell option to "control=freeze".
4414 PH/27 If a host name lookup failed very early in a connection, for example, if
4415 the IP address matched host_lookup and the reverse lookup yielded a name
4416 that did not have a forward lookup, an error message of the form "no IP
4417 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4418 could be logged. Now it outputs the IP address instead of "NULL".
4420 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4421 allows the sender and the authenticated sender to be set when
4422 submitting a message from within Exim. Since child_open_exim() is
4423 documented for local_scan(), the new function should be too.
4425 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4426 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4427 results in an empty string is now treated as unset.
4429 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4431 PH/31 Added sender_verify_fail logging option.
4433 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4434 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4435 tidied the source and removed it altogether.
4437 PH/33 When a queue run was abandoned because the load average was too high, a
4438 log line was always written; now it is written only if the queue_run log
4439 selector is set. In addition, the log line for abandonment now contains
4440 information about the queue run such as the pid. This is always present
4441 in "start" and "stop" lines but was omitted from the "abandon" line.
4443 PH/34 Omit spaces between a header name and the colon in the error message that
4444 is given when verify = headers_syntax fails (if there are lots of them,
4445 the message gets confusing).
4447 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4448 names, as there are now some PTR records that contain slashes. This check
4449 is only to protect against broken name servers that fall over on strange
4450 characters, so the fact that it applies to all lookups doesn't matter.
4452 PH/36 Now that the new test suite is complete, we can remove some of the
4453 special code in Exim that was needed for the old test suite. For example,
4454 sorting DNS records because real resolvers return them in an arbitrary
4455 order. The new test suite's fake resolver always returns records in the
4458 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4459 bounces) except when queue_only is set, to avoid logging races between
4460 the different processes.
4462 PH/38 Panic-die if .include specifies a non-absolute path.
4464 PH/39 A tweak to the "H" retry rule from its user.
4466 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4467 a label. They prevented compilation on older perls.
4469 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4470 a warning to be raised on newish perls.
4472 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4473 on queue. Changes to match documented behaviour of showing count of
4474 messages matching specified criteria.
4476 PH/40 Changed the default ident timeout from 30s to 5s.
4478 PH/41 Added support for the use of login_cap features, on those BSD systems
4479 that have them, for controlling the resources used by pipe deliveries.
4481 PH/42 The content-scanning code uses fopen() to create files in which to put
4482 message data. Previously it was not paying any attention to the mode of
4483 the files. Exim runs with umask(0) because the rest of the code creates
4484 files with open(), and sets the required mode explicitly. Thus, these
4485 files were ending up world-writeable. This was not a big issue, because,
4486 being within the spool directory, they were not world-accessible. I have
4487 created a function called modefopen, which takes an additional mode
4488 argument. It sets umask(777), creates the file, chmods it to the required
4489 mode, then resets the umask. All the relevant calls to fopen() in the
4490 content scanning code have been changed to use this function.
4492 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4493 to 24 hours. This avoids potential overflow problems when processing G
4494 and H retry rules. I suspect nobody ever tinkers with this value.
4496 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4498 PH/45 When the plaintext authenticator is running as a client, the server's
4499 challenges are checked to ensure they are valid base64 strings. By
4500 default, the authentication attempt is cancelled if an invalid string is
4501 received. Setting client_ignore_invalid_base64 true ignores these errors.
4502 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4503 they are received. Thus, the responses can be made to depend on the
4504 challenges. If an invalid string is ignored, an empty string is placed in
4507 PH/46 Messages that are created by the autoreply transport now contains a
4508 References: header, in accordance with RFCs 2822 and 3834.
4510 PH/47 Added authenticated_sender_force to the smtp transport.
4512 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4514 PH/49 Installed latest patch from the Sieve maintainer.
4516 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4517 was also set, the appendfile transport was unnecessarily scanning a
4518 directory of message files (e.g. for maildir delivery) to find the count
4519 of files (along with the size), even though it did not need this
4520 information. It now does the scan only if it needs to find either the
4521 size of the count of files.
4523 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4525 PH/52 Two bugs concerned with error handling when the smtp transport is
4528 (i) Exim was not creating retry information for temporary errors given
4529 for individual recipients after the DATA command when the smtp transport
4530 was used in LMTP mode. This meant that they could be retried too
4531 frequently, and not timed out correctly.
4533 (ii) Exim was setting the flag that allows error details to be returned
4534 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4535 recipients that were returned after the DATA command.
4537 PH/53 This is related to PH/52, but is more general: for any failing address,
4538 when detailed error information was permitted to be returned to the
4539 sender, but the error was temporary, then after the final timeout, only
4540 "retry timeout exceeded" was returned. Now it returns the full error as
4541 well as "retry timeout exceeded".
4543 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4544 do this, and (what is worse) MTAs that accept it.
4546 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4547 will now be deprecated.
4549 PH/56 New os.c-cygwin from the Cygwin maintainer.
4551 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4552 formats (previously only available in exim formats via -bpr, -bpru,
4553 and -bpra. Now also available in native and exiqgrep formats)
4555 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4556 with very large, slow to parse queues
4558 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4560 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4562 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4563 responses to authentication challenges, though it was showing the
4564 challenges; (ii) I've removed the CR characters from the debug output for
4567 PH/58 Allow for the insertion of a newline as well as a space when a string
4568 is turned into more than one encoded-word during RFC 2047 encoding. The
4569 Sieve code now uses this.
4571 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4572 data_4xx, lost_connection, tls_required.
4574 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4575 message was being sent as an SMTP response.
4577 PH/61 Add -l and -k options to exicyclog.
4579 PH/62 When verifying, if an address was redirected to one new address, so that
4580 verification continued, and the new address failed or deferred after
4581 having set something in $address_data, the value of $address_data was not
4582 passed back to the ACL. This was different to the case when no
4583 redirection occurred. The value is now passed back in both cases.
4585 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4586 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4587 use login_cap.h, so on its own it isn't the distinguishing feature. The
4588 new name refers directly to the setclassresources() function.
4590 PH/65 Added configuration files for NetBSD3.
4592 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4594 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4595 is preferred over IPv4.
4597 PH/68 The bounce_return_message and bounce_return_body options were not being
4598 honoured for bounces generated during the reception of non-SMTP messages.
4599 In particular, this applied to messages rejected by the ACL. This bug has
4600 been fixed. However, if bounce_return_message is true and bounce_return_
4601 body is false, the headers that are returned for a non-SMTP message
4602 include only those that have been read before the error was detected.
4603 (In the case of an ACL rejection, they have all been read.)
4605 PH/69 The HTML version of the specification is now built in a directory called
4606 spec_html instead of spec.html, because the latter looks like a path with
4607 a MIME-type, and this confuses some software.
4609 PH/70 Catch two compiler warnings in sieve.c.
4611 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4612 function verify_get_ident() calls ip_connect() to connect a socket, but
4613 if the "connect()" function timed out, ip_connect() used to close the
4614 socket. However, verify_get_ident() also closes the socket later, and in
4615 between Exim writes to the log, which may get opened at this point. When
4616 the socket was closed in ip_connect(), the log could get the same file
4617 descriptor number as the socket. This naturally causes chaos. The fix is
4618 not to close the socket in ip_connect(); the socket should be closed by
4619 the function that creates it. There was only one place in the code where
4620 this was missing, in the iplookup router, which I don't think anybody now
4621 uses, but I've fixed it anyway.
4623 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4624 well as to direct DNS lookups. Otherwise the handling of names in host
4625 lists is inconsistent and therefore confusing.
4631 PH/01 Two changes to the default runtime configuration:
4633 (1) Move the checks for relay_from_hosts and authenticated clients from
4634 after to before the (commented out) DNS black list checks.
4636 (2) Add control=submission to the relay_from_hosts and authenticated
4637 clients checks, on the grounds that messages accepted by these
4638 statements are most likely to be submissions.
4640 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4642 (1) Generate an error if the third argument for the ${prvs expansion is
4645 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4648 (3) Reset the variables that are obtained from the first argument of
4649 ${prvscheck and used in the second argument before leaving the code,
4650 because their memory is reclaimed, so using them afterwards may do
4653 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4654 one (it's much easier than Tom thought :-).
4656 (5) Because of (4), we can now allow for the use of $prvscheck_result
4657 inside the third argument.
4659 PH/03 For some reason, the default setting of PATH when running a command from
4660 a pipe transport was just "/usr/bin". I have changed it to
4663 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4664 anything to be listed in the output from -bV.
4666 PH/05 When a filter generated an autoreply, the entire To: header line was
4667 quoted in the delivery log line, like this:
4669 => >A.N.Other <ano@some.domain> <original@ddress> ...
4671 This has been changed so that it extracts the operative address. There
4672 may be more than one such address. If so, they are comma-separated, like
4675 => >ano@some.domain,ona@other.domain <original@ddress> ...
4677 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4678 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4679 not being looked up in the rDNS to get a host name, Exim was showing the
4680 IP address twice in Received: lines, even though the IP addresses were
4681 identical. For example:
4683 Received: from [1.2.3.4] (helo=[1.2.3.4])
4685 However, if the real host name was known, it was omitting the HELO data
4686 if it matched the actual IP address. This has been tidied up so that it
4687 doesn't show the same IP address twice.
4689 PH/07 When both +timestamp and +memory debugging was on, the value given by
4690 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4691 called by the debug printing, thereby overwriting the timestamp buffer.
4692 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4694 PH/08 When the original message was included in an autoreply transport, it
4695 always said "this is a copy of the message, including all the headers",
4696 even if body_only or headers_only was set. It now gives an appropriate
4699 PH/09 Applied a patch from the Sieve maintainer which:
4701 o fixes some comments
4702 o adds the (disabled) notify extension core
4703 o adds some debug output for the result of if/elsif tests
4704 o points to the current vacation draft in the documentation
4705 and documents the missing references header update
4709 o fixes a bug in processing the envelope test (when testing
4710 multiple envelope elements, the last element determined the
4713 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4714 Electronic Mail") by including:
4716 Auto-submitted: auto-generated
4718 in the messages that it generates (bounce messages and others, such as
4719 warnings). In the case of bounce messages for non-SMTP messages, there was
4720 also a typo: it was using "Auto_submitted" (underscore instead of
4721 hyphen). Since every message generated by Exim is necessarily in response
4722 to another message, thes have all been changed to:
4724 Auto-Submitted: auto-replied
4726 in accordance with these statements in the RFC:
4728 The auto-replied keyword:
4730 - SHOULD be used on messages sent in direct response to another
4731 message by an automatic process,
4733 - MUST NOT be used on manually-generated messages,
4735 - MAY be used on Delivery Status Notifications (DSNs) and Message
4736 Disposition Notifications (MDNs),
4738 - MUST NOT be used on messages generated by automatic or periodic
4739 processes, except for messages which are automatic responses to
4742 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4743 to the default Received: header definition.
4745 PH/12 Added log selector acl_warn_skipped (default on).
4747 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4748 variables because (a) they are in the wrong storage pool and (b) even if
4749 they were copied, it wouldn't work properly because of the caching.
4751 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4752 checking when decoding. Apparently there are clients that generate
4753 overlong encoded strings. Why am I not surprised?
4755 PH/15 If the first argument of "${if match_address" was not empty, but did not
4756 contain an "@" character, Exim crashed. Now it writes a panic log message
4757 and treats the condition as false.
4759 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4761 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4762 extension "envelope-auth". The code is finished and in agreement with
4763 other implementations, but there is no documentation so far and in fact,
4764 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4765 not changing the active code.
4767 Print executed "if" and "elsif" statements when debugging is used. This
4768 helps a great deal to understand what a filter does.
4770 Document more things not specified clearly in RFC3028. I had all this
4771 sorted out, when out of a sudden new issues came to my mind. Oops."
4773 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4776 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4777 canonical form (as documented). However, after a host name lookup from
4778 the IP address, check_host() was doing a simple string comparison with
4779 addresses acquired from the DNS when checking that the found name did
4780 have the original IP as one of its addresses. Since any found IPv6
4781 addresses are likely to be in abbreviated form, the comparison could
4782 fail. Luckily, there already exists a function for doing the comparison
4783 by converting both addresses to binary, so now that is used instead of
4784 the text comparison.
4786 PH/20 There was another similar case to PH/19, when a complete host name was
4787 given in a host list; looking up its IP address could give an abbreviated
4788 form, whereas the current host's name might or might not be abbreviated.
4789 The same fix has been applied.
4795 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4796 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4799 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4800 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4802 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4804 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4805 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4806 Also, for "Auto-Submitted", treat anything other than "no" as
4807 non-personal, in accordance with RFC 3834. (Previously it treated
4808 anything starting "auto-" as non-personal.)
4810 TF/01 The control=submission/name=... option had a problem with syntax
4811 errors if the name included a slash character. The /name= option
4812 now slurps the rest of the string, so it can include any characters
4813 but it must come last in the list of options (after /sender_retain
4816 PH/05 Some modifications to the interface to the fake nameserver for the new
4824 TK/01 Added the "success_on_redirect" address verification option. See
4825 NewStuff for rationale and an example.
4827 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4829 PH/02 Patch to exigrep to allow it to work on syslog lines.
4831 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4832 fread() to skip over the body file's header line, because in Cygwin the
4833 header line is locked and is inaccessible.
4835 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4836 co-exist for some time) to make it clear that it is the Exim ID that is
4837 referenced, not the Message-ID: header line.
4839 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4840 string_format() function, because snprintf() does not exist on all
4843 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4844 ${stat: expansion item.
4846 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4847 protocol synchronization error", to keep the pedants happy.
4849 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4850 well as for IRIX systems, when gcc is being used. See the host.c source
4853 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4855 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4858 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4859 transport and to the smtp transport in LMTP mode.
4861 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4863 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4864 run a filter in a subprocess. This could lead to confusion in subsequent
4865 lookups in the parent process. There should also be a search_tidyup() at
4866 the end of the subprocess.
4868 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4869 only if the host matched helo_try_verify_hosts, which caused the
4870 verification to occur when the EHLO/HELO command was issued. The ACL just
4871 tested the remembered result. Now, if a previous verification attempt has
4872 not happened, "verify = helo" does it there and then.
4874 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4876 TK/03 Fix log output including CR from clamd.
4878 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4879 exist provoked a memory error which could cause a segfault.
4881 PH/15 Installed PCRE 6.2
4883 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4885 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4886 of the problem. Specifically, suggested +O2 rather than +O1 for the
4889 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4891 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4892 record, the port was not correctly logged when the outgoing_port log
4893 selector was set (it logged the transort's default port).
4895 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4896 fallback_hosts, and "hosts" in the smtp transport.
4898 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4899 host errors such as "Connection refused".
4901 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4902 authentication with radiusclient 0.4.9:
4904 - Error returned from rc_read_config was caught wrongly
4905 - Username/password not passed on to radius server due to wrong length.
4907 The presumption is that some radiusclient API changes for 4.51/PH/17
4908 were not taken care of correctly. The code is still untested by me (my
4909 Linux distribution still has 0.3.2 of radiusclient), but it was
4910 contributed by a Radius user.
4912 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4913 expanding the "port" option of the smtp transport.
4915 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4916 while reading a MIME header. Thanks to Tom Hughes for a patch.
4918 PH/24 Include config.h inside local_scan.h so that configuration settings are
4921 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4922 that in an ACL for RCPT (for example), you can examine exactly what was
4925 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4926 commands, but it was not correctly comparing the address with the actual
4927 client host address. Thus, it would show the EHLO address in Received:
4928 header lines when this was not necessary.
4930 PH/27 Added the % operator to ${eval:}.
4932 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4933 it should be ignoring failures (because with -C, for example, it has lost
4934 privilege). It wasn't ignoring creation failures other than "already
4937 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4940 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4941 because an input line was too long, either on its own, or by virtue of
4942 too many continuations, the temporary file was not being removed, and the
4943 return code was incorrect.
4945 PH/31 Missing "BOOL" in function definition in filtertest.c.
4947 PH/32 Applied Sieve patches from the maintainer.
4949 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4951 PH/33 Added "verify = not_blind".
4953 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4954 Local/Makefile (with some defaults set). These are used in built scripts
4955 such as exicyclog, but they have never been used in the exim_install
4956 script (though there are many overriding facilities there). I have
4957 arranged that the exim_install script now takes note of these two
4960 PH/35 Installed configuration files for Dragonfly.
4962 PH/36 When a locally submitted message by a trusted user did not contain a
4963 From: header, and the sender address was obtained from -f or from an SMTP
4964 MAIL command, and the trusted user did not use -F to supply a sender
4965 name, $originator_name was incorrectly used when constructing a From:
4966 header. Furthermore, $originator_name was used for submission mode
4967 messages from external hosts without From: headers in a similar way,
4968 which is clearly wrong.
4970 PH/37 Added control=suppress_local_fixups.
4972 PH/38 When log_selector = +received_sender was set, and the addition of the
4973 sender made the log line's construction buffer exactly full, or one byte
4974 less than full, an overflow happened when the terminating "\n" was
4977 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4978 when the result of a list match is failure because a DNS lookup failed.
4980 PH/40 RM_COMMAND is now used in the building process.
4982 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4983 the "build-* directories that it finds.
4985 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4986 address in a domain literal was a prefix of an interface address.
4988 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4989 when verifying a sender address, unless rewrite_headers is false.
4991 PH/44 Wrote a long comment about why errors_to addresses are verified as
4992 recipients, not senders.
4994 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4995 the ratelimit ACL was added.
4997 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4999 PH/46 Added extra information about PostgreSQL errors to the error string.
5001 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5002 avoiding the need to install special zones in a real server. This is
5003 backwards compatible; if it can't find the fake resolver, it drops back.
5004 Thus, both old and new test suites can be run.
5006 TF/02 Added util/ratelimit.pl
5008 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5009 clock is set back in time.
5011 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5012 Candler <B.Candler@pobox.com>.
5014 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5015 OK for addresses that are the result of redirections.
5017 PH/48 A number of further additions for the benefit of the new test suite,
5018 including a fake gethostbyname() that interfaces to the fake DNS resolver
5021 TF/06 The fix for widen_domains has also been applied to qualify_single and
5022 search_parents which are the other dnslookup options that can cause
5025 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5028 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5030 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5031 service" if it encountered an MX record with an empty target hostname.
5032 The message is now "an MX or SRV record indicated no SMTP service".
5034 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5035 if the DNS of the sending site is misconfigured. This is quite a
5036 common situation. This change restores the behaviour of treating a
5037 helo verification defer as a failure.
5039 PH/51 If self=fail was set on a router, the bounce message did not include the
5040 actual error message.
5046 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5048 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5049 command itself ended in error, the underlying message about the transport
5050 filter timeout was being overwritten with the pipe command error. Now the
5051 underlying error message should be appended to the second error message.
5053 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5055 PH/02 Reduce the amount of output that "make" produces by default. Full output
5056 can still be requested.
5058 PH/03 The warning log line about a condition test deferring for a "warn" verb
5059 was being output only once per connection, rather than after each
5060 occurrence (because it was using the same function as for successful
5061 "warn" verbs). This seems wrong, so I have changed it.
5063 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5064 it should not have, which might have caused a crash in the right
5065 circumstances, but probably never did.
5067 PH/04 Installed a modified version of Tony Finch's patch to make submission
5068 mode fix the return path as well as the Sender: header line, and to
5069 add a /name= option so that you can make the user's friendly name appear
5072 TF/03 Added the control = fakedefer ACL modifier.
5074 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5075 Mark Lowes for thorough testing.
5077 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5079 TK/03 Merged latest SRS patch from Miles Wilton.
5081 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5082 with the definition in sysexits.h (which is #included earlier).
5083 Fortunately, Exim does not actually use EX_OK. The code used to try to
5084 preserve the sysexits.h value, by assuming that macro definitions were
5085 scanned for macro replacements. I have been disabused of this notion,
5086 so now the code just undefines EX_OK before #including unistd.h.
5088 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5089 in the smtp transport. When a block could not be written in a single
5090 write() function, the timeout was being re-applied to each part-write.
5091 This seems wrong - if the receiver was accepting one byte at a time it
5092 would take for ever. The timeout is now adjusted when this happens. It
5093 doesn't have to be particularly precise.
5095 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5096 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5098 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5099 without a domain if the check to <postmaster@domain> fails.
5101 SC/01 Eximstats: added -xls and the ability to specify output files
5102 (patch written by Frank Heydlauf).
5104 SC/02 Eximstats: use FileHandles for outputting results.
5106 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5108 SC/04 Eximstats: fixed display of large numbers with -nvr option
5110 SC/05 Eximstats: fixed merging of reports with empty tables.
5112 SC/06 Eximstats: added the -include_original_destination flag
5114 SC/07 Eximstats: removed tabs and trailing whitespace.
5116 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5118 TK/06 MBOX spool code: Add real "From " MBOX separator line
5119 so the .eml file is really in mbox format (even though
5120 most programs do not really care). Patch from Alex Miller.
5122 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5123 The latter is generated from $received_to and is only set if the
5124 message has one envelope recipient. SA can use these headers,
5125 obviously out-of-the-box. Patch from Alex Miller.
5127 PH/08 The ${def test on a variable was returning false if the variable's
5128 value was "0", contrary to what the specification has always said!
5129 The result should be true unless the variable is empty.
5131 PH/09 The syntax error of a character other than { following "${if
5132 def:variable_name" (after optional whitespace) was not being diagnosed.
5133 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5134 accidental colon was present, for example, could give incorrect results.
5136 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5137 result is used (not including appendfile, where other changes are about
5140 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5141 This involved changing a lot of size variables from int to off_t. It
5142 should work with maildirs and everything.
5144 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5145 spamd dying while we are connected to it.
5147 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5150 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5151 function for BDB 4.3.
5153 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5155 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5156 the sources for PCRE have been re-arranged and more files are now
5159 PH/15 The code I had for printing potentially long long variables in PH/11
5160 above was not the best (it lost precision). The length of off_t variables
5161 is now inspected at build time, and an appropriate printing format (%ld
5162 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5163 to be "long long int" or "long int". This is needed for the internal
5164 formatting function string_vformat().
5166 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5167 the configuration file to be ":syslog", then the script "guesses" where
5168 the logs files are, rather than using the compiled in default. In our
5169 case the guess is not the same as the compiled default, so the script
5170 suddenly stopped working when I started to use syslog. The patch checks
5171 to see if log_file_path is "". If so, it attempts to read it from exim
5172 with no configuration file to get the compiled in version, before it
5173 falls back to the previous guessing code."
5175 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5176 implementing BATV in an Exim configuration. See NewStuff for the gory
5179 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5180 Makefile that are specific to HP-UX.
5182 PH/18 If the "use_postmaster" option was set for a recipient callout together
5183 with the "random" option, the postmaster address was used as the MAIL
5184 FROM address for the random test, but not for the subsequent recipient
5185 test. It is now used for both.
5187 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5188 patch removes a few documentation additions to RFC 3028, because the
5189 latest draft now contains them. It adds the new en;ascii-case comparator
5190 and a new error check for 8bit text in MIME parts. Comparator and
5191 require names are now matched exactly. I enabled the subaddress
5192 extension, but it is not well tested yet (read: it works for me)."
5194 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5195 rework some of the code of TK/09 above to avoid the hardwired use of
5196 "%lld" and "long long". Replaced the call to snprintf() with a call to
5199 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5200 records point to non-existent hosts", "retry timeout exceeded", and
5201 "retry time not reached for any host after a long failure period".
5203 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5204 experimental DomainKeys support:
5206 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5207 (2) On an error such as an illegally used "control", the wrong name for
5208 the control was given.
5210 These problems did NOT occur unless DomainKeys support was compiled.
5212 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5214 PH/24 Added ${if match_ip condition.
5216 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5217 databases so that it will be absolutely obvious if a crash occurs in the
5218 DB library. This is a regular occurrence (often caused by mis-matched
5221 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5222 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5223 picked up on a user's system that detects such things. There doesn't seem
5224 to be a gcc warning option for this - only an attribute that has to be
5225 put on the function's prototype. It seems that in Fedora Core 4 they have
5226 set this on a number of new functions. No doubt there will be more in due
5229 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5230 specify a transport. However, if an address that was verified by such a
5231 router was the subject of a callout, Exim crashed because it tried to
5232 read the rcpt_include_affixes from the non-existent transport. Now it
5233 just assumes that the setting of that option is false. This bug was
5234 introduced by 4.51/PH/31.
5236 PH/28 Changed -d+all to exclude +memory, because that information is very
5237 rarely of interest, but it makes the output a lot bigger. People tend to
5238 do -d+all out of habit.
5240 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5241 code in os-type was giving problems when libc.so lives in lib64, like on
5244 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5245 aren't the modern standard, and it seems that some systems' include files
5246 don't always have them. Exim was already checking for some of the newer
5247 ones like T_AAAA, and defining it itself. I've added checks for all the
5248 record types that Exim uses.
5250 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5251 not automatically generating a new one, as it is supposed to. This
5252 prevented TLS from working. If the file did exist, but contained invalid
5253 data, a new version was generated, as expected. It was only the case of a
5254 non-existent file that was broken.
5256 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5257 with a change in libdomainkeys > 0.64.
5259 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5260 from DNS. If the selector record carries the flag, it now has
5261 precedence over the domain-wide flag.
5263 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5265 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5266 the use of a port name that isn't defined in /etc/services) occurred, the
5267 message was deferred as in a normal delivery, and thus remained on the
5268 spool, instead of being failed because of the mua_wrapper setting. This
5269 is now fixed, and I tidied up some of the mua_wrapper messages at the
5272 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5273 the messages in a hash of arrays rather than using individual hashes.
5274 This is a bit cleaner and results in dramatic memory savings, albeit
5275 at a slight CPU cost.
5277 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5278 as requested by Marc Sherman.
5280 SC/10 Eximstats: added histograms for user specified patterns as requested
5283 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5285 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5286 fopen() in the content-scanning modules that did not already have it.
5292 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5293 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5295 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5297 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5299 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5300 to test Sieve filters that use "vacation".
5302 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5303 that changes the way the GnuTLS parameters are stored in the cache file.
5304 The new format can be generated externally. For backward compatibility,
5305 if the data in the cache doesn't make sense, Exim assumes it has read an
5306 old-format file, and it generates new data and writes a new file. This
5307 means that you can't go back to an older release without removing the
5310 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5311 work if there are any delivery delays because "one_time" forces the
5312 parent to be marked "delivered", so its unseen clone is never tried
5313 again. For this reason, Exim now forbids the simultaneous setting of
5316 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5317 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5318 entry if you want to know the details. The fix, however, neglected to
5319 consider the case when local delivery batching is involved. The test for
5320 "previously delivered" was not happening when checking to see if an
5321 address could be batched with a previous (undelivered) one; under
5322 certain circumstances this could lead to multiple deliveries to the same
5325 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5326 in its include files, and this causes problems building Exim.
5328 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5329 header_syntax) but Exim was just ignoring anything given after a slash.
5330 In particular, this caused confusion with an attempt to use "verify =
5331 reverse_host_lookup/defer_ok". An error is now given when options are
5332 supplied for verify items that do not have them. (Maybe reverse_host_
5333 lookup should have a defer_ok option, but that's a different point.)
5335 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5336 defined by RFC 821) to 2048, because there were problems with some AUTH
5337 commands, and RFC 1869 says the size should be increased for extended
5338 SMTP commands that take arguments.
5340 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5343 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5344 "unknown" error; now it says that the functionality isn't in the binary.
5346 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5347 an address' error message when a string expansion fails (syntax or
5348 whatever). Otherwise the password may appear in the log. Following change
5349 PH/42 below, there is no longer a chance of it appearing in a bounce
5352 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5354 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5355 of its addresses was ever tried. (Bugzilla bug #2.)
5357 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5358 the result incorrectly in the debug output. (It correctly added a newline
5359 to what was transported.)
5361 TF/01 Added $received_time.
5363 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5364 commented out examples of how to interface to a virus scanner and to
5365 SpamAssassin. Also added commented examples of av_scanner and
5366 spamd_address settings.
5368 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5369 and controls are allowed in which ACLs. There were a couple of minor
5370 errors. Some of the entries in the conditions table (which is a table of
5371 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5372 negation of where the condition IS allowed.
5374 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5376 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5377 header file does not have a version number, so I've had to invent a new
5378 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5379 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5380 radiusclient), but it was contributed by a Radius user.
5382 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5383 files or directories, for OpenSSL.
5385 PH/19 When an Exim process that is running as root has to create an Exim log
5386 file, it does so in a subprocess that runs as exim:exim so as to get the
5387 ownership right at creation (otherwise, other Exim processes might see
5388 the file with the wrong ownership). There was no test for failure of this
5389 fork() call, which would lead to the process getting stuck as it waited
5390 for a non-existent subprocess. Forks do occasionally fail when resources
5391 run out. I reviewed all the other calls to fork(); they all seem to check
5394 PH/20 When checking for unexpected SMTP input at connect time (before writing
5395 the banner), Exim was not dealing correctly with a non-positive return
5396 from the read() function. If the client had disconnected by this time,
5397 the result was a log entry for a synchronization error with an empty
5398 string after "input=" when read() returned zero. If read() returned -1
5399 (an event I could not check), uninitialized data bytes were printed.
5400 There were reports of junk text (parts of files, etc) appearing after
5403 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5405 PH/22 Added support for macro redefinition, and (re)definition in between
5406 driver and ACL definitions.
5408 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5409 forgetting to use the resulting value; it was using the unexpanded value.
5411 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5412 hadn't been configured. The fix is from Juergen Kreileder, who
5413 understands it better than I do:
5415 "Here's what I see happening with three configured cyrus_sasl
5416 authenticators configured (plain, login, cram-md5):
5418 On startup auth_cyrus_sasl_init() gets called for each of these.
5419 This means three calls to sasl_listmech() without a specified mech_list.
5420 => SASL tests which mechs of all available mechs actually work
5421 => three warnings about OTP not working
5422 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5424 With the patch, sasl_listmech() also gets called three times. But now
5425 SASL's mech_list option is set to the server_mech specified in the the
5426 authenticator. Or in other words, the answer from sasl_listmech()
5427 gets limited to just the mech you're testing for (which is different
5429 => the return list contains just 'plain' or 'login', 'cram-md5' or
5430 nothing depending on the value of ob->server_mech.
5432 I've just tested the patch: Authentication still works fine,
5433 unavailable mechs specified in the exim configuration are still
5434 caught, and the auth.log warnings about OTP are gone."
5436 PH/25 When debugging is enabled, the contents of the command line are added
5437 to the debugging output, even when log_selector=+arguments is not
5440 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5441 answer is "GNU", and only if the return is "GNU/something" is the answer
5444 PH/27 $acl_verify_message is now set immediately after the failure of a
5445 verification in an ACL, and so is available in subsequent modifiers. In
5446 particular, the message can be preserved by coding like this:
5448 warn !verify = sender
5449 set acl_m0 = $acl_verify_message
5451 Previously, $acl_verify_message was set only while expanding "message"
5452 and "log_message" when a very denied access.
5454 PH/28 Modified OS/os.c-Linux with
5456 -#ifndef OS_LOAD_AVERAGE
5457 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5459 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5460 nomenclature these days.)
5462 PH/29 Installed patch from the Sieve maintainer that adds the options
5463 sieve_useraddress and sieve_subaddress to the redirect router.
5465 PH/30 In these circumstances:
5466 . Two addresses routed to the same list of hosts;
5467 . First host does not offer TLS;
5468 . First host accepts first address;
5469 . First host gives temporary error to second address;
5470 . Second host offers TLS and a TLS session is established;
5471 . Second host accepts second address.
5472 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5473 and peerdn, if requested) that were in fact used only for the second
5476 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5477 attention to any local part prefix or suffix that was matched by the
5478 router that accepted the address. It now behaves in the same way as it
5479 does for delivery: the affixes are removed from the local part unless
5480 rcpt_include_affixes is set on the transport.
5482 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5483 timeout during the DATA phase of an incoming message.
5485 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5486 applied a patch sanctioned by the Sieve maintainer.
5488 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5489 the uid or gid is negative. A case of a negative gid caused this to be
5490 noticed. The fix allows for either to be negative.
5492 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5493 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5496 PH/36 The change PH/12 above was broken. Fixed it.
5498 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5499 the grounds that routing the same address twice would always produce the
5500 same answer. This might have been true once, but it is certainly no
5501 longer true now. Routing a child address may depend on the previous
5502 routing that produced that child. Some complicated redirection strategies
5503 went wrong when messages had multiple recipients, and made Exim's
5504 behaviour dependent on the order in which the addresses were given.
5506 I have moved the duplicate checking until after the routing is complete.
5507 Exim scans the addresses that are assigned to local and remote
5508 transports, and removes any duplicates. This means that more work will be
5509 done, as duplicates will always all be routed, but duplicates are
5510 presumably rare, so I don't expect this is of any significance.
5512 For deliveries to pipes, files, and autoreplies, the duplicate checking
5513 still happens during the routing process, since they are not going to be
5516 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5517 It corrects a timeout issue with spamd. This is Ian's comment: "The
5518 background is that sometimes spamd either never reads data from a
5519 connection it has accepted, or it never writes response data. The exiscan
5520 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5521 blindly assumes that writes won't block so it may never time out."
5523 PH/39 Allow G after quota size as well as K and M.
5525 PH/40 The value set for $authenticated_id in an authenticator may not contain
5526 binary zeroes or newlines because the value is written to log lines and
5527 to spool files. There was no check on this. Now the value is run through
5528 the string_printing() function so that such characters are converted to
5529 printable escape sequences.
5531 PH/41 $message_linecount is a new variable that contains the total number of
5532 lines in the message. Compare $body_linecount, which is the count for the
5535 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5536 bounce and delay warning messages, except in certain special cases, which
5539 (a) An SMTP error message from a remote host;
5540 (b) A message specified in a :fail: redirection;
5541 (c) A message specified in a "fail" command in a system filter;
5542 (d) A message specified in a FAIL return from the queryprogram router;
5543 (e) A message specified by the cannot_route_message router option.
5545 In these cases only, Exim does include the error details in bounce and
5546 warning messages. There are also a few cases where bland messages such
5547 as "unrouteable address" or "local delivery error" are given.
5549 PH/43 $value is now also set for the "else" part of a ${run expansion.
5551 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5552 being worked on, but at least Exim now implements the latest version to
5555 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5556 process to complete was treated as a delivery failure, a timeout while
5557 writing the message to the pipe was logged, but erroneously treated as a
5558 successful delivery. Such timeouts include transport filter timeouts. For
5559 consistency with the overall process timeout, these timeouts are now
5560 treated as errors, giving rise to delivery failures by default. However,
5561 there is now a new Boolean option for the pipe transport called
5562 timeout_defer, which, if set TRUE, converts the failures into defers for
5563 both kinds of timeout. A transport filter timeout is now identified in
5566 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5567 systems where "make" and "gmake" are different, calling "gmake" at top
5568 level broke things. I've arranged for the value of $(MAKE) to be passed
5569 from the Makefile to this script so that it can call the same version of
5573 A note about Exim versions 4.44 and 4.50
5574 ----------------------------------------
5576 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5577 changes of various kinds. As a consequence, a big documentation update was
5578 needed. This delayed the release for rather longer than seemed good, especially
5579 in the light of a couple of (minor) security issues. Therefore, the changes
5580 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5581 release. So 4.44 and 4.50 are in effect two different branches that both start
5584 I have left the 4.50 change log unchanged; it contains all the changes since
5585 4.43. The change log for 4.44 is below; many of its items are identical to
5586 those for 4.50. This seems to be the most sensible way to preserve the
5587 historical information.
5593 1. Minor wording change to the doc/README.SIEVE file.
5595 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5596 computation of the current number of files was incorrect.
5598 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5599 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5602 4. Give more explanation in the error message when the command for a transport
5603 filter fails to execute.
5605 5. There are several places where Exim runs a non-Exim command in a
5606 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5607 was being done only for the command run by the queryprogram router. It is
5608 now done for all such subprocesses. The other cases are: ${run, transport
5609 filters, and the commands run by the lmtp and pipe transports.
5611 6. Added CONFIGURE_GROUP build-time option.
5613 7. Some older OS have a limit of 256 on the maximum number of file
5614 descriptors. Exim was using setrlimit() to set 1000 as a large value
5615 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5616 systems. I've change it so that if it can't get 1000, it tries for 256.
5618 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5619 was an oversight, and furthermore, ever since the addition of extra
5620 controls (e.g. 4.43/32), the checks on when to allow different forms of
5621 "control" were broken. There should now be diagnostics for all cases when a
5622 control that does not make sense is encountered.
5624 9. Added the /retain_sender option to "control=submission".
5626 10. $recipients is now available in the predata ACL (oversight).
5628 11. Tidy the search cache before the fork to do a delivery from a message
5629 received from the command line. Otherwise the child will trigger a lookup
5630 failure and thereby defer the delivery if it tries to use (for example) a
5631 cached ldap connection that the parent has called unbind on.
5633 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5634 of $address_data from the recipient verification was clobbered by the
5635 sender verification.
5637 13. The value of address_data from a sender verification is now available in
5638 $sender_address_data in subsequent conditions in the ACL statement.
5640 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5642 15. Added a new option "connect=<time>" to callout options, to set a different
5645 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5646 was its contents. (It was OK if the option was not defined at all.)
5648 17. A "Completed" log line is now written for messages that are removed from
5649 the spool by the -Mrm option.
5651 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5652 information about exactly what failed.
5654 19. Added -dd to debug only the daemon process.
5656 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5657 handles timeouts, both on the server side and network timeouts. Renamed the
5658 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5660 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5661 It is now set to "smtps".
5663 22. $host_address is now set to the target address during the checking of
5664 ignore_target_hosts.
5666 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5667 being passed; this would have caused $sender_host_name to have been used if
5668 matching the list had actually called for a host name (not very likely,
5669 since this list is usually IP addresses). A host name is now passed as
5672 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5673 code that specifies a non-restarting handler (typically sigaction() in
5674 modern systems) in an attempt to fix a rare and obscure crash bug.
5676 25. Narrowed the window for a race in the daemon that could cause it to ignore
5677 SIGCHLD signals. This is not a major problem, because they are used only to
5678 wake it up if nothing else does.
5680 26. A malformed maildirsize file could cause Exim to calculate negative values
5681 for the mailbox size or file count. Odd effects could occur as a result.
5682 The maildirsize information is now recalculated if the size or filecount
5685 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5686 support for a long time. Removed HAVE_SYS_VFS_H.
5688 28. Installed the latest version of exipick from John Jetmore.
5690 29. In an address list, if the pattern was not a regular expression, an empty
5691 subject address (from a bounce message) matched only if the pattern was an
5692 empty string. Non-empty patterns were not even tested. This was the wrong
5693 because it is perfectly reasonable to use an empty address as part of a
5694 database query. An empty address is now tested by patterns that are
5695 lookups. However, all the other forms of pattern expect the subject to
5696 contain a local part and a domain, and therefore, for them, an empty
5697 address still always fails if the pattern is not itself empty.
5699 30. Exim went into a mad DNS loop when attempting to do a callout where the
5700 host was specified on an smtp transport, and looking it up yielded more
5701 than one IP address.
5703 31. Re-factored the code for checking spool and log partition space into a
5704 function that finds that data and another that does the check. The former
5705 is then used to implement four new variables: $spool_space, $log_space,
5706 $spool_inodes, and $log_inodes.
5708 32. The RFC2047 encoding function was originally intended for short strings
5709 such as real names; it was not keeping to the 75-character limit for
5710 encoded words that the RFC imposes. It now respects the limit, and
5711 generates multiple encoded words if necessary. To be on the safe side, I
5712 have increased the buffer size for the ${rfc2047: expansion operator from
5715 33. It is now permitted to omit both strings after an "if" condition; if the
5716 condition is true, the result is "true". As before, when the second string
5717 is omitted, a false condition yields an empty string. This makes it less
5718 cumbersome to write custom ACL and router conditions.
5720 34. Failure to deliver a bounce message always caused it to be frozen, even if
5721 there was an errors_to setting on the router. The errors_to setting is now
5724 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5725 canonical form (fully expanded) before being placed in
5726 $sender_host_address.
5728 36. The table in the code that translates DNS record types into text (T_A to
5729 "A" for instance) was missing entries for NS and CNAME. It is just possible
5730 that this could have caused confusion if both these types were looked up
5731 for the same domain, because the text type is used as part of Exim's
5732 per-process caching. But the chance of anyone hitting this buglet seems
5735 37. The dnsdb lookup has been extended in a number of ways.
5737 (1) There is a new type, "zns", which walks up the domain tree until it
5738 finds some nameserver records. It should be used with care.
5740 (2) There is a new type, "mxh", which is like "mx" except that it returns
5741 just the host names, not the priorities.
5743 (3) It is now possible to give a list of domains (or IP addresses) to be
5744 looked up. The behaviour when one of the lookups defers can be
5745 controlled by a keyword.
5747 (4) It is now possible to specify the separator character for use when
5748 multiple records are returned.
5750 38. The dnslists ACL condition has been extended: it is now possible to supply
5751 a list of IP addresses and/or domains to be looked up in a particular DNS
5754 39. Added log_selector=+queue_time_overall.
5756 40. When running the queue in the test harness, wait just a tad after forking a
5757 delivery process, to get repeatability of debugging output.
5759 41. Include certificate and key file names in error message when GnuTLS fails
5760 to set them up, because the GnuTLS error message doesn't include the name
5761 of the failing file when there is a problem reading it.
5763 42. Allow both -bf and -bF in the same test run.
5765 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5767 44. The "Exiscan patch" is now merged into the mainline Exim source.
5769 45. Sometimes the final signoff response after QUIT could fail to get
5770 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5771 < 0 before doing a fflush(). This bug looks as though it goes back to the
5772 introduction of TLS in release 3.20, but "sometimes" must have been rare
5773 because the tests only now provoked it.
5775 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5776 (this can affect the format of dates).
5778 47. exim_tidydb, when checking for the continued existence of a message for
5779 which it has found a message-specific retry record, was not finding
5780 messages that were in split spool directories. Consequently, it was
5781 deleting retry records that should have stayed in existence.
5783 48. Steve fixed some bugs in eximstats.
5785 49. The SPA authentication driver was not abandoning authentication and moving
5786 on to the next authenticator when an expansion was forced to fail,
5787 contradicting the general specification for all authenticators. Instead it
5788 was generating a temporary error. It now behaves as specified.
5790 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5791 (the order specifies the preference for clients). The order is now AES256,
5792 AES128, 3DES, ARCFOUR128.
5794 51. Small patch to Sieve code - explicitly set From: when generating an
5797 52. Exim crashed if a remote delivery caused a very long error message to be
5798 recorded - for instance if somebody sent an entire SpamAssassin report back
5799 as a large number of 550 error lines. This bug was coincidentally fixed by
5800 increasing the size of one of Exim's internal buffers (big_buffer) that
5801 happened as part of the Exiscan merge. However, to be on the safe side, I
5802 have made the code more robust (and fixed the comments that describe what
5805 53. Now that there can be additional text after "Completed" in log lines (if
5806 the queue_time_overall log selector is set), a one-byte patch to exigrep
5807 was needed to allow it to recognize "Completed" as not the last thing in
5810 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5811 patch that reportedly fixes this has been added. I am not expert enough to
5812 create a test for it. This is what the patch creator wrote:
5814 "I found a little strange behaviour of ldap code when working with
5815 Windows 2003 AD Domain, where users was placed in more than one
5816 Organization Units. When I tried to give exim partial DN, the exit code
5817 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5818 But simultaneously result of request was absolutely normal ldap result,
5819 so I produce this patch..."
5821 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5822 REFERENCE, so I have modified the code to exclude the patch when that macro
5825 55. Some experimental protocols are using DNS PTR records for new purposes. The
5826 keys for these records are domain names, not reversed IP addresses. The
5827 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5828 leaves it alone. Component reversal etc. now happens only for IP addresses.
5831 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5833 57. Double the size of the debug message buffer (to 2048) so that more of very
5834 long debug lines gets shown.
5836 58. The exicyclog utility now does better if the number of log files to keep
5837 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5839 59. Two changes related to the smtp_active_hostname option:
5841 (1) $smtp_active_hostname is now available as a variable.
5842 (2) The default for smtp_banner uses $smtp_active_hostname instead
5843 of $primary_hostname.
5845 60. The host_aton() function is supposed to be passed a string that is known
5846 to be a valid IP address. However, in the case of IPv6 addresses, it was
5847 not checking this. This is a hostage to fortune. Exim now panics and dies
5848 if the condition is not met. A case was found where this could be provoked
5849 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5850 components; fortuitously, this particular loophole had already been fixed
5851 by change 4.50/55 above.
5853 If there are any other similar loopholes, the new check in host_aton()
5854 itself should stop them being exploited. The report I received stated that
5855 data on the command line could provoke the exploit when Exim was running as
5856 exim, but did not say which command line option was involved. All I could
5857 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5858 running as the user.
5861 61. There was a buffer overflow vulnerability in the SPA authentication code
5862 (which came originally from the Samba project). I have added a test to the
5863 spa_base64_to_bits() function which I hope fixes it.
5866 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5867 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5869 63. The daemon start-up calls getloadavg() while still root for those OS that
5870 need the first call to be done as root, but it missed one case: when
5871 deliver_queue_load_max is set with deliver_drop_privilege. This is
5872 necessary for the benefit of the queue runner, because there is no re-exec
5873 when deliver_drop_privilege is set.
5875 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5876 This has been fixed.
5878 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5879 was in use, was not putting the data itself into the right store pool;
5880 consequently, it could be overwritten for a subsequent message in the same
5881 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5884 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5886 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5887 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5888 the function were treating the return as a boolean value, which happened to
5889 work because 0=false and not-0=true, but is not correct code.
5891 68. The host_aton() function was not handling scoped IPv6 addresses (those
5892 with, for example, "%eth0" on the end) correctly.
5894 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5895 negated items (that is, ~something) in unsigned ints. Some compilers
5896 apparently mutter when there is no cast.
5898 70. If an address verification called from an ACL failed, and did not produce a
5899 user-specific message (i.e. there was only a "system" message), nothing was
5900 put in $acl_verify_message. In this situation, it now puts the system
5903 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5904 session; change 4.31/43 added the unwanted input to the log line - except
5905 that it did not do this in the start of session case. It now does.
5907 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5908 This is wrong and can cause the other end to generate a synchronization
5909 error if it is another Exim or anything else that does the synchronization
5910 check. A QUIT command is no longer sent after a timeout.
5912 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5913 during host lookups.
5915 74. The defer_ok option of callout verification was not working if it was used
5916 when verifying addresses in header lines, that is, for this case:
5918 verify = header_sender/callout=defer_ok
5920 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5921 those file descriptors could be used for SMTP connections. If anything
5922 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5923 could be sent to the SMTP client, causing chaos. The daemon now opens
5924 stdin, stdout, and stderr to /dev/null when it puts itself into the
5927 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5928 log by default. The user can override this with suitable Perl magic.
5930 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5931 the log message when discard triggers, was not working for the DATA ACL or
5932 for the non-SMTP ACL.
5934 78. Error message wording change in sieve.c.
5936 79. If smtp_accept_max_per_host was set, the number of connections could be
5937 restricted to fewer than expected, because the daemon was trying to set up
5938 a new connection before checking whether the processes handling previous
5939 connections had finished. The check for completed processes is now done
5940 earlier. On busy systems, this bug wouldn't be noticed because something
5941 else would have woken the daemon, and it would have reaped the completed
5944 80. If a message was submitted locally by a user whose login name contained one
5945 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5946 It caused a spool format error. I have fixed the spool reading code. A
5947 related problem was that the "from" clause in the Received: line became
5948 illegal because of the space(s). It is now covered by ${quote_local_part.
5950 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5953 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5955 83. Patch from Sieve maintainer to fix unterminated string problem in
5956 "vacation" handling.
5958 84. Some minor changes to the Linux configuration files to help with other
5959 OS variants using glibc.
5961 85. One more patch for Sieve to update vacation handling to latest spec.
5964 ----------------------------------------------------
5965 See the note above about the 4.44 and 4.50 releases.
5966 ----------------------------------------------------
5972 1. Change 4.43/35 introduced a bug that caused file counts to be
5973 incorrectly computed when quota_filecount was set in an appendfile
5976 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5977 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5980 3. Give more explanation in the error message when the command for a transport
5981 filter fails to execute.
5983 4. There are several places where Exim runs a non-Exim command in a
5984 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5985 was being done only for the command run by the queryprogram router. It is
5986 now done for all such subprocesses. The other cases are: ${run, transport
5987 filters, and the commands run by the lmtp and pipe transports.
5989 5. Some older OS have a limit of 256 on the maximum number of file
5990 descriptors. Exim was using setrlimit() to set 1000 as a large value
5991 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5992 systems. I've change it so that if it can't get 1000, it tries for 256.
5994 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5995 was an oversight, and furthermore, ever since the addition of extra
5996 controls (e.g. 4.43/32), the checks on when to allow different forms of
5997 "control" were broken. There should now be diagnostics for all cases when a
5998 control that does not make sense is encountered.
6000 7. $recipients is now available in the predata ACL (oversight).
6002 8. Tidy the search cache before the fork to do a delivery from a message
6003 received from the command line. Otherwise the child will trigger a lookup
6004 failure and thereby defer the delivery if it tries to use (for example) a
6005 cached ldap connection that the parent has called unbind on.
6007 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6008 of $address_data from the recipient verification was clobbered by the
6009 sender verification.
6011 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6012 was its contents. (It was OK if the option was not defined at all.)
6014 11. A "Completed" log line is now written for messages that are removed from
6015 the spool by the -Mrm option.
6017 12. $host_address is now set to the target address during the checking of
6018 ignore_target_hosts.
6020 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6021 being passed; this would have caused $sender_host_name to have been used if
6022 matching the list had actually called for a host name (not very likely,
6023 since this list is usually IP addresses). A host name is now passed as
6026 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6027 code that specifies a non-restarting handler (typically sigaction() in
6028 modern systems) in an attempt to fix a rare and obscure crash bug.
6030 15. Narrowed the window for a race in the daemon that could cause it to ignore
6031 SIGCHLD signals. This is not a major problem, because they are used only to
6032 wake it up if nothing else does.
6034 16. A malformed maildirsize file could cause Exim to calculate negative values
6035 for the mailbox size or file count. Odd effects could occur as a result.
6036 The maildirsize information is now recalculated if the size or filecount
6039 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6040 support for a long time. Removed HAVE_SYS_VFS_H.
6042 18. Updated exipick to current release from John Jetmore.
6044 19. Allow an empty sender to be matched against a lookup in an address list.
6045 Previously the only cases considered were a regular expression, or an
6048 20. Exim went into a mad DNS lookup loop when doing a callout where the
6049 host was specified on the transport, if the DNS lookup yielded more than
6052 21. The RFC2047 encoding function was originally intended for short strings
6053 such as real names; it was not keeping to the 75-character limit for
6054 encoded words that the RFC imposes. It now respects the limit, and
6055 generates multiple encoded words if necessary. To be on the safe side, I
6056 have increased the buffer size for the ${rfc2047: expansion operator from
6059 22. Failure to deliver a bounce message always caused it to be frozen, even if
6060 there was an errors_to setting on the router. The errors_to setting is now
6063 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6064 canonical form (fully expanded) before being placed in
6065 $sender_host_address.
6067 24. Updated eximstats to version 1.33
6069 25. Include certificate and key file names in error message when GnuTLS fails
6070 to set them up, because the GnuTLS error message doesn't include the name
6071 of the failing file when there is a problem reading it.
6073 26. Expand error message when OpenSSL has problems setting up cert/key files.
6076 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6077 (this can affect the format of dates).
6079 28. exim_tidydb, when checking for the continued existence of a message for
6080 which it has found a message-specific retry record, was not finding
6081 messages that were in split spool directories. Consequently, it was
6082 deleting retry records that should have stayed in existence.
6084 29. eximstats updated to version 1.35
6085 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6086 1.35 - bugfix such that pie charts by volume are generated correctly
6088 30. The SPA authentication driver was not abandoning authentication and moving
6089 on to the next authenticator when an expansion was forced to fail,
6090 contradicting the general specification for all authenticators. Instead it
6091 was generating a temporary error. It now behaves as specified.
6093 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6094 (the order specifies the preference for clients). The order is now AES256,
6095 AES128, 3DES, ARCFOUR128.
6097 31. Small patch to Sieve code - explicitly set From: when generating an
6100 32. Exim crashed if a remote delivery caused a very long error message to be
6101 recorded - for instance if somebody sent an entire SpamAssassin report back
6102 as a large number of 550 error lines. This bug was coincidentally fixed by
6103 increasing the size of one of Exim's internal buffers (big_buffer) that
6104 happened as part of the Exiscan merge. However, to be on the safe side, I
6105 have made the code more robust (and fixed the comments that describe what
6108 33. Some experimental protocols are using DNS PTR records for new purposes. The
6109 keys for these records are domain names, not reversed IP addresses. The
6110 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6111 leaves it alone. Component reversal etc. now happens only for IP addresses.
6114 34. The host_aton() function is supposed to be passed a string that is known
6115 to be a valid IP address. However, in the case of IPv6 addresses, it was
6116 not checking this. This is a hostage to fortune. Exim now panics and dies
6117 if the condition is not met. A case was found where this could be provoked
6118 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6119 components; fortuitously, this particular loophole had already been fixed
6120 by change 4.50/55 or 4.44/33 above.
6122 If there are any other similar loopholes, the new check in host_aton()
6123 itself should stop them being exploited. The report I received stated that
6124 data on the command line could provoke the exploit when Exim was running as
6125 exim, but did not say which command line option was involved. All I could
6126 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6127 running as the user.
6130 35. There was a buffer overflow vulnerability in the SPA authentication code
6131 (which came originally from the Samba project). I have added a test to the
6132 spa_base64_to_bits() function which I hope fixes it.
6135 36. The daemon start-up calls getloadavg() while still root for those OS that
6136 need the first call to be done as root, but it missed one case: when
6137 deliver_queue_load_max is set with deliver_drop_privilege. This is
6138 necessary for the benefit of the queue runner, because there is no re-exec
6139 when deliver_drop_privilege is set.
6141 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6142 was in use, was not putting the data itself into the right store pool;
6143 consequently, it could be overwritten for a subsequent message in the same
6144 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6147 38. Sometimes the final signoff response after QUIT could fail to get
6148 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6149 < 0 before doing a fflush(). This bug looks as though it goes back to the
6150 introduction of TLS in release 3.20, but "sometimes" must have been rare
6151 because the tests only now provoked it.
6157 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6158 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6159 Now it returns an integer. A number of calls were still expecting a T/F
6160 return. Fortuitously, in all cases, the tests worked in OK situations,
6161 which is the norm. However, things would have gone wrong on any write
6162 failures on the smtp file descriptor. This function is used when sending
6163 messages over SMTP and also when doing verify callouts.
6165 2. When Exim is called to do synchronous delivery of a locally submitted
6166 message (the -odf or -odi options), it no longer closes stderr before doing
6169 3. Implemented the mua_wrapper option.
6171 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6173 5. Implemented the functions header_remove(), header_testname(),
6174 header_add_at_position(), and receive_remove_recipient(), and exported them
6177 6. If an ACL "warn" statement specified the addition of headers, Exim already
6178 inserted X-ACL-Warn: at the start if there was no header name. However, it
6179 was not making this test for the second and subsequent header lines if
6180 there were newlines in the string. This meant that an invalid header could
6181 be inserted if Exim was badly configured.
6183 7. Allow an ACL "warn" statement to add header lines at the start or after all
6184 the Received: headers, as well as at the end.
6186 8. Added the rcpt_4xx retry error code.
6188 9. Added postmaster_mailfrom=xxx to callout verification option.
6190 10. Added mailfrom=xxxx to the callout verification option, for verify=
6193 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6194 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6196 12. Inserted some casts to stop certain compilers warning when using pointer
6197 differences as field lengths or precisions in printf-type calls (mostly
6198 affecting debugging statements).
6200 13. Added optional readline() support for -be (dynamically loaded).
6202 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6203 same clock tick as a message's arrival, so that its received time was the
6204 same as the "first fail" time on the retry record, and that message
6205 remained on the queue past the ultimate address timeout, every queue runner
6206 would try a delivery (because it was past the ultimate address timeout) but
6207 after another failure, the ultimate address timeout, which should have then
6208 bounced the address, did not kick in. This was a "< instead of <=" error;
6209 in most cases the first failure would have been in the next clock tick
6210 after the received time, and all would be well.
6212 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6213 being recognized when the domain list was tested by the match_domain
6214 condition in an expansion string.
6216 16. Added the ${str2b64: operator.
6218 17. Exim was always calling setrlimit() to set a large limit for the number of
6219 processes, without checking whether the existing limit was already
6220 adequate. (It did check for the limit on file descriptors.) Furthermore,
6221 errors from getrlimit() and setrlimit() were being ignored. Now they are
6222 logged to the main and panic logs, but Exim does carry on, to try to do its
6223 job under whatever limits there are.
6225 18. Imported PCRE 5.0.
6227 19. Trivial typo in log message " temporarily refused connection" (the leading
6230 20. If the log selector return_path_on_delivery was set and an address was
6231 redirected to /dev/null, the delivery process crashed because it assumed
6232 that a return path would always be set for a "successful" delivery. In this
6233 case, the whole delivery is bypassed as an optimization, and therefore no
6236 21. Internal re-arrangement: the function for sending a challenge and reading
6237 a response while authentication was assuming a zero-terminated challenge
6238 string. It's now changed to take a pointer and a length, to allow for
6239 binary data in such strings.
6241 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6243 23. Exim was not respecting finduser_retries when seeking the login of the
6244 uid under which it was called; it was always trying 10 times. (The default
6245 setting of finduser_retries is zero.) Also, it was sleeping after the final
6246 failure, which is pointless.
6248 24. Implemented tls_on_connect_ports.
6250 25. Implemented acl_smtp_predata.
6252 26. If the domain in control=submission is set empty, Exim assumes that the
6253 authenticated id is a complete email address when it generates From: or
6254 Sender: header lines.
6256 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6257 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6258 chown and chgrp in /bin and hostname in /usr/bin.
6260 28. Exim was keeping the "process log" file open after each use, just as it
6261 does for the main log. This opens the possibility of it remaining open for
6262 long periods when the USR1 signal hits a daemon. Occasional processlog
6263 errors were reported, that could have been caused by this. Anyway, it seems
6264 much more sensible not to leave this file open at all, so that is what now
6267 29. The long-running daemon process does not normally write to the log once it
6268 has entered its main loop, and it closes the log before doing so. This is
6269 so that log files can straightforwardly be renamed and moved. However,
6270 there are a couple of unusual error situations where the daemon does write
6271 log entries, and I had neglected to close the log afterwards.
6273 30. The text of an SMTP error response that was received during a remote
6274 delivery was being truncated at 512 bytes. This is too short for some of
6275 the long messages that one sometimes sees. I've increased the limit to
6278 31. It is now possible to make retry rules that apply only when a message has a
6279 specific sender, in particular, an empty sender.
6281 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6282 it possible to be selective about when SMTP synchronization is enforced.
6284 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6286 32. Implemented hosts_connection_nolog.
6288 33. Added an ACL for QUIT.
6290 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6293 35. Added mailbox_size and mailbox_filecount to appendfile.
6295 36. Added control = no_multiline_responses to ACLs.
6297 37. There was a bug in the logic of the code that waits for the clock to tick
6298 in the case where the clock went backwards by a substantial amount such
6299 that the microsecond fraction of "now" was more than the microsecond
6300 fraction of "then" (but the whole seconds number was less).
6302 38. Added support for the libradius Radius client library this is found on
6303 FreeBSD (previously only the radiusclient library was supported).
6309 1. When certain lookups returned multiple values in the form name=value, the
6310 quoting of the values was not always being done properly. Specifically:
6311 (a) If the value started with a double quote, but contained no whitespace,
6313 (b) If the value contained whitespace other than a space character (i.e.
6314 tabs or newlines or carriage returns) it was not quoted.
6315 This fix has been applied to the mysql and pgsql lookups by writing a
6316 separate quoting function and calling it from the lookup code. The fix
6317 should probably also be applied to nisplus, ibase and oracle lookups, but
6318 since I cannot test any of those, I have not disturbed their existing code.
6320 2. A hit in the callout cache for a specific address caused a log line with no
6321 reason for rejecting RCPT. Now it says "Previous (cached) callout
6322 verification failure".
6324 3. There was an off-by-one bug in the queryprogram router. An over-long
6325 return line was truncated at 256 instead of 255 characters, thereby
6326 overflowing its buffer with the terminating zero. As well as fixing this, I
6327 have increased the buffer size to 1024 (and made a note to document this).
6329 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6330 when Exim is waiting for an SMTP response from a remote server, Exim
6331 restarts its select() call on the socket, thereby resetting its timeout.
6332 This is not a problem when such interrupts are rare. Somebody set up a cron
6333 job to run exiwhat every 2 minutes, which is less than the normal select()
6334 timeout (5 or 10 minutes). This meant that the select() timeout never
6335 kicked in because it was always reset. I have fixed this by comparing the
6336 time when an interrupt arrives with the time at the start of the first call
6337 to select(). If more time than the timeout has elapsed, the interrupt is
6338 treated as a timeout.
6340 5. Some internal re-factoring in preparation for the addition of Sieve
6341 extensions (by MH). In particular, the "personal" test is moved to a
6342 separate function, and given an option for scanning Cc: and Bcc: (which is
6343 not set for Exim filters).
6345 6. When Exim created an email address using the login of the caller as the
6346 local part (e.g. when creating a From: or Sender: header line), it was not
6347 quoting the local part when it contained special characters such as @.
6349 7. Installed new OpenBSD configuration files.
6351 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6352 try to make them clearer.
6354 9. Callout options, other than the timeout value, were being ignored when
6355 verifying sender addresses in header lines. For example, when using
6357 verify = header_sender/callout=no_cache
6359 the cache was (incorrectly) being used.
6361 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6362 only the Cygwin environment.
6364 11. When return_path_on_delivery was set as a log selector, if different remote
6365 addresses in the same message used different return paths and parallel
6366 remote delivery occurred, the wrong values would sometimes be logged.
6367 (Whenever a remote delivery process finished, the return path value from
6368 the most recently started remote delivery process was logged.)
6370 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6371 lines when AUTH and/or TLS are in use. This is the "received protocol"
6372 field. Exim used to use "asmtp" for authenticated SMTP, without any
6373 indication (in the protocol name) for TLS use. Now it follows the RFC and
6374 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6375 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6376 names appear in log lines as well as in Received: header lines.
6378 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6379 extensions, and comparison tests, and to fix some bugs.
6381 14. Changes to the "personal" filter test:
6383 (1) The test was buggy in that it was just doing the equivalent of
6384 "contains" tests on header lines. For example, if a user's address was
6385 anne@some.where, the "personal" test would incorrectly be true for
6387 To: susanne@some.where
6389 This test is now done by extracting each address from the header in turn,
6390 and checking the entire address. Other tests that are part of "personal"
6391 are now done using regular expressions (for example, to check local parts
6392 of addresses in From: header lines).
6394 (2) The list of non-personal local parts in From: addresses has been
6395 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6396 taken from the Sieve specification recommendations.
6398 (3) If the message contains any header line starting with "List-" it is
6399 treated as non-personal.
6401 (4) The test for "circular" in the Subject: header line has been removed
6402 because it now seems ill-conceived.
6404 15. Minor typos in src/EDITME comments corrected.
6406 16. Installed latest exipick from John Jetmore.
6408 17. If headers_add on a router specified a text string that was too long for
6409 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6410 of string_sprintf() is now avoided.
6412 18. $message_body_size was not set (it was always zero) when running the DATA
6413 ACL and the local_scan() function.
6415 19. For the "mail" command in an Exim filter, no default was being set for
6416 the once_repeat time, causing a random time value to be used if "once" was
6417 specified. (If the value happened to be <= 0, no repeat happened.) The
6418 default is now 0s, meaning "never repeat". The "vacation" command was OK
6419 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6420 (I found it when inspecting the code).
6422 20. There is now an overall timeout for performing a callout verification. It
6423 defaults to 4 times the callout timeout, which applies to individual SMTP
6424 commands during the callout. The overall timeout applies when there is more
6425 than one host that can be tried. The timeout is checked before trying the
6426 next host. This prevents very long delays if there are a large number of
6427 hosts and all are timing out (e.g. when the network connections are timing
6428 out). The value of the overall timeout can be changed by specifying an
6429 additional sub-option for "callout", called "maxwait". For example:
6431 verify = sender/callout=5s,maxwait=20s
6433 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6434 to the end before writing, but this should make it even safer).
6436 22. Exim was forgetting that it had advertised PIPELINING for the second and
6437 subsequent messages on an SMTP connection. It was also not resetting its
6438 memory on STARTTLS and an internal HELO.
6440 23. When Exim logs an SMTP synchronization error within a session, it now
6441 records whether PIPELINING has been advertised or not.
6443 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6444 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6445 rather than long int.
6447 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6449 26. Added the never_mail option to autoreply.
6455 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6456 crash if the getsockname() call failed; this can happen if a connection is
6457 closed very soon after it is established. The problem was simply in the
6458 order in which certain operations were done, causing Exim to try to write
6459 to the SMTP stream before it had set up the file descriptor. The bug has
6460 been fixed by making things happen in the correct order.
6466 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6467 before the connection was closed, thus losing the rejection response.
6469 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6470 some early Solaris releases, but causes trouble in current releases where
6471 socklen_t is defined.
6473 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6476 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6479 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6480 that releasing the top of it at the end releases what was used for sub-
6481 expansions (unless the block got too big). However, discard this block if
6482 the first thing is a variable or header, so that we can use its block when
6483 it is dynamic (useful for very large $message_headers, for example).
6485 6. Lookups now cache *every* query, not just the most recent. A new, separate
6486 store pool is used for this. It can be recovered when all lookup caches are
6487 flushed. Lookups now release memory at the end of their result strings.
6488 This has involved some general refactoring of the lookup sources.
6490 7. Some code has been added to the store_xxx() functions to reduce the amount
6491 of flapping under certain conditions.
6493 8. log_incoming_interface used to affect only the <= reception log lines. Now
6494 it causes the local interface and port to be added to several more SMTP log
6495 lines, for example "SMTP connection from", and rejection lines.
6497 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6499 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6501 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6502 could be overwritten at the end of the current message (or the start of a
6503 new message if it was set in a HELO ACL). The value is now preserved for
6504 the duration of the SMTP connection.
6506 12. If a transport had a headers_rewrite setting, and a matching header line
6507 contained an unqualified address, that address was qualified, even if it
6508 did not match any rewriting rules. The underlying bug was that the values
6509 of the flags that permit the existence of unqualified sender and recipient
6510 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6511 non-local messages, and by -bnq for local messages) were not being
6512 preserved with the message after it was received.
6514 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6515 "next input=" as part of the text comprising the host identity instead of
6516 the correct text. The code was using the same buffer for two different
6517 strings. However, depending on which order the printing function evaluated
6518 its arguments, the bug did not always show up. Under Linux, for example, my
6519 test suite worked just fine.
6521 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6522 doesn't work with some older versions of Perl. It has been changed to "my",
6523 which in any case is probably the better facility to use.
6525 15. A really picky compiler found some instances of statements for creating
6526 error messages that either had too many or two few arguments for the format
6529 16. The size of the buffer for calls to the DNS resolver has been increased
6530 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6531 for addresses that have a lot of PTR records. This alleviates a problem; it
6532 does not fully solve it.
6534 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6535 buffer now truncates the list and logs the incident, which is the same
6536 action as happens when Exim is looking up a host name and its aliases.
6537 Previously in this situation something unpredictable would happen;
6538 sometimes it was "internal error: store_reset failed".
6540 18. If a server dropped the connection unexpectedly when an Exim client was
6541 using GnuTLS and trying to read a response, the client delivery process
6542 crashed while trying to generate an error log message.
6544 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6545 string, for example:
6547 warn message = H1: something\nH2: something
6549 the text was added as a single header line from Exim's point of view
6550 though it ended up OK in the delivered message. However, searching for the
6551 second and subsequent header lines using $h_h2: did not work. This has been
6552 fixed. Similarly, if a system filter added multiple headers in this way,
6553 the routers could not see them.
6555 20. Expanded the error message when iplsearch is called with an invalid key to
6556 suggest using net-iplsearch in a host list.
6558 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6559 ACLs are no longer actually imposed (and a message to that effect is
6562 22. If a "gecos" field in a passwd entry contained escaped characters, in
6563 particular, if it contained a \" sequence, Exim got it wrong when building
6564 a From: or a Sender: header from that name. A second bug also caused
6565 incorrect handling when an unquoted " was present following a character
6566 that needed quoting.
6568 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6569 was not being matched caselessly.
6571 24. Arranged for all hyphens in the exim.8 source to be escaped with
6574 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6575 properly. Recipient callout cache records were still being keyed to include
6576 the sender, even when use_sender was set false. This led to far more
6577 callouts that were necessary. The sender is no longer included in the key
6578 when use_sender is false.
6580 26. Added "control = submission" modifier to ACLs.
6582 27. Added the ${base62d: operator to decode base 62 numbers.
6584 28. dnsdb lookups can now access SRV records.
6586 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6587 the configuration file.
6589 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6590 (-v) mode. This makes the output for a verbose queue run more intelligible.
6592 31. Added a use_postmaster feature to recipient callouts.
6594 32. Added the $body_zerocount variable, containing the number of binary zero
6595 bytes in the message body.
6597 33. The time of last modification of the "new" subdirectory is now used as the
6598 "mailbox time last read" when there is a quota error for a maildir
6601 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6603 35. Added +ignore_unknown as a special item in host lists.
6605 36. Code for decoding IPv6 addresses in host lists is now included, even if
6606 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6607 address was recognized as an IP address, but was then not correctly decoded
6608 into binary, causing unexpected and incorrect effects when compared with
6615 1. Very minor rewording of debugging text in manualroute to say "list of
6616 hosts" instead of "hostlist".
6618 2. If verify=header_syntax was set, and a header line with an unqualified
6619 address (no domain) and a large number of spaces between the end of the
6620 name and the colon was received, the reception process suffered a buffer
6621 overflow, and (when I tested it) crashed. This was caused by some obsolete
6622 code that should have been removed. The fix is to remove it!
6624 3. When running in the test harness, delay a bit after writing a bounce
6625 message to get a bit more predictability in the log output.
6627 4. Added a call to search_tidyup() just before forking a reception process. In
6628 theory, someone could use a lookup in the expansion of smtp_accept_max_
6629 per_host which, without the tidyup, could leave open a database connection.
6631 5. Added the variables $recipient_data and $sender_data which get set from a
6632 lookup success in an ACL "recipients" or "senders" condition, or a router
6633 "senders" option, similar to $domain_data and $local_part_data.
6635 6. Moved the writing of debug_print from before to after the "senders" test
6638 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6639 problems for message scanning, either using a data ACL, or using
6640 local_scan() because the Received: header was not generated till after they
6641 were called (in order to set the time as the time of reception completion).
6642 I have revised the way this works. The header is now generated after the
6643 body is received, but before the ACL or local_scan() are called. After they
6644 are run, the timestamp in the header is updated.
6650 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6651 before starting a queue runner without re-exec. This happened only when
6652 deliver_drop_privilege was set or when the Exim user was set to root. The
6653 effect of the bug was that timeouts during subsequent deliveries caused
6654 crashes instead of being properly handled. The handler is now left at its
6655 default (and expected) setting.
6657 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6658 message, again when deliver_drop_privilege is set or Exim is run as root.
6659 The bug described in (1) was not present in this case, but the tidying up
6660 of the other signals was missing. I have made the two cases consistent.
6662 3. The ignore_target_hosts setting on a manualroute router was being ignored
6663 for hosts that were looked up using the /MX notation.
6665 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6668 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6669 operated on the sender address. After changing the $sender_address to <>
6670 for the sender address verify, Exim was re-instated it as the original
6671 (before rewriting) address, but remembering that it had rewritten it, so it
6672 wasn't rewriting it again. This bug also had the effect of breaking the
6673 sender address verification caching when the sender address was rewritten.
6675 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6676 This has been changed so that if the ip literal address matches
6677 ignore_target_hosts, the router declines.
6679 7. Added expansion conditions match_domain, match_address, and match_local_
6680 part (NOT match_host).
6682 8. The placeholder for the Received: header didn't have a length field set.
6684 9. Added code to Exim itself and to exim_lock to test for a specific race
6685 condition that could lead to file corruption when using MBX delivery. The
6686 issue is with the lockfile that is created in /tmp. If this file is removed
6687 after a process has opened it but before that process has acquired a lock,
6688 there is the potential for a second process to recreate the file and also
6689 acquire a lock. This could lead to two Exim processes writing to the file
6690 at the same time. The added code performs the same test as UW imapd; it
6691 checks after acquiring the lock that its file descriptor still refers to
6692 the same named file.
6694 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6695 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6698 11. Added the smtp_active_hostname option. If used, this will typically be made
6699 to depend on the incoming interface address. Because $interface_address is
6700 not set up until the daemon has forked a reception process, error responses
6701 that can happen earlier (such as "too many connections") no longer contain
6704 12. If an expansion in a condition on a "warn" statement fails because a lookup
6705 defers, the "warn" statement is abandoned, and the next ACL statement is
6706 processed. Previously this caused the whole ACL to be aborted.
6708 13. Added the iplsearch lookup type.
6710 14. Added ident_timeout as a log selector.
6712 15. Added tls_certificate_verified as a log selector.
6714 16. Added a global option tls_require_ciphers (compare the smtp transport
6715 option of the same name). This controls incoming TLS connections.
6717 17. I finally figured out how to make tls_require_ciphers do a similar thing
6718 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6719 before starting the TLS session.
6721 18. Tabs are now shown as \t in -bP output.
6723 19. If the log selector return_path_on_delivery was set, Exim crashed when
6724 bouncing a message because it had too many Received: header lines.
6726 20. If two routers both had headers_remove settings, and the first one included
6727 a superfluous trailing colon, the final name in the first list and the
6728 first name in the second list were incorrectly joined into one item (with a
6729 colon in the middle).
6735 1. Added -C and -D options to the exinext utility, mainly to make it easier
6736 to include in the automated testing, but these could be helpful when
6737 multiple configurations are in use.
6739 2. The exinext utility was not formatting the output nicely when there was
6740 an alternate port involved in the retry record key, nor when there was a
6741 message id as well (for retries that were specific to a specific message
6742 and a specific host). It was also confused by IPv6 addresses, because of
6743 the additional colons they contain. I have fixed the IPv4 problem, and
6744 patched it up to do a reasonable job for IPv6.
6746 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6747 delivery, the log line now contains "pipelined" if PIPELINING was used.
6749 4. An SMTP transport process used to panic and die if the bind() call to set
6750 an explicit outgoing interface failed. This has been changed; it is now
6751 treated in the same way as a connect() failure.
6753 5. A reference to $sender_host_name in the part of a conditional expansion
6754 that was being skipped was still causing a DNS lookup. This no longer
6757 6. The def: expansion condition was not recognizing references to header lines
6758 that used bh_ and bheader_.
6760 7. Added the _cache feature to named lists.
6762 8. The code for checking quota_filecount in the appendfile transport was
6763 allowing one more file than it should have been.
6765 9. For compatibility with Sendmail, the command line option
6773 and sets the incoming protocol and host name (for trusted callers). The
6774 host name and its colon can be omitted when only the protocol is to be set.
6775 Note the Exim already has two private options, -pd and -ps, that refer to
6776 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6777 "s", but I don't think that's a major issue.
6779 10. A number of refactoring changes to the code, none of which should affect
6782 (a) The number of logging options was getting close to filling up the
6783 32-bit word that was used as a bit map. I have split them into two classes:
6784 those that are passed in the argument to log_write(), and those that are
6785 only ever tested independently outside of that function. These are now in
6786 separate 32-bit words, so there is plenty of room for expansion again.
6787 There is no change in the user interface or the logging behaviour.
6789 (b) When building, for example, log lines, the code previously used a
6790 macro that called string_cat() twice, in order to add two strings. This is
6791 not really sufficiently general. Furthermore, there was one instance where
6792 it was actually wrong because one of the argument was used twice, and in
6793 one call a function was used. (As it happened, calling the function twice
6794 did not affect the overall behaviour.) The macro has been replaced by a
6795 function that can join an arbitrary number of extra strings onto a growing
6798 (c) The code for expansion conditions now uses a table and a binary chop
6799 instead of a serial search (which was left over from when there were very
6800 few conditions). Also, it now recognizes conditions like "pam" even when
6801 the relevant support is not compiled in: a suitably worded error message is
6802 given if an attempt is made to use such a condition.
6804 11. Added ${time_interval:xxxxx}.
6806 12. A bug was causing one of the ddress fields not to be passed back correctly
6807 from remote delivery subprocesses. The field in question was not being
6808 subsequently used, so this caused to problems in practice.
6810 13. Added new log selectors queue_time and deliver_time.
6812 14. Might have fixed a bug in maildirsizefile handling that threw up
6813 "unexpected character" debug warnings, and recalculated the data
6814 unnecessarily. In any case, I expanded the warning message to give more
6817 15. Added the message "Restricted characters in address" to the statements in
6818 the default ACL that block characters like @ and % in local parts.
6820 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6821 Three changes have been made:
6823 (a) There was a serious bug; a negative response to MAIL caused the whole
6824 recipient domain to be cached as invalid, thereby blocking all messages
6825 to all local parts at the same domain, from all senders. This bug has
6826 been fixed. The domain is no longer cached after a negative response to
6827 MAIL if the sender used is not empty.
6829 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6832 (c) A new callout option, "use_sender" has been added for people who want
6833 the modified behaviour.
6839 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6842 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6843 indeed breaks things for older releases.
6845 3. Added additional logging to the case where there is a problem reading data
6846 from a filter that is running in a subprocess using a pipe, in order to
6847 try to track down a specific problem.
6849 4. Testing facility fudge: when running in the test harness and attempting
6850 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6851 getting "No route to host". Convert this to a timeout.
6853 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6856 6. Some OS don't have socklen_t but use size_t instead. This affects the
6857 fifth argument of getsockopt() amongst other things. This is now
6858 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6859 can be set for individual OS. I have set it for SunOS5, OSF1, and
6860 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6861 some earlier ones do not.
6863 7. Change 4.30/15 was not doing the test caselessly.
6865 8. The standard form for an IPv6 address literal was being rejected by address
6866 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6867 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6868 this, as well as the form without the "IPv6" on the front (but only when
6869 address literals are enabled, of course).
6871 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6873 10. Exim crashed if a message with an empty sender address specified by -f
6874 encountered a router with an errors_to setting. This could be provoked only
6875 by a command such as
6879 where an empty string was supplied; "<>" did not hit this bug.
6881 11. Installed PCRE release 4.5.
6883 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6884 remained set. It is now erased.
6886 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6887 times from message ids (which are base 36 rather than the normal 62).
6889 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6890 were being counted as actual protocol errors, and logged if the log
6891 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6892 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6893 and DATA following a set of rejected RCPTs do not count as protocol errors.
6894 In other words, Exim assumes they were pipelined, though this may not
6895 actually be the case. Of course, in all cases the client gets an
6896 appropriate error code.
6898 15. If a lookup fails in an ACL condition, a message about the failure may
6899 be available; it is used if testing the ACL cannot continue, because most
6900 such messages specify what the cause of the deferral is. However, some
6901 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6902 that caused an old message to be retained and used if a later statement
6903 caused a defer, replacing the real cause of the deferral.
6905 16. If an IP address had so many PTR records that the DNS lookup buffer
6906 was not large enough to hold them, Exim could crash while trying to process
6907 the truncated data. It now detects and logs this case.
6909 17. Further to 4.21/58, another change has been made: if (and only if) the
6910 first line of a message (the first header line) ends with CRLF, a bare LF
6911 in a subsequent header line has a space inserted after it, so as not to
6912 terminate the header.
6914 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6915 unnecessarily, used atoi() instead of strtol(), and didn't check the
6916 termination when getting file sizes from file names by regex.
6918 19. Completely re-implemented the support for maildirsize files, in the light
6919 of a number of problems with the previous contributed implementation
6920 (4.30/29). In particular:
6922 . If the quota is zero, the maildirsize file is maintained, but no quota is
6925 . If the maildir directory does not exist, it is created before any attempt
6926 to write a maildirsize file.
6928 . The quota value in the file is just a cache; if the quota is changed in
6929 the transport, the new value overrides.
6931 . A regular expression is available for excluding directories from the
6934 20. The autoreply transport checks the characters in options that define the
6935 message's headers; it allows continued headers, but it was checking with
6936 isspace() after an embedded newline instead of explicitly looking for a
6939 21. If all the "regular" hosts to which an address was routed had passed their
6940 expiry times, and had not reached their retry times, the address was
6941 bounced, even if fallback hosts were defined. Now Exim should go on to try
6944 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6945 equivalent code in the SMTP transport. Some hosts send humungous responses
6946 to HELO/EHLO, more than 1024 it seems.
6948 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6949 gives compiler warnings in some environments. I've now done it "properly",
6952 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6953 (because of problems with the built-in one) was declared to return uschar *
6954 instead of char *, causing compiler failure.
6956 25. Fixed a file descriptor leak when processing alias/forward files.
6958 26. Fixed a minor format string issue in dbfn.c.
6960 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6962 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6963 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6966 29. When a sender address is verified, it is cached, to save repeating the test
6967 when there is more than one recipient in a message. However, when the
6968 verification involves a callout, it is possible for different callout
6969 options to be set for different recipients. It is too complicated to keep
6970 track of this in the cache, so now Exim always runs a verification when a
6971 callout is required, relying on the callout cache for the optimization.
6972 The overhead is duplication of the address routing, but this should not be
6975 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6976 to be verified with callout=postmaster, and the main callout worked but the
6977 postmaster check failed, the verification correctly failed. However, if a
6978 subsequent RCPT command asked for sender verification *without* the
6979 postmaster check, incorrect caching caused this verification also to fail,
6982 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6983 it was not caching the DNS options (qualify_single, search_parents) that
6984 were used when the lookup failed. A subsequent lookup with different
6985 options therefore always gave the same answer, though there were cases
6986 where it should not have. (Example: a "domains = !$mx_any" option on a
6987 dnslookup router: the "domains" option is always processed without any
6988 widening, but the router might have qualify_single set.) Now Exim uses the
6989 cached value only when the same options are set.
6991 32. Added John Jetmore's "exipick" utility to the distribution.
6993 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6994 than a timeout (e.g. a certificate is required, and is not provided), an
6995 Exim server now closes the connection immediately. Previously it waited for
6996 the client to close - but if the client is SSL, it seems that they each
6997 wait for each other, leading to a delay before one of them times out.
6999 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7000 maintained 0.8.x compatibility because I don't think many are using it, and
7001 it is clearly obsolete.
7003 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7006 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7007 client certificate was expired. A simple patch fixes this, though I don't
7008 understand the full logic of why the verify callback is called multiple
7011 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7012 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7013 which causes problems with some clients (such as the Certicom SSL Plus
7014 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7015 disables the coutermeasure allowing Eudora to connect."
7017 38. Exim was not checking that a write() to a log file succeeded. This could
7018 lead to Bad Things if a log got too big, in particular if it hit a file
7019 size limit. Exim now panics and dies if it cannot write to a log file, just
7020 as it does if it cannot open a log file.
7022 39. Modified OS/Makefile-Linux so that it now contains
7024 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7026 The two -D definitions ensure that Exim is compiled with large file
7027 support, which makes it possible to handle log files that are bigger than
7030 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7031 instance) a domain was checked against a named list that involved a lookup,
7032 causing $domain_data to be set, then another domain was checked against the
7033 same list, then the first domain was re-checked, the value of $domain_data
7034 after the final check could be wrong. In particular, if the second check
7035 failed, it could be set empty. This bug probably also applied to
7038 41. The strip_trailing_dot option was not being applied to the address given
7039 with the -f command-line option.
7041 42. The code for reading a message's header from the spool was incrementing
7042 $received_count, but never initializing it. This meant that the value was
7043 incorrect (doubled) while delivering a message in the same process in which
7044 it was received. In the most common configuration of Exim, this never
7045 happens - a fresh exec is done - but it can happen when
7046 deliver_drop_privilege is set.
7048 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7049 it now includes up to 150 characters of the unexpected data in the log
7052 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7053 and building data strings. The size of both of these buffers was 10 000
7054 bytes - far larger than anybody would *ever* want, thought I. Needless to
7055 say, somebody hit the limit. I have increased the maximum line length to
7056 20 000 and the maximum data length of concatenated lines to 100 000. I have
7057 also fixed two bugs, because there was no checking on these buffers. Tsk,
7058 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7059 buffer is too small.
7061 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7062 lsearch lookups. Now it does.
7064 46. When parsing a route_list item in a manualroute router, a fixed-length
7065 buffer was used for the list of hosts. I made this 1024 bytes long,
7066 thinking that nobody would ever have a list of hosts that long. Wrong.
7067 Somebody had a whole pile of complicated expansion conditions, and the
7068 string was silently truncated, leading to an expansion error. It turns out
7069 that it is easier to change to an unlimited length (owing to other changes
7070 that have happened since this code was originally written) than to build
7071 structure for giving a limitation error. The length of the item that
7072 expands into the list of hosts is now unlimited.
7074 47. The lsearch lookup could not handle data where the length of text line was
7075 more than 4095 characters. Such lines were truncated, leading to shortened
7076 data being returned. It should now handle lines of any length.
7078 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7079 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7082 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7083 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7084 no longer happens in comment lines. A list of replacements is now placed
7085 at the head of all of the source files, except those whose only change is
7086 to replace PERL_COMMAND in the very first #! line.
7088 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7089 messages on the queue, with a bottom-up merge sort, using code contributed
7090 by Michael Haardt. This should make operations like -bp somewhat faster on
7091 large queues. It won't affect queue runners, except when queue_run_in_order
7094 51. Installed eximstats 1.31 in the distribution.
7096 52. Added support for SRV lookups to the dnslookup router.
7098 53. If an ACL referred to $message_body or $message_body_end, the value was not
7099 reset for any messages that followed in the same SMTP session.
7101 54. The store-handling optimization for building very long strings was not
7102 differentiating between the different store pools. I don't think this
7103 actually made any difference in practice, but I've tidied it.
7105 55. While running the routers to verify a sender address, $sender_address
7106 was still set to the sender address. This is wrong, because when routing to
7107 send a bounce to the sender, it would be empty. Therefore, I have changed
7108 it so that, while verifying a sender address, $sender_address is set to <>.
7109 (There is no change to what happens when verifying a recipient address.)
7111 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7112 target A or AAAA records (if not already returned) without resetting the
7113 qualify_single or search_parents options of the DNS resolver. These are
7114 inappropriate in this case because the targets of MX and SRV records must
7115 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7116 target that, when qualified, matched something in the local domain. These
7117 two options are now turned off when doing these lookups.
7119 57. It seems that at least some releases of Reiserfs (which does not have the
7120 concept of a fixed number of inodes) returns zero and not -1 for the
7121 number of available inodes. This interacted badly with check_spool_inodes,
7122 which assumed that -1 was the "no such thing" setting. What I have done is
7123 to check that the total number of inodes is greater than zero before doing
7124 the test of how many are available.
7126 58. When a "warn" ACL statement has a log_message modifier, the message is
7127 remembered, and not repeated. This is to avoid a lot of repetition when a
7128 message has many recipients that cause the same warning to be written.
7129 However, Exim was preserving the list of already written lines for an
7130 entire SMTP session, which doesn't seem right. The memory is now reset if a
7131 new message is started.
7133 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7134 debugging output unless log_rewrite was also set.
7136 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7137 of (int)(handle) when we know that handle contains (void *)(-1).
7139 61. The Exim daemon panic-logs an error return when it closes the incoming
7140 connection. However "connection reset by peer" seems to be common, and
7141 isn't really an error worthy of noting specially, so that particular error
7144 62. When Exim is trying to find all the local interfaces, it used to panic and
7145 die if the ioctl to get the interface flags failed. However, it seems that
7146 on at least one OS (Solaris 9) it is possible to have an interface that is
7147 included in the list of interfaces, but for which you get a failure error
7148 for this call. This happens when the interface is not "plumbed" into a
7149 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7150 failure of the "get flags" call assumes that the interface is down.
7152 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7153 makes life easier for people who are doing arithmetic on fields extracted
7154 from dates, where you often get leading zeros that should not be
7155 interpreted as octal.
7157 64. Added qualify_domain to the redirect router, to override the global
7160 65. If a pathologically long header line contained very many addresses (the
7161 report of this problem mentioned 10 000) and each of them was rewritten,
7162 Exim could use up a very large amount of memory. (It kept on making new
7163 copies of the header line as it rewrote, and never released the old ones.)
7164 At the expense of a bit more processing, the header rewriting function has
7165 been changed so that it no longer eats memory in this way.
7167 66. The generation of the Received: header has been moved from the time that a
7168 message starts to be received, to the time that it finishes. The timestamp
7169 in the Received: header should now be very close to that of the <= log
7170 line. There are two side-effects of this change:
7172 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7173 logged header lines no longer include the local Received: line, because
7174 it has not yet been created. The same applies to a copy of the message
7175 that is returned to a non-SMTP sender when a message is rejected.
7177 (b) When a filter file is tested using -bf, no additional Received: header
7178 is added to the test message. After some thought, I decided that this
7181 This change does not affect the value of $received_for. It is still set
7182 after address rewriting, but before local_scan() is called.
7184 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7186 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7187 gave an unhelpful panic error message, and a defer error. I have managed to
7188 change this behaviour so that it now rejects any supplied certificate,
7189 which seems right, as the list of acceptable certificates is empty.
7191 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7192 gave an unhelpful defer error. I have not managed to make this reject any
7193 supplied certificates, but the error message it gives is "no certificate
7194 supplied", which is not helpful.
7196 70. exigrep's output now also includes lines that are not associated with any
7197 message, but which match the given pattern. Implemented by a patch from
7198 Martin Sluka, which also tidied up the Perl a bit.
7200 71. Recipient callout verification, like sender verification, was using <> in
7201 the MAIL FROM command. This isn't really the right thing, since the actual
7202 sender may affect whether the remote host accepts the recipient or not. I
7203 have changed it to use the actual sender in the callout; this means that
7204 the cache record is now keyed on a recipient/sender pair, not just the
7205 recipient address. There doesn't seem to be a real danger of callout loops,
7206 since a callout by the remote host to check the sender would use <>.
7207 [SEE ABOVE: changed after hitting problems.]
7209 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7210 temporary errors. However, in the case of such a code being given after
7211 the end of a data transmission (i.e. after ".") Exim was failing to write
7212 a retry record for the message. (Yes, there was some broken host that was
7213 actually sending 8xx at this point.)
7215 73. An unknown lookup type in a host list could cause Exim to panic-die when
7216 the list was checked. (An example that provoked this was putting <; in the
7217 middle of a list instead of at the start.) If this happened during a DATA
7218 ACL check, a -D file could be left lying around. This kind of configuration
7219 error no longer causes Exim to die; instead it causes a defer error. The
7220 incident is still logged to the main and panic logs.
7222 74. Buglet left over from Exim 3 conversion. The message "too many messages
7223 in one connection" was written to the rejectlog but not the mainlog, except
7224 when address rewriting (yes!) was being logged.
7226 75. Added write_rejectlog option.
7228 76. When a system filter was run not as root (that is, when system_filter_user
7229 was set), the values of the $n variables were not being returned to the
7230 main process; thus, they were not subsequently available in the $sn
7233 77. Added +return_path_on_delivery log selector.
7235 78. A connection timeout was being treated differently from recipients deferred
7236 when testing hosts_max_try with a message that was older than the host's
7237 retry timeout. (The host should not be counted, thus allowing all hosts to
7238 be tried at least once before bouncing.) This may have been the cause of an
7239 occasionally reported bug whereby a message would remain on the queue
7240 longer than the retry timeout, but would be bounced if a delivery was
7241 forced. I say "may" because I never totally pinned down the problem;
7242 setting up timeout/retry tests is difficult. See also the next item.
7244 79. The ultimate address timeout was not being applied to errors that involved
7245 a combination of host plus message (for example, a timeout on a MAIL
7246 command). When an address resolved to a number of possible hosts, and they
7247 were not all tried for each delivery (e.g. because of hosts_max_try), a
7248 message could remain on the queue longer than the retry timeout.
7250 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7253 81. Fixed an obscure SMTP outgoing bug which required at least the following
7254 conditions: (a) there was another message waiting for the same server;
7255 (b) the server returned 5xx to all RCPT commands in the first message so
7256 that the message was not completed; (c) the server dropped the connection
7257 or gave a negative response to the RSET that Exim sends to abort the
7258 transaction. The observed case was a dropped connection after DATA that had
7259 been sent in pipelining mode. That is, the server had advertised PIPELINING
7260 but was not implementing it correctly. The effect of the bug was incorrect
7261 behaviour, such as trying another host, and this could lead to a crash.
7267 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7268 and daemon.c were passed as pointers to ints; they should have been
7269 pointers to socklen_t variables (which are typically unsigned ints).
7271 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7274 3. Fixed a really odd bug that affected only the testing scheme; patching a
7275 certain fixed string in the binary changed the value of another string that
7276 happened to be identical to the end of the original first string.
7278 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7279 name", it returns that address as the IP address. On some operating
7280 systems (e.g. Solaris), it also passes back the IP address string as the
7281 "host name". However, on others (e.g. Linux), it passes back an empty
7282 string. Exim wasn't checking for this, and was changing the host name to an
7283 empty string, assuming it had been canonicalized.
7285 5. Although rare, it is permitted to have more than one PTR record for a given
7286 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7287 all the names associated with an address, because they do in Solaris.
7288 However, it seems that they do not in Linux for data that comes from the
7289 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7290 I found this out when I moved to a new Linux workstation and tried to run
7291 the Exim test suite.
7293 To get round this problem I have changed the code so that it now does its
7294 own call to the DNS to look up PTR records when searching for a host name.
7295 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7296 addresses that are only in /etc/hosts are still found.
7298 This behaviour is, however, controlled by an option called host_lookup_
7299 order, which defaults to "bydns:byaddr". If people want to use the other
7300 order, or indeed, just use one or the other means of lookup, they can
7301 specify it in this variable.
7303 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7304 some operating systems, this comes back from gethostbyaddr() as an empty
7305 string, and this is what Exim used to test for. However, it seems that in
7306 other systems, "." is yielded. Exim now tests for this case too.
7308 7. The values of check_spool_space and check_log_space are now held internally
7309 as a number of kilobytes instead of an absolute number of bytes. If a
7310 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7311 kilobyte. This means that much larger values can be stored.
7313 8. Exim monitor: an attempt to get the action menu when not actually pointing
7314 at a message produces an empty menu entitled "No message selected". This
7315 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7316 no entries in it ("Shell widget menu has zero width and/or height"). So I
7317 have added a single, blank menu entry in this case.
7319 9. Added ${quote_local_part.
7321 10. MIME decoding is now applied to the contents of Subject: header lines when
7324 11. Now that a reference to $sender_host_address automatically causes a reverse
7325 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7326 host lookup before query-style lookups in lists that might use this
7327 variable. This has therefore been abolished, and the "net-" prefix is no
7328 longer necessary for query-style lookups.
7330 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7331 have been a typo for LFLAGS, so it has been changed.
7333 13. The install script calls Exim with "-C /dev/null" in order to find the
7334 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7335 to be output. However, since Exim outputs its version number before the
7336 error, it didn't break the script. It just looked ugly. I fixed this by
7337 always allowing "-C /dev/null" if the caller is root.
7339 14. Ignore overlarge ACL variable number when reading spool file - insurance
7340 against a later release with more variables having written the file.
7342 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7343 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7344 this, as well as the form without the "IPv6" on the front.
7346 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7347 OS/Makefile-Darwin file.
7349 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7350 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7352 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7353 variable. However, some OS use ints for inodes. I've added cast to long int
7354 to get rid of the compiler warning.
7356 19. I had forgotten to lock out "/../" in configuration file names when
7357 ALT_CONFIG_PREFIX was set.
7359 20. Routers used for verification do not need to specify transports. However,
7360 if such a router generated a host list, and callout was configured, Exim
7361 crashed, because it could not find a port number from the (non-existent)
7362 transport. It now assumes port 25 in this circumstance.
7364 21. Added the -t option to exigrep.
7366 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7367 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7368 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7369 them is set, LOOKUP_LSEARCH is forced.
7371 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7373 24. Added sender and host information to the "rejected by local_scan()" log
7374 line; previously there was no indication of these.
7376 25. Added .include_if_exists.
7378 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7379 Linux. It turns out that not all file systems support this. Apparently some
7380 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7381 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7382 not supported on the file descriptor, is now ignored when Exim is trying to
7383 sync a directory. This applies only to Linux.
7385 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7387 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7388 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7391 29. Added support for maildirsize files from supplied patch (modified a bit).
7393 30. The use of :fail: followed by an empty string could lead Exim to respond to
7394 sender verification failures with (e.g.):
7396 550 Verification failed for <xxx>
7397 550 Sender verify failed
7399 where the first response line was missing the '-' that indicates it is not
7400 the final line of the response.
7402 31. The loop for finding the name of the user that called Exim had a hardwired
7403 limit of 10; it now uses the value of finduser_retries, which is used for
7404 all other user lookups.
7406 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7409 33. Exim was neglecting to zero errno before one call of strtol() when
7410 expanding a string and expecting an integer value. On some systems this
7411 resulted in spurious "integer overflow" errors. Also, it was casting the
7412 result into an int without checking.
7414 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7415 did not work. The code looks as if it has *never* worked, though it appears
7416 to have been documented since at least release 1.62. I have made it work.
7418 35. The "timeout_DNS" error in retry rules, also documented since at least
7419 1.62, also never worked. As it isn't clear exactly what this means, and
7420 clearly it isn't a major issue, I have abolished the feature by treating it
7421 as "timeout", and writing a warning to the main and panic logs.
7423 36. The display of retry rules for -brt wasn't always showing the error code
7426 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7427 timeout_connect_A, timeout_connect_MX.
7429 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7430 to the empty sender.
7432 39. The daemon was not analysing the content of -oX till after it had closed
7433 stderr and disconnected from the controlling terminal. This meant that any
7434 syntax errors were only noted on the panic log, and the return code from
7435 the command was 0. By re-arranging the code a little, I've made the
7436 decoding happen first, so such errors now appear on stderr, and the return
7437 code is 1. However, the actual setting up of the sockets still happens in
7438 the disconnected process, so errors there are still only recorded on the
7441 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7442 connections (as happens on some IP stacks) was logged at start up time as
7443 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7444 it from "IPv6 and IPv4", which means that two separate sockets are being
7447 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7448 says whether AF_INET or AF_INET6 was passed as an argument.
7450 42. Exiwhat output was messed up when time zones were included in log
7453 43. Exiwhat now gives more information about the daemon's listening ports,
7454 and whether -tls-on-connect was used.
7456 44. The "port" option of the smtp transport is now expanded.
7458 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7459 silently ignored. Now an error message is written to the main and panic
7462 46. There's a new ACL modifier called "logwrite" which writes to a log file
7463 as soon as it is encountered.
7465 47. Added $local_user_uid and $local_user_gid at routing time.
7467 48. Exim crashed when trying to verify a sender address that was being
7470 49. Exim was recognizing only a space character after ".include". It now also
7471 recognizes a tab character.
7473 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7474 extracting the relevant information from the specification. The man page no
7475 longer contains scrambled data for the -d option, and I've added a section
7476 at the front about calling Exim under different names.
7478 51. Added "extra_headers" argument to the "mail" command in filter files.
7480 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7483 53. Installed eximstats 1.29.
7485 54. Added transport_filter_timeout as a generic transport option.
7487 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7488 Cc: header lines. This was required by RFC 822, but it not required by RFC
7491 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7492 incoming messages that did not have them. Now it does so only if the
7493 message originates locally, that is, if there is no associated remote host
7494 address. When Resent- header lines are present, this applies to the Resent-
7495 lines rather than the non-Resent- lines.
7497 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7498 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7500 58. Messages for configuration errors now include the name of the main
7501 configuration file - useful now that there may be more than one file in a
7502 list (.included file names were always shown).
7504 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7505 for those rare installations that do not start the daemon as root or run it
7506 setuid root. I've cut out the call to initgroups() if the daemon is not
7509 60. The Exim user and group can now be bound into the binary as text strings
7510 that are looked up at the start of Exim's processing.
7512 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7514 62. Added $mailstore_basename variable.
7516 63. Installed patch to sieve.c from Michael Haardt.
7518 64. When Exim failed to open the panic log after failing to open the main log,
7519 the original message it was trying to log was written to stderr and debug
7520 output, but if they were not available (the usual case in production), it
7521 was lost. Now it is written to syslog before the two lines that record the
7522 failures to open the logs.
7524 65. Users' Exim filters run in subprocesses under the user's uid. It is
7525 possible for a "deliver" command or an alias in a "personal" command to
7526 provoke an address rewrite. If logging of address rewriting is configured,
7527 this fails because the process is not running as root or exim. There may be
7528 a better way of dealing with this, but for the moment (because 4.30 needs
7529 to be released), I have disabled address rewrite logging when running a
7530 filter in a non-root, non-exim process.
7536 1. The buildconfig auxiliary program wasn't quoting the value set for
7537 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7538 not defined. This bug was masked in 4.22 by the effect that was fixed in
7541 2. Some messages that were rejected after a message id was allocated were
7542 shown as "incomplete" by exigrep. It no longer does this for messages that
7543 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7545 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7546 have allow_domain_literals set, the ID did not get logged in the <= line.
7547 Domain literals are now always recognized in Message-ID: header lines.
7549 4. The first argument for a ${extract expansion item is the key name or field
7550 number. Leading and trailing spaces in this item were not being ignored,
7551 causing some misleading effects.
7553 5. When deliver_drop_privilege was set, single queue runner processes started
7554 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7555 same command in the process it spins off) were not dropping privilege.
7557 6. When the daemon running as "exim" started a queue runner, it always
7558 re-executed Exim in the spun-off process. This is a waste of effort when
7559 deliver_drop_privilege is set. The new process now just calls the
7560 queue-runner function directly.
7566 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7569 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7570 ignored. Though the use of -oP was forcing the writing of a pid file, it
7571 was always written to the default place.
7573 3. If the message "no IP address found for host xxxx" is generated during
7574 incoming verification, it is now followed by identification of the incoming
7575 connection (so you can more easily find what provoked it).
7577 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7579 5. Added some features to "harden" Exim a bit more against certain attacks:
7581 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7582 be put in Local/Makefile. This is like the never_users runtime option,
7583 but it cannot be overridden. The default setting is "root".
7585 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7586 prefix string with which any file named in a -C command line option
7589 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7590 is retained for -C and -D only if the caller of Exim is root. Without
7591 it, the exim user may also use -C and -D and retain privilege.
7593 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7594 command line option is disabled.
7596 6. Macro names set by the -D option must start with an upper case letter, just
7597 like macro names defined in the configuration file.
7599 7. Added "dereference=" facility to LDAP.
7601 8. Two instances of the typo "uknown" in the source files are fixed.
7603 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7604 the Configure-Makefile script screwed up while processing it.
7606 10. Incorporated PCRE 4.4.
7608 11. The SMTP synchronization check was not operating right at the start of an
7609 SMTP session. For example, it could not catch a HELO sent before the client
7610 waited for the greeting. There is now a check for outstanding input at the
7611 point when the greeting is written. Because of the duplex, asynchronous
7612 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7613 way, but not yet received, when the check is performed.
7615 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7616 on TCP/IP sockets, because this apparently causes some broken clients to
7619 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7620 unchanged) from the Cygwin maintainer.
7622 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7623 was supported instead of testing for mbx. Effectively a typo.
7625 15. The spa authenticator server code was not checking that the input it
7626 received was valid base64.
7628 16. The debug output line for the "set" modifier in ACLs was not showing the
7629 name of the variable that was being set.
7631 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7633 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7634 lookup on its own. Something else was needed to trigger the lookup. For
7635 example, a match in host_lookup or the need for a host name in a host list.
7636 Now, if $sender_host_name is referenced and the host name has not yet been
7637 looked up, a lookup is performed. If the lookup fails, the variable remains
7638 empty, and $host_lookup_failed is set to "1".
7640 19. Added "eqi" as a case-independent comparison operator.
7642 20. The saslauthd authentication condition could segfault if neither service
7643 nor realm was specified.
7645 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7646 error message that was logged was misleading, and incoming SMTP
7647 connections were dropped. The message is now more accurate, and temporary
7648 errors are given to SMTP connections.
7650 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7651 (except RSET) with a 421 error, until QUIT is received. However, it was
7652 failing to send a response to QUIT.
7654 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7655 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7656 ACL is now run after the helo_try_verify_hosts code.
7658 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7659 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7660 case-independent, so other case variants are also recognized). Apparently
7661 some systems use these upper case variants.
7663 25. If more than two messages were waiting for the same host, and a transport
7664 filter was specified for the transport, Exim sent two messages over the
7665 same TCP/IP connection, and then failed with "socket operation on non-
7666 socket" when it tried to send the third.
7668 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7670 27. The extern definition of crypt16() in expand.c was not being excluded when
7671 the OS had its own crypt16() function.
7673 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7674 as a preferred synonym for return_size_limit, both as an option and as an
7677 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7679 30. Changed the default configuration ACL to relax the local part checking rule
7680 for addresses that are not in any local domains. For these addresses,
7681 slashes and pipe symbols are allowed within local parts, but the sequence
7682 /../ is explicitly forbidden.
7684 31. SPA server authentication was not clearing the challenge buffer before
7687 32. log_message in a "warn" ACL statement was writing to the reject log as
7688 well as to the main log, which contradicts the documentation and doesn't
7689 seem right (because no rejection is happening). So I have stopped it.
7691 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7692 However, I am unable to do any testing of this.
7694 34. Fixed an infelicity in the appendfile transport. When checking directories
7695 for a mailbox, to see if any needed to be created, it was accidentally
7696 using path names with one or more superfluous leading slashes; tracing
7697 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7699 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7700 contents are added to the log line that is written for every discarded
7701 recipient. (Previously a log_message setting was ignored.)
7703 36. The ${quote: operator now quotes the string if it is empty.
7705 37. The install script runs exim in order to find its version number. If for
7706 some reason other than non-existence or emptiness, which it checks, it
7707 could not run './exim', it was installing it with an empty version number,
7708 i.e. as "exim-". This error state is now caught, and the installation is
7711 38. An argument was missing from the function that creates an error message
7712 when Exim fails to connect to the socket for saslauthd authentication.
7713 This could cause Exim to crash, or give a corrupted message.
7715 39. Added isip, isip4, and isip6 to ${if conditions.
7717 40. The ACL variables $acl_xx are now saved with the message, and can be
7718 accessed later in routers, transports, and filters.
7720 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7721 strings in the file are not string-expanded.
7723 42. If a MAIL command specified a SIZE value that was too large to fit into an
7724 int variable, the check against message_size_limit failed. Such values are
7725 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7726 day this will have to be increased, but I don't think I want to be around
7727 when emails are that large.
7734 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7735 iconv() is not standard in FreeBSD.
7737 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7738 IPv6 enabled. The observed symptom was a segmentation fault on return from
7739 the function os_common_find_running_interfaces() in src/os.c.
7741 3. In the check_special_case() function in daemon.c I had used "errno" as an
7742 argument name, which causes warnings on some systems. This was basically a
7743 typo, since it was named "eno" in the comments!
7745 4. The code that waits for the clock to tick (at a resolution of some fraction
7746 of a second) so as to ensure message-id uniqueness was always waiting for
7747 at least one whole tick, when it could have waited for less. [This is
7748 almost certainly not relevant at current processor speeds, where it is
7749 unlikely to ever wait at all. But we try to future-proof.]
7751 5. The function that sleeps for a time interval that includes fractions of a
7752 second contained a race. It did not block SIGALRM between setting the
7753 timer, and suspending (a couple of lines later). If the interval was short
7754 and the sigsuspend() was delayed until after it had expired, the suspension
7755 never ended. On busy systems this could lead to processes getting stuck for
7758 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7759 process, before it forks any delivery processes. The open lookup caching
7760 mechanism meant that the open file or database connection was passed into
7761 the delivery process. The problem was that delivery processes always tidy
7762 up cached lookup data. This could cause a problem for the next delivery
7763 process started by the queue runner, because the external queue runner
7764 process does not know about the closure. So the next delivery process
7765 still has data in the lookup cache. In the case of a file lookup, there was
7766 no problem because closing a file descriptor in a subprocess doesn't affect
7767 the parent. However, if the lookup was caching a connection to a database,
7768 the connection was closed, and the second delivery process was likely to
7769 see errors such as "PGSQL: query failed: server closed the connection
7770 unexpectedly". The problem has been fixed by closing all cached lookups
7771 in a queue runner before running a delivery process.
7773 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7774 seem to have the "const" qualifier which it has on other OS. I've
7777 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7778 *of the same type* (client or server) with the same public name that an
7779 error should be diagnosed.
7781 9. When Exim looked up a host name for an IP address, but failed to find the
7782 original IP address when looking up the host name (a safety check), it
7783 output the message "<ip address> does not match any IP for NULL", which was
7784 confusing, to say the least. The bug was that the host name should have
7785 appeared instead of "NULL".
7787 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7788 user that is built into the binary, and the -C or -D options is used, root
7789 privilege is dropped before the configuration file is read. In addition,
7790 logging is switched to stderr instead of the normal log files. If the
7791 configuration then re-defines the Exim user, the unprivileged environment
7792 is probably not what is expected, so Exim logs a panic warning message (but
7795 However, if deliver_drop_privilege is set, the unprivileged state may well
7796 be exactly what is intended, so the warning has been cut out in that case,
7797 and Exim is allowed to try to write to its normal log files.
7803 1. smtp_return_error_details was not giving details for temporary sender
7804 or receiver verification errors.
7806 2. Diagnose a configuration error if two authenticators have the same public
7809 3. Exim used not to create the message log file for a message until the first
7810 delivery attempt. This could be confusing when incoming messages were held
7811 for policy or load reasons. The message log file is now created at the time
7812 the message is received, and an initial "Received" line is written to it.
7814 4. The automatically generated man page for command line options had a minor
7815 bug that caused no ill effects; however, a more serious problem was that
7816 the procedure for building the man page automatically didn't always
7817 operate. Consequently, release 4.20 contains an out-of-date version. This
7818 shouldn't happen again.
7820 5. When building Exim with embedded Perl support, the script that builds the
7821 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7822 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7824 6. The freeze_tell option was not being used for messages that were frozen on
7825 arrival, either by an ACL or by local_scan().
7827 7. Added the smtp_incomplete_transaction log selector.
7829 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7830 was accepting AUTH without a new EHLO.
7832 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7833 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7836 10. Logging of TCP/IP connections (when configured) now happens in the main
7837 daemon process instead of the child process, so that the TCP/IP connection
7838 count is more accurate (but it can never be perfect).
7840 11. The use of "drop" in a nested ACL was not being handled correctly in the
7841 outer ACL. Now, if condition failure induced by the nested "drop" causes
7842 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7843 or "require"), the connection is dropped.
7845 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7846 that yield "discard" can now be used with an "accept" or a "discard" verb,
7847 but an error is generated for any others (because I can't see a useful way
7848 to define what should happen).
7850 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7851 are now processed in the same way as lines in the Exim configuration file.
7852 In particular, continuation lines are supported.
7854 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7856 15. Added -ti meaning -t -i.
7858 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7859 domains, and warn by logging if others are found.
7861 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7862 returned from GIFCONF when Exim is trying to find the list of interfaces on
7863 a host. The code in os.c has been modified to copy each ifreq to an aligned
7864 structure in all cases.
7866 Also, in some cases, the returned ifreq's were being copied to a 'struct
7867 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7868 means the last couple of bytes of an IPv6 address could be chopped if the
7869 ifreq contained only a normal sockaddr (14 bytes storage).
7871 18. Named domain lists were not supported in the hosts_treat_as_local option.
7872 An entry such as +xxxx was not recognized, and was treated as a literal
7875 19. Ensure that header lines added by a DATA ACL are included in the reject log
7876 if the ACL subsequently rejects the message.
7878 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7879 MD5 (which is deprecated).
7881 21. When testing a filter file using -bf, Exim was writing a message when it
7882 took the sender from a "From " line in the message, but it was not doing so
7883 when it took $return_path from a Return-Path: header line. It now does.
7885 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7886 with a valid header line field name (a series of printing characters
7887 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7889 23. Changed "disc" in the source to "disk" to conform to the documentation and
7890 the book and for uniformity.
7892 24. Ignore Sendmail's -Ooption=value command line item.
7894 25. When execve() failed while trying to run a command in a pipe transport,
7895 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7896 could be confused with a return value of 69 from the command itself. This
7897 has been changed to 127, the value the shell returns if it is asked to run
7898 a non-existent command. The wording for the related log line suggests a
7899 non-existent command as the problem.
7901 26. If received_header_text expands to an empty string, do not add a Received:
7902 header line to the message. (Well, it adds a token one on the spool, but
7903 marks it "old" so that it doesn't get used or transmitted.)
7905 27. Installed eximstats 1.28 (addition of -nt option).
7907 28. There was no check for failure on the call to getsockname() in the daemon
7908 code. This can fail if there is a shortage of resources on the system, with
7909 ENOMEM, for example. A temporary error is now given on failure.
7911 29. Contrary to the C standard, it seems that in some environments, the
7912 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7913 program. Exim now does this explicitly; it affects the formatting of
7914 timestamps using strftime().
7916 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7917 complaints. I've now initialized all the variables, to avoid this.
7919 32. Header lines added by a system filter were not being "seen" during
7920 transport-time rewrites.
7922 33. The info_callback() function passed to OpenSSL is set up with type void
7923 (*)(SSL *, int, int), as described somewhere. However, when calling the
7924 function (actually a macro) that sets it up, the type void(*)() is
7925 expected. I've put in a cast to prevent warnings from picky compilers.
7927 34. If a DNS black list lookup found a CNAME record, but there were no A
7928 records associated with the domain it pointed at, Exim crashed.
7930 35. If a DNS black list lookup returned more than one A record, Exim ignored
7931 all but the first. It now scans all returned addresses if a particular IP
7932 value is being sought. In this situation, the contents of the
7933 $dnslist_value variable are a list of all the addresses, separated by a
7936 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7937 to accept a host name and all its aliases if the forward lookup for any of
7938 them yielded the IP address of the incoming connection. Now it accepts only
7939 those names whose forward lookup yields the correct IP address. Any other
7940 names are discarded. This closes a loophole whereby a rogue DNS
7941 administrator could create reverse DNS records to break through a
7942 wildcarded host restriction in an ACL.
7944 37. If a user filter or a system filter that ran in a subprocess used any of
7945 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7946 the wrong values were passed to the pipe command ($thisaddress had the
7947 value of $0, $0 had the value of $1, etc). This bug was introduced by
7948 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7950 38. Improved the line breaking for long SMTP error messages from ACLs.
7951 Previously, if there was no break point between 40 and 75 characters, Exim
7952 left the rest of the message alone. Two changes have been made: (a) I've
7953 reduced the minimum length to 35 characters; (b) if it can't find a break
7954 point between 35 and 75 characters, it looks ahead and uses the first one
7955 that it finds. This may give the occasional overlong line, but at least the
7956 remaining text gets split now.
7958 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7959 file descriptors might be low, and that setting 1000 would always raise it.
7960 It turns out that in some environments, the limit is already over 1000 and
7961 that lowering it causes trouble. So now Exim takes care not to decrease it.
7963 40. When delivering a message, the value of $return_path is set to $sender_
7964 address at the start of routing (routers may change the value). By an
7965 oversight, this default was not being set up when an address was tested by
7966 -bt or -bv, which affected the outcome if any router or filter referred to
7969 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7970 log only when "message" or "log_message" are set. However, if one of the
7971 conditions was an address verification, or a call to a nested ACL, the
7972 messages generated by the underlying test were being passed through. This
7973 no longer happens. The underlying message is available in $acl_verify_
7974 message for both "message" and "log_message" expansions, so it can be
7975 passed through if needed.
7977 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7978 new expansion $bh_ to give the encoded byte string without charset
7979 translation. Translation happens only if iconv() is available; HAVE_ICONV
7980 indicates this at build time. HEADERS_CHARSET gives the charset to
7981 translate to; headers_charset can change it in the configuration, and
7982 "headers charset" can change it in an individual filter file.
7984 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7985 that creates RFC 2047 encoded "words" labels them as that charset instead
7986 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7987 expansion operator; (ii) when Exim creates a From: line for a local
7988 message; (iii) when a header line is rewritten to include a "phrase" part.
7990 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7991 buggy, causing it to skip the first lines of messages whose message ID
7992 ended in 'D'. This would not have bitten before Exim release 4.14, because
7993 message IDs were unlikely to end in 'D' before then. The effect was to have
7994 incorrect size information for certain domains.
7996 45. #include "config.h" was missing at the start of the crypt16.c module. This
7997 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8000 46. If there was a timeout during a "random" callout check, Exim treated it as
8001 a failure of the random address, and carried on sending RSET and the real
8002 address. If the delay was just some slowness somewhere, the response to the
8003 original RCPT would be taken as a response to RSET and so on, causing
8004 mayhem of various kinds.
8006 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8007 when I implemented it. It didn't allow for the fact that some option values
8008 may legitimately be negative (e.g. size_addition), and it didn't even do
8009 the right test for positive values.
8011 48. Domain names in DNS records are case-independent. Exim always looks them up
8012 in lower case. Some resolvers return domain names in exactly the case they
8013 appear in the zone file, that is, they may contain uppercase letters. Not
8014 all resolvers do this - some return always lower case. Exim was treating a
8015 change of case by a resolver as a change of domain, similar to a widening
8016 of a domain abbreviation. This triggered its re-routing code and so it was
8017 trying to route what was effectively the same domain again. This normally
8018 caused routing to fail (because the router wouldn't handle the domain
8019 twice). Now Exim checks for this case specially, and just changes the
8020 casing of the domain that it ultimately uses when it transmits the message
8023 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8026 50. If a filter generated a file delivery with a non-absolute name (possible if
8027 no home directory exists for the router), the forbid_file option was not
8030 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8031 the existing equality matching.
8033 52. Exim was using ints instead of ino_t variables in some places where it was
8034 dealing with inode numbers.
8036 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8037 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8038 called TMPDIR, and if it finds it is different, it changes its value.
8040 54. The smtp_printf() function is now made available to local_scan() so
8041 additional output lines can be written before returning. There is also an
8042 smtp_fflush() function to enable the detection of a dropped connection.
8043 The variables smtp_input and smtp_batched_input are exported to
8046 55. Changed the default runtime configuration: the message "Unknown user"
8047 has been removed from the ACL, and instead placed on the localuser router,
8048 using the cannot_route_message feature. This means that any verification
8049 failures that generate their own messages won't get overridden. Similarly,
8050 the "Unrouteable address" message that was in the ACL for unverifiable
8051 relay addresses has also been removed.
8053 56. Added hosts_avoid_esmtp to the smtp transport.
8055 57. The exicyclog script was not checking for the esoteric option
8056 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8057 will work only if exicyclog is run under the appropriate euid.
8059 58. Following a discussion on the list, the rules by which Exim recognises line
8060 endings on incoming messages have been changed. The -dropcr and drop_cr
8061 options are now no-ops, retained only for backwards compatibility. The
8062 following line terminators are recognized: LF CRLF CR. However, special
8063 processing applies to CR:
8065 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8066 nor a local message in the state where . is a terminator.
8068 (ii) If a bare CR is encountered in a header line, an extra space is added
8069 after the line terminator so as not to end the header. The reasoning
8070 behind this is that bare CRs in header lines are most likely either
8071 to be mistakes, or people trying to play silly games.
8073 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8074 was being incorrectly given as 18 bytes larger than it should have been.
8075 This is a VOB (very old bug).
8077 60. This may never have affected anything current, but just in case it has:
8078 When the local host is found other than at the start of a list of hosts,
8079 the local host, those with the same MX, and any that follow, are discarded.
8080 When the list in question was part of a longer list of hosts, the following
8081 hosts (not currently being processed) were also being discarded. This no
8082 longer happens. I'm not sure if this situation could ever has previously
8085 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8088 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8089 header. This is something that is recommended in a new Internet Draft, and
8090 is something that is documented as being done by Sendmail. There are two
8091 possible values. For messages generated by the autoreply transport, Exim
8094 Auto-Submitted: auto-replied
8096 whereas for all other generated messages (e.g. bounces) it adds
8098 Auto-Submitted: auto-generated
8100 63. The "personal" condition in filters now includes a test for the
8101 Auto-Submitted: header. If it contains the string "auto-" the message it
8102 not considered personal.
8104 64. Added rcpt_include_affixes as a generic transport option.
8106 65. Added queue_only_override (default true).
8108 66. Added the syslog_duplication option.
8110 67. If what should have been the first header line of a message consisted of
8111 a space followed by a colon, Exim was mis-interpreting it as a header line.
8112 It isn't of course - it is syntactically invalid and should therefore be
8113 treated as the start of the message body. The misbehaviour could have
8114 caused a number of strange effects, including loss of data in subsequent
8115 header lines, and spool format errors.
8117 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8118 client host had authenticated. This control can now be exercised by an ACL
8119 for more flexibility.
8121 69. By default, callouts do not happen when testing with -bh. There is now a
8122 variant, -bhc, which does actually run the callout code, including
8123 consulting and updating the callout cache.
8125 70. Added support for saslauthd authentication, courtesy of Alexander
8128 71. If statvfs() failed on the spool or log directories while checking their
8129 size for availability, Exim confusingly gave the error "space shortage".
8130 Furthermore, in debugging mode it crashed with a floating point exception.
8131 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8132 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8133 serious problem, Exim now writes to the main and panic logs when this
8134 happens, with details of the failure. It then refuses to accept the
8135 incoming message, giving the message "spool directory problem" or "log
8136 directory problem" with a 421 code for SMTP messages.
8138 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8139 0, 1, and 2 exist, because some OS complain for execs without them (see
8140 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8141 descriptors. However, the code omitted to check that the open succeeded,
8142 causing mysterious errors if for some reason the permissions on /dev/null
8143 got screwed. Now Exim writes a message to the main and panic logs, and
8144 bombs out if it can't open /dev/null.
8146 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8147 interact so that it is all more flexible. It is supposed to remain
8148 backwards compatible. Also added extra_local_interfaces.
8150 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8151 to bomb out with an assertion failure - to the client this appears as a
8152 connection drop. This problem occurs in the part of the code that was taken
8153 from the Samba project. Fortunately, the assertion is in a very simple
8154 function, so I have fixed this by reproducing the function inline in the
8155 one place where it is called, and arranging for authentication to fail
8156 instead of killing the process with assert().
8158 75. The SPA client code was not working when the server requested OEM rather
8159 than Unicode encoding.
8161 76. Added code to make require_files with a specific uid setting more usable in
8162 the case where statting the file as root fails - usually a non-root-mounted
8163 NFS file system. When this happens and the failure is EACCES, Exim now
8164 forks a subprocess and does the per-uid checking as the relevant uid.
8166 77. Added process_log_path.
8168 78. If log_file_path was not explicitly set, a setting of check_log_space or
8169 check_log_inodes was ignored.
8171 79. If a space check for the spool or log partitions fails, the incident is now
8172 logged. Of course, in the latter case the data may get lost...
8174 80. Added the %p formatting code to string_format() so that it can be used to
8175 print addresses in debug_print(). Adjusted all the address printing in the
8176 debugging in store.c to use %p rather than %d.
8178 81. There was a concern that a line of code in smtp_in.c could overflow a
8179 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8180 initially expressed, the concern was not well-founded, because trailing
8181 spaces are removed early. However, if the trailing spaces were followed by
8182 a NULL, they did not get removed, so the overflow was possible. Two fixes
8185 (a) I re-wrote the offending code in a cleaner fashion.
8186 (b) If an incoming SMTP command contains a NULL character, it is rejected
8189 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8190 runs initgroups(), so that if the Exim user is in any additional groups,
8191 they will be used during message reception.
8197 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob