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
76 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
77 related applications. Previously an "H" was used where available info
78 says that "M" should be, so change to match.
80 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
81 as arguments, so an implementation trying to copy these into a local
82 buffer was taking a taint-enforcement trap. Fix by using dynamically
83 created buffers. Similar fix for radius expansion condition.
85 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
86 reasonable, eg. to count headers. Fix by using dynamically created
87 buffers rather than a local. Do similar fixes for ACL actions "dcc",
88 "log_reject_target", "malware" and "spam"; the arguments are expanded
89 so could be handling tainted values.
91 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
92 broken the (no-op) support for this sendmail command. Restore it
93 to doing nothing, silently, and returning good status.
95 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
96 record path was given (or the default used) without a leading directory
97 path, an error occurred on trying to open it. Use the transport's working
100 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
101 transport. Previously it was the name on the DNS A-record; use instead
102 the head of the CNAME chain leading there (if there is one). This seems
103 to align better with RFC 6125.
105 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
106 smtp_accept_max_per_host allocated resources which were not released
107 when the limit was exceeded. This eventually crashed the daemon. Fix
108 by adding a release action in that path.
110 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
111 expanded; previously using tainted values was rejected. Fix by using
112 dynamically-created buffers.
114 JH/09 Relax restrictions on ACL verify condition needing access to message
115 headers. Previously they were only permitted in data and non-smtp ACLs;
116 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
117 not_blind, header_sender and header_names_ascii verification.
119 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
120 Previously a macro used one argument twice; when called with the
121 argument as an expression having side-effects, incorrect operation
122 resulted. Use an inlineable function.
124 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
125 held open for a verify callout. Previously this wan not accounted for
126 and a corrupt onward SMTP conversation resulted.
128 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
129 Previously they were embedded naively in the constructed address; when
130 needed, strip the quoting and quote the entire local_part.
131 Also make the inbound_srs expansion condition handle quoting.
133 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
134 excluded, not matching the documentation.
136 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
137 was given for the sqlite_dbfile a trap resulted.
139 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
140 "name" argument resulted in a trap. There is no reason to disallow such;
141 this was a coding error.
143 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
144 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
145 spent suspended, ignoring the POSIX definition. Previously we assumed
146 it did and a constant offset from real time could be used as a correction.
147 Change to using the same clock source for the start-of-message and the
148 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
149 exists, just to get a clock slightly more aligned to reality.
151 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
152 RFC says it is optional some validators care. The missing char was not
153 intended but triggered by a line-wrap alignment. Discovery and fix by
154 Guillaume Outters, hacked on by JH.
156 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
157 name being quoted was tainted a trap would be taken. Fix by using
158 dynamically created buffers. The routine could have been called by a
159 rewrite with the "h" flag, by using the "-F" command-line option, or
160 by using a "name=" option on a control=submission ACL modifier.
162 JH/19 SPF: change the Authentication-Results expansion component to give
163 smtp.helo when the sender domain is empty. Previously it gave
166 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
167 not in 127.0.0.0/8 to help in spotting list domains taken over by a
168 domain-parking registrar.
170 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
171 Previously when a whitespace character was specified it was not inserted
172 after removing the newline.
174 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
175 the domain part of the recipient address. This overrides any tls_sni
176 option set, which was previously used.
178 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
181 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
182 is_tainted() had an off-by-one error in the overenthusiastic direction.
183 Find and fix by Gavan. Although NetBSD is not a supported platform for
184 4.94 this bug could affect other platforms.
186 PP/01 Fix default prime selection to be consistent.
187 One path used ike23 still, instead of exim.dev.20160529.3; now both
188 execution flows will use the same DH primes (currently
189 exim.dev.20160529.3).
191 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
192 option in smtp transport, to match the documentation. Previously
193 verification was not being done in some cases where it should have been.
195 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
196 than one server was defined and depending on the platform memory layout
197 details, an internal consistency trap could be hit while walking the list
200 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
201 files. The value is used by the authresults expansion item. Previously
202 if this was used in a router or transport, a crash could result.
204 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
205 have a sufficient incidence of EINTR returns from select that an
206 interaction with other operations done by the main daemon loop exposed
207 a bug in the error-handling. This was benign apart from the log
210 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
211 for consistency with delivered "=>" and failed "**" lines. While we're
212 there, handle PRX and TFO.
214 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
215 applied. This resulted, if any header-line rewrite rules were configured,
216 in a panic-log triggerable by sending a message with a long address in
217 a header. Fix by increasing the arbitrary limit to larger than a single
218 (dewrapped) 5322 header line maximum size.
220 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
221 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
222 RFC 6648 which deprecates X- options in protocols as a general practice.
223 Changeover between the implementations is handled by the mechanisms
226 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
227 a remote which uses callout/hold. Previously the local was queued.
229 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
232 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
233 log line. The asterisk is now consistently appended to the remote IP
234 (and port, if given), and will also be provided on defer and fail log
235 lines. Previously it could be placed on the local IP if that was being
236 logged, and was only provided on delivery lines.
238 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
240 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
241 authenticator client_send option. Previously the next char, after a pair
242 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
243 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
244 documentation. There is still no way to get a leading ^ immediately
245 after a NUL (ie. for the password of a PLAIN method authenticator.
247 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
248 files. For bad sizes read, delete the record and whine to paniclog.
250 JH/38 When logging an AUTH failure, as server, do not include sensitive
251 information. Previously, the credentials would be included if given
252 as part of the AUTH command line and an ACL denied authentication.
254 JH/39 Bug 2691: fix $local_part_data. When the matching list element
255 referred to a file, bad data was returned. This likely also affected
258 JH/40 The gsasl authenticator now supports caching of the salted password
259 generated by the client-side implementation. This required the addition
260 of a new variable: $auth4.
262 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
263 left undeleted; the attempt to re-create it then failed - resulting in
264 the usual "SIGHUP tp have daemon reload configuration" to not work.
265 This affected any platform not supporting "abstract" Unix-domain
266 sockets (i.e. not Linux).
268 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
269 recipients" response to RCPT in a later response, with a 250. The
270 previous coding assumed this would not happen, and under PIPELINING
271 would result in both lost and duplicate recipients for a message.
273 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
274 Previously the weighting was incorrectly applied. Similar fix for socks
275 proxies. Found and fixed by Heiko Schlichting.
277 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
278 not handle sub-lists included using the +namedlist syntax. While
279 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
280 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
283 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
284 Previously the permanent pool was used, so the sore could not be freed.
285 This meant a connection with many messages would use continually-growing
288 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
289 per-pool so as not to waste too much space. Previously a constant size
290 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
291 DOS attacks harder. The cost is wasted memory use in the larger blocks.
293 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
294 same space repeatedly, and a smaller process footprint.
296 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
297 for GnuTLS. Previously the name started "%s" which, while not a bug,
298 looked as if if might be one.
300 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
301 "continued connection") make the $proxy_* variables available. Previously
302 the information was not passed across the exec() call for subsequent
303 transport executions. This also mean that the log lines for the
304 messages can show the proxy information.
306 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
307 receive time. With modern systems the difference is significant.
308 The historical behaviour can be restored by disabling (a new) log_selector
309 "queue_time_exclusive".
311 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
312 content) were specifically tested for. Now, also cover expansions
313 resulting in ACL names and inline ACL content.
315 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
316 making it unusable in complex expressions.
318 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
319 fail with a taint-check on the spoolfile name, and leave the message
322 HS/01 Enforce absolute PID file path name.
324 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
326 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
328 PP/02 Bug 2643: Correct TLS DH constants.
329 A missing NUL termination in our code-generation tool had led to some
330 incorrect Diffie-Hellman constants in the Exim source.
331 Reported by kylon94, code-gen tool fix by Simon Arlott.
333 PP/03 Impose security length checks on various command-line options.
334 Fixes CVE-2020-SPRSS reported by Qualys.
336 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
337 better. Reported by Qualys.
339 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
340 providing a particularly obnoxious sender full name.
343 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
345 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
348 PP/08 Change default for recipients_max from unlimited to 50,000.
350 PP/09 Fix security issue with too many recipients on a message (to remove a
351 known security problem if someone does set recipients_max to unlimited,
352 or if local additions add to the recipient list).
353 Fixes CVE-2020-RCPTL reported by Qualys.
355 PP/10 Fix security issue in SMTP verb option parsing
356 Fixes CVE-2020-EXOPT reported by Qualys.
358 PP/11 Fix security issue in BDAT state confusion.
359 Ensure we reset known-good where we know we need to not be reading BDAT
360 data, as a general case fix, and move the places where we switch to BDAT
361 mode until after various protocol state checks.
362 Fixes CVE-2020-BDATA reported by Qualys.
364 HS/03 Die on "/../" in msglog file names
366 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
367 the Exim runtime user are allowed to create files.
369 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
372 QS/03 When reading the output from interpreted forward files we do not
373 pass the pipe between the parent and the interpreting process to
374 executed child processes (if any).
376 QS/04 Always die if requested from internal logging, even is logging is
379 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
380 the API; compilation noo longer completes with DMARC support included.
381 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
382 been reported on other platforms.
384 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
386 JH/56 Make the majority of info read from config files readonly, for defence-in-
387 depth against exploits. Suggestion by Qualys.
388 Not supported on Solaris 10.
390 JH/57 Fix control=fakreject for a custom message containing tainted data.
391 Previously this resulted in a log complaint, due to a re-expansion present
392 since fakereject was originally introduced.
394 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
395 modifier was given, a loop resulted.
397 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
398 pipelined SMTP command followed the BDAT LAST then it would be
399 incorrectly treated as part of the message body, causing a verification
402 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
403 require looser validation rules than those for 821-level addresses,
404 which only permit IP addresses.
410 JH/01 Avoid costly startup code when not strictly needed. This reduces time
411 for some exim process initialisations. It does mean that the logging
412 of TLS configuration problems is only done for the daemon startup.
414 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
416 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
417 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
420 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
421 DKIM signing is being done. Previously a transport_filter would always
422 disable CHUNKING, falling back to traditional DATA.
424 JH/05 Regard command-line recipients as tainted.
426 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
428 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
429 PAM library frees one of the arguments given to it, despite the
430 documentation. Therefore a plain malloc must be used.
432 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
433 on-stack buffers were used, resulting in a taint trap when DSN information
434 copied from a received message was written into the buffer.
436 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
437 the ordering of its ARC headers. This caused a crash.
439 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
440 a new record was being constructed with information from the peer, a trap
443 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
444 installation would get error messages from DMARC verify, when it hit the
445 nonexistent file indicated by the default. Distros wanting DMARC enabled
446 should both provide the file and set the option.
447 Also enforce no DMARC verification for command-line sourced messages.
449 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
450 could, depending on the platform, hang at the STARTTLS response.
452 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
453 message on a connection. Previously if one message had ARC headers and
454 the following one did not, a crash could result when adding an
455 Authentication-Results: header.
457 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
458 Exim main code and Exim-related utities. The introduction of taint
459 tracking also did many adjustments to string handling. Since then, eximon
460 frequently terminated with an assert failure.
462 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
463 check for 452 responses. This slightly helps the inefficieny of doing
464 a large alias-expansion into a recipient-limited target. The max_rcpt
465 transport option still applies (and at the current default, will override
466 the new feature). The check is done for either cause of synch, and forces
467 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
468 connection. The new facility is not tunable at this time.
470 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
471 library live data was being used, so the results became garbage. Make
472 copies while it is still usable.
474 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
475 on delivery deferred (==) and failed (**) lines (if a delivery was
476 attemtped). Previously it was only on completion (=>) lines.
478 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
479 for the expansion of the server_scram_iter and server_scram_salt options.
481 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
482 are now specifically given a NO_DATA response without hitting the system
483 resolver. The library goes on to do the now-standard TXT lookup.
484 Use of dnsdb lookups is not affected.
486 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
487 only retrieve the errormessage once. Previously two calls to dlerror()
488 were used, and the second one (for mainlog/paniclog) retrieved null
491 JH/20 Taint checking: disallow use of tainted data for
492 - the appendfile transport file and directory options
493 - the pipe transport command
494 - the autoreply transport file, log and once options
495 - file names used by the redirect router (including filter files)
497 - paths used by single-key lookups
498 Previously this was permitted.
500 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
501 adjusted the size of a major service buffer; this failed because the
502 buffer was in use at the time. Change to a compile-time increase in the
503 buffer size, when this authenticator is compiled into exim.
505 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
506 previous fast-mode was untenable in the face of glibs using mmap to
507 support larger malloc requests.
509 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
510 New values supported, if defined on system where compiled:
511 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
512 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
514 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
515 running a limited number of proceses in parallel, a benefit is gained. The
516 amount varies with the platform hardware and load. The use of the option
517 queue_run_in_order means we cannot do this, as ordering becomes
520 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
521 had introduced a string-copy (for ensuring NUL-termination) which was not
522 appropriate for that case, which can include embedded NUL bytes in the
523 block of data. Investigation showed the copy to actually be needless, the
524 data being length-specified.
526 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
527 done during a receiving connection, and both used TLS, global info was
528 used rather than per-connection info for tracking the state of data
529 queued for transmission. This could result in a connection hang.
531 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
532 Previously, when delivering serveral messages down a single connection
533 only the first would provide a SIZE. This was due to the size information
534 not being properly tracked.
536 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
537 TAI (at 37 seconds currently), pretend to be in UTC for time-related
538 expansion and logging. Previously, spurious values such as a future
539 minute could be seen.
541 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
542 it could crash from a null-deref. This could also affect the
543 ${addresses: } operator and ${readsock } item.
545 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
546 message following a mime one, the variable was not reset.
548 JH/30 When an pipelined-connect fails at the first response, assume incorrect
549 cached capability (perhaps the peer reneged?) and immediately retry in
552 JH/31 Fix spurious detection of timeout while writing to transport filter.
554 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
555 an attempt to copy the string was made before checking it.
557 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
558 taint of the lookup key was maintained; we now regard the presence in the
559 filesystem as sufficient validation.
561 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
562 argument is supplied.
564 JH/35 The dsearch lookup now requires that the directory is an absolute path.
565 Previously this was not checked, and nonempty relative paths made an
566 access under Exim's current working directory.
568 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
569 Previously no event was raised.
571 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
572 parameter supplied by the sender MAIL FROM command. Previously it was
573 ignored, and only the check_spool_space option value for the required
576 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
577 the size of the signing public-key. Previously it was instead giving
578 the size of the signature hash.
580 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
581 the default. See the (new) dkim_verify_min_keysizes option.
583 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
584 and an ACL use a lookup for checking either the local_part or domain,
585 stale data could be accessed. Ensure that variable references are
586 dropped between messages.
588 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
589 by the client was not checked as pointing within response data before
590 being used. A malicious client could thus cause an out-of-bounds read and
591 possibly gain authentication. Fix by adding the check.
593 JH/42 Internationalisation: change the default for downconversion in the smtp
594 transport to be "if needed". Previously it was "as previously set" for
595 the message, which usually meant "if needed" for message-submission but
596 "no" for everything else. However, MTAs have been seen using SMTPUTF8
597 even when the envelope addresses did not need it, resulting in forwarding
598 failures to non-supporting MTAs. A downconvert in such cases will be
599 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
600 transport. The change does mean that addresses needing conversion will
601 be converted when previously a delivery failure would occur.
603 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
604 response was received it would be used unchecked in a fail-DSN, violating
605 standards on line-length limits. Truncate if needed.
607 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
608 doesn't work. (Additionally add a new main config option to configure the
615 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
616 side, to decode a TLS 1.3 packet capture.
618 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
619 Previously the default library behaviour applied, sending two, each in
622 JH/03 Debug output for ACL now gives the config file name and line number for
625 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
627 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
629 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
630 buffer overrun for (non-chunking) other transports.
632 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
633 TLS1.3, means that a server rejecting a client certificate is not visible
634 to the client until the first read of encrypted data (typically the
635 response to EHLO). Add detection for that case and treat it as a failed
636 TLS connection attempt, so that the normal retry-in-clear can work (if
637 suitably configured).
639 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
640 and/or domain. Found and fixed by Jason Betts.
642 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
643 configuration). If a CNAME target was not a wellformed name pattern, a
646 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
647 the OS reports them interleaved with other addresses.
649 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
650 used both for input and for a verify callout, both encrypted, SMTP
651 responses being sent by the server could be lost. This resulted in
652 dropped connections and sometimes bounces generated by a peer sending
655 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
656 string. Previously it was possible to cause undefined behaviour in a
657 library routine (usually a crash). Found by "zerons".
659 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
662 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
663 API was removed, so update to use the newer ones.
665 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
666 any timeout set, is taking a long time. Previously we would hang on to a
667 rotated logfile "forever" if the input was arriving with long gaps
668 (a previous attempt to fix addressed lack, for a long time, of initial
671 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
672 shared (NFS) environment. The length of the tempfile name is now
673 4 + 16 ("hdr.$message_exim_id") which might break on file
674 systems which restrict the file name length to lower values.
677 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
678 shared (NFS) environment.
680 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
681 did for all versions <4.90). Notably -M, -m, --invert, -I may be
684 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
685 on some platforms for bit 31.
687 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
688 to changes apparently associated with TLS1.3 handling some of the APIs
689 previously used were either nonfunctional or inappropriate. Strings
690 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
691 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
692 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
693 This affects log line X= elements, the $tls_{in,out}_cipher variables,
694 and the use of specific cipher names in the encrypted= ACL condition.
696 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
698 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
699 verification result was not updated unless hosts_require_ocsp applied.
701 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
702 queue_list_requires_admin set to false, non-admin users were denied the
705 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
706 directory-of-certs mode. Previously they were advertised despite the
709 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
710 A single TCP connection by a client will now hold a TLS connection open
711 for multiple message deliveries, by default. Previously the default was to
714 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
715 default. If built with the facility, DANE will be used. The facility
716 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
718 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
719 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
720 must be defined and you must still, unless you define DISABLE_TLS, manage
721 the the include-dir and library-file requirements that go with that
722 choice. Non-TLS builds are still supported.
724 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
727 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
728 default. If the platform supports and has the facility enabled, it will
729 be requested on all coneections.
731 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
732 controlled by the build-time option SUPPORT_PIPE_CONNECT.
734 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
736 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
737 success-DSN messages. Previously the From: header was always the default
738 one for these; the option was ignored.
740 JH/28 Fix the timeout on smtp response to apply to the whole response.
741 Previously it was reset for every read, so a teergrubing peer sending
742 single bytes within the time limit could extend the connection for a
743 long time. Credit to Qualsys Security Advisory Team for the discovery.
745 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
746 delivery address, which leaked information of the results of local
747 forwarding. Change to the original envelope recipient address, per
750 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
751 requested. Previously not bounce was generated and a log entry of
752 error ignored was made.
754 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
756 JH/32 Introduce a general tainting mechanism for values read from the input
757 channel, and values derived from them. Refuse to expand any tainted
758 values, to catch one form of exploit.
760 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
761 was unused and the unexpanded text used for the test. Found and
762 fixed by Ruben Jenster.
764 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
765 an attempt to use a TLS library read routine dereffed a nul pointer,
768 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
769 them in our smtp response.
771 JH/36 Have the general router option retry_use_local_part default to true when
772 any of the restrictive preconditions are set (to anything). Previously it
773 was only for check_local user. The change removes one item of manual
774 configuration which is required for proper retries when a remote router
775 handles a subset of addresses for a domain.
777 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
778 link count into consideration.
780 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
781 caused the extension of big_buffer, the following lines were ignored.
783 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
784 accordance with RFC 2308. Previously there was no expiry, so a longlived
785 receive process (eg. due to ACL delays) versus a short SOA value could
788 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
790 JH/39 Promote DMARC support to mainline.
792 JH/40 Bug 2452: Add a References: header to DSNs.
794 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
795 parameters. The relevant library call is documented as "Deprecated: This
796 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
797 3.6.0, DH parameters are negotiated following RFC7919."
799 HS/06 Change the default of dnssec_request_domains to "*"
801 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
802 carried on and emitted a BDAT command, even when PIPELINING was not
805 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
806 buffer was used for the filename, resulting in a trap when tainted
807 arguments (eg. $domain) were used.
809 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
810 recommended to avoid a possible server-load attack. The feature can be
811 re-enabled via the openssl_options main cofiguration option.
813 JH/45 local_scan API: documented the current smtp_printf() call. This changed
814 for version 4.90 - adding a "more data" boolean to the arguments.
815 Bumped the ABI version number also, this having been missed previously;
816 release versions 4.90 to 4.92.3 inclusive were effectively broken in
817 respect of usage of smtp_printf() by either local_scan code or libraries
818 accessed via the ${dlfunc } expansion item. Both will need coding
819 adjustment for any calls to smtp_printf() to match the new function
820 signature; a FALSE value for the new argument is always safe.
822 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
823 the file-offset (which the Linux syscall does, and exim expects); this
824 resulted in an indefinite loop.
826 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
827 to do ARC verification. The Authentication-Results: header line added
828 by the configuration then had no ARC item.
830 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
831 an incorrect mode was used for reading the result, resulting in it being
838 JH/01 Remove code calling the customisable local_scan function, unless a new
839 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
841 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
842 non-signal-safe functions being used.
844 JH/03 Bug 2269: When presented with a received message having a stupidly large
845 number of DKIM-Signature headers, disable DKIM verification to avoid
846 a resource-consumption attack. The limit is set at twenty.
848 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
849 report of oldest_pass in ${authres } in consequence, and separate out
850 some descriptions of reasons for verification fail.
852 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
853 files in the spool were present and unlocked. A queue-runner could spot
854 them, resulting in a duplicate delivery. Fix that by doing the unlock
855 after the unlink. Investigation by Tim Stewart. Take the opportunity to
856 add more error-checking on spoolfile handling while that code is being
859 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
860 No known attacks, no CVE, this is defensive hardening.
862 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
863 a queue-runner could start a delivery while other operations were ongoing.
864 Cutthrough delivery was a common victim, resulting in duplicate delivery.
865 Found and investigated by Tim Stewart. Fix by using the open message data
866 file handle rather than opening another, and not locally closing it (which
867 releases a lock) for that case, while creating the temporary .eml format
868 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
870 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
871 $sender_verify_failure/$recipient_verify_failure to "random".
873 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
876 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
877 Previously this would segfault.
879 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
882 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
883 like zero, since the resolver should be doing this for us, But we need one
884 as a CNAME but no MX presence gets the CNAME returned; we need to check
885 that doesn't point to an MX to declare it "no MX returned" rather than
886 "error, loop". A new main option is added so the older capability of
887 following some limited number of chain links is maintained.
889 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
891 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
892 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
893 now-deprecated earlier definitions used only specified the range up to TLS
894 1.2 (in the older-version library docs).
896 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
898 JH/15 Rework TLS client-side context management. Stop using a global, and
899 explicitly pass a context around. This enables future use of TLS for
900 connections to service-daemons (eg. malware scanning) while a client smtp
901 connection is using TLS; with cutthrough connections this is quite likely.
903 JH/16 Fix ARC verification to do AS checks in reverse order.
905 JH/17 Support a "tls" option on the ${readsocket } expansion item.
907 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
908 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
909 Previously the "utf8" would be re-prepended for every additional message.
911 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
912 Previously thery were accepted, resulting in issues when attempting to
913 forward messages to a non-supporting MTA.
915 PP/02 Let -n work with printing macros too, not just options.
917 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
918 one parent address was copied, and bogus data was used at delivery-logging
919 time. Either a crash (after delivery) or bogus log data could result.
920 Discovery and analysis by Tim Stewart.
922 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
923 Previously if the string ended mid-character, we did not insert the
924 promised '?' replacement.
926 PP/04 Documentation: current string operators work on bytes, not codepoints.
928 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
929 should pack well giving a smaller memory footprint so better caching and
930 therefore performance. Group the declarations where this can't be done so
931 that the byte-sized flag variables are not interspersed among pointer
932 variables, giving a better chance of good packing by the compiler.
934 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
935 non-null, to avoid issues with sites running BATV. Previously reports were
936 sent with an empty envelope sender so looked like bounces.
938 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
939 The ignore_error flag wasn't being returned from the filter subprocess so
940 was not set for later routers. Investigation and fix by Matthias Kurz.
942 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
943 and a msg:complete for the whole, when a message is manually removed using
944 -Mrm. Developement by Matthias Kurz, hacked on by JH.
946 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
947 a "Gnu special" function, asprintf() in the DB utility binary builds; I
948 hope that is portable enough.
950 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
951 requiring a known-CA anchor certificate; make it now rely entirely on the
952 TLSA as an anchor. Checking the name on the leaf cert against the name
953 on the A-record for the host is still done for TA (but not for EE mode).
955 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
956 included in delivery lines for non-proxied connections, when compiled with
957 SUPPORT_SOCKS and running with proxy logging enabled.
959 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
960 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
961 move the existing event to fire before the normal logging of message
962 failure so that custom logging is bracketed by normal logging.
964 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
965 msg:fail:internal event. Developement by Matthias Kurz.
967 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
968 far too small for todays use of crypto signatures stored there. Go all
969 the way to the max DNS message size of 64kB, even though this might be
970 overmuch for IOT constrained device use.
972 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
973 copy a string over itself. The library routine is documented as not
974 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
976 JH/32 For main options check_spool_space and check_inode_space, where the
977 platform supports 64b integers, support more than the previous 2^31 kB
978 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
979 the previous G, M, k.
981 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
982 $authenticated_fail_id variable on authentication failure. Previously
985 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
986 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
987 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
988 GNUTLS_SEC_PARAM_MEDIUM.
990 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
991 side. Previously we would continue as if no SNI had been received.
993 JH/36 Harden the handling of string-lists. When a list consisted of a sole
994 "<" character, which should be a list-separator specification, we walked
995 off past the nul-terimation.
997 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
998 causes) even when the retry time is not yet met. Previously they were
999 not, meaning that when (say) an account was over-quota and temp-rejecting,
1000 and multiple senders' messages were queued, only one sender would get
1001 notified on each configured delay_warning cycle.
1003 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1005 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1006 cipher-suites, an error can be left on the stack even for a succeeding
1007 accept; this results in impossible error messages when a later operation
1010 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1011 return error codes indicating retry. Under TLS1.3 this becomes required.
1013 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1014 it only wrote the new authenticators, resulting in a lack of tracking of
1015 peer changes of ESMTP extensions until the next cache flush.
1017 JH/41 Fix the loop reading a message header line to check for integer overflow,
1018 and more-often against header_maxsize. Previously a crafted message could
1019 induce a crash of the recive process; now the message is cleanly rejected.
1021 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1022 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1028 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1029 When redis_servers is set to a list of > 1 element, and the Redis servers
1030 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1031 case of MOVED into a DEFER case instead, thus moving the query onto the
1032 next server in the list. For a cluster of N elements, all N servers must
1033 be defined in redis_servers.
1035 GF/02 Catch and remove uninitialized value warning in exiqsumm
1036 Check for existence of @ARGV before looking at $ARGV[0]
1038 JH/01 Replace the store_release() internal interface with store_newblock(),
1039 which internalises the check required to safely use the old one, plus
1040 the allocate and data copy operations duplicated in both (!) of the
1041 extant use locations.
1043 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1044 modifier. This matches the restriction on the commandline.
1046 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1047 Previously only the last row was returned.
1049 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1050 we assumed that tags in the header were well-formed, and parsed the
1051 element content after inspecting only the first char of the tag.
1052 Assumptions at that stage could crash the receive process on malformed
1055 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1056 While running the DKIM ACL we operate on the Permanent memory pool so that
1057 variables created with "set" persist to the DATA ACL. Also (at any time)
1058 DNS lookups that fail create cache records using the Permanent pool. But
1059 expansions release any allocations made on the current pool - so a dnsdb
1060 lookup expansion done in the DKIM ACL releases the memory used for the
1061 DNS negative-cache, and bad things result. Solution is to switch to the
1062 Main pool for expansions.
1063 While we're in that code, add checks on the DNS cache during store_reset,
1064 active in the testsuite.
1065 Problem spotted, and debugging aided, by Wolfgang Breyha.
1067 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1068 When none of the hosts presented to a transport match an already-open
1069 connection, close it and proceed with the list. Previously we would
1070 queue the message. Spotted by Lena with Yahoo, probably involving
1073 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1074 Previously a spurious "250 OK id=" response was appended to the proper
1077 JH/08 The "support for" informational output now, which built with Content
1078 Scanning support, has a line for the malware scanner interfaces compiled
1079 in. Interface can be individually included or not at build time.
1081 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1082 by the template makefile "src/EDITME". The "STREAM" support for an older
1083 ClamAV interface method is removed.
1085 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1086 rows affected is given instead).
1088 JH/11 The runtime Berkeley DB library version is now additionally output by
1089 "exim -d -bV". Previously only the compile-time version was shown.
1091 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1092 SMTP connection. Previously, when one had more recipients than the
1093 first, an abortive onward connection was made. Move to full support for
1094 multiple onward connections in sequence, handling cutthrough connection
1095 for all multi-message initiating connections.
1097 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1098 routers. Previously, a multi-recipient message would fail to match the
1099 onward-connection opened for the first recipient, and cause its closure.
1101 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1102 a timeout on read on a GnuTLS initiating connection, resulting in the
1103 initiating connection being dropped. This mattered most when the callout
1104 was marked defer_ok. Fix to keep the two timeout-detection methods
1107 JH/15 Relax results from ACL control request to enable cutthrough, in
1108 unsupported situations, from error to silently (except under debug)
1109 ignoring. This covers use with PRDR, frozen messages, queue-only and
1112 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1114 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1115 metadata, resulting in a crash in free().
1117 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1118 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1119 Broken also in d185889f4, with init system revamp.
1121 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1122 Previously we abruptly closed the connection after reading a malware-
1123 found indication; now we go on to read the "scan ok" response line,
1126 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1127 ACL. Previously, a crash would result.
1129 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1130 macro text after a replacement (previously it was only once per line) and
1131 by skipping builtin macros when searching for an uppercase lead character.
1133 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1134 for the build is renamed.
1136 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1137 was allocated for every new TLS startup, meaning one per message. Fix
1138 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1140 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1141 reported the original. Fix to report (as far as possible) the ACL
1142 result replacing the original.
1144 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1145 OpenSSL. Certificate information is loaded for every new TLS startup,
1146 and the resources needed to be freed.
1148 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1150 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1153 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1154 DATA response info to the (existing) per-recipient response info for
1155 the "C=" log element. It can have useful tracking info from the
1156 destination system. Patch from Simon Arlott.
1158 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1159 length value. Previously this would segfault.
1161 HS/02 Support Avast multiline protoocol, this allows passing flags to
1162 newer versions of the scanner.
1164 JH/28 Ensure that variables possibly set during message acceptance are marked
1165 dead before release of memory in the daemon loop. This stops complaints
1166 about them when the debug_store option is enabled. Discovered specifically
1167 for sender_rate_period, but applies to a whole set of variables.
1168 Do the same for the queue-runner and queue-list loops, for variables set
1169 from spool message files. Do the same for the SMTP per-message loop, for
1170 certain variables indirectly set in ACL operations.
1172 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1173 as a multi-recipient message from a mailinglist manager). The coding had
1174 an arbitrary cutoff number of characters while checking for more input;
1175 enforced by writing a NUL into the buffer. This corrupted long / fast
1176 input. The problem was exposed more widely when more pipelineing of SMTP
1177 responses was introduced, and one Exim system was feeding another.
1178 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1179 receiving system, and refused recipients seen by the sending system
1180 (propating to people being dropped from mailing lists).
1181 Discovered and pinpointed by David Carter.
1183 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1184 replaced by the ${authresults } expansion.
1186 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1188 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1189 allows proper process termination in container environments.
1191 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1192 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1194 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1195 and "err_perm", deprecated since 4.83 when the RFC-defined words
1196 "temperror" and "permerror" were introduced.
1198 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1199 transport-filters or DKIM-signing. The restriction was lost in the
1200 consolidation of verify-callout and delivery SMTP handling.
1201 Extend the restriction to also cover ARC-signing.
1203 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1204 in defer=pass mode supply a 450 to the initiator. Previously the message
1207 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1208 tls_require_ciphers is used as before.
1210 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1211 "pass_unscanned". Only tmpfails from the scanner are written to
1212 the paniclog, as they may require admin intervention (permission
1213 denied, license issues). Other scanner errors (like decompression
1214 bombs) do not cause a paniclog entry.
1216 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1217 Previously it was possible to log spurious information in receive log
1220 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1221 triggered odd behaviour from Outlook Express clients.
1223 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1226 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1227 since the IETF WG has not yet settled on that versus the original
1228 "bare" representation.
1230 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1231 Previously the millisecond value corrupted the output.
1232 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1233 corrupted the output.
1239 JH/01 Rework error string handling in TLS interface so that the caller in
1240 more cases is responsible for logging. This permits library-sourced
1241 string to be attached to addresses during delivery, and collapses
1242 pairs of long lines into single ones.
1244 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1245 during configuration. Wildcards are allowed and expanded.
1247 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1248 This permits better logging.
1250 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1251 do not close it down and have the child transport start it up again on
1252 the passed-on TCP connection. Instead, proxy the child (and any
1253 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1254 affected: the continued delivery log lines do not have any DNSSEC, TLS
1255 Certificate or OCSP information. TLS cipher information is still logged.
1257 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1258 identical IP addresses on different listening ports. Will also affect
1261 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1262 add noisy ifdef guards to special-case this sillyness.
1263 Patch from Bernd Kuhls.
1265 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1266 than 255 are no longer allowed.
1268 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1269 Disable the session-cache too, which might reduce our load. Since we
1270 currrectly use a new context for every connection, both as server and
1271 client, there is no benefit for these.
1272 GnuTLS appears to not support tickets server-side by default (we don't
1273 call gnutls_session_ticket_enable_server()) but client side is enabled
1274 by default on recent versions (3.1.3 +) unless the PFS priority string
1277 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1278 <https://reproducible-builds.org/specs/source-date-epoch/>.
1280 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1281 the check for any unsuccessful recipients did not notice the limit, and
1282 erroneously found still-pending ones.
1284 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1285 MSG_MORE. Only in-clear (not on TLS connections).
1287 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1288 transport-filter is involved we can buffer the headers in memory for
1289 creating the signature, and read the spool data file once for the
1290 signature and again for transmission.
1292 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1293 4.77 as the kernel support then wasn't solid, having issues in 64bit
1294 mode. Now, it's been long enough. Add support for FreeBSD also.
1296 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1297 case where the routing stage had gathered several addresses to send to
1298 a host before calling the transport for the first, we previously failed
1299 to close down TLS in the old transport process before passing the TCP
1300 connection to the new process. The new one sent a STARTTLS command
1301 which naturally failed, giving a failed delivery and bloating the retry
1302 database. Investigation and fix prototype from Wolfgang Breyha.
1304 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1305 false-negatives in the check that the sender had not preempted a response
1306 or prompt from Exim (running as a server), due to that code's lack of
1307 awareness of the SMTP input buffering.
1309 PP/04 Add commandline_checks_require_admin option.
1310 Exim drops privileges sanely, various checks such as -be aren't a
1311 security problem, as long as you trust local users with access to their
1312 own account. When invoked by services which pass untrusted data to
1313 Exim, this might be an issue. Set this option in main configuration
1314 AND make fixes to the calling application, such as using `--` to stop
1317 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1318 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1319 responses to those, into a single TLS record each way (this usually means
1320 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1323 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1324 affects you only if you're dancing at the edge of the param size limits.
1325 If you are, and this message makes sense to you, then: raise the
1326 configured limit or use OpenSSL 1.1. Nothing we can do for older
1329 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1330 cmdline element to get the documented default one. Previously it was
1333 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1334 are unsuable for non-message-specific reasons - eg. connection timeout,
1335 banner-time rejection.
1337 JH/16 Fix logging of delivery remote port, when specified by router, under
1340 PP/06 Repair manualroute's ability to take options in any order, even if one
1341 is the name of a transport.
1344 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1346 JH/17 Change the list-building routines interface to use the expanding-string
1347 triplet model, for better allocation and copying behaviour.
1349 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1350 Previously it was constructed the first time a possibly-matching string
1351 was met in the configuration file input during startup; now it is done
1354 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1355 compatible one, to avoid the (poorly documented) possibility of a config
1356 file in the working directory redirecting the DB files, possibly correpting
1357 some existing file. CVE-2017-10140 assigned for BDB.
1359 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1360 cache-hot. Previously, although the result was properly cached, the
1361 initial verify call returned a defer.
1363 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1364 the main verify for receipient in uncached-mode.
1366 JH/22 Retire historical build files to an "unsupported" subdir. These are
1367 defined as "ones for which we have no current evidence of testing".
1369 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1370 if present. Previously it was ignored.
1372 JH/24 Start using specified-initialisers in C structure init coding. This is
1373 a C99 feature (it's 2017, so now considered safe).
1375 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1376 if was a fixed-sized field and bitmask ops via macros; it is now more
1379 PP/07 GitHub PR 56: Apply MariaDB build fix.
1380 Patch provided by Jaroslav Škarvada.
1382 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1383 during Coverity cleanups [4.87 JH/47]
1384 Diagnosis and fix provided by Michael Fischer v. Mollard.
1386 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1387 the right size to place the terminating semicolon on its own folded
1388 line, the header hash was calculated to an incorrect value thanks to
1389 the (relaxed) space the fold became.
1391 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1392 and confused the parent.
1394 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1395 which could crash as a result. This could lead to undeliverable messages.
1397 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1400 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1401 matters on fast-turnover and PID-randomising systems, which were getting
1402 out-of-order delivery.
1404 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1405 a possibly-overlapping copy. The symptom was that "Remote host closed
1406 connection in response to HELO" was logged instead of the actual 4xx
1409 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1410 Previously only that bufferd was discarded, resulting in SYMTP command
1413 JH/32 DKIM: when a message has multiple signatures matching an identity given
1414 in dkim_verify_signers, run the dkim acl once for each. Previously only
1415 one run was done. Bug 2189.
1417 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1418 "panic the current process" to "deliberately defer". The panic log is
1419 still written with the problem list name; the mail and reject logs now
1420 get a temp-reject line for the message that was being handled, saying
1421 something like "domains check lookup or other defer". The SMTP 451
1422 message is still "Temporary local problem".
1424 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1425 A crafted sequence of BDAT commands could result in in-use memory beeing
1426 freed. CVE-2017-16943.
1428 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1429 from SMTP input. Previously it was always done; now only done for DATA
1430 and not BDAT commands. CVE-2017-16944.
1432 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1433 to the message (such as an overlong header line). Previously this was
1434 not done and we did not exit BDAT mode. Followon from the previous item
1435 though a different problem.
1441 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1442 than -2003 did; needs libidn2 in addition to libidn.
1444 JH/02 The path option on a pipe transport is now expanded before use.
1446 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1447 Patch provided by "Björn", documentation fix added too.
1449 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1450 missing a wire-to-host endian conversion.
1452 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1453 close after a BDAT command line could be taken as a following command,
1454 giving a synch failure. Fix by only checking for synch immediately
1455 before acknowledging the chunk.
1457 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1458 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1459 macro. Patches provided by Josh Soref.
1461 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1462 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1463 by IANA as a service extension. However, John Klensin suggests that we
1466 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1467 the dkim code may be unix-mode line endings rather than smtp wire-format
1468 CRLF, so prepend a CR to any bare LF.
1470 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1471 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1473 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1474 the first were themselves being wrongly included in the feed into dkim
1475 processing; with most chunk sizes in use this resulted in an incorrect
1476 body hash calculated value.
1478 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1479 DKIM signature block, for verification. Although advised against by
1480 standards it is specifically not ruled illegal.
1482 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1484 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1485 missing a body hash (the bh= tag).
1487 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1488 It seems that HAProxy sends the Proxy Protocol information in clear and
1489 only then does a TLS startup, so do the same.
1491 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1492 TCP connections (such as for Spamd) unless the daemon successfully set
1493 Fast Open mode on its listening sockets. This fixes breakage seen on
1494 too-old kernels or those not configured for Fast Open, at the cost of
1495 requiring both directions being enabled for TFO, and TFO never being used
1496 by non-daemon-related Exim processes.
1498 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1499 endings, at least on the first header line. Try to canonify any that get
1500 past that check, despite the cost.
1502 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1503 now limited to an arbitrary five deep, while parsing addresses with the
1504 strip_excess_angle_brackets option enabled.
1506 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1507 instead leave the unprompted TLS handshake in socket buffer for the
1508 TLS library to consume.
1510 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1512 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1514 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1515 region in one operation, for speed, and this leaves assigned pointers
1516 dangling. Add checks run only under the testsuite which checks all
1517 variables at a store-reset and panics on a dangling pointer; add code
1518 explicitly nulling out all the variables discovered. Fixes one known
1519 bug: a transport crash, where a dangling pointer for $sending_ip_address
1520 originally assigned in a verify callout, is re-used.
1522 PP/06 Drop '.' from @INC in various Perl scripts.
1524 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1526 PP/08 Reduce a number of compilation warnings under clang; building with
1527 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1528 should be warning-free.
1530 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1532 HS/01 Fix portability problems introduced by PP/08 for platforms where
1533 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1535 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1536 chunk. This allows us to accept broken chunked messages. We need a more
1537 general solution here.
1539 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1540 already-broken messages in the queue.
1542 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1544 JH/19 Fix reference counting bug in routing-generated-address tracking.
1550 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1551 supports it and a size is available (ie. the sending peer gave us one).
1553 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1554 years of being deprecated). The replacements are the ACLs
1555 acl_smtp_mime and acl_not_smtp_mime.
1557 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1558 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1559 or even an in-clear connection were permitted. Now, if the host lookup
1560 was dnssec and dane was requested then the host is only used if the
1561 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1562 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1563 if one fails this test.
1564 This means that a poorly-configured remote DNS will make it incommunicado;
1565 but it protects against a DNS-interception attack on it.
1567 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1568 non-noisy when a race steals the message being considered.
1570 JH/05 If main configuration option tls_certificate is unset, generate a
1571 self-signed certificate for inbound TLS connections.
1573 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1574 in rewrites and routers.
1576 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1577 and logged a warning sing 4.83; now they are a configuration file error.
1579 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1580 (lacking @domain). Apply the same qualification processing as RCPT.
1582 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1584 JH/10 Support ${sha256:} applied to a string (as well as the previous
1587 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1588 a cutthrough deliver is pending, as we always want to make a connection.
1589 This also avoids re-routing the message when later placing the cutthrough
1590 connection after a verify cache hit.
1591 Do not update it with the verify result either.
1593 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1594 when routing results in more than one destination address.
1596 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1597 signing (which inhibits the cutthrough capability). Previously only
1598 the presence of an option was tested; now an expansion evaluating as
1599 empty is permissible (obviously it should depend only on data available
1600 when the cutthrough connection is made).
1602 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1603 the relevant preceding SMTP command did not note the pipelining mode.
1605 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1606 Previously they were not counted.
1608 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1609 as one having no matching records. Previously we deferred the message
1610 that needed the lookup.
1612 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1613 distinguished as "(=".
1615 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1616 for missing MX records. Previously it only worked for missing A records.
1618 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1620 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1621 after the data-go-ahead and data-ack. Patch from Jason Betts.
1623 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1624 even for a "none" policy. Patch from Tony Meyer.
1626 JH/22 Fix continued use of a connection for further deliveries. If a port was
1627 specified by a router, it must also match for the delivery to be
1630 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1631 When one of the recipients of a message was unsuitable for the connection
1632 (has no matching addresses), we lost track of needing to mark it
1633 deferred. As a result mail would be lost.
1635 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1637 JH/25 Decoding ACL controls is now done using a binary search; the source code
1638 takes up less space and should be simpler to maintain. Merge the ACL
1639 condition decode tables also, with similar effect.
1641 JH/26 Fix problem with one_time used on a redirect router which returned the
1642 parent address unchanged. A retry would see the parent address marked as
1643 delivered, so not attempt the (identical) child. As a result mail would
1646 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1647 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1648 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1651 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1652 Main config options check_{log,spool}_{inodes,space} are now
1653 100 inodes, 10MB unless set otherwise in the configuration.
1655 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1656 Previously it only applied to the main-section connection policy
1659 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1661 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1662 by me. Added RFC7919 DH primes as an alternative.
1664 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1665 are not in the system include path.
1667 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1668 GnuTLS, when a session startup failed (eg because the client disconnected)
1669 Exim did stdio operations after fclose. This was exposed by a recent
1670 change which nulled out the file handle after the fclose.
1672 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1673 signed directly by the cert-signing cert, rather than an intermediate
1674 OCSP-signing cert. This is the model used by LetsEncrypt.
1676 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1678 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1679 an incoming connection.
1681 HS/02 Bug 1802: Do not half-close the connection after sending a request
1684 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1685 fallback to "prime256v1".
1687 JH/34 SECURITY: Use proper copy of DATA command in error message.
1688 Could leak key material. Remotely exploitable. CVE-2016-9963.
1694 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1695 and 3.4.4 - once the server is enabled to respond to an OCSP request
1696 it does even when not requested, resulting in a stapling non-aware
1697 client dropping the TLS connection.
1699 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1700 support variable-length bit vectors. No functional change.
1702 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1703 The I= interface field on outgoing lines is now after the H= remote
1704 host field, same as incoming lines. There is a separate
1705 outgoing_interface log selector which allows you to disable the
1708 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1709 If not running log_selector +smtp_connection the mainlog would be held
1710 open indefinitely after a "too many connections" event, including to a
1711 deleted file after a log rotate. Leave the per net connection logging
1712 leaving it open for efficiency as that will be quickly detected by the
1713 check on the next write.
1715 HS/01 Bug 1671: Fix post transport crash.
1716 Processing the wait-<transport> messages could crash the delivery
1717 process if the message IDs didn't exist for some reason. When
1718 using 'split_spool_directory=yes' the construction of the spool
1719 file name failed already, exposing the same netto behaviour.
1721 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1722 mime_regex ACL conditions.
1724 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1725 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1726 to HELO, local diagnostic string.
1728 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1729 log line to debug. Even when configured with a tls authenticator many
1730 client connections are expected to not authenticate in this way, so
1731 an authenticate fail is not an error.
1733 HS/02 Add the Exim version string to the process info. This way exiwhat
1734 gives some more detail about the running daemon.
1736 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1737 matter for fast-change records such as DNSBLs.
1739 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1740 retry record, even if constant. There may be multiple transports with
1741 different interface settings and the retry behaviour needs to be kept
1744 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1746 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1748 JH/10 Bug 840: fix log_defer_output option of pipe transport
1750 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1751 hosts_randomize. This matters a lot when combined with mua_wrapper.
1753 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1754 ${quote_pgsql:<string>} operator.
1756 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1758 JH/14 Change header file name for freeradius-client. Relevant if compiling
1759 with Radius support; from the Gentoo tree and checked under Fedora.
1761 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1763 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1764 option transport_filter, meaning no filtering.
1766 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1768 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1769 defaults to "*" (all hosts). The variable is now available when not built
1770 with TLS, default unset, mainly to enable keeping the testsuite sane.
1771 If a server certificate is not supplied (via tls_certificate) an error is
1772 logged, and clients will find TLS connections fail on startup. Presumably
1773 they will retry in-clear.
1774 Packagers of Exim are strongly encouraged to create a server certificate
1775 at installation time.
1777 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1778 with the $config_file variable.
1780 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1781 in transport context, after the attempt, and per-recipient. The latter type
1782 is per host attempted. The event data is the error message, and the errno
1783 information encodes the lookup type (A vs. MX) used for the (first) host,
1784 and the trailing two digits of the smtp 4xx response.
1786 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1787 to write to mainlog (or rejectlog, paniclog) in the window between file
1788 creation and permissions/ownership being changed. Particularly affects
1789 installations where exicyclog is run as root, rather than exim user;
1790 result is that the running daemon panics and dies.
1792 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1794 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1795 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1796 "pri" and "weight". Note that the previous implicit priority given by the
1797 list order is no longer honoured.
1799 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1800 for DKIM processing.
1802 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1803 by defining SUPPORT_SOCKS.
1805 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1806 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1807 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1808 variables are renamed to proxy_{local,external}_{address,port}.
1810 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1811 for a build by defining SUPPORT_I18N
1813 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1814 of the query string, and make ${quote_redis:} do that quoting.
1816 JH/29 Move Events support from Experimental to mainline, enabled by default
1817 and removable for a build by defining DISABLE_EVENT.
1819 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1821 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1822 cached by the daemon.
1824 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1825 by defining LOOKUP_REDIS. The libhiredis library is required.
1827 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1828 keys are given for lookup.
1830 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1831 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1832 only supported when built with TLS support. The PolarSSL SHA routines
1833 are still used when the TLS library is too old for convenient support.
1835 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1836 openssl_options), for security. OpenSSL forces this from version 1.1.0
1837 server-side so match that on older versions.
1839 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1840 allocation for $value could be released as the expansion processing
1841 concluded, but leaving the global pointer active for it.
1843 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1844 and to use the domains and local_parts ACL conditions.
1846 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1847 incorrectly not doubled on cutthrough transmission, hence seen as a
1848 body-termination at the receiving system - resulting in truncated mails.
1849 Commonly the sender saw a TCP-level error, and retransmitted the message
1850 via the normal store-and-forward channel. This could result in duplicates
1851 received - but deduplicating mailstores were liable to retain only the
1852 initial truncated version.
1854 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1856 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1858 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1859 we're in there, support oversigning also; bug 1309.
1861 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1863 HS/04 Add support for keep_environment and add_environment options.
1865 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1866 either intentional arithmetic overflow during PRNG, or testing config-
1869 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1870 delivery resulted in actual delivery. Cancel cutthrough before DATA
1873 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1874 rejects a recipient: pass the reject to the originator.
1876 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1877 Many were false-positives and ignorable, but it's worth fixing the
1880 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1881 for the new environment-manipulation done at startup. Move the routines
1882 from being local to tls.c to being global via the os.c file.
1884 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1885 an extract embedded as result-arg for a map, the first arg for extract
1886 is unavailable so we cannot tell if this is a numbered or keyed
1887 extraction. Accept either.
1893 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1896 JH/02 The smtp transport option "multi_domain" is now expanded.
1898 JH/03 The smtp transport now requests PRDR by default, if the server offers
1901 JH/04 Certificate name checking on server certificates, when exim is a client,
1902 is now done by default. The transport option tls_verify_cert_hostnames
1903 can be used to disable this per-host. The build option
1904 EXPERIMENTAL_CERTNAMES is withdrawn.
1906 JH/05 The value of the tls_verify_certificates smtp transport and main options
1907 default to the word "system" to access the system default CA bundle.
1908 For GnuTLS, only version 3.0.20 or later.
1910 JH/06 Verification of the server certificate for a TLS connection is now tried
1911 (but not required) by default. The verification status is now logged by
1912 default, for both outbound TLS and client-certificate supplying inbound
1915 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1918 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1919 Status Notification (bounce) messages are now MIME format per RFC 3464.
1920 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1921 under the control of the dsn_advertise_hosts option, and routers may
1922 have a dsn_lasthop option.
1924 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1925 default, modifiable by a malware= option. The list separator for
1926 the options can now be changed in the usual way. Bug 68.
1928 JH/10 The smtp_receive_timeout main option is now expanded before use.
1930 JH/11 The incoming_interface log option now also enables logging of the
1931 local interface on delivery outgoing connections.
1933 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1934 if the interface and destination host and port all match.
1936 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1939 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1940 Patch from Andrew Lewis.
1942 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1943 now supports optional time-restrictions, weighting, and priority
1944 modifiers per server. Patch originally by <rommer@active.by>.
1946 JH/16 The spamd_address main option now supports a mixed list of local
1947 and remote servers. Remote servers can be IPv6 addresses, and
1948 specify a port-range.
1950 JH/17 Bug 68: The spamd_address main option now supports an optional
1951 timeout value per server.
1953 JH/18 Bug 1581: Router and transport options headers_add/remove can
1954 now have the list separator specified.
1956 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1959 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1962 JH/21 Support for the A6 type of dns record is withdrawn.
1964 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1965 rather than the verbs used.
1967 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1968 from 255 to 1024 chars.
1970 JH/24 Verification callouts now attempt to use TLS by default.
1972 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1973 are generic router options now. The defaults didn't change.
1975 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1976 Original patch from Alexander Shikoff, worked over by JH.
1978 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1979 files if ZCAT_COMMAND is not executable.
1981 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1983 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1985 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1986 Normally benign, it bites when the pair was led to by a CNAME;
1987 modern usage is to not canonicalize the domain to a CNAME target
1988 (and we were inconsistent anyway for A-only vs AAAA+A).
1990 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1992 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1993 when evaluating $sender_host_dnssec.
1995 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1996 $sender_helo_dnssec variable.
1998 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2000 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2002 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2004 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2005 documented as working, but never had. Support all but $spam_report.
2007 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2008 added for tls authenticator.
2010 HS/03 Add perl_taintmode main config option
2016 TL/01 When running the test suite, the README says that variables such as
2017 no_msglog_check are global and can be placed anywhere in a specific
2018 test's script, however it was observed that placement needed to be near
2019 the beginning for it to behave that way. Changed the runtest perl
2020 script to read through the entire script once to detect and set these
2021 variables, reset to the beginning of the script, and then run through
2022 the script parsing/test process like normal.
2024 TL/02 The BSD's have an arc4random API. One of the functions to induce
2025 adding randomness was arc4random_stir(), but it has been removed in
2026 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2027 function when detected.
2029 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2030 cause callback expansion.
2032 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2033 syntax errors in an expansion can be treated as a string instead of
2034 logging or causing an error, due to the internal use of bool_lax
2035 instead of bool when processing it.
2037 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2038 server certificates when making smtp deliveries.
2040 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2042 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2044 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2046 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2047 Merged patch from Sebastian Wiedenroth.
2049 JH/05 Fix results-pipe from transport process. Several recipients, combined
2050 with certificate use, exposed issues where response data items split
2051 over buffer boundaries were not parsed properly. This eventually
2052 resulted in duplicates being sent. This issue only became common enough
2053 to notice due to the introduction of connection certificate information,
2054 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2056 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2057 size buffer was used, resulting in syntax errors when an expansion
2060 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2061 version 3.3.6 or later.
2063 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2064 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2065 both become "event_action", the variables become $event_name, $event_data
2066 and $event_defer_errno. There is a new variable $verify_mode, usable in
2067 routers, transports and related events. The tls:cert event is now also
2068 raised for inbound connections, if the main configuration event_action
2071 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2072 early OCSP support, but no stapling (appears to be less than 1.0.0).
2074 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2075 server certificate names available under the smtp transport option
2076 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2079 JH/10 Time-related extraction expansions from certificates now use the main
2080 option "timezone" setting for output formatting, and are consistent
2081 between OpenSSL and GnuTLS compilations. Bug 1541.
2083 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2084 encoded parameter in the incoming message. Bug 1558.
2086 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2087 include certificate info, eximon was claiming there were spoolfile
2090 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2092 JH/14 Log delivery-related information more consistently, using the sequence
2093 "H=<name> [<ip>]" wherever possible.
2095 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2096 are problematic for Debian distribution, omit them from the release
2099 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2101 JH/16 Fix string representation of time values on 64bit time_t architectures.
2104 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2105 output list separator was used.
2110 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2111 checkers that were complaining about end of non-void function with no
2114 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2115 This was a regression introduced in 4.83 by another bugfix.
2117 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2119 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2120 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2126 TF/01 Correctly close the server side of TLS when forking for delivery.
2128 When a message was received over SMTP with TLS, Exim failed to clear up
2129 the incoming connection properly after forking off the child process to
2130 deliver the message. In some situations the subsequent outgoing
2131 delivery connection happened to have the same fd number as the incoming
2132 connection previously had. Exim would try to use TLS and fail, logging
2133 a "Bad file descriptor" error.
2135 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2136 utilities have not been installed.
2138 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2139 temporary space as the ACL may create new global variables.
2141 TL/01 LDAP support uses per connection or global context settings, depending
2142 upon the detected version of the libraries at build time.
2144 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2145 to extract and use the src ip:port in logging and expansions as if it
2146 were a direct connection from the outside internet. PPv2 support was
2147 updated based on HAProxy spec change in May 2014.
2149 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2151 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2152 Properly escape header and check for NULL return.
2154 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2157 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2159 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2160 characters in header names, implemented as a verify condition.
2161 Contributed by Michael Fischer v. Mollard.
2163 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2164 results permerror and temperror. Previous values are deprecated but
2165 still accepted. In a future release, err_perm and err_temp will be
2166 completely removed, which will be a backward incompatibility if the
2167 ACL tests for either of these two old results. Patch contributed by
2168 user bes-internal on the mailing list.
2170 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2172 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2173 selectors, in both main and reject logs.
2175 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2178 JH/07 Add malware type "sock" for talking to simple daemon.
2180 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2182 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2183 routers/transports under cutthrough routing.
2185 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2186 numbers. Touch up "bool" conditional to keep the same definition.
2188 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2190 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2192 JH/12 Expand items in router/transport headers_add or headers_remove lists
2193 individually rather than the list as a whole. Bug 1452.
2195 Required for reasonable handling of multiple headers_ options when
2196 they may be empty; requires that headers_remove items with embedded
2197 colons must have them doubled (or the list-separator changed).
2199 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2200 view the policy declared in the DMARC record. Currently, $dmarc_status
2201 is a combined value of both the record presence and the result of the
2204 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2206 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2207 dnslookup router and the smtp transport (applying to the forward
2210 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2211 of ldap servers used for a specific lookup. Patch provided by Heiko
2214 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2215 New variable $lookup_dnssec_authenticated for observability.
2217 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2218 Patch submitted by Lars Timman.
2220 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2222 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2223 Requires trusted mode and valid format message id, aborts otherwise.
2224 Patch contributed by Heiko Schlichting.
2226 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2227 certextract with support for various fields. Bug 1358.
2229 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2230 is requested by default, modifiable by smtp transport option
2233 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2234 operate on certificate variables to give certificate fingerprints
2235 Also new ${sha256:cert_variable}.
2237 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2239 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2240 Christian Aistleitner.
2242 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2244 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2245 file. Patch from Wolfgang Breyha.
2247 JH/25 Expand the coverage of the delivery $host and $host_address to
2248 client authenticators run in verify callout. Bug 1476.
2250 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2251 align with daemon_smtp_ports. Bug 72.
2253 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2254 support and error reporting did not work properly.
2256 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2257 and is readable. Patch from Andrew Colin Kissa.
2259 TL/14 Enhance documentation of ${run expansion and how it parses the
2260 commandline after expansion, particularly in the case when an
2261 unquoted variable expansion results in an empty value.
2263 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2265 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2266 by renaming away from it. Addresses GH issue 15, reported by
2269 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2271 TL/15 SECURITY: prevent double expansion in math comparison functions
2272 (can expand unsanitized data). Not remotely exploitable.
2279 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2281 PP/02 Make -n do something, by making it not do something.
2282 When combined with -bP, the name of an option is not output.
2284 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2287 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2288 $sender_host_name and config options to manage this, and basic check
2291 PP/05 DSCP support for outbound connections and control modifier for inbound.
2293 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2294 (Only plugin which currently uses this is kerberos4, which nobody should
2295 be using, but we should make it available and other future plugins might
2296 conceivably use it, even though it would break NAT; stuff *should* be
2297 using channel bindings instead).
2299 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2300 name; added for Sendmail compatibility; requires admin caller.
2301 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2302 just ignore it); requires trusted caller.
2303 Also parse but ignore: -Ac -Am -X<logfile>
2306 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2308 TL/02 Add +smtp_confirmation as a default logging option.
2310 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2311 Patch by Magnus Holmgren from 2007-02-20.
2313 TL/04 Bugzilla 1281 - Spec typo.
2314 Bugzilla 1283 - Spec typo.
2315 Bugzilla 1290 - Spec grammar fixes.
2317 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2319 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2321 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2322 mailing list by by Dmitry Isaikin.
2324 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2326 JH/02 Support "G" suffix to numbers in ${if comparisons.
2328 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2330 NM/01 Bugzilla 1197 - Spec typo
2331 Bugzilla 1196 - Spec examples corrections
2333 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2335 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2336 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2339 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2340 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2342 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2343 "acl {{name}{arg}...}", and optional args on acl condition
2346 JH/05 Permit multiple router/transport headers_add/remove lines.
2348 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2350 JH/07 Avoid using a waiting database for a single-message-only transport.
2351 Performance patch from Paul Fisher. Bugzilla 1262.
2353 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2356 JH/09 Add $headers_added variable, with content from use of ACL modifier
2357 add_header (but not yet added to the message). Bugzilla 199.
2359 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2360 Pulled from Bugzilla 817 by Wolfgang Breyha.
2362 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2364 (nb: this is the same fix as in Exim 4.80.1)
2366 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2369 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2372 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2374 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2375 advertise SMTP AUTH mechanism to us, instead of a generic
2376 protocol violation error. Also, make Exim more robust to bad
2377 data from the Dovecot auth socket.
2379 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2381 When a queue runner is handling a message, Exim first routes the
2382 recipient addresses, during which it prunes them based on the retry
2383 hints database. After that it attempts to deliver the message to
2384 any remaining recipients. It then updates the hints database using
2387 So if a recipient address works intermittently, it can get repeatedly
2388 deferred at routing time. The retry hints record remains fresh so the
2389 address never reaches the final cutoff time.
2391 This is a fairly common occurrence when a user is bumping up against
2392 their storage quota. Exim had some logic in its local delivery code
2393 to deal with this. However it did not apply to per-recipient defers
2394 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2396 This change adds a proper retry rule check during routing so that the
2397 final cutoff time is checked against the message's age. We only do
2398 this check if there is an address retry record and there is not a
2399 domain retry record; this implies that previous attempts to handle
2400 the address had the retry_use_local_parts option turned on. We use
2401 this as an approximation for the destination being like a local
2402 delivery, as in LMTP.
2404 I suspect this new check makes the old local delivery cutoff check
2405 redundant, but I have not verified this so I left the code in place.
2407 TF/02 Correct gecos expansion when From: is a prefix of the username.
2409 Test 0254 submits a message to Exim with the header
2413 When I ran the test suite under the user fanf2, Exim expanded
2414 the header to contain my full name, whereas it should have added
2415 a Resent-Sender: header. It erroneously treats any prefix of the
2416 username as equal to the username.
2418 This change corrects that bug.
2420 GF/01 DCC debug and logging tidyup
2421 Error conditions log to paniclog rather than rejectlog.
2422 Debug lines prefixed by "DCC: " to remove any ambiguity.
2424 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2426 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2427 Bug spotted by Jeremy Harris; was flawed since initial commit.
2428 Would have resulted in OCSP responses post-SNI triggering an Exim
2429 NULL dereference and crash.
2431 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2433 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2434 Bug detection, analysis and fix by Samuel Thibault.
2435 Bugzilla 1331, Debian bug #698092.
2437 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2439 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2440 Server implementation by Todd Lyons, client by JH.
2441 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2442 config variable "prdr_enable" controls whether the server
2443 advertises the facility. If the client requests PRDR a new
2444 acl_data_smtp_prdr ACL is called once for each recipient, after
2445 the body content is received and before the acl_smtp_data ACL.
2446 The client is controlled by both of: a hosts_try_prdr option
2447 on the smtp transport, and the server advertisement.
2448 Default client logging of deliveries and rejections involving
2449 PRDR are flagged with the string "PRDR".
2451 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2452 fclose(). Diagnosis by Todd Lyons.
2454 PP/17 Update configure.default to handle IPv6 localhost better.
2455 Patch by Alain Williams (plus minor tweaks).
2458 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2459 This is now consistent with GnuTLS, and is now documented: the
2460 previous undocumented portable approach to treating the option as
2461 unset was to force an expansion failure. That still works, and
2462 an empty string is now equivalent.
2464 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2465 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2466 not performing validation itself.
2468 PP/20 Added force_command boolean option to pipe transport.
2469 Patch from Nick Koston, of cPanel Inc.
2471 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2474 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2476 PP/21 Fix eximon continuous updating with timestamped log-files.
2477 Broken in a format-string cleanup in 4.80, missed when I repaired the
2478 other false fix of the same issue.
2479 Report and fix from Heiko Schlichting.
2482 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2483 Report from Prashanth Katuri.
2485 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2486 It's SecureTransport, so affects any MacOS clients which use the
2487 system-integrated TLS libraries, including email clients.
2489 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2490 using a MIME ACL for non-SMTP local injection.
2491 Report and assistance in diagnosis by Warren Baker.
2493 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2495 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2497 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2498 last id that failed so it may be referenced in subsequent ACL's.
2500 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2503 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2504 ldap library initialization, allowing self-signed CA's to be
2505 used. Also properly sets require_cert option later in code by
2506 using NULL (global ldap config) instead of ldap handle (per
2507 session). Bug diagnosis and testing by alxgomz.
2509 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2510 the src/util/ subdirectory.
2512 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2513 renamed to Transport Post Delivery Action by Jeremy Harris, as
2516 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2517 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2518 redis_servers = needs to be configured which will be used by the redis
2519 lookup. Patch from Warren Baker, of The Packet Hub.
2521 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2523 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2524 hostname or reverse DNS when processing a host list. Used suggestions
2525 from multiple comments on this bug.
2527 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2529 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2530 Missed a few lines, added it to make the runtest require no keyboard
2533 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2534 contains upper case chars. Make router use caseful_local_part.
2536 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2537 support when GnuTLS has been built with p11-kit.
2543 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2545 This, or similar/improved, will also be change PP/11 of 4.82.
2551 PP/01 Handle short writes when writing local log-files.
2552 In practice, only affects FreeBSD (8 onwards).
2553 Bugzilla 1053, with thanks to Dmitry Isaikin.
2555 NM/01 Bugzilla 949 - Documentation tweak
2557 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2560 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2562 PP/02 Implemented gsasl authenticator.
2564 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2566 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2567 `pkg-config foo` for cflags/libs.
2569 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2570 with rest of GSASL and with heimdal_gssapi.
2572 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2573 `pkg-config foo` for cflags/libs for the TLS implementation.
2575 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2576 properties get this fed in as external SSF. A number of robustness
2577 and debugging improvements to the cyrus_sasl authenticator.
2579 PP/08 cyrus_sasl server now expands the server_realm option.
2581 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2582 Patch by Jeremy Harris.
2584 PP/10 Added dbmjz lookup type.
2586 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2588 PP/12 MAIL args handles TAB as well as SP, for better interop with
2589 non-compliant senders.
2590 Analysis and variant patch by Todd Lyons.
2592 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2593 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2594 Patch from Dirk Mueller <dmueller@suse.com>
2596 PP/13 tls_peerdn now print-escaped for spool files.
2597 Observed some $tls_peerdn in wild which contained \n, which resulted
2598 in spool file corruption.
2600 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2601 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2602 or write after TLS renegotiation, which otherwise led to messages
2605 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2606 as a tracking header (ie: a signed header comes before the signature).
2607 Patch from Wolfgang Breyha.
2609 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2610 comma-sep list; embedded commas doubled.
2612 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2614 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2616 Report and patch from Dmitry Banschikov.
2618 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2619 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2620 failures. We appear to now support TLS1.1+ with Exim.
2622 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2623 lets Exim select keys and certificates based upon TLS SNI from client.
2624 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2625 before an outbound SMTP session. New log_selector, +tls_sni.
2627 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2628 NULL dereference. Report and patch from Alun Jones.
2630 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2631 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2632 Not seeing resolver debug output on NetBSD, but suspect this is a
2633 resolver implementation change.
2635 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2636 Left warnings. Added "eximon gdb" invocation mode.
2638 PP/21 Defaulting "accept_8bitmime" to true, not false.
2640 PP/22 Added -bw for inetd wait mode support.
2642 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2643 locate the relevant includes and libraries. Made this the default.
2645 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2646 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2648 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2649 This may cause build issues on older platforms.
2651 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2652 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2653 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2654 Added SNI support via GnuTLS too.
2655 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2657 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2659 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2660 Applied second patch from Janne, implementing suggestion to default
2661 multiple-strings-in-record handling to match SPF spec.
2663 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2665 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2666 read-only, out of scope).
2667 Patch from Wolfgang Breyha, report from Stuart Northfield.
2669 PP/29 Fix three issues highlighted by clang analyser static analysis.
2670 Only crash-plausible issue would require the Cambridge-specific
2671 iplookup router and a misconfiguration.
2672 Report from Marcin Mirosław.
2674 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2676 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2677 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2678 As part of this, removing so much warning spew let me fix some minor
2679 real issues in debug logging.
2681 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2682 assignment on my part. Fixed.
2684 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2685 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2686 Janne Snabb (who went above and beyond: thank you).
2688 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2689 string otherwise requires a connection and a bunch more work and it's
2690 relatively easy to get wrong. Should also expose TLS library linkage
2693 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2694 64-bit ${eval} (JH/03).
2696 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2697 GNU libc to support some of the 64-bit stuff, should not lead to
2698 conflicts. Defined before os.h is pulled in, so if a given platform
2699 needs to override this, it can.
2701 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2702 protection layer was required, which is not implemented.
2703 Bugzilla 1254, patch from Wolfgang Breyha.
2705 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2706 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2707 tls_dhparam take prime identifiers. Also unbreak combination of
2708 OpenSSL+DH_params+TLSSNI.
2710 PP/39 Disable SSLv2 by default in OpenSSL support.
2716 PP/01 Solaris build fix for Oracle's LDAP libraries.
2717 Bugzilla 1109, patch from Stephen Usher.
2719 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2721 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2724 TF/02 Fix a couple more cases where we did not log the error message
2725 when unlink() failed. See also change 4.74-TF/03.
2727 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2728 lock up or crash if it happened to be inside a call to libc when it
2729 got a SIGUSR1 from exiwhat.
2731 The SIGUSR1 handler appends the current process status to the process
2732 log which is later printed by exiwhat. It used to use the general
2733 purpose logging code to do this, but several functions it calls are
2734 not safe for signals.
2736 The new output code in the SIGUSR1 handler is specific to the process
2737 log, and simple enough that it's easy to inspect for signal safety.
2738 Removing some special cases also simplifies the general logging code.
2739 Removing the spurious timestamps from the process log simplifies
2742 TF/04 Improved ratelimit ACL condition.
2744 The /noupdate option has been deprecated in favour of /readonly which
2745 has clearer semantics. The /leaky, /strict, and /readonly update modes
2746 are mutually exclusive. The update mode is no longer included in the
2747 database key; it just determines when the database is updated. (This
2748 means that when you upgrade Exim will forget old rate measurements.)
2750 Exim now checks that the per_* options are used with an update mode that
2751 makes sense for the current ACL. For example, when Exim is processing a
2752 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2753 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2754 must specify per_mail/readonly. If you omit the update mode it defaults to
2755 /leaky where that makes sense (as before) or /readonly where required.
2757 The /noupdate option is now undocumented but still supported for
2758 backwards compatibility. It is equivalent to /readonly except that in
2759 ACLs where /readonly is required you may specify /leaky/noupdate or
2760 /strict/noupdate which are treated the same as /readonly.
2762 A useful new feature is the /count= option. This is a generalization
2763 of the per_byte option, so that you can measure the throughput of other
2764 aggregate values. For example, the per_byte option is now equivalent
2765 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2767 The per_rcpt option has been generalized using the /count= mechanism
2768 (though it's more complicated than the per_byte equivalence). When it is
2769 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2770 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2771 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2772 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2773 Note that using per_rcpt with a non-readonly update mode in more than
2774 one ACL will cause the recipients to be double-counted. (The per_mail
2775 and per_byte options don't have this problem.)
2777 The handling of very low rates has changed slightly. If the computed rate
2778 is less than the event's count (usually one) then this event is the first
2779 after a long gap. In this case the rate is set to the same as this event's
2780 count, so that the first message of a spam run is counted properly.
2782 The major new feature is a mechanism for counting the rate of unique
2783 events. The new per_addr option counts the number of different
2784 recipients that someone has sent messages to in the last time period. It
2785 behaves like per_rcpt if all the recipient addresses are different, but
2786 duplicate recipient addresses do not increase the measured rate. Like
2787 the /count= option this is a general mechanism, so the per_addr option
2788 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2789 example, measure the rate that a client uses different sender addresses
2790 with the options per_mail/unique=$sender_address. There are further
2791 details in the main documentation.
2793 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2795 TF/06 Removed a few PCRE remnants.
2797 TF/07 Automatically extract Exim's version number from tags in the git
2798 repository when doing development or release builds.
2800 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2801 Bugzilla 879. Patch from Paul Fisher.
2803 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2804 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2807 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2809 PP/05 Variable $av_failed, true if the AV scanner deferred.
2810 Bugzilla 1078. Patch from John Horne.
2812 PP/06 Stop make process more reliably on build failure.
2813 Bugzilla 1087. Patch from Heiko Schlittermann.
2815 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2816 Bugzilla 1089. Patch from Heiko Schlittermann.
2818 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2819 Bugzilla 1131. Patch from Holger Weiß.
2821 PP/09 Handle IPv6 addresses with SPF.
2822 Bugzilla 860. Patch from Wolfgang Breyha.
2824 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2826 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2829 PP/11 match_* no longer expand right-hand-side by default.
2830 New compile-time build option, EXPAND_LISTMATCH_RHS.
2831 New expansion conditions, "inlist", "inlisti".
2833 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2835 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2837 PP/14 fix log_write() format string regression from TF/03.
2838 Bugzilla 1152. Patch from Dmitry Isaikin.
2844 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2846 PP/02 Harmonised TLS library version reporting; only show if debugging.
2847 Layout now matches that introduced for other libraries in 4.74 PP/03.
2849 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2851 PP/04 New "dns_use_edns0" global option.
2853 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2856 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2857 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2859 TK/01 Updated PolarSSL code to 0.14.2.
2860 Bugzilla 1097. Patch from Andreas Metzler.
2862 PP/07 Catch divide-by-zero in ${eval:...}.
2863 Fixes bugzilla 1102.
2865 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2868 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2869 format-string attack -- SECURITY: remote arbitrary code execution.
2871 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2872 time unintentionally subject to list matching rules, letting the header
2873 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2874 arbitrary string expansion). This allowed for information disclosure.
2876 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2877 INT_MIN/-1 -- value coerced to INT_MAX.
2883 NM/01 Workaround for PCRE version dependency in version reporting
2886 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2887 This fixes portability to compilers other than gcc, notably
2888 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2890 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2891 makefiles for portability to HP-UX and POSIX correctness.
2893 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2894 Also via indented variable definition in the Makefile.
2895 (Debugging by Oliver Heesakkers).
2897 PP/02 Restore caching of spamd results with expanded spamd_address.
2898 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2900 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2901 Improves build reliability. Fix from: Frank Elsner
2903 NM/02 Fix wide character breakage in the rfc2047 coding
2904 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2906 NM/03 Allow underscore in dnslist lookups
2907 Fixes bug 1026. Patch from Graeme Fowler
2909 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2910 Code patches from Adam Ciarcinski of NetBSD.
2912 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2915 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2916 is logged, to avoid truncation. Patch from John Horne.
2918 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2919 Patch from Jakob Hirsch.
2921 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2922 SQL string expansion failure details.
2923 Patch from Andrey Oktyabrski.
2925 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2926 Patch from Simon Arlott.
2928 PP/09 New lookups functionality failed to compile on old gcc which rejects
2929 extern declarations in function scope.
2930 Patch from Oliver Fleischmann
2932 PP/10 Use sig_atomic_t for flags set from signal handlers.
2933 Check getgroups() return and improve debugging.
2934 Fixed developed for diagnosis in bug 927 (which turned out to be
2937 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2938 Patch from Mark Zealey.
2940 PP/12 Bugzilla 1056: Improved spamd server selection.
2941 Patch from Mark Zealey.
2943 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2944 Based on patch from Heiko Schlittermann.
2946 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2947 Patch from Uwe Doering, sign-off by Michael Haardt.
2949 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2950 variable declaration deep within a block. Bug and patch from
2953 PP/15 lookups-Makefile IRIX compatibility coercion.
2955 PP/16 Make DISABLE_DKIM build knob functional.
2957 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2958 Patch by Simon Arlott
2960 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2961 variable argument macros. Our copy now differs from upstream.
2967 TF/01 Failure to get a lock on a hints database can have serious
2968 consequences so log it to the panic log.
2970 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2971 controlled using the smtp_confirmation log selector.
2973 TF/03 Include the error message when we fail to unlink a spool file.
2975 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2976 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2977 for maintaining out-of-tree patches for some time.
2979 PP/01 Bugzilla 139: Documentation and portability issues.
2980 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2981 Handle per-OS dynamic-module compilation flags.
2983 PP/02 Let /dev/null have normal permissions.
2984 The 4.73 fixes were a little too stringent and complained about the
2985 permissions on /dev/null. Exempt it from some checks.
2986 Reported by Andreas M. Kirchwitz.
2988 PP/03 Report version information for many libraries, including
2989 Exim version information for dynamically loaded libraries. Created
2990 version.h, now support a version extension string for distributors
2991 who patch heavily. Dynamic module ABI change.
2993 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2994 privilege escalation vulnerability whereby the Exim run-time user
2995 can cause root to append content of the attacker's choosing to
2998 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3001 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3002 If dropping privileges for untrusted macros, we disabled normal logging
3003 on the basis that it would fail; for the Exim run-time user, this is not
3004 the case, and it resulted in successful deliveries going unlogged.
3005 Fixed. Reported by Andreas Metzler.
3011 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3012 only prepend for the Resent-* case. Fixes regression introduced in
3013 Exim 4.70 by NM/22 for Bugzilla 607.
3015 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3016 increasing numbers of administrators be bitten by this.
3018 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3020 PP/03 Bugzilla 994: added openssl_options main configuration option.
3022 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3024 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3026 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3028 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3029 without a peer certificate, leading to a segfault because of an
3030 assumption that peers always have certificates. Be a little more
3031 paranoid. Problem reported by Martin Tscholak.
3033 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3034 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3035 NB: ClamAV planning to remove STREAM in "middle of 2010".
3036 CL also introduces -bmalware, various -d+acl logging additions and
3037 more caution in buffer sizes.
3039 PP/09 Implemented reverse_ip expansion operator.
3041 PP/10 Bugzilla 937: provide a "debug" ACL control.
3043 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3045 PP/12 Bugzilla 973: Implement --version.
3047 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3049 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3051 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3053 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3054 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3055 ignore trailing whitespace.
3057 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3059 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3062 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3063 Notification from Dr Andrew Aitchison.
3065 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3066 ExtendedDetectionInfo response format.
3067 Notification from John Horne.
3069 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3072 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3073 XSL and documented dependency on system catalogs, with examples of how
3076 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3079 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3080 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3081 configuration file which is writeable by the Exim user or group.
3083 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3084 of configuration files to cover files specified with the -C option if
3085 they are going to be used with root privileges, not just the default
3088 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3089 option (effectively making it always true).
3091 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3092 files to be used while preserving root privileges.
3094 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3095 that rogue child processes cannot use them.
3097 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3098 run-time user, instead of root.
3100 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3101 Exim run-time user without dropping privileges.
3103 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3104 result string, instead of calling string_vformat() twice with the same
3107 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3108 for other users. Others should always drop root privileges if they use
3109 -C on the command line, even for a whitelisted configure file.
3111 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3113 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3119 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3120 $data_path, and $header_path variables; fixed documentation bugs and
3123 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3124 exipick to access non-standard spools, including the "frozen" queue
3127 NM/01 Bugzilla 965: Support mysql stored procedures.
3128 Patch from Alain Williams
3130 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3132 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3133 Patch from Andreas Metzler
3135 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3136 Patch from Kirill Miazine
3138 NM/05 Bugzilla 671: Added umask to procmail example.
3140 JJ/03 installed exipick 20100323.0, fixing doc bug
3142 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3143 directory. Notification and patch from Dan Rosenberg.
3145 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3147 TK/02 Improve log output when DKIM signing operation fails.
3149 MH/01 Treat the transport option dkim_domain as a colon separated
3150 list, not as a single string, and sign the message with each element,
3151 omitting multiple occurences of the same signer.
3153 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3154 Bugzilla 985, 986. Patch by Simon Arlott
3156 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3157 Patch by Simon Arlott
3159 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3160 MBX locking. Notification from Dan Rosenberg.
3166 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3168 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3170 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3172 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3174 NM/04 Bugzilla 915: Flag broken perl installation during build.
3180 TK/01 Added patch by Johannes Berg that expands the main option
3181 "spamd_address" if it starts with a dollar sign.
3183 TK/02 Write list of recipients to X-Envelope-Sender header when building
3184 the mbox-format spool file for content scanning (suggested by Jakob
3187 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3188 (http://www.dcc-servers.net/) support via dccifd. Activated by
3189 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3191 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3192 by Mark Daniel Reidel <mr@df.eu>.
3194 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3195 When building exim an external PCRE library is now needed -
3196 PCRE is a system library on the majority of modern systems.
3197 See entry on PCRE_LIBS in EDITME file.
3199 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3200 conversation. Added nologin parameter to request.
3201 Patch contributed by Kirill Miazine.
3203 TF/01 Do not log submission mode rewrites if they do not change the address.
3205 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3207 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3208 log files in place. Contributed by Roberto Lima.
3210 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3212 TF/03 Bugzilla 615: When checking the local_parts router precondition
3213 after a local_part_suffix or local_part_prefix option, Exim now
3214 does not use the address's named list lookup cache, since this
3215 contains cached lookups for the whole local part.
3217 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3218 Robert Millan. Documentation is in experimental-spec.txt.
3220 TF/04 Bugzilla 668: Fix parallel build (make -j).
3222 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3224 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3225 Patch provided by Jan Srzednicki.
3227 TF/05 Leading white space used to be stripped from $spam_report which
3228 wrecked the formatting. Now it is preserved.
3230 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3231 that they are available at delivery time.
3233 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3235 TF/08 TLS error reporting now respects the incoming_interface and
3236 incoming_port log selectors.
3238 TF/09 Produce a more useful error message if an SMTP transport's hosts
3239 setting expands to an empty string.
3241 NM/06 Bugzilla 744: EXPN did not work under TLS.
3242 Patch provided by Phil Pennock.
3244 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3245 Patch provided by Richard Godbee.
3247 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3248 acl_smtp_notquit, added index entry.
3250 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3251 Patch provided by Eugene Bujak.
3253 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3254 accept(). Patch provided by Maxim Dounin.
3256 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3257 Patch provided by Phil Pennock.
3259 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3261 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3262 Patch provided by Brad "anomie" Jorsch.
3264 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3265 Patch provided by Dean Brooks.
3267 TK/05 Add native DKIM support (does not depend on external libraries).
3269 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3270 Patch provided by Graeme Fowler.
3272 NM/16 Bugzilla 851: Documentation example syntax fix.
3274 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3276 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3279 NM/19 Bugzilla 745: TLS version reporting.
3280 Patch provided by Phil Pennock.
3282 NM/20 Bugzilla 167: bool: condition support.
3283 Patch provided by Phil Pennock.
3285 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3286 clients. Patch provided by Phil Pennock.
3288 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3289 Patch provided by Brad "anomie" Jorsch.
3291 NM/23 Bugzilla 687: Fix misparses in eximstats.
3292 Patch provided by Heiko Schlittermann.
3294 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3295 Patch provided by Heiko Schlittermann.
3297 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3298 plus update to original patch.
3300 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3302 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3303 Patch provided by David Brownlee.
3305 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3307 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3309 NM/30 Bugzilla 888: TLS documentation bugfixes.
3311 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3313 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3314 Unlike the original bugzilla I have changed all shell scripts in src tree.
3316 NM/33 Bugzilla 898: Transport filter timeout fix.
3317 Patch by Todd Rinaldo.
3319 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3320 Patch by Serge Demonchaux.
3322 NM/35 Bugzilla 39: Base64 decode bug fixes.
3323 Patch by Jakob Hirsch.
3325 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3327 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3329 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3331 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3337 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3338 ALT-N's libdkim that I have put here:
3339 http://duncanthrax.net/exim-experimental/
3341 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3342 were called 'true' and it seems that C99 defines that as a reserved
3343 keyword to be used with 'bool' variable types. That means you could
3344 not include C99-style headers which use bools without triggering
3345 build errors in sieve.c.
3347 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3348 as mailq or other aliases. Changed the --help handling significantly
3349 to do whats expected. exim_usage() emits usage/help information.
3351 SC/01 Added the -bylocaldomain option to eximstats.
3353 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3355 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3357 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3363 PH/01 Another patch from the Sieve maintainer.
3365 PH/02 When an IPv6 address is converted to a string for single-key lookup
3366 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3367 dots are used instead of colons so that keys in lsearch files need not
3368 contain colons. This was done some time before quoting was made available
3369 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3370 (notated using the quote facility) so as to distinguish them from IPv4
3371 keys. This meant that lookups for IP addresses in host lists did not work
3372 for iplsearch lookups.
3374 This has been fixed by arranging for IPv6 addresses to be expressed with
3375 colons if the lookup type is iplsearch. This is not incompatible, because
3376 previously such lookups could never work.
3378 The situation is now rather anomalous, since one *can* have colons in
3379 ordinary lsearch keys. However, making the change in all cases is
3380 incompatible and would probably break a number of configurations.
3382 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3385 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3386 conversion specification without a maximum field width, thereby enabling
3387 a rogue spamd server to cause a buffer overflow. While nobody in their
3388 right mind would setup Exim to query an untrusted spamd server, an
3389 attacker that gains access to a server running spamd could potentially
3390 exploit this vulnerability to run arbitrary code as the Exim user.
3392 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3393 $primary_hostname instead of what libspf2 thinks the hosts name is.
3395 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3396 a directory entry by the name of the lookup key. Previously, if a
3397 symlink pointed to a non-existing file or a file in a directory that
3398 Exim lacked permissions to read, a lookup for a key matching that
3399 symlink would fail. Now it is enough that a matching directory entry
3400 exists, symlink or not. (Bugzilla 503.)
3402 PH/03 The body_linecount and body_zerocount variables are now exported in the
3405 PH/04 Added the $dnslist_matched variable.
3407 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3408 This means they are set thereafter only if the connection becomes
3411 PH/06 Added the client_condition to authenticators so that some can be skipped
3412 by clients under certain conditions.
3414 PH/07 The error message for a badly-placed control=no_multiline_responses left
3415 "_responses" off the end of the name.
3417 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3419 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3420 (without spaces) instead of just copying the configuration text.
3422 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3424 PH/11 Added $max_received_linelength.
3426 PH/12 Added +ignore_defer and +include_defer to host lists.
3428 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3429 way in which PCRE > 7.0 is built.
3431 PH/14 Implemented queue_only_load_latch.
3433 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3434 MAIL command. The effect was to mangle the value on 64-bit systems.
3436 PH/16 Another patch from the Sieve maintainer.
3438 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3440 PH/18 If a system quota error occurred while trying to create the file for
3441 a maildir delivery, the message "Mailbox is full" was not appended to the
3442 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3443 only to a quota excession during the actual writing of the file.
3445 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3446 characters?) which causes problems in log lines. The DN values are now
3447 passed through string_printing() before being added to log lines.
3449 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3450 and InterBase are left for another time.)
3452 PH/21 Added message_body_newlines option.
3454 PH/22 Guard against possible overflow in moan_check_errorcopy().
3456 PH/23 POSIX allows open() to be a macro; guard against that.
3458 PH/24 If the recipient of an error message contained an @ in the local part
3459 (suitably quoted, of course), incorrect values were put in $domain and
3460 $local_part during the evaluation of errors_copy.
3466 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3467 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3470 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3471 issue a MAIL command.
3473 PH/02 In an ACL statement such as
3475 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3477 if a client was not listed at all, or was listed with a value other than
3478 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3479 the condition was not true (as it should be), so access was not denied.
3480 The bug was that the ! inversion was incorrectly passed on to the second
3481 item. This has been fixed.
3483 PH/03 Added additional dnslists conditions == and =& which are different from
3484 = and & when the dns lookup returns more than one IP address.
3486 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3487 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3489 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3490 FSYNC, which compiles an option called disable_fsync that allows for
3491 bypassing fsync(). The documentation is heavily laced with warnings.
3493 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3495 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3496 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3497 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3498 including adding "make clean"; (3) Added -fPIC when compiling the test
3499 dynamically loaded module, to get rid of a warning.
3501 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3502 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3503 = 0s. The bug is otherwise harmless.
3505 PH/07 There was a bug in the dovecot authenticator such that the value of
3506 $auth1 could be overwritten, and so not correctly preserved, after a
3507 successful authentication. This usually meant that the value preserved by
3508 the server_setid option was incorrect.
3510 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3512 PH/09 Installed PCRE release 7.0.
3514 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3515 run for batched SMTP input. It is now run at the start of every message
3516 in the batch. While fixing this I discovered that the process information
3517 (output by running exiwhat) was not always getting set for -bs and -bS
3518 input. This is fixed, and it now also says "batched" for BSMTP.
3520 PH/11 Added control=no_pipelining.
3522 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3523 patch, slightly modified), and move the expansion of helo_data till after
3524 the connection is made in the smtp transport (so it can use these
3527 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3529 PH/14 Added log_selector = +pid.
3531 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3533 PH/16 Add ${if forany and ${if forall.
3535 PH/17 Added dsn_from option to vary the From: line in DSNs.
3537 PH/18 Flush SMTP output before performing a callout, unless control =
3538 no_callout_flush is set.
3540 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3541 was true (the default) a successful delivery failed to delete the retry
3542 item, thus causing premature timeout of the address. The bug is now
3545 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3547 PH/21 Long custom messages for fakedefer and fakereject are now split up
3548 into multiline responses in the same way that messages for "deny" and
3549 other ACL rejections are.
3551 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3552 with slight modification.
3554 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3555 draft, changing the syntax and factoring some duplicate code".
3557 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3558 for deliveries of the second and subsequent messages over the same SMTP
3561 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3562 ${reduce, with only minor "tidies".
3564 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3566 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3567 expansion side effects.
3569 PH/27 When a message times out after an over-quota error from an Exim-imposed
3570 quota, the bounce message says "mailbox is full". This message was not
3571 being given when it was a system quota that was exceeded. It now should
3574 MH/03 Made $recipients available in local_scan(). local_scan() already has
3575 better access to the recipient list through recipients_list[], but
3576 $recipients can be useful in postmaster-provided expansion strings.
3578 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3579 in the case of a MAIL command with additional options following the
3580 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3581 were accidentally chopped off.
3583 PH/29 SMTP synchronization checks are implemented when a command is read -
3584 there is a check that no more input is waiting when there shouldn't be
3585 any. However, for some commands, a delay in an ACL can mean that it is
3586 some time before the response is written. In this time, more input might
3587 arrive, invalidly. So now there are extra checks after an ACL has run for
3588 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3589 pipelining has not been advertised.
3591 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3593 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3594 returned in the SMTP response when smtp_return_error_details was false.
3595 This has been fixed.
3597 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3598 instead of the C I/O that was originally supplied, because problems were
3599 reported on Solaris.
3601 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3602 Exim which did not show up earlier: it was assuming that a call to
3603 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3604 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3605 was a macro that became an assignment, so it seemed to work. This has
3606 changed to a proper function call with a void return, hence the compile
3607 error. Exim's code has been fixed.
3609 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3612 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3614 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3616 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3617 because it turns out to be too big or there is a timeout) there is no
3618 "Completed" line in the log. When some messages of this type were
3619 selected by exigrep, they were listed as "not completed". Others were
3620 picked up by some special patterns. I have improved the selection
3621 criteria to be more general.
3623 PH/38 The host_find_failed option in the manualroute router can now be set
3624 to "ignore", to completely ignore a host whose IP address cannot be
3625 found. If all hosts are ignored, the behaviour is controlled by the new
3626 host_all_ignored option.
3628 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3629 homing or because of multiple MX records with /mx) generated more than
3630 one IP address, and the following item turned out to be the local host,
3631 all the secondary addresses of the first item were incorrectly removed
3632 from the list, along with the local host and any following hosts (which
3633 is what is supposed to happen).
3635 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3636 whoever called Exim into the -H file. In the case of the daemon it was
3637 behaving confusingly. When first started, it used values for whoever
3638 started the daemon, but after a SIGHUP it used the Exim user (because it
3639 calls itself on a restart). I have changed the code so that it now always
3642 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3643 message are rejected with the same error (e.g. no authentication or bad
3644 sender address), and a DATA command is nevertheless sent (as can happen
3645 with PIPELINING or a stupid MUA), the error message that was given to the
3646 RCPT commands is included in the rejection of the DATA command. This is
3647 intended to be helpful for MUAs that show only the final error to their
3650 PH/42 Another patch from the Sieve maintainer.
3652 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3653 Eximstats - Fixed some broken HTML links and added missing column headers
3655 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3656 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3658 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3660 PH/43 Yet another patch from the Sieve maintainer.
3662 PH/44 I found a way to check for a TCP/IP connection going away before sending
3663 the response to the final '.' that terminates a message, but only in the
3664 case where the client has not sent further data following the '.'
3665 (unfortunately, this is allowed). However, in many cases there won't be
3666 any further data because there won't be any more messages to send. A call
3667 to select() can be used: if it shows that the input is "ready", there is
3668 either input waiting, or the socket has been closed. An attempt to read
3669 the next input character can distinguish the two cases. Previously, Exim
3670 would have sent an OK response which the client would never have see.
3671 This could lead to message repetition. This fix should cure that, at
3672 least in a lot of common cases.
3674 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3675 advertised in response to EHLO.
3681 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3682 fixed by 4.65/MH/01 (is this a record?) are fixed:
3684 (i) An empty string was always treated as zero by the numeric comparison
3685 operators. This behaviour has been restored.
3687 (ii) It is documented that the numeric comparison operators always treat
3688 their arguments as decimal numbers. This was broken in that numbers
3689 starting with 0 were being interpreted as octal.
3691 While fixing these problems I realized that there was another issue that
3692 hadn't been noticed. Values of message_size_limit (both the global option
3693 and the transport option) were treated as octal if they started with 0.
3694 The documentation was vague. These values are now always treated as
3695 decimal, and I will make that clear in the documentation.
3701 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3702 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3705 MH/01 Don't check that the operands of numeric comparison operators are
3706 integers when their expansion is in "skipping" mode (fixes bug
3707 introduced by 4.64-PH/07).
3709 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3710 child addresses, Exim now panics and dies. Previously, because the count
3711 is held in a short int, deliveries were likely to be lost. As such a
3712 large number of recipients for a single message is ridiculous
3713 (performance will be very, very poor), I have chosen to impose a limit
3714 rather than extend the field.
3720 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3721 leftover -K file (the existence of which was triggered by #402).
3722 While we were at it, introduced process PID as part of the -K
3723 filename. This should rule out race conditions when creating
3726 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3727 processing considerably. Previous code took too long for large mails,
3728 triggering a timeout which in turn triggers #401.
3730 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3731 in the DK code in transports.c. sendfile() is not really portable,
3732 hence the _LINUX specificness.
3734 TF/01 In the add_headers option to the mail command in an Exim filter,
3735 there was a bug that Exim would claim a syntax error in any
3736 header after the first one which had an odd number of characters
3739 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3740 callout verification, Exim cached a "reject" for the entire domain. This
3741 is correct for most verifications, but it is not correct for a recipient
3742 verification with use_sender or use_postmaster set, because in that case
3743 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3744 case of MAIL FROM:<> rejection from other early rejections (e.g.
3745 rejection of HELO). When verifying a recipient using a non-null MAIL
3746 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3747 Whatever the result of the callout, the value of the domain cache is
3748 left unchanged (for any other kind of callout, getting as far as trying
3749 RCPT means that the domain itself is ok).
3751 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3754 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3755 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3756 session, the daemon ignores ECONNECT errors and logs others; it now
3757 ignores EPIPE as well.
3759 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3760 (quoted-printable decoding).
3762 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3763 later the small subsequent patch to fix an introduced bug.
3765 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3767 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3769 PH/08 An error is now given if message_size_limit is specified negative.
3771 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3772 to be given (somewhat) arbitrary names.
3774 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3777 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3778 miscellaneous code fixes
3780 PH/10 Added the log_reject_target ACL modifier to specify where to log
3783 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3784 hostname. This is wrong, because it relates to the incoming message (and
3785 probably the interface on which it is arriving) and not to the outgoing
3786 callout (which could be using a different interface). This has been
3787 changed to use the value of the helo_data option from the smtp transport
3788 instead - this is what is used when a message is actually being sent. If
3789 there is no remote transport (possible with a router that sets up host
3790 addresses), $smtp_active_hostname is used.
3792 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3793 tweaks were necessary in order to get it to work (see also 21 below):
3794 (a) The code assumed that strncpy() returns a negative number on buffer
3795 overflow, which isn't the case. Replaced with Exim's string_format()
3797 (b) There were several signed/unsigned issues. I just did the minimum
3798 hacking in of casts. There is scope for a larger refactoring.
3799 (c) The code used strcasecmp() which is not a standard C function.
3800 Replaced with Exim's strcmpic() function.
3801 (d) The code set only $1; it now sets $auth1 as well.
3802 (e) A simple test gave the error "authentication client didn't specify
3803 service in request". It would seem that Dovecot has changed its
3804 interface. Fortunately there's a specification; I followed it and
3805 changed what the client sends and it appears to be working now.
3807 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3810 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3811 address A is aliased to B and C, where B exists and C does not. Without
3812 -v the output is "A verified" because verification stops after a
3813 successful redirection if more than one address is generated. However,
3814 with -v the child addresses are also verified. Exim was outputting "A
3815 failed to verify" and then showing the successful verification for C,
3816 with its parentage. It now outputs "B failed to verify", showing B's
3817 parentage before showing the successful verification of C.
3819 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3820 look up a TXT record in a specific list after matching in a combined
3823 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3824 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3825 they consult the DNS. I had assumed they would set it the way they
3826 wanted; and indeed my experiments on Linux seem to show that in some
3827 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3828 To be on the safe side, however, I have now made the interface to
3829 host_find_byname() similar to host_find_bydns(), with an argument
3830 containing the DNS resolver options. The host_find_byname() function now
3831 sets these options at its start, just as host_find_bydns() does. The smtp
3832 transport options dns_qualify_single and dns_search_parents are passed to
3833 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3834 of host_find_byname() use the default settings of RES_DEFNAMES
3835 (qualify_single) but not RES_DNSRCH (search_parents).
3837 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3838 spool_read_header() do less string testing, by means of a preliminary
3839 switch on the second character of optional "-foo" lines. (This is
3840 overdue, caused by the large number of possibilities that now exist.
3841 Originally there were few.) While I was there, I also converted the
3842 str(n)cmp tests so they don't re-test the leading "-" and the first
3843 character, in the hope this might squeeze out yet more improvement.
3845 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3846 flag allowing group syntax was set by the header_syntax check but not
3847 turned off, possible causing trouble later; (2) The flag was not being
3848 set at all for the header_verify test, causing "group"-style headers to
3849 be rejected. I have now set it in this case, and also caused header_
3850 verify to ignore an empty address taken from a group. While doing this, I
3851 came across some other cases where the code for allowing group syntax
3852 while scanning a header line wasn't quite right (mostly, not resetting
3853 the flag correctly in the right place). These bugs could have caused
3854 trouble for malformed header lines. I hope it is now all correct.
3856 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3857 with the "reply" argument non-NULL. The code, however (which originally
3858 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3859 but it didn't always do it. This confused somebody who was copying the
3860 code for some other use. I have removed all the tests.
3862 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3863 feature that was used to support insecure browsers during the U.S. crypto
3864 embargo. It requires special client support, and Exim is probably the
3865 only MTA that supported it -- and would never use it because real RSA is
3866 always available. This code has been removed, because it had the bad
3867 effect of slowing Exim down by computing (never used) parameters for the
3868 RSA_EXPORT functionality.
3870 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3871 authenticator to fail if there's a tab character in the incoming data
3872 (there should never be unless someone is messing about, as it's supposed
3873 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3874 if the connection is using TLS or if the remote IP is the same as the
3875 local IP, and the "valid-client-cert option" if a client certificate has
3878 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3879 authenticators. This can be used for authorization after authentication
3880 succeeds. (In the case of plaintext, it servers for both authentication
3883 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3884 if any retry times were supplied.
3886 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3887 connection, where there is no client IP address to check. In this
3888 situation, the verify now always succeeds.
3890 PH/25 Applied John Jetmore's -Mset patch.
3892 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3894 PH/27 In a string expansion for a processed (not raw) header when multiple
3895 headers of the same name were present, leading whitespace was being
3896 removed from all of them, but trailing whitespace was being removed only
3897 from the last one. Now trailing whitespace is removed from each header
3898 before concatenation. Completely empty headers in a concatenation (as
3899 before) are ignored.
3901 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3902 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3904 PH/29 [Removed. This was a change that I later backed out, and forgot to
3905 correct the ChangeLog entry (that I had efficiently created) before
3906 committing the later change.]
3908 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3909 address errors (4xx response to RCPT) over the same connection as other
3910 messages routed to the same hosts. Such deliveries are always "forced",
3911 so retry times are not inspected. This resulted in far too many retries
3912 for the affected addresses. The effect occurred only when there were more
3913 hosts than the hosts_max_try setting in the smtp transport when it had
3914 the 4xx errors. Those hosts that it had tried were not added to the list
3915 of hosts for which the message was waiting, so if all were tried, there
3916 was no problem. Two fixes have been applied:
3918 (i) If there are any address or message errors in an SMTP delivery, none
3919 of the hosts (tried or untried) are now added to the list of hosts
3920 for which the message is waiting, so the message should not be a
3921 candidate for sending over the same connection that was used for a
3922 successful delivery of some other message. This seems entirely
3923 reasonable: after all the message is NOT "waiting for some host".
3924 This is so "obvious" that I'm not sure why it wasn't done
3925 previously. Hope I haven't missed anything, but it can't do any
3926 harm, as the worst effect is to miss an optimization.
3928 (ii) If, despite (i), such a delivery is accidentally attempted, the
3929 routing retry time is respected, so at least it doesn't keep
3930 hammering the server.
3932 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3933 in ${readsocket because some servers need this prod.
3935 PH/32 Added some extra debug output when updating a wait-xxx database.
3937 PH/33 The hint "could be header name not terminated by colon", which has been
3938 given for certain expansion errors for a long time, was not being given
3939 for the ${if def:h_colon_omitted{... case.
3941 PH/34 The spec says: "With one important exception, whenever a domain list is
3942 being scanned, $domain contains the subject domain." There was at least
3943 one case where this was not true.
3945 PH/35 The error "getsockname() failed: connection reset by peer" was being
3946 written to the panic log as well as the main log, but it isn't really
3947 panic-worthy as it just means the connection died rather early on. I have
3948 removed the panic log writing for the ECONNRESET error when getsockname()
3951 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3952 runs only) independently of the message's sender address. This meant
3953 that, if the 4xx error was in fact related to the sender, a different
3954 message to the same recipient with a different sender could confuse
3955 things. In particular, this can happen when sending to a greylisting
3956 server, but other circumstances could also provoke similar problems.
3957 I have changed the default so that the retry time for these errors is now
3958 based a combination of the sender and recipient addresses. This change
3959 can be overridden by setting address_retry_include_sender=false in the
3962 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3963 remote server are returned as part of bounce messages. This was not
3964 happening for LMTP over a pipe (the lmtp transport), but now it is the
3965 same for both kinds of LMTP.
3967 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3968 in header lines that were in fact CNAMEs. This is no longer the case.
3970 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3971 and queue runs started by the daemon processed all messages. This has
3972 been fixed so that -R and -S can now usefully be given with -q<time>.
3974 PH/40 Import PCRE release 6.7 (fixes some bugs).
3976 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3978 PH/42 Give an error if -q is specified more than once.
3980 PH/43 Renamed the variables $interface_address and $interface_port as
3981 $received_ip_address and $received_port, to make it clear that these
3982 values apply to message reception, and not to the outgoing interface when
3983 a message is delivered. (The old names remain recognized, of course.)
3985 PH/44 There was no timeout on the connect() call when using a Unix domain
3986 socket in the ${readsocket expansion. There now is.
3988 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3989 be meaningful with "accept".
3991 SC/01 Eximstats V1.43
3992 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3994 SC/02 Eximstats V1.44
3995 Use a glob alias rather than an array ref in the generated
3996 parser. This improves both readability and performance.
3998 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3999 Collect SpamAssassin and rejection statistics.
4000 Don't display local sender or destination tables unless
4001 there is data to show.
4002 Added average volumes into the top table text output.
4004 SC/04 Eximstats V1.46
4005 Collect data on the number of addresses (recipients)
4006 as well as the number of messages.
4008 SC/05 Eximstats V1.47
4009 Added 'Message too big' to the list of mail rejection
4010 reasons (thanks to Marco Gaiarin).
4012 SC/06 Eximstats V1.48
4013 Mainlog lines which have GMT offsets and are too short to
4014 have a flag are now skipped.
4016 SC/07 Eximstats V1.49 (Alain Williams)
4017 Added the -emptyok flag.
4019 SC/08 Eximstats V1.50
4020 Fixes for obtaining the IP address from reject messages.
4022 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4023 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4024 whitespace changes from 4.64-PH/27
4026 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4029 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4030 are found, allow negative numbers in numeric criteria)
4032 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4034 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4035 to match changes made in 4.64-PH/43
4037 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4039 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4040 "domain" is actually an IP address, and force a failure. This locks out
4041 those revolvers/nameservers that support "A-for-A" lookups, in
4042 contravention of the specifications.
4044 PH/48 When a host name was looked up from an IP address, and the subsequent
4045 forward lookup of the name timed out, the host name was left in
4046 $sender_host_name, contrary to the specification.
4048 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4049 restricted to single-key lookups, Exim was not diagnosing an error if
4050 * or *@ was used with a query-style lookup.
4052 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4054 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4055 long ago, but noone interested enough thought of it. Let's just say that
4056 the "1.1" means that there are some new functions that weren't there at
4057 some point in the past.
4059 PH/51 Error processing for expansion failure of helo_data from an smtp
4060 transport during callout processing was broken.
4062 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4063 tested/used via the -bh/-bhc/-bs options.
4065 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4066 bug, fixed in subsequent PCRE releases).
4068 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4069 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4071 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4077 SC/01 Use a glob alias rather than an array ref in eximstats generated
4078 parser. This improves both readability and performance.
4080 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4081 Don't display local sender or destination tables in eximstats unless
4082 there is data to show.
4083 Added average volumes into the eximstats top table text output.
4085 SC/03 Collect data on the number of addresses (recipients) as well
4086 as the number of messages in eximstats.
4088 TF/01 Correct an error in the documentation for the redirect router. Exim
4089 does (usually) call initgroups() when daemonizing.
4091 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4092 with consistent privilege compared to when running as a daemon.
4094 TF/03 Note in the spec that $authenticated_id is not set for local
4095 submissions from trusted users.
4097 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4098 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4100 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4101 by adding some example configuration directives to the default
4102 configuration file. A little bit of work is required to uncomment the
4103 directives and define how usernames and passwords are checked, but
4104 there is now a framework to start from.
4106 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4107 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4108 without this. I don't know how relevant this is to other LDAP libraries.
4110 PH/02 Add the verb name to the "unknown ACL verb" error.
4112 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4114 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4116 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4117 directory not expanded when it should be if an expanded home directory
4118 was set for the address (which is overridden by the transport).
4120 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4123 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4124 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4125 because it is too late at that time, and has no effect.
4127 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4128 security issue with \' (bugzilla #107). I could not use the
4129 PQescapeStringConn() function, because it needs a PGconn value as one of
4132 PH/08 When testing addresses using -bt, indicate those final addresses that
4133 are duplicates that would not cause an additional delivery. At least one
4134 person was confused, thinking that -bt output corresponded to deliveries.
4135 (Suppressing duplicates isn't a good idea as you lose the information
4136 about possibly different redirections that led to the duplicates.)
4138 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4139 systems where poll() doesn't work, in particular OS X.
4141 PH/10 Added more information to debugging output for retry time not reached.
4143 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4144 operations in malware.c.
4146 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4149 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4150 syslog_duplication set false, log lines that would normally be written
4151 both the the main log and to the reject log were not written to syslog at
4154 PH/14 In the default configuration, change the use of "message" in ACL warn
4155 statements to "add_header".
4157 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4158 not followed by a command (e.g. "seen endif").
4160 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4161 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4164 PH/17 Added extra conditions to the default value of delay_warning_condition
4168 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4169 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4170 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4173 The Auto-Submitted: and various List- headers are standardised, whereas I
4174 don't think Precedence: ever was.
4176 PH/18 Refactored debugging code in route_finduser() to show more information,
4177 in particular, the error code if getpwnam() issues one.
4179 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4180 This is apparently needed in addition to the PH/07 change above to avoid
4181 any possible encoding problems.
4183 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4184 but not after initializing Perl.
4186 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4187 output them only if debugging. By default they are written stderr,
4188 apparently, which is not desirable.
4190 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4193 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4196 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4198 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4199 authenticated or an ident call has been made. Suppress the default
4200 values for $authenticated_id and $authenticated_sender (but permit -oMai
4201 and -oMas) when testing with -bh.
4203 PH/24 Re-jigged the order of the tests in the default configuration so that the
4204 tests for valid domains and recipients precede the DNS black list and CSA
4205 tests, on the grounds that those ones are more expensive.
4207 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4208 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4209 This bug exists in every version of Exim that I still have, right back to
4212 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4213 However, an attempt to turn on case-sensitivity in a regex key by
4214 including (?-i) didn't work because the subject string was already
4215 lowercased, and the effects were non-intuitive. It turns out that a
4216 one-line patch can be used to allow (?-i) to work as expected.
4222 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4223 other effects) broke the use of negated acl sub-conditions.
4225 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4228 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4229 "Deny" causes Exim to reject the incoming connection with a 554 error.
4230 Unfortunately, if there is a major crisis, such as a disk failure,
4231 tcp-wrappers gives "deny", whereas what one would like would be some
4232 kind of temporary error. A kludge has been added to help with this.
4233 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4234 554 error is used if errno is still zero or contains ENOENT (which occurs
4235 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4238 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4240 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4241 errors. Otherwise a message that provokes a temporary error (when other
4242 messages do not) can cause a whole host to time out.
4244 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4245 addresses were routed directly to files or pipes from a redirect router.
4246 File deliveries just didn't batch; pipe deliveries might have suffered
4249 PH/06 A failure to get a lock for a hints database would erroneously always say
4250 "Failed to get write lock", even when it was really a read lock.
4252 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4253 of 0600. This has been changed to use the value of the lockfile_mode
4254 option (which defaults to 0600).
4256 PH/08 Applied small patch from the Sieve maintainer.
4258 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4259 folder from quota calculations, a direct delivery into this folder messed
4260 up the contents of the maildirsize file. This was because the regex was
4261 used only to exclude .Trash (or whatever) when the size of the mailbox
4262 was calculated. There was no check that a delivery was happening into an
4263 excluded directory. This bug has been fixed by ignoring all quota
4264 processing for deliveries into excluded directories.
4266 PH/10 Added the maildirfolder_create_regex option to appendfile.
4272 PH/01 The code for finding all the local interface addresses on a FreeBSD
4273 system running IPv6 was broken. This may well have applied to all BSD
4274 systems, as well as to others that have similar system calls. The broken
4275 code found IPv4 interfaces correctly, but gave incorrect values for the
4276 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4277 that it would not match correctly against @[] and not recognize the IPv6
4280 PH/02 The ipliteral router was not recognizing addresses of the form user@
4281 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4283 PH/03 Added disable_ipv6.
4285 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4286 decoded form, because it is most often used to construct To: headers
4287 lines in autoreplies, and the decoded form may well be syntactically
4288 invalid. However, $reply_address has leading white space removed, and all
4289 newlines turned into spaces so that the autoreply transport does not
4292 PH/05 If group was specified without a user on a router, and no group or user
4293 was specified on a transport, the group from the router was ignored.
4295 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4296 for visible compile-time settings that can be used to change these
4297 numbers, for those that want even more. Backwards compatibility with old
4298 spool files has been maintained. However, going back to a previous Exim
4299 release will lost any variables that are in spool files.
4301 PH/07 Two small changes when running in the test harness: increase delay when
4302 passing a TCP/IP connection to a new process, in case the original
4303 process has to generate a bounce, and remove special handling of
4304 127.0.0.2 (sic), which is no longer necessary.
4306 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4307 be the same on different OS.
4309 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4312 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4313 whether --show-vars was specified or not
4315 JJ/02 exipick: Added support for new ACL variable spool format introduced
4318 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4319 syntactically invalid From: or Reply-to: line, and a filter used this to
4320 generate an autoreply, and therefore failed to obtain an address for the
4321 autoreply, Exim could try to deliver to a non-existent relative file
4322 name, causing unrelated and misleading errors. What now happens is that
4323 it logs this as a hard delivery error, but does not attempt to create a
4326 PH/11 The exinext utility has a -C option for testing purposes, but although
4327 the given file was scanned by exinext itself; it wasn't being passed on
4328 when Exim was called.
4330 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4331 an end-of-file indication when reading a command response.
4333 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4334 compiled. In many other places in Exim, IPv6 addresses are always
4335 recognized, so I have changed this. It also means that IPv4 domain
4336 literals of the form [IPV4:n.n.n.n] are now always recognized.
4338 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4339 used if the router is not running as root, for example, when verifying at
4340 ACL time, or when using -bh. The debugging output from this situation was
4341 non-existent - all you got was a failure to exec. I have made two
4344 (a) Failures to set uid/gid, the current directory, or a process leader
4345 in a subprocess such as that created by queryprogram now generate
4346 suitable debugging output when -d is set.
4348 (b) The queryprogram router detects when it is not running as root,
4349 outputs suitable debugging information if -d is set, and then runs
4350 the subprocess without attempting to change uid/gid.
4352 PH/15 Minor change to Makefile for building test_host (undocumented testing
4355 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4356 additional section of a DNS packet that returns MX or SRV records.
4357 Instead, it always explicitly searches for A/AAAA records. This avoids
4358 major problems that occur when a DNS server includes only records of one
4359 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4360 fixed another bug: if SRV records were looked up and the corresponding
4361 address records were *not* found in the additional section, the port
4362 values from the SRV records were lost.
4364 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4365 using the correct key (the original address) when searching the retry
4366 rules in order to find which one to use for generating the retry hint.
4368 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4369 adding the default one. Similarly, if it contains a Reply-To: header, the
4370 errors_reply_to option, if set, is not used.
4372 PH/19 When calculating a retry time, Exim used to measure the "time since
4373 failure" by looking at the "first failed" field in the retry record. Now
4374 it does not use this if it is later than than the arrival time of the
4375 message. Instead it uses the arrival time. This makes for better
4376 behaviour in cases where some deliveries succeed, thus re-setting the
4377 "first failed" field. An example is a quota failure for a huge message
4378 when small messages continue to be delivered. Without this change, the
4379 "time since failure" will always be short, possible causing more frequent
4380 delivery attempts for the huge message than are intended.
4381 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4383 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4384 $1, $2, $3) because the numerical variables can be reset during some
4385 expansion items (e.g. "match"), thereby losing the authentication data.
4387 PH/21 Make -bV show the size of off_t variables so that the test suite can
4388 decide whether to run tests for quotas > 2G.
4390 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4391 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4392 filecount value is greater than 2G or if a quota value is greater than 2G
4393 on a system where the size of off_t is not greater than 4, a panic error
4396 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4397 never match. The debug and -bh output now contains an explicit error
4398 message indicating a malformed IPv4 address or mask.
4400 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4401 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4402 PH/23 above applies.
4404 PH/25 Do not write to syslog when running in the test harness. The only
4405 occasion when this arises is a failure to open the main or panic logs
4406 (for which there is an explicit test).
4408 PH/26 Added the /no_tell option to "control=freeze".
4410 PH/27 If a host name lookup failed very early in a connection, for example, if
4411 the IP address matched host_lookup and the reverse lookup yielded a name
4412 that did not have a forward lookup, an error message of the form "no IP
4413 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4414 could be logged. Now it outputs the IP address instead of "NULL".
4416 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4417 allows the sender and the authenticated sender to be set when
4418 submitting a message from within Exim. Since child_open_exim() is
4419 documented for local_scan(), the new function should be too.
4421 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4422 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4423 results in an empty string is now treated as unset.
4425 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4427 PH/31 Added sender_verify_fail logging option.
4429 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4430 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4431 tidied the source and removed it altogether.
4433 PH/33 When a queue run was abandoned because the load average was too high, a
4434 log line was always written; now it is written only if the queue_run log
4435 selector is set. In addition, the log line for abandonment now contains
4436 information about the queue run such as the pid. This is always present
4437 in "start" and "stop" lines but was omitted from the "abandon" line.
4439 PH/34 Omit spaces between a header name and the colon in the error message that
4440 is given when verify = headers_syntax fails (if there are lots of them,
4441 the message gets confusing).
4443 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4444 names, as there are now some PTR records that contain slashes. This check
4445 is only to protect against broken name servers that fall over on strange
4446 characters, so the fact that it applies to all lookups doesn't matter.
4448 PH/36 Now that the new test suite is complete, we can remove some of the
4449 special code in Exim that was needed for the old test suite. For example,
4450 sorting DNS records because real resolvers return them in an arbitrary
4451 order. The new test suite's fake resolver always returns records in the
4454 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4455 bounces) except when queue_only is set, to avoid logging races between
4456 the different processes.
4458 PH/38 Panic-die if .include specifies a non-absolute path.
4460 PH/39 A tweak to the "H" retry rule from its user.
4462 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4463 a label. They prevented compilation on older perls.
4465 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4466 a warning to be raised on newish perls.
4468 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4469 on queue. Changes to match documented behaviour of showing count of
4470 messages matching specified criteria.
4472 PH/40 Changed the default ident timeout from 30s to 5s.
4474 PH/41 Added support for the use of login_cap features, on those BSD systems
4475 that have them, for controlling the resources used by pipe deliveries.
4477 PH/42 The content-scanning code uses fopen() to create files in which to put
4478 message data. Previously it was not paying any attention to the mode of
4479 the files. Exim runs with umask(0) because the rest of the code creates
4480 files with open(), and sets the required mode explicitly. Thus, these
4481 files were ending up world-writeable. This was not a big issue, because,
4482 being within the spool directory, they were not world-accessible. I have
4483 created a function called modefopen, which takes an additional mode
4484 argument. It sets umask(777), creates the file, chmods it to the required
4485 mode, then resets the umask. All the relevant calls to fopen() in the
4486 content scanning code have been changed to use this function.
4488 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4489 to 24 hours. This avoids potential overflow problems when processing G
4490 and H retry rules. I suspect nobody ever tinkers with this value.
4492 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4494 PH/45 When the plaintext authenticator is running as a client, the server's
4495 challenges are checked to ensure they are valid base64 strings. By
4496 default, the authentication attempt is cancelled if an invalid string is
4497 received. Setting client_ignore_invalid_base64 true ignores these errors.
4498 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4499 they are received. Thus, the responses can be made to depend on the
4500 challenges. If an invalid string is ignored, an empty string is placed in
4503 PH/46 Messages that are created by the autoreply transport now contains a
4504 References: header, in accordance with RFCs 2822 and 3834.
4506 PH/47 Added authenticated_sender_force to the smtp transport.
4508 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4510 PH/49 Installed latest patch from the Sieve maintainer.
4512 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4513 was also set, the appendfile transport was unnecessarily scanning a
4514 directory of message files (e.g. for maildir delivery) to find the count
4515 of files (along with the size), even though it did not need this
4516 information. It now does the scan only if it needs to find either the
4517 size of the count of files.
4519 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4521 PH/52 Two bugs concerned with error handling when the smtp transport is
4524 (i) Exim was not creating retry information for temporary errors given
4525 for individual recipients after the DATA command when the smtp transport
4526 was used in LMTP mode. This meant that they could be retried too
4527 frequently, and not timed out correctly.
4529 (ii) Exim was setting the flag that allows error details to be returned
4530 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4531 recipients that were returned after the DATA command.
4533 PH/53 This is related to PH/52, but is more general: for any failing address,
4534 when detailed error information was permitted to be returned to the
4535 sender, but the error was temporary, then after the final timeout, only
4536 "retry timeout exceeded" was returned. Now it returns the full error as
4537 well as "retry timeout exceeded".
4539 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4540 do this, and (what is worse) MTAs that accept it.
4542 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4543 will now be deprecated.
4545 PH/56 New os.c-cygwin from the Cygwin maintainer.
4547 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4548 formats (previously only available in exim formats via -bpr, -bpru,
4549 and -bpra. Now also available in native and exiqgrep formats)
4551 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4552 with very large, slow to parse queues
4554 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4556 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4558 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4559 responses to authentication challenges, though it was showing the
4560 challenges; (ii) I've removed the CR characters from the debug output for
4563 PH/58 Allow for the insertion of a newline as well as a space when a string
4564 is turned into more than one encoded-word during RFC 2047 encoding. The
4565 Sieve code now uses this.
4567 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4568 data_4xx, lost_connection, tls_required.
4570 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4571 message was being sent as an SMTP response.
4573 PH/61 Add -l and -k options to exicyclog.
4575 PH/62 When verifying, if an address was redirected to one new address, so that
4576 verification continued, and the new address failed or deferred after
4577 having set something in $address_data, the value of $address_data was not
4578 passed back to the ACL. This was different to the case when no
4579 redirection occurred. The value is now passed back in both cases.
4581 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4582 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4583 use login_cap.h, so on its own it isn't the distinguishing feature. The
4584 new name refers directly to the setclassresources() function.
4586 PH/65 Added configuration files for NetBSD3.
4588 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4590 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4591 is preferred over IPv4.
4593 PH/68 The bounce_return_message and bounce_return_body options were not being
4594 honoured for bounces generated during the reception of non-SMTP messages.
4595 In particular, this applied to messages rejected by the ACL. This bug has
4596 been fixed. However, if bounce_return_message is true and bounce_return_
4597 body is false, the headers that are returned for a non-SMTP message
4598 include only those that have been read before the error was detected.
4599 (In the case of an ACL rejection, they have all been read.)
4601 PH/69 The HTML version of the specification is now built in a directory called
4602 spec_html instead of spec.html, because the latter looks like a path with
4603 a MIME-type, and this confuses some software.
4605 PH/70 Catch two compiler warnings in sieve.c.
4607 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4608 function verify_get_ident() calls ip_connect() to connect a socket, but
4609 if the "connect()" function timed out, ip_connect() used to close the
4610 socket. However, verify_get_ident() also closes the socket later, and in
4611 between Exim writes to the log, which may get opened at this point. When
4612 the socket was closed in ip_connect(), the log could get the same file
4613 descriptor number as the socket. This naturally causes chaos. The fix is
4614 not to close the socket in ip_connect(); the socket should be closed by
4615 the function that creates it. There was only one place in the code where
4616 this was missing, in the iplookup router, which I don't think anybody now
4617 uses, but I've fixed it anyway.
4619 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4620 well as to direct DNS lookups. Otherwise the handling of names in host
4621 lists is inconsistent and therefore confusing.
4627 PH/01 Two changes to the default runtime configuration:
4629 (1) Move the checks for relay_from_hosts and authenticated clients from
4630 after to before the (commented out) DNS black list checks.
4632 (2) Add control=submission to the relay_from_hosts and authenticated
4633 clients checks, on the grounds that messages accepted by these
4634 statements are most likely to be submissions.
4636 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4638 (1) Generate an error if the third argument for the ${prvs expansion is
4641 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4644 (3) Reset the variables that are obtained from the first argument of
4645 ${prvscheck and used in the second argument before leaving the code,
4646 because their memory is reclaimed, so using them afterwards may do
4649 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4650 one (it's much easier than Tom thought :-).
4652 (5) Because of (4), we can now allow for the use of $prvscheck_result
4653 inside the third argument.
4655 PH/03 For some reason, the default setting of PATH when running a command from
4656 a pipe transport was just "/usr/bin". I have changed it to
4659 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4660 anything to be listed in the output from -bV.
4662 PH/05 When a filter generated an autoreply, the entire To: header line was
4663 quoted in the delivery log line, like this:
4665 => >A.N.Other <ano@some.domain> <original@ddress> ...
4667 This has been changed so that it extracts the operative address. There
4668 may be more than one such address. If so, they are comma-separated, like
4671 => >ano@some.domain,ona@other.domain <original@ddress> ...
4673 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4674 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4675 not being looked up in the rDNS to get a host name, Exim was showing the
4676 IP address twice in Received: lines, even though the IP addresses were
4677 identical. For example:
4679 Received: from [1.2.3.4] (helo=[1.2.3.4])
4681 However, if the real host name was known, it was omitting the HELO data
4682 if it matched the actual IP address. This has been tidied up so that it
4683 doesn't show the same IP address twice.
4685 PH/07 When both +timestamp and +memory debugging was on, the value given by
4686 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4687 called by the debug printing, thereby overwriting the timestamp buffer.
4688 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4690 PH/08 When the original message was included in an autoreply transport, it
4691 always said "this is a copy of the message, including all the headers",
4692 even if body_only or headers_only was set. It now gives an appropriate
4695 PH/09 Applied a patch from the Sieve maintainer which:
4697 o fixes some comments
4698 o adds the (disabled) notify extension core
4699 o adds some debug output for the result of if/elsif tests
4700 o points to the current vacation draft in the documentation
4701 and documents the missing references header update
4705 o fixes a bug in processing the envelope test (when testing
4706 multiple envelope elements, the last element determined the
4709 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4710 Electronic Mail") by including:
4712 Auto-submitted: auto-generated
4714 in the messages that it generates (bounce messages and others, such as
4715 warnings). In the case of bounce messages for non-SMTP messages, there was
4716 also a typo: it was using "Auto_submitted" (underscore instead of
4717 hyphen). Since every message generated by Exim is necessarily in response
4718 to another message, thes have all been changed to:
4720 Auto-Submitted: auto-replied
4722 in accordance with these statements in the RFC:
4724 The auto-replied keyword:
4726 - SHOULD be used on messages sent in direct response to another
4727 message by an automatic process,
4729 - MUST NOT be used on manually-generated messages,
4731 - MAY be used on Delivery Status Notifications (DSNs) and Message
4732 Disposition Notifications (MDNs),
4734 - MUST NOT be used on messages generated by automatic or periodic
4735 processes, except for messages which are automatic responses to
4738 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4739 to the default Received: header definition.
4741 PH/12 Added log selector acl_warn_skipped (default on).
4743 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4744 variables because (a) they are in the wrong storage pool and (b) even if
4745 they were copied, it wouldn't work properly because of the caching.
4747 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4748 checking when decoding. Apparently there are clients that generate
4749 overlong encoded strings. Why am I not surprised?
4751 PH/15 If the first argument of "${if match_address" was not empty, but did not
4752 contain an "@" character, Exim crashed. Now it writes a panic log message
4753 and treats the condition as false.
4755 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4757 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4758 extension "envelope-auth". The code is finished and in agreement with
4759 other implementations, but there is no documentation so far and in fact,
4760 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4761 not changing the active code.
4763 Print executed "if" and "elsif" statements when debugging is used. This
4764 helps a great deal to understand what a filter does.
4766 Document more things not specified clearly in RFC3028. I had all this
4767 sorted out, when out of a sudden new issues came to my mind. Oops."
4769 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4772 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4773 canonical form (as documented). However, after a host name lookup from
4774 the IP address, check_host() was doing a simple string comparison with
4775 addresses acquired from the DNS when checking that the found name did
4776 have the original IP as one of its addresses. Since any found IPv6
4777 addresses are likely to be in abbreviated form, the comparison could
4778 fail. Luckily, there already exists a function for doing the comparison
4779 by converting both addresses to binary, so now that is used instead of
4780 the text comparison.
4782 PH/20 There was another similar case to PH/19, when a complete host name was
4783 given in a host list; looking up its IP address could give an abbreviated
4784 form, whereas the current host's name might or might not be abbreviated.
4785 The same fix has been applied.
4791 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4792 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4795 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4796 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4798 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4800 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4801 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4802 Also, for "Auto-Submitted", treat anything other than "no" as
4803 non-personal, in accordance with RFC 3834. (Previously it treated
4804 anything starting "auto-" as non-personal.)
4806 TF/01 The control=submission/name=... option had a problem with syntax
4807 errors if the name included a slash character. The /name= option
4808 now slurps the rest of the string, so it can include any characters
4809 but it must come last in the list of options (after /sender_retain
4812 PH/05 Some modifications to the interface to the fake nameserver for the new
4820 TK/01 Added the "success_on_redirect" address verification option. See
4821 NewStuff for rationale and an example.
4823 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4825 PH/02 Patch to exigrep to allow it to work on syslog lines.
4827 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4828 fread() to skip over the body file's header line, because in Cygwin the
4829 header line is locked and is inaccessible.
4831 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4832 co-exist for some time) to make it clear that it is the Exim ID that is
4833 referenced, not the Message-ID: header line.
4835 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4836 string_format() function, because snprintf() does not exist on all
4839 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4840 ${stat: expansion item.
4842 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4843 protocol synchronization error", to keep the pedants happy.
4845 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4846 well as for IRIX systems, when gcc is being used. See the host.c source
4849 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4851 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4854 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4855 transport and to the smtp transport in LMTP mode.
4857 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4859 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4860 run a filter in a subprocess. This could lead to confusion in subsequent
4861 lookups in the parent process. There should also be a search_tidyup() at
4862 the end of the subprocess.
4864 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4865 only if the host matched helo_try_verify_hosts, which caused the
4866 verification to occur when the EHLO/HELO command was issued. The ACL just
4867 tested the remembered result. Now, if a previous verification attempt has
4868 not happened, "verify = helo" does it there and then.
4870 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4872 TK/03 Fix log output including CR from clamd.
4874 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4875 exist provoked a memory error which could cause a segfault.
4877 PH/15 Installed PCRE 6.2
4879 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4881 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4882 of the problem. Specifically, suggested +O2 rather than +O1 for the
4885 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4887 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4888 record, the port was not correctly logged when the outgoing_port log
4889 selector was set (it logged the transort's default port).
4891 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4892 fallback_hosts, and "hosts" in the smtp transport.
4894 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4895 host errors such as "Connection refused".
4897 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4898 authentication with radiusclient 0.4.9:
4900 - Error returned from rc_read_config was caught wrongly
4901 - Username/password not passed on to radius server due to wrong length.
4903 The presumption is that some radiusclient API changes for 4.51/PH/17
4904 were not taken care of correctly. The code is still untested by me (my
4905 Linux distribution still has 0.3.2 of radiusclient), but it was
4906 contributed by a Radius user.
4908 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4909 expanding the "port" option of the smtp transport.
4911 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4912 while reading a MIME header. Thanks to Tom Hughes for a patch.
4914 PH/24 Include config.h inside local_scan.h so that configuration settings are
4917 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4918 that in an ACL for RCPT (for example), you can examine exactly what was
4921 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4922 commands, but it was not correctly comparing the address with the actual
4923 client host address. Thus, it would show the EHLO address in Received:
4924 header lines when this was not necessary.
4926 PH/27 Added the % operator to ${eval:}.
4928 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4929 it should be ignoring failures (because with -C, for example, it has lost
4930 privilege). It wasn't ignoring creation failures other than "already
4933 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4936 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4937 because an input line was too long, either on its own, or by virtue of
4938 too many continuations, the temporary file was not being removed, and the
4939 return code was incorrect.
4941 PH/31 Missing "BOOL" in function definition in filtertest.c.
4943 PH/32 Applied Sieve patches from the maintainer.
4945 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4947 PH/33 Added "verify = not_blind".
4949 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4950 Local/Makefile (with some defaults set). These are used in built scripts
4951 such as exicyclog, but they have never been used in the exim_install
4952 script (though there are many overriding facilities there). I have
4953 arranged that the exim_install script now takes note of these two
4956 PH/35 Installed configuration files for Dragonfly.
4958 PH/36 When a locally submitted message by a trusted user did not contain a
4959 From: header, and the sender address was obtained from -f or from an SMTP
4960 MAIL command, and the trusted user did not use -F to supply a sender
4961 name, $originator_name was incorrectly used when constructing a From:
4962 header. Furthermore, $originator_name was used for submission mode
4963 messages from external hosts without From: headers in a similar way,
4964 which is clearly wrong.
4966 PH/37 Added control=suppress_local_fixups.
4968 PH/38 When log_selector = +received_sender was set, and the addition of the
4969 sender made the log line's construction buffer exactly full, or one byte
4970 less than full, an overflow happened when the terminating "\n" was
4973 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4974 when the result of a list match is failure because a DNS lookup failed.
4976 PH/40 RM_COMMAND is now used in the building process.
4978 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4979 the "build-* directories that it finds.
4981 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4982 address in a domain literal was a prefix of an interface address.
4984 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4985 when verifying a sender address, unless rewrite_headers is false.
4987 PH/44 Wrote a long comment about why errors_to addresses are verified as
4988 recipients, not senders.
4990 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4991 the ratelimit ACL was added.
4993 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4995 PH/46 Added extra information about PostgreSQL errors to the error string.
4997 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4998 avoiding the need to install special zones in a real server. This is
4999 backwards compatible; if it can't find the fake resolver, it drops back.
5000 Thus, both old and new test suites can be run.
5002 TF/02 Added util/ratelimit.pl
5004 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5005 clock is set back in time.
5007 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5008 Candler <B.Candler@pobox.com>.
5010 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5011 OK for addresses that are the result of redirections.
5013 PH/48 A number of further additions for the benefit of the new test suite,
5014 including a fake gethostbyname() that interfaces to the fake DNS resolver
5017 TF/06 The fix for widen_domains has also been applied to qualify_single and
5018 search_parents which are the other dnslookup options that can cause
5021 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5024 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5026 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5027 service" if it encountered an MX record with an empty target hostname.
5028 The message is now "an MX or SRV record indicated no SMTP service".
5030 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5031 if the DNS of the sending site is misconfigured. This is quite a
5032 common situation. This change restores the behaviour of treating a
5033 helo verification defer as a failure.
5035 PH/51 If self=fail was set on a router, the bounce message did not include the
5036 actual error message.
5042 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5044 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5045 command itself ended in error, the underlying message about the transport
5046 filter timeout was being overwritten with the pipe command error. Now the
5047 underlying error message should be appended to the second error message.
5049 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5051 PH/02 Reduce the amount of output that "make" produces by default. Full output
5052 can still be requested.
5054 PH/03 The warning log line about a condition test deferring for a "warn" verb
5055 was being output only once per connection, rather than after each
5056 occurrence (because it was using the same function as for successful
5057 "warn" verbs). This seems wrong, so I have changed it.
5059 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5060 it should not have, which might have caused a crash in the right
5061 circumstances, but probably never did.
5063 PH/04 Installed a modified version of Tony Finch's patch to make submission
5064 mode fix the return path as well as the Sender: header line, and to
5065 add a /name= option so that you can make the user's friendly name appear
5068 TF/03 Added the control = fakedefer ACL modifier.
5070 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5071 Mark Lowes for thorough testing.
5073 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5075 TK/03 Merged latest SRS patch from Miles Wilton.
5077 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5078 with the definition in sysexits.h (which is #included earlier).
5079 Fortunately, Exim does not actually use EX_OK. The code used to try to
5080 preserve the sysexits.h value, by assuming that macro definitions were
5081 scanned for macro replacements. I have been disabused of this notion,
5082 so now the code just undefines EX_OK before #including unistd.h.
5084 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5085 in the smtp transport. When a block could not be written in a single
5086 write() function, the timeout was being re-applied to each part-write.
5087 This seems wrong - if the receiver was accepting one byte at a time it
5088 would take for ever. The timeout is now adjusted when this happens. It
5089 doesn't have to be particularly precise.
5091 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5092 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5094 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5095 without a domain if the check to <postmaster@domain> fails.
5097 SC/01 Eximstats: added -xls and the ability to specify output files
5098 (patch written by Frank Heydlauf).
5100 SC/02 Eximstats: use FileHandles for outputting results.
5102 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5104 SC/04 Eximstats: fixed display of large numbers with -nvr option
5106 SC/05 Eximstats: fixed merging of reports with empty tables.
5108 SC/06 Eximstats: added the -include_original_destination flag
5110 SC/07 Eximstats: removed tabs and trailing whitespace.
5112 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5114 TK/06 MBOX spool code: Add real "From " MBOX separator line
5115 so the .eml file is really in mbox format (even though
5116 most programs do not really care). Patch from Alex Miller.
5118 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5119 The latter is generated from $received_to and is only set if the
5120 message has one envelope recipient. SA can use these headers,
5121 obviously out-of-the-box. Patch from Alex Miller.
5123 PH/08 The ${def test on a variable was returning false if the variable's
5124 value was "0", contrary to what the specification has always said!
5125 The result should be true unless the variable is empty.
5127 PH/09 The syntax error of a character other than { following "${if
5128 def:variable_name" (after optional whitespace) was not being diagnosed.
5129 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5130 accidental colon was present, for example, could give incorrect results.
5132 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5133 result is used (not including appendfile, where other changes are about
5136 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5137 This involved changing a lot of size variables from int to off_t. It
5138 should work with maildirs and everything.
5140 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5141 spamd dying while we are connected to it.
5143 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5146 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5147 function for BDB 4.3.
5149 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5151 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5152 the sources for PCRE have been re-arranged and more files are now
5155 PH/15 The code I had for printing potentially long long variables in PH/11
5156 above was not the best (it lost precision). The length of off_t variables
5157 is now inspected at build time, and an appropriate printing format (%ld
5158 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5159 to be "long long int" or "long int". This is needed for the internal
5160 formatting function string_vformat().
5162 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5163 the configuration file to be ":syslog", then the script "guesses" where
5164 the logs files are, rather than using the compiled in default. In our
5165 case the guess is not the same as the compiled default, so the script
5166 suddenly stopped working when I started to use syslog. The patch checks
5167 to see if log_file_path is "". If so, it attempts to read it from exim
5168 with no configuration file to get the compiled in version, before it
5169 falls back to the previous guessing code."
5171 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5172 implementing BATV in an Exim configuration. See NewStuff for the gory
5175 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5176 Makefile that are specific to HP-UX.
5178 PH/18 If the "use_postmaster" option was set for a recipient callout together
5179 with the "random" option, the postmaster address was used as the MAIL
5180 FROM address for the random test, but not for the subsequent recipient
5181 test. It is now used for both.
5183 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5184 patch removes a few documentation additions to RFC 3028, because the
5185 latest draft now contains them. It adds the new en;ascii-case comparator
5186 and a new error check for 8bit text in MIME parts. Comparator and
5187 require names are now matched exactly. I enabled the subaddress
5188 extension, but it is not well tested yet (read: it works for me)."
5190 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5191 rework some of the code of TK/09 above to avoid the hardwired use of
5192 "%lld" and "long long". Replaced the call to snprintf() with a call to
5195 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5196 records point to non-existent hosts", "retry timeout exceeded", and
5197 "retry time not reached for any host after a long failure period".
5199 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5200 experimental DomainKeys support:
5202 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5203 (2) On an error such as an illegally used "control", the wrong name for
5204 the control was given.
5206 These problems did NOT occur unless DomainKeys support was compiled.
5208 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5210 PH/24 Added ${if match_ip condition.
5212 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5213 databases so that it will be absolutely obvious if a crash occurs in the
5214 DB library. This is a regular occurrence (often caused by mis-matched
5217 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5218 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5219 picked up on a user's system that detects such things. There doesn't seem
5220 to be a gcc warning option for this - only an attribute that has to be
5221 put on the function's prototype. It seems that in Fedora Core 4 they have
5222 set this on a number of new functions. No doubt there will be more in due
5225 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5226 specify a transport. However, if an address that was verified by such a
5227 router was the subject of a callout, Exim crashed because it tried to
5228 read the rcpt_include_affixes from the non-existent transport. Now it
5229 just assumes that the setting of that option is false. This bug was
5230 introduced by 4.51/PH/31.
5232 PH/28 Changed -d+all to exclude +memory, because that information is very
5233 rarely of interest, but it makes the output a lot bigger. People tend to
5234 do -d+all out of habit.
5236 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5237 code in os-type was giving problems when libc.so lives in lib64, like on
5240 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5241 aren't the modern standard, and it seems that some systems' include files
5242 don't always have them. Exim was already checking for some of the newer
5243 ones like T_AAAA, and defining it itself. I've added checks for all the
5244 record types that Exim uses.
5246 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5247 not automatically generating a new one, as it is supposed to. This
5248 prevented TLS from working. If the file did exist, but contained invalid
5249 data, a new version was generated, as expected. It was only the case of a
5250 non-existent file that was broken.
5252 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5253 with a change in libdomainkeys > 0.64.
5255 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5256 from DNS. If the selector record carries the flag, it now has
5257 precedence over the domain-wide flag.
5259 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5261 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5262 the use of a port name that isn't defined in /etc/services) occurred, the
5263 message was deferred as in a normal delivery, and thus remained on the
5264 spool, instead of being failed because of the mua_wrapper setting. This
5265 is now fixed, and I tidied up some of the mua_wrapper messages at the
5268 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5269 the messages in a hash of arrays rather than using individual hashes.
5270 This is a bit cleaner and results in dramatic memory savings, albeit
5271 at a slight CPU cost.
5273 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5274 as requested by Marc Sherman.
5276 SC/10 Eximstats: added histograms for user specified patterns as requested
5279 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5281 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5282 fopen() in the content-scanning modules that did not already have it.
5288 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5289 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5291 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5293 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5295 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5296 to test Sieve filters that use "vacation".
5298 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5299 that changes the way the GnuTLS parameters are stored in the cache file.
5300 The new format can be generated externally. For backward compatibility,
5301 if the data in the cache doesn't make sense, Exim assumes it has read an
5302 old-format file, and it generates new data and writes a new file. This
5303 means that you can't go back to an older release without removing the
5306 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5307 work if there are any delivery delays because "one_time" forces the
5308 parent to be marked "delivered", so its unseen clone is never tried
5309 again. For this reason, Exim now forbids the simultaneous setting of
5312 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5313 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5314 entry if you want to know the details. The fix, however, neglected to
5315 consider the case when local delivery batching is involved. The test for
5316 "previously delivered" was not happening when checking to see if an
5317 address could be batched with a previous (undelivered) one; under
5318 certain circumstances this could lead to multiple deliveries to the same
5321 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5322 in its include files, and this causes problems building Exim.
5324 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5325 header_syntax) but Exim was just ignoring anything given after a slash.
5326 In particular, this caused confusion with an attempt to use "verify =
5327 reverse_host_lookup/defer_ok". An error is now given when options are
5328 supplied for verify items that do not have them. (Maybe reverse_host_
5329 lookup should have a defer_ok option, but that's a different point.)
5331 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5332 defined by RFC 821) to 2048, because there were problems with some AUTH
5333 commands, and RFC 1869 says the size should be increased for extended
5334 SMTP commands that take arguments.
5336 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5339 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5340 "unknown" error; now it says that the functionality isn't in the binary.
5342 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5343 an address' error message when a string expansion fails (syntax or
5344 whatever). Otherwise the password may appear in the log. Following change
5345 PH/42 below, there is no longer a chance of it appearing in a bounce
5348 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5350 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5351 of its addresses was ever tried. (Bugzilla bug #2.)
5353 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5354 the result incorrectly in the debug output. (It correctly added a newline
5355 to what was transported.)
5357 TF/01 Added $received_time.
5359 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5360 commented out examples of how to interface to a virus scanner and to
5361 SpamAssassin. Also added commented examples of av_scanner and
5362 spamd_address settings.
5364 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5365 and controls are allowed in which ACLs. There were a couple of minor
5366 errors. Some of the entries in the conditions table (which is a table of
5367 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5368 negation of where the condition IS allowed.
5370 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5372 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5373 header file does not have a version number, so I've had to invent a new
5374 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5375 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5376 radiusclient), but it was contributed by a Radius user.
5378 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5379 files or directories, for OpenSSL.
5381 PH/19 When an Exim process that is running as root has to create an Exim log
5382 file, it does so in a subprocess that runs as exim:exim so as to get the
5383 ownership right at creation (otherwise, other Exim processes might see
5384 the file with the wrong ownership). There was no test for failure of this
5385 fork() call, which would lead to the process getting stuck as it waited
5386 for a non-existent subprocess. Forks do occasionally fail when resources
5387 run out. I reviewed all the other calls to fork(); they all seem to check
5390 PH/20 When checking for unexpected SMTP input at connect time (before writing
5391 the banner), Exim was not dealing correctly with a non-positive return
5392 from the read() function. If the client had disconnected by this time,
5393 the result was a log entry for a synchronization error with an empty
5394 string after "input=" when read() returned zero. If read() returned -1
5395 (an event I could not check), uninitialized data bytes were printed.
5396 There were reports of junk text (parts of files, etc) appearing after
5399 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5401 PH/22 Added support for macro redefinition, and (re)definition in between
5402 driver and ACL definitions.
5404 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5405 forgetting to use the resulting value; it was using the unexpanded value.
5407 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5408 hadn't been configured. The fix is from Juergen Kreileder, who
5409 understands it better than I do:
5411 "Here's what I see happening with three configured cyrus_sasl
5412 authenticators configured (plain, login, cram-md5):
5414 On startup auth_cyrus_sasl_init() gets called for each of these.
5415 This means three calls to sasl_listmech() without a specified mech_list.
5416 => SASL tests which mechs of all available mechs actually work
5417 => three warnings about OTP not working
5418 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5420 With the patch, sasl_listmech() also gets called three times. But now
5421 SASL's mech_list option is set to the server_mech specified in the the
5422 authenticator. Or in other words, the answer from sasl_listmech()
5423 gets limited to just the mech you're testing for (which is different
5425 => the return list contains just 'plain' or 'login', 'cram-md5' or
5426 nothing depending on the value of ob->server_mech.
5428 I've just tested the patch: Authentication still works fine,
5429 unavailable mechs specified in the exim configuration are still
5430 caught, and the auth.log warnings about OTP are gone."
5432 PH/25 When debugging is enabled, the contents of the command line are added
5433 to the debugging output, even when log_selector=+arguments is not
5436 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5437 answer is "GNU", and only if the return is "GNU/something" is the answer
5440 PH/27 $acl_verify_message is now set immediately after the failure of a
5441 verification in an ACL, and so is available in subsequent modifiers. In
5442 particular, the message can be preserved by coding like this:
5444 warn !verify = sender
5445 set acl_m0 = $acl_verify_message
5447 Previously, $acl_verify_message was set only while expanding "message"
5448 and "log_message" when a very denied access.
5450 PH/28 Modified OS/os.c-Linux with
5452 -#ifndef OS_LOAD_AVERAGE
5453 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5455 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5456 nomenclature these days.)
5458 PH/29 Installed patch from the Sieve maintainer that adds the options
5459 sieve_useraddress and sieve_subaddress to the redirect router.
5461 PH/30 In these circumstances:
5462 . Two addresses routed to the same list of hosts;
5463 . First host does not offer TLS;
5464 . First host accepts first address;
5465 . First host gives temporary error to second address;
5466 . Second host offers TLS and a TLS session is established;
5467 . Second host accepts second address.
5468 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5469 and peerdn, if requested) that were in fact used only for the second
5472 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5473 attention to any local part prefix or suffix that was matched by the
5474 router that accepted the address. It now behaves in the same way as it
5475 does for delivery: the affixes are removed from the local part unless
5476 rcpt_include_affixes is set on the transport.
5478 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5479 timeout during the DATA phase of an incoming message.
5481 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5482 applied a patch sanctioned by the Sieve maintainer.
5484 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5485 the uid or gid is negative. A case of a negative gid caused this to be
5486 noticed. The fix allows for either to be negative.
5488 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5489 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5492 PH/36 The change PH/12 above was broken. Fixed it.
5494 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5495 the grounds that routing the same address twice would always produce the
5496 same answer. This might have been true once, but it is certainly no
5497 longer true now. Routing a child address may depend on the previous
5498 routing that produced that child. Some complicated redirection strategies
5499 went wrong when messages had multiple recipients, and made Exim's
5500 behaviour dependent on the order in which the addresses were given.
5502 I have moved the duplicate checking until after the routing is complete.
5503 Exim scans the addresses that are assigned to local and remote
5504 transports, and removes any duplicates. This means that more work will be
5505 done, as duplicates will always all be routed, but duplicates are
5506 presumably rare, so I don't expect this is of any significance.
5508 For deliveries to pipes, files, and autoreplies, the duplicate checking
5509 still happens during the routing process, since they are not going to be
5512 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5513 It corrects a timeout issue with spamd. This is Ian's comment: "The
5514 background is that sometimes spamd either never reads data from a
5515 connection it has accepted, or it never writes response data. The exiscan
5516 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5517 blindly assumes that writes won't block so it may never time out."
5519 PH/39 Allow G after quota size as well as K and M.
5521 PH/40 The value set for $authenticated_id in an authenticator may not contain
5522 binary zeroes or newlines because the value is written to log lines and
5523 to spool files. There was no check on this. Now the value is run through
5524 the string_printing() function so that such characters are converted to
5525 printable escape sequences.
5527 PH/41 $message_linecount is a new variable that contains the total number of
5528 lines in the message. Compare $body_linecount, which is the count for the
5531 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5532 bounce and delay warning messages, except in certain special cases, which
5535 (a) An SMTP error message from a remote host;
5536 (b) A message specified in a :fail: redirection;
5537 (c) A message specified in a "fail" command in a system filter;
5538 (d) A message specified in a FAIL return from the queryprogram router;
5539 (e) A message specified by the cannot_route_message router option.
5541 In these cases only, Exim does include the error details in bounce and
5542 warning messages. There are also a few cases where bland messages such
5543 as "unrouteable address" or "local delivery error" are given.
5545 PH/43 $value is now also set for the "else" part of a ${run expansion.
5547 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5548 being worked on, but at least Exim now implements the latest version to
5551 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5552 process to complete was treated as a delivery failure, a timeout while
5553 writing the message to the pipe was logged, but erroneously treated as a
5554 successful delivery. Such timeouts include transport filter timeouts. For
5555 consistency with the overall process timeout, these timeouts are now
5556 treated as errors, giving rise to delivery failures by default. However,
5557 there is now a new Boolean option for the pipe transport called
5558 timeout_defer, which, if set TRUE, converts the failures into defers for
5559 both kinds of timeout. A transport filter timeout is now identified in
5562 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5563 systems where "make" and "gmake" are different, calling "gmake" at top
5564 level broke things. I've arranged for the value of $(MAKE) to be passed
5565 from the Makefile to this script so that it can call the same version of
5569 A note about Exim versions 4.44 and 4.50
5570 ----------------------------------------
5572 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5573 changes of various kinds. As a consequence, a big documentation update was
5574 needed. This delayed the release for rather longer than seemed good, especially
5575 in the light of a couple of (minor) security issues. Therefore, the changes
5576 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5577 release. So 4.44 and 4.50 are in effect two different branches that both start
5580 I have left the 4.50 change log unchanged; it contains all the changes since
5581 4.43. The change log for 4.44 is below; many of its items are identical to
5582 those for 4.50. This seems to be the most sensible way to preserve the
5583 historical information.
5589 1. Minor wording change to the doc/README.SIEVE file.
5591 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5592 computation of the current number of files was incorrect.
5594 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5595 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5598 4. Give more explanation in the error message when the command for a transport
5599 filter fails to execute.
5601 5. There are several places where Exim runs a non-Exim command in a
5602 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5603 was being done only for the command run by the queryprogram router. It is
5604 now done for all such subprocesses. The other cases are: ${run, transport
5605 filters, and the commands run by the lmtp and pipe transports.
5607 6. Added CONFIGURE_GROUP build-time option.
5609 7. Some older OS have a limit of 256 on the maximum number of file
5610 descriptors. Exim was using setrlimit() to set 1000 as a large value
5611 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5612 systems. I've change it so that if it can't get 1000, it tries for 256.
5614 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5615 was an oversight, and furthermore, ever since the addition of extra
5616 controls (e.g. 4.43/32), the checks on when to allow different forms of
5617 "control" were broken. There should now be diagnostics for all cases when a
5618 control that does not make sense is encountered.
5620 9. Added the /retain_sender option to "control=submission".
5622 10. $recipients is now available in the predata ACL (oversight).
5624 11. Tidy the search cache before the fork to do a delivery from a message
5625 received from the command line. Otherwise the child will trigger a lookup
5626 failure and thereby defer the delivery if it tries to use (for example) a
5627 cached ldap connection that the parent has called unbind on.
5629 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5630 of $address_data from the recipient verification was clobbered by the
5631 sender verification.
5633 13. The value of address_data from a sender verification is now available in
5634 $sender_address_data in subsequent conditions in the ACL statement.
5636 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5638 15. Added a new option "connect=<time>" to callout options, to set a different
5641 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5642 was its contents. (It was OK if the option was not defined at all.)
5644 17. A "Completed" log line is now written for messages that are removed from
5645 the spool by the -Mrm option.
5647 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5648 information about exactly what failed.
5650 19. Added -dd to debug only the daemon process.
5652 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5653 handles timeouts, both on the server side and network timeouts. Renamed the
5654 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5656 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5657 It is now set to "smtps".
5659 22. $host_address is now set to the target address during the checking of
5660 ignore_target_hosts.
5662 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5663 being passed; this would have caused $sender_host_name to have been used if
5664 matching the list had actually called for a host name (not very likely,
5665 since this list is usually IP addresses). A host name is now passed as
5668 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5669 code that specifies a non-restarting handler (typically sigaction() in
5670 modern systems) in an attempt to fix a rare and obscure crash bug.
5672 25. Narrowed the window for a race in the daemon that could cause it to ignore
5673 SIGCHLD signals. This is not a major problem, because they are used only to
5674 wake it up if nothing else does.
5676 26. A malformed maildirsize file could cause Exim to calculate negative values
5677 for the mailbox size or file count. Odd effects could occur as a result.
5678 The maildirsize information is now recalculated if the size or filecount
5681 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5682 support for a long time. Removed HAVE_SYS_VFS_H.
5684 28. Installed the latest version of exipick from John Jetmore.
5686 29. In an address list, if the pattern was not a regular expression, an empty
5687 subject address (from a bounce message) matched only if the pattern was an
5688 empty string. Non-empty patterns were not even tested. This was the wrong
5689 because it is perfectly reasonable to use an empty address as part of a
5690 database query. An empty address is now tested by patterns that are
5691 lookups. However, all the other forms of pattern expect the subject to
5692 contain a local part and a domain, and therefore, for them, an empty
5693 address still always fails if the pattern is not itself empty.
5695 30. Exim went into a mad DNS loop when attempting to do a callout where the
5696 host was specified on an smtp transport, and looking it up yielded more
5697 than one IP address.
5699 31. Re-factored the code for checking spool and log partition space into a
5700 function that finds that data and another that does the check. The former
5701 is then used to implement four new variables: $spool_space, $log_space,
5702 $spool_inodes, and $log_inodes.
5704 32. The RFC2047 encoding function was originally intended for short strings
5705 such as real names; it was not keeping to the 75-character limit for
5706 encoded words that the RFC imposes. It now respects the limit, and
5707 generates multiple encoded words if necessary. To be on the safe side, I
5708 have increased the buffer size for the ${rfc2047: expansion operator from
5711 33. It is now permitted to omit both strings after an "if" condition; if the
5712 condition is true, the result is "true". As before, when the second string
5713 is omitted, a false condition yields an empty string. This makes it less
5714 cumbersome to write custom ACL and router conditions.
5716 34. Failure to deliver a bounce message always caused it to be frozen, even if
5717 there was an errors_to setting on the router. The errors_to setting is now
5720 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5721 canonical form (fully expanded) before being placed in
5722 $sender_host_address.
5724 36. The table in the code that translates DNS record types into text (T_A to
5725 "A" for instance) was missing entries for NS and CNAME. It is just possible
5726 that this could have caused confusion if both these types were looked up
5727 for the same domain, because the text type is used as part of Exim's
5728 per-process caching. But the chance of anyone hitting this buglet seems
5731 37. The dnsdb lookup has been extended in a number of ways.
5733 (1) There is a new type, "zns", which walks up the domain tree until it
5734 finds some nameserver records. It should be used with care.
5736 (2) There is a new type, "mxh", which is like "mx" except that it returns
5737 just the host names, not the priorities.
5739 (3) It is now possible to give a list of domains (or IP addresses) to be
5740 looked up. The behaviour when one of the lookups defers can be
5741 controlled by a keyword.
5743 (4) It is now possible to specify the separator character for use when
5744 multiple records are returned.
5746 38. The dnslists ACL condition has been extended: it is now possible to supply
5747 a list of IP addresses and/or domains to be looked up in a particular DNS
5750 39. Added log_selector=+queue_time_overall.
5752 40. When running the queue in the test harness, wait just a tad after forking a
5753 delivery process, to get repeatability of debugging output.
5755 41. Include certificate and key file names in error message when GnuTLS fails
5756 to set them up, because the GnuTLS error message doesn't include the name
5757 of the failing file when there is a problem reading it.
5759 42. Allow both -bf and -bF in the same test run.
5761 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5763 44. The "Exiscan patch" is now merged into the mainline Exim source.
5765 45. Sometimes the final signoff response after QUIT could fail to get
5766 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5767 < 0 before doing a fflush(). This bug looks as though it goes back to the
5768 introduction of TLS in release 3.20, but "sometimes" must have been rare
5769 because the tests only now provoked it.
5771 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5772 (this can affect the format of dates).
5774 47. exim_tidydb, when checking for the continued existence of a message for
5775 which it has found a message-specific retry record, was not finding
5776 messages that were in split spool directories. Consequently, it was
5777 deleting retry records that should have stayed in existence.
5779 48. Steve fixed some bugs in eximstats.
5781 49. The SPA authentication driver was not abandoning authentication and moving
5782 on to the next authenticator when an expansion was forced to fail,
5783 contradicting the general specification for all authenticators. Instead it
5784 was generating a temporary error. It now behaves as specified.
5786 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5787 (the order specifies the preference for clients). The order is now AES256,
5788 AES128, 3DES, ARCFOUR128.
5790 51. Small patch to Sieve code - explicitly set From: when generating an
5793 52. Exim crashed if a remote delivery caused a very long error message to be
5794 recorded - for instance if somebody sent an entire SpamAssassin report back
5795 as a large number of 550 error lines. This bug was coincidentally fixed by
5796 increasing the size of one of Exim's internal buffers (big_buffer) that
5797 happened as part of the Exiscan merge. However, to be on the safe side, I
5798 have made the code more robust (and fixed the comments that describe what
5801 53. Now that there can be additional text after "Completed" in log lines (if
5802 the queue_time_overall log selector is set), a one-byte patch to exigrep
5803 was needed to allow it to recognize "Completed" as not the last thing in
5806 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5807 patch that reportedly fixes this has been added. I am not expert enough to
5808 create a test for it. This is what the patch creator wrote:
5810 "I found a little strange behaviour of ldap code when working with
5811 Windows 2003 AD Domain, where users was placed in more than one
5812 Organization Units. When I tried to give exim partial DN, the exit code
5813 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5814 But simultaneously result of request was absolutely normal ldap result,
5815 so I produce this patch..."
5817 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5818 REFERENCE, so I have modified the code to exclude the patch when that macro
5821 55. Some experimental protocols are using DNS PTR records for new purposes. The
5822 keys for these records are domain names, not reversed IP addresses. The
5823 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5824 leaves it alone. Component reversal etc. now happens only for IP addresses.
5827 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5829 57. Double the size of the debug message buffer (to 2048) so that more of very
5830 long debug lines gets shown.
5832 58. The exicyclog utility now does better if the number of log files to keep
5833 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5835 59. Two changes related to the smtp_active_hostname option:
5837 (1) $smtp_active_hostname is now available as a variable.
5838 (2) The default for smtp_banner uses $smtp_active_hostname instead
5839 of $primary_hostname.
5841 60. The host_aton() function is supposed to be passed a string that is known
5842 to be a valid IP address. However, in the case of IPv6 addresses, it was
5843 not checking this. This is a hostage to fortune. Exim now panics and dies
5844 if the condition is not met. A case was found where this could be provoked
5845 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5846 components; fortuitously, this particular loophole had already been fixed
5847 by change 4.50/55 above.
5849 If there are any other similar loopholes, the new check in host_aton()
5850 itself should stop them being exploited. The report I received stated that
5851 data on the command line could provoke the exploit when Exim was running as
5852 exim, but did not say which command line option was involved. All I could
5853 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5854 running as the user.
5857 61. There was a buffer overflow vulnerability in the SPA authentication code
5858 (which came originally from the Samba project). I have added a test to the
5859 spa_base64_to_bits() function which I hope fixes it.
5862 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5863 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5865 63. The daemon start-up calls getloadavg() while still root for those OS that
5866 need the first call to be done as root, but it missed one case: when
5867 deliver_queue_load_max is set with deliver_drop_privilege. This is
5868 necessary for the benefit of the queue runner, because there is no re-exec
5869 when deliver_drop_privilege is set.
5871 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5872 This has been fixed.
5874 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5875 was in use, was not putting the data itself into the right store pool;
5876 consequently, it could be overwritten for a subsequent message in the same
5877 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5880 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5882 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5883 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5884 the function were treating the return as a boolean value, which happened to
5885 work because 0=false and not-0=true, but is not correct code.
5887 68. The host_aton() function was not handling scoped IPv6 addresses (those
5888 with, for example, "%eth0" on the end) correctly.
5890 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5891 negated items (that is, ~something) in unsigned ints. Some compilers
5892 apparently mutter when there is no cast.
5894 70. If an address verification called from an ACL failed, and did not produce a
5895 user-specific message (i.e. there was only a "system" message), nothing was
5896 put in $acl_verify_message. In this situation, it now puts the system
5899 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5900 session; change 4.31/43 added the unwanted input to the log line - except
5901 that it did not do this in the start of session case. It now does.
5903 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5904 This is wrong and can cause the other end to generate a synchronization
5905 error if it is another Exim or anything else that does the synchronization
5906 check. A QUIT command is no longer sent after a timeout.
5908 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5909 during host lookups.
5911 74. The defer_ok option of callout verification was not working if it was used
5912 when verifying addresses in header lines, that is, for this case:
5914 verify = header_sender/callout=defer_ok
5916 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5917 those file descriptors could be used for SMTP connections. If anything
5918 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5919 could be sent to the SMTP client, causing chaos. The daemon now opens
5920 stdin, stdout, and stderr to /dev/null when it puts itself into the
5923 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5924 log by default. The user can override this with suitable Perl magic.
5926 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5927 the log message when discard triggers, was not working for the DATA ACL or
5928 for the non-SMTP ACL.
5930 78. Error message wording change in sieve.c.
5932 79. If smtp_accept_max_per_host was set, the number of connections could be
5933 restricted to fewer than expected, because the daemon was trying to set up
5934 a new connection before checking whether the processes handling previous
5935 connections had finished. The check for completed processes is now done
5936 earlier. On busy systems, this bug wouldn't be noticed because something
5937 else would have woken the daemon, and it would have reaped the completed
5940 80. If a message was submitted locally by a user whose login name contained one
5941 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5942 It caused a spool format error. I have fixed the spool reading code. A
5943 related problem was that the "from" clause in the Received: line became
5944 illegal because of the space(s). It is now covered by ${quote_local_part.
5946 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5949 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5951 83. Patch from Sieve maintainer to fix unterminated string problem in
5952 "vacation" handling.
5954 84. Some minor changes to the Linux configuration files to help with other
5955 OS variants using glibc.
5957 85. One more patch for Sieve to update vacation handling to latest spec.
5960 ----------------------------------------------------
5961 See the note above about the 4.44 and 4.50 releases.
5962 ----------------------------------------------------
5968 1. Change 4.43/35 introduced a bug that caused file counts to be
5969 incorrectly computed when quota_filecount was set in an appendfile
5972 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5973 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5976 3. Give more explanation in the error message when the command for a transport
5977 filter fails to execute.
5979 4. There are several places where Exim runs a non-Exim command in a
5980 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5981 was being done only for the command run by the queryprogram router. It is
5982 now done for all such subprocesses. The other cases are: ${run, transport
5983 filters, and the commands run by the lmtp and pipe transports.
5985 5. Some older OS have a limit of 256 on the maximum number of file
5986 descriptors. Exim was using setrlimit() to set 1000 as a large value
5987 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5988 systems. I've change it so that if it can't get 1000, it tries for 256.
5990 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5991 was an oversight, and furthermore, ever since the addition of extra
5992 controls (e.g. 4.43/32), the checks on when to allow different forms of
5993 "control" were broken. There should now be diagnostics for all cases when a
5994 control that does not make sense is encountered.
5996 7. $recipients is now available in the predata ACL (oversight).
5998 8. Tidy the search cache before the fork to do a delivery from a message
5999 received from the command line. Otherwise the child will trigger a lookup
6000 failure and thereby defer the delivery if it tries to use (for example) a
6001 cached ldap connection that the parent has called unbind on.
6003 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6004 of $address_data from the recipient verification was clobbered by the
6005 sender verification.
6007 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6008 was its contents. (It was OK if the option was not defined at all.)
6010 11. A "Completed" log line is now written for messages that are removed from
6011 the spool by the -Mrm option.
6013 12. $host_address is now set to the target address during the checking of
6014 ignore_target_hosts.
6016 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6017 being passed; this would have caused $sender_host_name to have been used if
6018 matching the list had actually called for a host name (not very likely,
6019 since this list is usually IP addresses). A host name is now passed as
6022 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6023 code that specifies a non-restarting handler (typically sigaction() in
6024 modern systems) in an attempt to fix a rare and obscure crash bug.
6026 15. Narrowed the window for a race in the daemon that could cause it to ignore
6027 SIGCHLD signals. This is not a major problem, because they are used only to
6028 wake it up if nothing else does.
6030 16. A malformed maildirsize file could cause Exim to calculate negative values
6031 for the mailbox size or file count. Odd effects could occur as a result.
6032 The maildirsize information is now recalculated if the size or filecount
6035 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6036 support for a long time. Removed HAVE_SYS_VFS_H.
6038 18. Updated exipick to current release from John Jetmore.
6040 19. Allow an empty sender to be matched against a lookup in an address list.
6041 Previously the only cases considered were a regular expression, or an
6044 20. Exim went into a mad DNS lookup loop when doing a callout where the
6045 host was specified on the transport, if the DNS lookup yielded more than
6048 21. The RFC2047 encoding function was originally intended for short strings
6049 such as real names; it was not keeping to the 75-character limit for
6050 encoded words that the RFC imposes. It now respects the limit, and
6051 generates multiple encoded words if necessary. To be on the safe side, I
6052 have increased the buffer size for the ${rfc2047: expansion operator from
6055 22. Failure to deliver a bounce message always caused it to be frozen, even if
6056 there was an errors_to setting on the router. The errors_to setting is now
6059 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6060 canonical form (fully expanded) before being placed in
6061 $sender_host_address.
6063 24. Updated eximstats to version 1.33
6065 25. Include certificate and key file names in error message when GnuTLS fails
6066 to set them up, because the GnuTLS error message doesn't include the name
6067 of the failing file when there is a problem reading it.
6069 26. Expand error message when OpenSSL has problems setting up cert/key files.
6072 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6073 (this can affect the format of dates).
6075 28. exim_tidydb, when checking for the continued existence of a message for
6076 which it has found a message-specific retry record, was not finding
6077 messages that were in split spool directories. Consequently, it was
6078 deleting retry records that should have stayed in existence.
6080 29. eximstats updated to version 1.35
6081 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6082 1.35 - bugfix such that pie charts by volume are generated correctly
6084 30. The SPA authentication driver was not abandoning authentication and moving
6085 on to the next authenticator when an expansion was forced to fail,
6086 contradicting the general specification for all authenticators. Instead it
6087 was generating a temporary error. It now behaves as specified.
6089 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6090 (the order specifies the preference for clients). The order is now AES256,
6091 AES128, 3DES, ARCFOUR128.
6093 31. Small patch to Sieve code - explicitly set From: when generating an
6096 32. Exim crashed if a remote delivery caused a very long error message to be
6097 recorded - for instance if somebody sent an entire SpamAssassin report back
6098 as a large number of 550 error lines. This bug was coincidentally fixed by
6099 increasing the size of one of Exim's internal buffers (big_buffer) that
6100 happened as part of the Exiscan merge. However, to be on the safe side, I
6101 have made the code more robust (and fixed the comments that describe what
6104 33. Some experimental protocols are using DNS PTR records for new purposes. The
6105 keys for these records are domain names, not reversed IP addresses. The
6106 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6107 leaves it alone. Component reversal etc. now happens only for IP addresses.
6110 34. The host_aton() function is supposed to be passed a string that is known
6111 to be a valid IP address. However, in the case of IPv6 addresses, it was
6112 not checking this. This is a hostage to fortune. Exim now panics and dies
6113 if the condition is not met. A case was found where this could be provoked
6114 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6115 components; fortuitously, this particular loophole had already been fixed
6116 by change 4.50/55 or 4.44/33 above.
6118 If there are any other similar loopholes, the new check in host_aton()
6119 itself should stop them being exploited. The report I received stated that
6120 data on the command line could provoke the exploit when Exim was running as
6121 exim, but did not say which command line option was involved. All I could
6122 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6123 running as the user.
6126 35. There was a buffer overflow vulnerability in the SPA authentication code
6127 (which came originally from the Samba project). I have added a test to the
6128 spa_base64_to_bits() function which I hope fixes it.
6131 36. The daemon start-up calls getloadavg() while still root for those OS that
6132 need the first call to be done as root, but it missed one case: when
6133 deliver_queue_load_max is set with deliver_drop_privilege. This is
6134 necessary for the benefit of the queue runner, because there is no re-exec
6135 when deliver_drop_privilege is set.
6137 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6138 was in use, was not putting the data itself into the right store pool;
6139 consequently, it could be overwritten for a subsequent message in the same
6140 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6143 38. Sometimes the final signoff response after QUIT could fail to get
6144 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6145 < 0 before doing a fflush(). This bug looks as though it goes back to the
6146 introduction of TLS in release 3.20, but "sometimes" must have been rare
6147 because the tests only now provoked it.
6153 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6154 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6155 Now it returns an integer. A number of calls were still expecting a T/F
6156 return. Fortuitously, in all cases, the tests worked in OK situations,
6157 which is the norm. However, things would have gone wrong on any write
6158 failures on the smtp file descriptor. This function is used when sending
6159 messages over SMTP and also when doing verify callouts.
6161 2. When Exim is called to do synchronous delivery of a locally submitted
6162 message (the -odf or -odi options), it no longer closes stderr before doing
6165 3. Implemented the mua_wrapper option.
6167 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6169 5. Implemented the functions header_remove(), header_testname(),
6170 header_add_at_position(), and receive_remove_recipient(), and exported them
6173 6. If an ACL "warn" statement specified the addition of headers, Exim already
6174 inserted X-ACL-Warn: at the start if there was no header name. However, it
6175 was not making this test for the second and subsequent header lines if
6176 there were newlines in the string. This meant that an invalid header could
6177 be inserted if Exim was badly configured.
6179 7. Allow an ACL "warn" statement to add header lines at the start or after all
6180 the Received: headers, as well as at the end.
6182 8. Added the rcpt_4xx retry error code.
6184 9. Added postmaster_mailfrom=xxx to callout verification option.
6186 10. Added mailfrom=xxxx to the callout verification option, for verify=
6189 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6190 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6192 12. Inserted some casts to stop certain compilers warning when using pointer
6193 differences as field lengths or precisions in printf-type calls (mostly
6194 affecting debugging statements).
6196 13. Added optional readline() support for -be (dynamically loaded).
6198 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6199 same clock tick as a message's arrival, so that its received time was the
6200 same as the "first fail" time on the retry record, and that message
6201 remained on the queue past the ultimate address timeout, every queue runner
6202 would try a delivery (because it was past the ultimate address timeout) but
6203 after another failure, the ultimate address timeout, which should have then
6204 bounced the address, did not kick in. This was a "< instead of <=" error;
6205 in most cases the first failure would have been in the next clock tick
6206 after the received time, and all would be well.
6208 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6209 being recognized when the domain list was tested by the match_domain
6210 condition in an expansion string.
6212 16. Added the ${str2b64: operator.
6214 17. Exim was always calling setrlimit() to set a large limit for the number of
6215 processes, without checking whether the existing limit was already
6216 adequate. (It did check for the limit on file descriptors.) Furthermore,
6217 errors from getrlimit() and setrlimit() were being ignored. Now they are
6218 logged to the main and panic logs, but Exim does carry on, to try to do its
6219 job under whatever limits there are.
6221 18. Imported PCRE 5.0.
6223 19. Trivial typo in log message " temporarily refused connection" (the leading
6226 20. If the log selector return_path_on_delivery was set and an address was
6227 redirected to /dev/null, the delivery process crashed because it assumed
6228 that a return path would always be set for a "successful" delivery. In this
6229 case, the whole delivery is bypassed as an optimization, and therefore no
6232 21. Internal re-arrangement: the function for sending a challenge and reading
6233 a response while authentication was assuming a zero-terminated challenge
6234 string. It's now changed to take a pointer and a length, to allow for
6235 binary data in such strings.
6237 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6239 23. Exim was not respecting finduser_retries when seeking the login of the
6240 uid under which it was called; it was always trying 10 times. (The default
6241 setting of finduser_retries is zero.) Also, it was sleeping after the final
6242 failure, which is pointless.
6244 24. Implemented tls_on_connect_ports.
6246 25. Implemented acl_smtp_predata.
6248 26. If the domain in control=submission is set empty, Exim assumes that the
6249 authenticated id is a complete email address when it generates From: or
6250 Sender: header lines.
6252 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6253 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6254 chown and chgrp in /bin and hostname in /usr/bin.
6256 28. Exim was keeping the "process log" file open after each use, just as it
6257 does for the main log. This opens the possibility of it remaining open for
6258 long periods when the USR1 signal hits a daemon. Occasional processlog
6259 errors were reported, that could have been caused by this. Anyway, it seems
6260 much more sensible not to leave this file open at all, so that is what now
6263 29. The long-running daemon process does not normally write to the log once it
6264 has entered its main loop, and it closes the log before doing so. This is
6265 so that log files can straightforwardly be renamed and moved. However,
6266 there are a couple of unusual error situations where the daemon does write
6267 log entries, and I had neglected to close the log afterwards.
6269 30. The text of an SMTP error response that was received during a remote
6270 delivery was being truncated at 512 bytes. This is too short for some of
6271 the long messages that one sometimes sees. I've increased the limit to
6274 31. It is now possible to make retry rules that apply only when a message has a
6275 specific sender, in particular, an empty sender.
6277 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6278 it possible to be selective about when SMTP synchronization is enforced.
6280 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6282 32. Implemented hosts_connection_nolog.
6284 33. Added an ACL for QUIT.
6286 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6289 35. Added mailbox_size and mailbox_filecount to appendfile.
6291 36. Added control = no_multiline_responses to ACLs.
6293 37. There was a bug in the logic of the code that waits for the clock to tick
6294 in the case where the clock went backwards by a substantial amount such
6295 that the microsecond fraction of "now" was more than the microsecond
6296 fraction of "then" (but the whole seconds number was less).
6298 38. Added support for the libradius Radius client library this is found on
6299 FreeBSD (previously only the radiusclient library was supported).
6305 1. When certain lookups returned multiple values in the form name=value, the
6306 quoting of the values was not always being done properly. Specifically:
6307 (a) If the value started with a double quote, but contained no whitespace,
6309 (b) If the value contained whitespace other than a space character (i.e.
6310 tabs or newlines or carriage returns) it was not quoted.
6311 This fix has been applied to the mysql and pgsql lookups by writing a
6312 separate quoting function and calling it from the lookup code. The fix
6313 should probably also be applied to nisplus, ibase and oracle lookups, but
6314 since I cannot test any of those, I have not disturbed their existing code.
6316 2. A hit in the callout cache for a specific address caused a log line with no
6317 reason for rejecting RCPT. Now it says "Previous (cached) callout
6318 verification failure".
6320 3. There was an off-by-one bug in the queryprogram router. An over-long
6321 return line was truncated at 256 instead of 255 characters, thereby
6322 overflowing its buffer with the terminating zero. As well as fixing this, I
6323 have increased the buffer size to 1024 (and made a note to document this).
6325 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6326 when Exim is waiting for an SMTP response from a remote server, Exim
6327 restarts its select() call on the socket, thereby resetting its timeout.
6328 This is not a problem when such interrupts are rare. Somebody set up a cron
6329 job to run exiwhat every 2 minutes, which is less than the normal select()
6330 timeout (5 or 10 minutes). This meant that the select() timeout never
6331 kicked in because it was always reset. I have fixed this by comparing the
6332 time when an interrupt arrives with the time at the start of the first call
6333 to select(). If more time than the timeout has elapsed, the interrupt is
6334 treated as a timeout.
6336 5. Some internal re-factoring in preparation for the addition of Sieve
6337 extensions (by MH). In particular, the "personal" test is moved to a
6338 separate function, and given an option for scanning Cc: and Bcc: (which is
6339 not set for Exim filters).
6341 6. When Exim created an email address using the login of the caller as the
6342 local part (e.g. when creating a From: or Sender: header line), it was not
6343 quoting the local part when it contained special characters such as @.
6345 7. Installed new OpenBSD configuration files.
6347 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6348 try to make them clearer.
6350 9. Callout options, other than the timeout value, were being ignored when
6351 verifying sender addresses in header lines. For example, when using
6353 verify = header_sender/callout=no_cache
6355 the cache was (incorrectly) being used.
6357 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6358 only the Cygwin environment.
6360 11. When return_path_on_delivery was set as a log selector, if different remote
6361 addresses in the same message used different return paths and parallel
6362 remote delivery occurred, the wrong values would sometimes be logged.
6363 (Whenever a remote delivery process finished, the return path value from
6364 the most recently started remote delivery process was logged.)
6366 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6367 lines when AUTH and/or TLS are in use. This is the "received protocol"
6368 field. Exim used to use "asmtp" for authenticated SMTP, without any
6369 indication (in the protocol name) for TLS use. Now it follows the RFC and
6370 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6371 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6372 names appear in log lines as well as in Received: header lines.
6374 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6375 extensions, and comparison tests, and to fix some bugs.
6377 14. Changes to the "personal" filter test:
6379 (1) The test was buggy in that it was just doing the equivalent of
6380 "contains" tests on header lines. For example, if a user's address was
6381 anne@some.where, the "personal" test would incorrectly be true for
6383 To: susanne@some.where
6385 This test is now done by extracting each address from the header in turn,
6386 and checking the entire address. Other tests that are part of "personal"
6387 are now done using regular expressions (for example, to check local parts
6388 of addresses in From: header lines).
6390 (2) The list of non-personal local parts in From: addresses has been
6391 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6392 taken from the Sieve specification recommendations.
6394 (3) If the message contains any header line starting with "List-" it is
6395 treated as non-personal.
6397 (4) The test for "circular" in the Subject: header line has been removed
6398 because it now seems ill-conceived.
6400 15. Minor typos in src/EDITME comments corrected.
6402 16. Installed latest exipick from John Jetmore.
6404 17. If headers_add on a router specified a text string that was too long for
6405 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6406 of string_sprintf() is now avoided.
6408 18. $message_body_size was not set (it was always zero) when running the DATA
6409 ACL and the local_scan() function.
6411 19. For the "mail" command in an Exim filter, no default was being set for
6412 the once_repeat time, causing a random time value to be used if "once" was
6413 specified. (If the value happened to be <= 0, no repeat happened.) The
6414 default is now 0s, meaning "never repeat". The "vacation" command was OK
6415 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6416 (I found it when inspecting the code).
6418 20. There is now an overall timeout for performing a callout verification. It
6419 defaults to 4 times the callout timeout, which applies to individual SMTP
6420 commands during the callout. The overall timeout applies when there is more
6421 than one host that can be tried. The timeout is checked before trying the
6422 next host. This prevents very long delays if there are a large number of
6423 hosts and all are timing out (e.g. when the network connections are timing
6424 out). The value of the overall timeout can be changed by specifying an
6425 additional sub-option for "callout", called "maxwait". For example:
6427 verify = sender/callout=5s,maxwait=20s
6429 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6430 to the end before writing, but this should make it even safer).
6432 22. Exim was forgetting that it had advertised PIPELINING for the second and
6433 subsequent messages on an SMTP connection. It was also not resetting its
6434 memory on STARTTLS and an internal HELO.
6436 23. When Exim logs an SMTP synchronization error within a session, it now
6437 records whether PIPELINING has been advertised or not.
6439 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6440 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6441 rather than long int.
6443 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6445 26. Added the never_mail option to autoreply.
6451 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6452 crash if the getsockname() call failed; this can happen if a connection is
6453 closed very soon after it is established. The problem was simply in the
6454 order in which certain operations were done, causing Exim to try to write
6455 to the SMTP stream before it had set up the file descriptor. The bug has
6456 been fixed by making things happen in the correct order.
6462 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6463 before the connection was closed, thus losing the rejection response.
6465 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6466 some early Solaris releases, but causes trouble in current releases where
6467 socklen_t is defined.
6469 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6472 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6475 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6476 that releasing the top of it at the end releases what was used for sub-
6477 expansions (unless the block got too big). However, discard this block if
6478 the first thing is a variable or header, so that we can use its block when
6479 it is dynamic (useful for very large $message_headers, for example).
6481 6. Lookups now cache *every* query, not just the most recent. A new, separate
6482 store pool is used for this. It can be recovered when all lookup caches are
6483 flushed. Lookups now release memory at the end of their result strings.
6484 This has involved some general refactoring of the lookup sources.
6486 7. Some code has been added to the store_xxx() functions to reduce the amount
6487 of flapping under certain conditions.
6489 8. log_incoming_interface used to affect only the <= reception log lines. Now
6490 it causes the local interface and port to be added to several more SMTP log
6491 lines, for example "SMTP connection from", and rejection lines.
6493 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6495 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6497 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6498 could be overwritten at the end of the current message (or the start of a
6499 new message if it was set in a HELO ACL). The value is now preserved for
6500 the duration of the SMTP connection.
6502 12. If a transport had a headers_rewrite setting, and a matching header line
6503 contained an unqualified address, that address was qualified, even if it
6504 did not match any rewriting rules. The underlying bug was that the values
6505 of the flags that permit the existence of unqualified sender and recipient
6506 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6507 non-local messages, and by -bnq for local messages) were not being
6508 preserved with the message after it was received.
6510 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6511 "next input=" as part of the text comprising the host identity instead of
6512 the correct text. The code was using the same buffer for two different
6513 strings. However, depending on which order the printing function evaluated
6514 its arguments, the bug did not always show up. Under Linux, for example, my
6515 test suite worked just fine.
6517 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6518 doesn't work with some older versions of Perl. It has been changed to "my",
6519 which in any case is probably the better facility to use.
6521 15. A really picky compiler found some instances of statements for creating
6522 error messages that either had too many or two few arguments for the format
6525 16. The size of the buffer for calls to the DNS resolver has been increased
6526 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6527 for addresses that have a lot of PTR records. This alleviates a problem; it
6528 does not fully solve it.
6530 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6531 buffer now truncates the list and logs the incident, which is the same
6532 action as happens when Exim is looking up a host name and its aliases.
6533 Previously in this situation something unpredictable would happen;
6534 sometimes it was "internal error: store_reset failed".
6536 18. If a server dropped the connection unexpectedly when an Exim client was
6537 using GnuTLS and trying to read a response, the client delivery process
6538 crashed while trying to generate an error log message.
6540 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6541 string, for example:
6543 warn message = H1: something\nH2: something
6545 the text was added as a single header line from Exim's point of view
6546 though it ended up OK in the delivered message. However, searching for the
6547 second and subsequent header lines using $h_h2: did not work. This has been
6548 fixed. Similarly, if a system filter added multiple headers in this way,
6549 the routers could not see them.
6551 20. Expanded the error message when iplsearch is called with an invalid key to
6552 suggest using net-iplsearch in a host list.
6554 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6555 ACLs are no longer actually imposed (and a message to that effect is
6558 22. If a "gecos" field in a passwd entry contained escaped characters, in
6559 particular, if it contained a \" sequence, Exim got it wrong when building
6560 a From: or a Sender: header from that name. A second bug also caused
6561 incorrect handling when an unquoted " was present following a character
6562 that needed quoting.
6564 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6565 was not being matched caselessly.
6567 24. Arranged for all hyphens in the exim.8 source to be escaped with
6570 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6571 properly. Recipient callout cache records were still being keyed to include
6572 the sender, even when use_sender was set false. This led to far more
6573 callouts that were necessary. The sender is no longer included in the key
6574 when use_sender is false.
6576 26. Added "control = submission" modifier to ACLs.
6578 27. Added the ${base62d: operator to decode base 62 numbers.
6580 28. dnsdb lookups can now access SRV records.
6582 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6583 the configuration file.
6585 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6586 (-v) mode. This makes the output for a verbose queue run more intelligible.
6588 31. Added a use_postmaster feature to recipient callouts.
6590 32. Added the $body_zerocount variable, containing the number of binary zero
6591 bytes in the message body.
6593 33. The time of last modification of the "new" subdirectory is now used as the
6594 "mailbox time last read" when there is a quota error for a maildir
6597 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6599 35. Added +ignore_unknown as a special item in host lists.
6601 36. Code for decoding IPv6 addresses in host lists is now included, even if
6602 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6603 address was recognized as an IP address, but was then not correctly decoded
6604 into binary, causing unexpected and incorrect effects when compared with
6611 1. Very minor rewording of debugging text in manualroute to say "list of
6612 hosts" instead of "hostlist".
6614 2. If verify=header_syntax was set, and a header line with an unqualified
6615 address (no domain) and a large number of spaces between the end of the
6616 name and the colon was received, the reception process suffered a buffer
6617 overflow, and (when I tested it) crashed. This was caused by some obsolete
6618 code that should have been removed. The fix is to remove it!
6620 3. When running in the test harness, delay a bit after writing a bounce
6621 message to get a bit more predictability in the log output.
6623 4. Added a call to search_tidyup() just before forking a reception process. In
6624 theory, someone could use a lookup in the expansion of smtp_accept_max_
6625 per_host which, without the tidyup, could leave open a database connection.
6627 5. Added the variables $recipient_data and $sender_data which get set from a
6628 lookup success in an ACL "recipients" or "senders" condition, or a router
6629 "senders" option, similar to $domain_data and $local_part_data.
6631 6. Moved the writing of debug_print from before to after the "senders" test
6634 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6635 problems for message scanning, either using a data ACL, or using
6636 local_scan() because the Received: header was not generated till after they
6637 were called (in order to set the time as the time of reception completion).
6638 I have revised the way this works. The header is now generated after the
6639 body is received, but before the ACL or local_scan() are called. After they
6640 are run, the timestamp in the header is updated.
6646 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6647 before starting a queue runner without re-exec. This happened only when
6648 deliver_drop_privilege was set or when the Exim user was set to root. The
6649 effect of the bug was that timeouts during subsequent deliveries caused
6650 crashes instead of being properly handled. The handler is now left at its
6651 default (and expected) setting.
6653 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6654 message, again when deliver_drop_privilege is set or Exim is run as root.
6655 The bug described in (1) was not present in this case, but the tidying up
6656 of the other signals was missing. I have made the two cases consistent.
6658 3. The ignore_target_hosts setting on a manualroute router was being ignored
6659 for hosts that were looked up using the /MX notation.
6661 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6664 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6665 operated on the sender address. After changing the $sender_address to <>
6666 for the sender address verify, Exim was re-instated it as the original
6667 (before rewriting) address, but remembering that it had rewritten it, so it
6668 wasn't rewriting it again. This bug also had the effect of breaking the
6669 sender address verification caching when the sender address was rewritten.
6671 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6672 This has been changed so that if the ip literal address matches
6673 ignore_target_hosts, the router declines.
6675 7. Added expansion conditions match_domain, match_address, and match_local_
6676 part (NOT match_host).
6678 8. The placeholder for the Received: header didn't have a length field set.
6680 9. Added code to Exim itself and to exim_lock to test for a specific race
6681 condition that could lead to file corruption when using MBX delivery. The
6682 issue is with the lockfile that is created in /tmp. If this file is removed
6683 after a process has opened it but before that process has acquired a lock,
6684 there is the potential for a second process to recreate the file and also
6685 acquire a lock. This could lead to two Exim processes writing to the file
6686 at the same time. The added code performs the same test as UW imapd; it
6687 checks after acquiring the lock that its file descriptor still refers to
6688 the same named file.
6690 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6691 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6694 11. Added the smtp_active_hostname option. If used, this will typically be made
6695 to depend on the incoming interface address. Because $interface_address is
6696 not set up until the daemon has forked a reception process, error responses
6697 that can happen earlier (such as "too many connections") no longer contain
6700 12. If an expansion in a condition on a "warn" statement fails because a lookup
6701 defers, the "warn" statement is abandoned, and the next ACL statement is
6702 processed. Previously this caused the whole ACL to be aborted.
6704 13. Added the iplsearch lookup type.
6706 14. Added ident_timeout as a log selector.
6708 15. Added tls_certificate_verified as a log selector.
6710 16. Added a global option tls_require_ciphers (compare the smtp transport
6711 option of the same name). This controls incoming TLS connections.
6713 17. I finally figured out how to make tls_require_ciphers do a similar thing
6714 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6715 before starting the TLS session.
6717 18. Tabs are now shown as \t in -bP output.
6719 19. If the log selector return_path_on_delivery was set, Exim crashed when
6720 bouncing a message because it had too many Received: header lines.
6722 20. If two routers both had headers_remove settings, and the first one included
6723 a superfluous trailing colon, the final name in the first list and the
6724 first name in the second list were incorrectly joined into one item (with a
6725 colon in the middle).
6731 1. Added -C and -D options to the exinext utility, mainly to make it easier
6732 to include in the automated testing, but these could be helpful when
6733 multiple configurations are in use.
6735 2. The exinext utility was not formatting the output nicely when there was
6736 an alternate port involved in the retry record key, nor when there was a
6737 message id as well (for retries that were specific to a specific message
6738 and a specific host). It was also confused by IPv6 addresses, because of
6739 the additional colons they contain. I have fixed the IPv4 problem, and
6740 patched it up to do a reasonable job for IPv6.
6742 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6743 delivery, the log line now contains "pipelined" if PIPELINING was used.
6745 4. An SMTP transport process used to panic and die if the bind() call to set
6746 an explicit outgoing interface failed. This has been changed; it is now
6747 treated in the same way as a connect() failure.
6749 5. A reference to $sender_host_name in the part of a conditional expansion
6750 that was being skipped was still causing a DNS lookup. This no longer
6753 6. The def: expansion condition was not recognizing references to header lines
6754 that used bh_ and bheader_.
6756 7. Added the _cache feature to named lists.
6758 8. The code for checking quota_filecount in the appendfile transport was
6759 allowing one more file than it should have been.
6761 9. For compatibility with Sendmail, the command line option
6769 and sets the incoming protocol and host name (for trusted callers). The
6770 host name and its colon can be omitted when only the protocol is to be set.
6771 Note the Exim already has two private options, -pd and -ps, that refer to
6772 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6773 "s", but I don't think that's a major issue.
6775 10. A number of refactoring changes to the code, none of which should affect
6778 (a) The number of logging options was getting close to filling up the
6779 32-bit word that was used as a bit map. I have split them into two classes:
6780 those that are passed in the argument to log_write(), and those that are
6781 only ever tested independently outside of that function. These are now in
6782 separate 32-bit words, so there is plenty of room for expansion again.
6783 There is no change in the user interface or the logging behaviour.
6785 (b) When building, for example, log lines, the code previously used a
6786 macro that called string_cat() twice, in order to add two strings. This is
6787 not really sufficiently general. Furthermore, there was one instance where
6788 it was actually wrong because one of the argument was used twice, and in
6789 one call a function was used. (As it happened, calling the function twice
6790 did not affect the overall behaviour.) The macro has been replaced by a
6791 function that can join an arbitrary number of extra strings onto a growing
6794 (c) The code for expansion conditions now uses a table and a binary chop
6795 instead of a serial search (which was left over from when there were very
6796 few conditions). Also, it now recognizes conditions like "pam" even when
6797 the relevant support is not compiled in: a suitably worded error message is
6798 given if an attempt is made to use such a condition.
6800 11. Added ${time_interval:xxxxx}.
6802 12. A bug was causing one of the ddress fields not to be passed back correctly
6803 from remote delivery subprocesses. The field in question was not being
6804 subsequently used, so this caused to problems in practice.
6806 13. Added new log selectors queue_time and deliver_time.
6808 14. Might have fixed a bug in maildirsizefile handling that threw up
6809 "unexpected character" debug warnings, and recalculated the data
6810 unnecessarily. In any case, I expanded the warning message to give more
6813 15. Added the message "Restricted characters in address" to the statements in
6814 the default ACL that block characters like @ and % in local parts.
6816 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6817 Three changes have been made:
6819 (a) There was a serious bug; a negative response to MAIL caused the whole
6820 recipient domain to be cached as invalid, thereby blocking all messages
6821 to all local parts at the same domain, from all senders. This bug has
6822 been fixed. The domain is no longer cached after a negative response to
6823 MAIL if the sender used is not empty.
6825 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6828 (c) A new callout option, "use_sender" has been added for people who want
6829 the modified behaviour.
6835 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6838 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6839 indeed breaks things for older releases.
6841 3. Added additional logging to the case where there is a problem reading data
6842 from a filter that is running in a subprocess using a pipe, in order to
6843 try to track down a specific problem.
6845 4. Testing facility fudge: when running in the test harness and attempting
6846 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6847 getting "No route to host". Convert this to a timeout.
6849 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6852 6. Some OS don't have socklen_t but use size_t instead. This affects the
6853 fifth argument of getsockopt() amongst other things. This is now
6854 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6855 can be set for individual OS. I have set it for SunOS5, OSF1, and
6856 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6857 some earlier ones do not.
6859 7. Change 4.30/15 was not doing the test caselessly.
6861 8. The standard form for an IPv6 address literal was being rejected by address
6862 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6863 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6864 this, as well as the form without the "IPv6" on the front (but only when
6865 address literals are enabled, of course).
6867 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6869 10. Exim crashed if a message with an empty sender address specified by -f
6870 encountered a router with an errors_to setting. This could be provoked only
6871 by a command such as
6875 where an empty string was supplied; "<>" did not hit this bug.
6877 11. Installed PCRE release 4.5.
6879 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6880 remained set. It is now erased.
6882 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6883 times from message ids (which are base 36 rather than the normal 62).
6885 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6886 were being counted as actual protocol errors, and logged if the log
6887 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6888 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6889 and DATA following a set of rejected RCPTs do not count as protocol errors.
6890 In other words, Exim assumes they were pipelined, though this may not
6891 actually be the case. Of course, in all cases the client gets an
6892 appropriate error code.
6894 15. If a lookup fails in an ACL condition, a message about the failure may
6895 be available; it is used if testing the ACL cannot continue, because most
6896 such messages specify what the cause of the deferral is. However, some
6897 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6898 that caused an old message to be retained and used if a later statement
6899 caused a defer, replacing the real cause of the deferral.
6901 16. If an IP address had so many PTR records that the DNS lookup buffer
6902 was not large enough to hold them, Exim could crash while trying to process
6903 the truncated data. It now detects and logs this case.
6905 17. Further to 4.21/58, another change has been made: if (and only if) the
6906 first line of a message (the first header line) ends with CRLF, a bare LF
6907 in a subsequent header line has a space inserted after it, so as not to
6908 terminate the header.
6910 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6911 unnecessarily, used atoi() instead of strtol(), and didn't check the
6912 termination when getting file sizes from file names by regex.
6914 19. Completely re-implemented the support for maildirsize files, in the light
6915 of a number of problems with the previous contributed implementation
6916 (4.30/29). In particular:
6918 . If the quota is zero, the maildirsize file is maintained, but no quota is
6921 . If the maildir directory does not exist, it is created before any attempt
6922 to write a maildirsize file.
6924 . The quota value in the file is just a cache; if the quota is changed in
6925 the transport, the new value overrides.
6927 . A regular expression is available for excluding directories from the
6930 20. The autoreply transport checks the characters in options that define the
6931 message's headers; it allows continued headers, but it was checking with
6932 isspace() after an embedded newline instead of explicitly looking for a
6935 21. If all the "regular" hosts to which an address was routed had passed their
6936 expiry times, and had not reached their retry times, the address was
6937 bounced, even if fallback hosts were defined. Now Exim should go on to try
6940 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6941 equivalent code in the SMTP transport. Some hosts send humungous responses
6942 to HELO/EHLO, more than 1024 it seems.
6944 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6945 gives compiler warnings in some environments. I've now done it "properly",
6948 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6949 (because of problems with the built-in one) was declared to return uschar *
6950 instead of char *, causing compiler failure.
6952 25. Fixed a file descriptor leak when processing alias/forward files.
6954 26. Fixed a minor format string issue in dbfn.c.
6956 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6958 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6959 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6962 29. When a sender address is verified, it is cached, to save repeating the test
6963 when there is more than one recipient in a message. However, when the
6964 verification involves a callout, it is possible for different callout
6965 options to be set for different recipients. It is too complicated to keep
6966 track of this in the cache, so now Exim always runs a verification when a
6967 callout is required, relying on the callout cache for the optimization.
6968 The overhead is duplication of the address routing, but this should not be
6971 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6972 to be verified with callout=postmaster, and the main callout worked but the
6973 postmaster check failed, the verification correctly failed. However, if a
6974 subsequent RCPT command asked for sender verification *without* the
6975 postmaster check, incorrect caching caused this verification also to fail,
6978 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6979 it was not caching the DNS options (qualify_single, search_parents) that
6980 were used when the lookup failed. A subsequent lookup with different
6981 options therefore always gave the same answer, though there were cases
6982 where it should not have. (Example: a "domains = !$mx_any" option on a
6983 dnslookup router: the "domains" option is always processed without any
6984 widening, but the router might have qualify_single set.) Now Exim uses the
6985 cached value only when the same options are set.
6987 32. Added John Jetmore's "exipick" utility to the distribution.
6989 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6990 than a timeout (e.g. a certificate is required, and is not provided), an
6991 Exim server now closes the connection immediately. Previously it waited for
6992 the client to close - but if the client is SSL, it seems that they each
6993 wait for each other, leading to a delay before one of them times out.
6995 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6996 maintained 0.8.x compatibility because I don't think many are using it, and
6997 it is clearly obsolete.
6999 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7002 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7003 client certificate was expired. A simple patch fixes this, though I don't
7004 understand the full logic of why the verify callback is called multiple
7007 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7008 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7009 which causes problems with some clients (such as the Certicom SSL Plus
7010 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7011 disables the coutermeasure allowing Eudora to connect."
7013 38. Exim was not checking that a write() to a log file succeeded. This could
7014 lead to Bad Things if a log got too big, in particular if it hit a file
7015 size limit. Exim now panics and dies if it cannot write to a log file, just
7016 as it does if it cannot open a log file.
7018 39. Modified OS/Makefile-Linux so that it now contains
7020 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7022 The two -D definitions ensure that Exim is compiled with large file
7023 support, which makes it possible to handle log files that are bigger than
7026 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7027 instance) a domain was checked against a named list that involved a lookup,
7028 causing $domain_data to be set, then another domain was checked against the
7029 same list, then the first domain was re-checked, the value of $domain_data
7030 after the final check could be wrong. In particular, if the second check
7031 failed, it could be set empty. This bug probably also applied to
7034 41. The strip_trailing_dot option was not being applied to the address given
7035 with the -f command-line option.
7037 42. The code for reading a message's header from the spool was incrementing
7038 $received_count, but never initializing it. This meant that the value was
7039 incorrect (doubled) while delivering a message in the same process in which
7040 it was received. In the most common configuration of Exim, this never
7041 happens - a fresh exec is done - but it can happen when
7042 deliver_drop_privilege is set.
7044 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7045 it now includes up to 150 characters of the unexpected data in the log
7048 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7049 and building data strings. The size of both of these buffers was 10 000
7050 bytes - far larger than anybody would *ever* want, thought I. Needless to
7051 say, somebody hit the limit. I have increased the maximum line length to
7052 20 000 and the maximum data length of concatenated lines to 100 000. I have
7053 also fixed two bugs, because there was no checking on these buffers. Tsk,
7054 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7055 buffer is too small.
7057 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7058 lsearch lookups. Now it does.
7060 46. When parsing a route_list item in a manualroute router, a fixed-length
7061 buffer was used for the list of hosts. I made this 1024 bytes long,
7062 thinking that nobody would ever have a list of hosts that long. Wrong.
7063 Somebody had a whole pile of complicated expansion conditions, and the
7064 string was silently truncated, leading to an expansion error. It turns out
7065 that it is easier to change to an unlimited length (owing to other changes
7066 that have happened since this code was originally written) than to build
7067 structure for giving a limitation error. The length of the item that
7068 expands into the list of hosts is now unlimited.
7070 47. The lsearch lookup could not handle data where the length of text line was
7071 more than 4095 characters. Such lines were truncated, leading to shortened
7072 data being returned. It should now handle lines of any length.
7074 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7075 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7078 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7079 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7080 no longer happens in comment lines. A list of replacements is now placed
7081 at the head of all of the source files, except those whose only change is
7082 to replace PERL_COMMAND in the very first #! line.
7084 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7085 messages on the queue, with a bottom-up merge sort, using code contributed
7086 by Michael Haardt. This should make operations like -bp somewhat faster on
7087 large queues. It won't affect queue runners, except when queue_run_in_order
7090 51. Installed eximstats 1.31 in the distribution.
7092 52. Added support for SRV lookups to the dnslookup router.
7094 53. If an ACL referred to $message_body or $message_body_end, the value was not
7095 reset for any messages that followed in the same SMTP session.
7097 54. The store-handling optimization for building very long strings was not
7098 differentiating between the different store pools. I don't think this
7099 actually made any difference in practice, but I've tidied it.
7101 55. While running the routers to verify a sender address, $sender_address
7102 was still set to the sender address. This is wrong, because when routing to
7103 send a bounce to the sender, it would be empty. Therefore, I have changed
7104 it so that, while verifying a sender address, $sender_address is set to <>.
7105 (There is no change to what happens when verifying a recipient address.)
7107 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7108 target A or AAAA records (if not already returned) without resetting the
7109 qualify_single or search_parents options of the DNS resolver. These are
7110 inappropriate in this case because the targets of MX and SRV records must
7111 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7112 target that, when qualified, matched something in the local domain. These
7113 two options are now turned off when doing these lookups.
7115 57. It seems that at least some releases of Reiserfs (which does not have the
7116 concept of a fixed number of inodes) returns zero and not -1 for the
7117 number of available inodes. This interacted badly with check_spool_inodes,
7118 which assumed that -1 was the "no such thing" setting. What I have done is
7119 to check that the total number of inodes is greater than zero before doing
7120 the test of how many are available.
7122 58. When a "warn" ACL statement has a log_message modifier, the message is
7123 remembered, and not repeated. This is to avoid a lot of repetition when a
7124 message has many recipients that cause the same warning to be written.
7125 However, Exim was preserving the list of already written lines for an
7126 entire SMTP session, which doesn't seem right. The memory is now reset if a
7127 new message is started.
7129 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7130 debugging output unless log_rewrite was also set.
7132 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7133 of (int)(handle) when we know that handle contains (void *)(-1).
7135 61. The Exim daemon panic-logs an error return when it closes the incoming
7136 connection. However "connection reset by peer" seems to be common, and
7137 isn't really an error worthy of noting specially, so that particular error
7140 62. When Exim is trying to find all the local interfaces, it used to panic and
7141 die if the ioctl to get the interface flags failed. However, it seems that
7142 on at least one OS (Solaris 9) it is possible to have an interface that is
7143 included in the list of interfaces, but for which you get a failure error
7144 for this call. This happens when the interface is not "plumbed" into a
7145 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7146 failure of the "get flags" call assumes that the interface is down.
7148 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7149 makes life easier for people who are doing arithmetic on fields extracted
7150 from dates, where you often get leading zeros that should not be
7151 interpreted as octal.
7153 64. Added qualify_domain to the redirect router, to override the global
7156 65. If a pathologically long header line contained very many addresses (the
7157 report of this problem mentioned 10 000) and each of them was rewritten,
7158 Exim could use up a very large amount of memory. (It kept on making new
7159 copies of the header line as it rewrote, and never released the old ones.)
7160 At the expense of a bit more processing, the header rewriting function has
7161 been changed so that it no longer eats memory in this way.
7163 66. The generation of the Received: header has been moved from the time that a
7164 message starts to be received, to the time that it finishes. The timestamp
7165 in the Received: header should now be very close to that of the <= log
7166 line. There are two side-effects of this change:
7168 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7169 logged header lines no longer include the local Received: line, because
7170 it has not yet been created. The same applies to a copy of the message
7171 that is returned to a non-SMTP sender when a message is rejected.
7173 (b) When a filter file is tested using -bf, no additional Received: header
7174 is added to the test message. After some thought, I decided that this
7177 This change does not affect the value of $received_for. It is still set
7178 after address rewriting, but before local_scan() is called.
7180 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7182 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7183 gave an unhelpful panic error message, and a defer error. I have managed to
7184 change this behaviour so that it now rejects any supplied certificate,
7185 which seems right, as the list of acceptable certificates is empty.
7187 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7188 gave an unhelpful defer error. I have not managed to make this reject any
7189 supplied certificates, but the error message it gives is "no certificate
7190 supplied", which is not helpful.
7192 70. exigrep's output now also includes lines that are not associated with any
7193 message, but which match the given pattern. Implemented by a patch from
7194 Martin Sluka, which also tidied up the Perl a bit.
7196 71. Recipient callout verification, like sender verification, was using <> in
7197 the MAIL FROM command. This isn't really the right thing, since the actual
7198 sender may affect whether the remote host accepts the recipient or not. I
7199 have changed it to use the actual sender in the callout; this means that
7200 the cache record is now keyed on a recipient/sender pair, not just the
7201 recipient address. There doesn't seem to be a real danger of callout loops,
7202 since a callout by the remote host to check the sender would use <>.
7203 [SEE ABOVE: changed after hitting problems.]
7205 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7206 temporary errors. However, in the case of such a code being given after
7207 the end of a data transmission (i.e. after ".") Exim was failing to write
7208 a retry record for the message. (Yes, there was some broken host that was
7209 actually sending 8xx at this point.)
7211 73. An unknown lookup type in a host list could cause Exim to panic-die when
7212 the list was checked. (An example that provoked this was putting <; in the
7213 middle of a list instead of at the start.) If this happened during a DATA
7214 ACL check, a -D file could be left lying around. This kind of configuration
7215 error no longer causes Exim to die; instead it causes a defer error. The
7216 incident is still logged to the main and panic logs.
7218 74. Buglet left over from Exim 3 conversion. The message "too many messages
7219 in one connection" was written to the rejectlog but not the mainlog, except
7220 when address rewriting (yes!) was being logged.
7222 75. Added write_rejectlog option.
7224 76. When a system filter was run not as root (that is, when system_filter_user
7225 was set), the values of the $n variables were not being returned to the
7226 main process; thus, they were not subsequently available in the $sn
7229 77. Added +return_path_on_delivery log selector.
7231 78. A connection timeout was being treated differently from recipients deferred
7232 when testing hosts_max_try with a message that was older than the host's
7233 retry timeout. (The host should not be counted, thus allowing all hosts to
7234 be tried at least once before bouncing.) This may have been the cause of an
7235 occasionally reported bug whereby a message would remain on the queue
7236 longer than the retry timeout, but would be bounced if a delivery was
7237 forced. I say "may" because I never totally pinned down the problem;
7238 setting up timeout/retry tests is difficult. See also the next item.
7240 79. The ultimate address timeout was not being applied to errors that involved
7241 a combination of host plus message (for example, a timeout on a MAIL
7242 command). When an address resolved to a number of possible hosts, and they
7243 were not all tried for each delivery (e.g. because of hosts_max_try), a
7244 message could remain on the queue longer than the retry timeout.
7246 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7249 81. Fixed an obscure SMTP outgoing bug which required at least the following
7250 conditions: (a) there was another message waiting for the same server;
7251 (b) the server returned 5xx to all RCPT commands in the first message so
7252 that the message was not completed; (c) the server dropped the connection
7253 or gave a negative response to the RSET that Exim sends to abort the
7254 transaction. The observed case was a dropped connection after DATA that had
7255 been sent in pipelining mode. That is, the server had advertised PIPELINING
7256 but was not implementing it correctly. The effect of the bug was incorrect
7257 behaviour, such as trying another host, and this could lead to a crash.
7263 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7264 and daemon.c were passed as pointers to ints; they should have been
7265 pointers to socklen_t variables (which are typically unsigned ints).
7267 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7270 3. Fixed a really odd bug that affected only the testing scheme; patching a
7271 certain fixed string in the binary changed the value of another string that
7272 happened to be identical to the end of the original first string.
7274 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7275 name", it returns that address as the IP address. On some operating
7276 systems (e.g. Solaris), it also passes back the IP address string as the
7277 "host name". However, on others (e.g. Linux), it passes back an empty
7278 string. Exim wasn't checking for this, and was changing the host name to an
7279 empty string, assuming it had been canonicalized.
7281 5. Although rare, it is permitted to have more than one PTR record for a given
7282 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7283 all the names associated with an address, because they do in Solaris.
7284 However, it seems that they do not in Linux for data that comes from the
7285 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7286 I found this out when I moved to a new Linux workstation and tried to run
7287 the Exim test suite.
7289 To get round this problem I have changed the code so that it now does its
7290 own call to the DNS to look up PTR records when searching for a host name.
7291 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7292 addresses that are only in /etc/hosts are still found.
7294 This behaviour is, however, controlled by an option called host_lookup_
7295 order, which defaults to "bydns:byaddr". If people want to use the other
7296 order, or indeed, just use one or the other means of lookup, they can
7297 specify it in this variable.
7299 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7300 some operating systems, this comes back from gethostbyaddr() as an empty
7301 string, and this is what Exim used to test for. However, it seems that in
7302 other systems, "." is yielded. Exim now tests for this case too.
7304 7. The values of check_spool_space and check_log_space are now held internally
7305 as a number of kilobytes instead of an absolute number of bytes. If a
7306 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7307 kilobyte. This means that much larger values can be stored.
7309 8. Exim monitor: an attempt to get the action menu when not actually pointing
7310 at a message produces an empty menu entitled "No message selected". This
7311 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7312 no entries in it ("Shell widget menu has zero width and/or height"). So I
7313 have added a single, blank menu entry in this case.
7315 9. Added ${quote_local_part.
7317 10. MIME decoding is now applied to the contents of Subject: header lines when
7320 11. Now that a reference to $sender_host_address automatically causes a reverse
7321 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7322 host lookup before query-style lookups in lists that might use this
7323 variable. This has therefore been abolished, and the "net-" prefix is no
7324 longer necessary for query-style lookups.
7326 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7327 have been a typo for LFLAGS, so it has been changed.
7329 13. The install script calls Exim with "-C /dev/null" in order to find the
7330 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7331 to be output. However, since Exim outputs its version number before the
7332 error, it didn't break the script. It just looked ugly. I fixed this by
7333 always allowing "-C /dev/null" if the caller is root.
7335 14. Ignore overlarge ACL variable number when reading spool file - insurance
7336 against a later release with more variables having written the file.
7338 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7339 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7340 this, as well as the form without the "IPv6" on the front.
7342 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7343 OS/Makefile-Darwin file.
7345 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7346 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7348 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7349 variable. However, some OS use ints for inodes. I've added cast to long int
7350 to get rid of the compiler warning.
7352 19. I had forgotten to lock out "/../" in configuration file names when
7353 ALT_CONFIG_PREFIX was set.
7355 20. Routers used for verification do not need to specify transports. However,
7356 if such a router generated a host list, and callout was configured, Exim
7357 crashed, because it could not find a port number from the (non-existent)
7358 transport. It now assumes port 25 in this circumstance.
7360 21. Added the -t option to exigrep.
7362 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7363 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7364 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7365 them is set, LOOKUP_LSEARCH is forced.
7367 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7369 24. Added sender and host information to the "rejected by local_scan()" log
7370 line; previously there was no indication of these.
7372 25. Added .include_if_exists.
7374 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7375 Linux. It turns out that not all file systems support this. Apparently some
7376 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7377 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7378 not supported on the file descriptor, is now ignored when Exim is trying to
7379 sync a directory. This applies only to Linux.
7381 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7383 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7384 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7387 29. Added support for maildirsize files from supplied patch (modified a bit).
7389 30. The use of :fail: followed by an empty string could lead Exim to respond to
7390 sender verification failures with (e.g.):
7392 550 Verification failed for <xxx>
7393 550 Sender verify failed
7395 where the first response line was missing the '-' that indicates it is not
7396 the final line of the response.
7398 31. The loop for finding the name of the user that called Exim had a hardwired
7399 limit of 10; it now uses the value of finduser_retries, which is used for
7400 all other user lookups.
7402 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7405 33. Exim was neglecting to zero errno before one call of strtol() when
7406 expanding a string and expecting an integer value. On some systems this
7407 resulted in spurious "integer overflow" errors. Also, it was casting the
7408 result into an int without checking.
7410 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7411 did not work. The code looks as if it has *never* worked, though it appears
7412 to have been documented since at least release 1.62. I have made it work.
7414 35. The "timeout_DNS" error in retry rules, also documented since at least
7415 1.62, also never worked. As it isn't clear exactly what this means, and
7416 clearly it isn't a major issue, I have abolished the feature by treating it
7417 as "timeout", and writing a warning to the main and panic logs.
7419 36. The display of retry rules for -brt wasn't always showing the error code
7422 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7423 timeout_connect_A, timeout_connect_MX.
7425 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7426 to the empty sender.
7428 39. The daemon was not analysing the content of -oX till after it had closed
7429 stderr and disconnected from the controlling terminal. This meant that any
7430 syntax errors were only noted on the panic log, and the return code from
7431 the command was 0. By re-arranging the code a little, I've made the
7432 decoding happen first, so such errors now appear on stderr, and the return
7433 code is 1. However, the actual setting up of the sockets still happens in
7434 the disconnected process, so errors there are still only recorded on the
7437 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7438 connections (as happens on some IP stacks) was logged at start up time as
7439 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7440 it from "IPv6 and IPv4", which means that two separate sockets are being
7443 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7444 says whether AF_INET or AF_INET6 was passed as an argument.
7446 42. Exiwhat output was messed up when time zones were included in log
7449 43. Exiwhat now gives more information about the daemon's listening ports,
7450 and whether -tls-on-connect was used.
7452 44. The "port" option of the smtp transport is now expanded.
7454 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7455 silently ignored. Now an error message is written to the main and panic
7458 46. There's a new ACL modifier called "logwrite" which writes to a log file
7459 as soon as it is encountered.
7461 47. Added $local_user_uid and $local_user_gid at routing time.
7463 48. Exim crashed when trying to verify a sender address that was being
7466 49. Exim was recognizing only a space character after ".include". It now also
7467 recognizes a tab character.
7469 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7470 extracting the relevant information from the specification. The man page no
7471 longer contains scrambled data for the -d option, and I've added a section
7472 at the front about calling Exim under different names.
7474 51. Added "extra_headers" argument to the "mail" command in filter files.
7476 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7479 53. Installed eximstats 1.29.
7481 54. Added transport_filter_timeout as a generic transport option.
7483 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7484 Cc: header lines. This was required by RFC 822, but it not required by RFC
7487 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7488 incoming messages that did not have them. Now it does so only if the
7489 message originates locally, that is, if there is no associated remote host
7490 address. When Resent- header lines are present, this applies to the Resent-
7491 lines rather than the non-Resent- lines.
7493 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7494 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7496 58. Messages for configuration errors now include the name of the main
7497 configuration file - useful now that there may be more than one file in a
7498 list (.included file names were always shown).
7500 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7501 for those rare installations that do not start the daemon as root or run it
7502 setuid root. I've cut out the call to initgroups() if the daemon is not
7505 60. The Exim user and group can now be bound into the binary as text strings
7506 that are looked up at the start of Exim's processing.
7508 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7510 62. Added $mailstore_basename variable.
7512 63. Installed patch to sieve.c from Michael Haardt.
7514 64. When Exim failed to open the panic log after failing to open the main log,
7515 the original message it was trying to log was written to stderr and debug
7516 output, but if they were not available (the usual case in production), it
7517 was lost. Now it is written to syslog before the two lines that record the
7518 failures to open the logs.
7520 65. Users' Exim filters run in subprocesses under the user's uid. It is
7521 possible for a "deliver" command or an alias in a "personal" command to
7522 provoke an address rewrite. If logging of address rewriting is configured,
7523 this fails because the process is not running as root or exim. There may be
7524 a better way of dealing with this, but for the moment (because 4.30 needs
7525 to be released), I have disabled address rewrite logging when running a
7526 filter in a non-root, non-exim process.
7532 1. The buildconfig auxiliary program wasn't quoting the value set for
7533 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7534 not defined. This bug was masked in 4.22 by the effect that was fixed in
7537 2. Some messages that were rejected after a message id was allocated were
7538 shown as "incomplete" by exigrep. It no longer does this for messages that
7539 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7541 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7542 have allow_domain_literals set, the ID did not get logged in the <= line.
7543 Domain literals are now always recognized in Message-ID: header lines.
7545 4. The first argument for a ${extract expansion item is the key name or field
7546 number. Leading and trailing spaces in this item were not being ignored,
7547 causing some misleading effects.
7549 5. When deliver_drop_privilege was set, single queue runner processes started
7550 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7551 same command in the process it spins off) were not dropping privilege.
7553 6. When the daemon running as "exim" started a queue runner, it always
7554 re-executed Exim in the spun-off process. This is a waste of effort when
7555 deliver_drop_privilege is set. The new process now just calls the
7556 queue-runner function directly.
7562 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7565 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7566 ignored. Though the use of -oP was forcing the writing of a pid file, it
7567 was always written to the default place.
7569 3. If the message "no IP address found for host xxxx" is generated during
7570 incoming verification, it is now followed by identification of the incoming
7571 connection (so you can more easily find what provoked it).
7573 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7575 5. Added some features to "harden" Exim a bit more against certain attacks:
7577 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7578 be put in Local/Makefile. This is like the never_users runtime option,
7579 but it cannot be overridden. The default setting is "root".
7581 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7582 prefix string with which any file named in a -C command line option
7585 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7586 is retained for -C and -D only if the caller of Exim is root. Without
7587 it, the exim user may also use -C and -D and retain privilege.
7589 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7590 command line option is disabled.
7592 6. Macro names set by the -D option must start with an upper case letter, just
7593 like macro names defined in the configuration file.
7595 7. Added "dereference=" facility to LDAP.
7597 8. Two instances of the typo "uknown" in the source files are fixed.
7599 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7600 the Configure-Makefile script screwed up while processing it.
7602 10. Incorporated PCRE 4.4.
7604 11. The SMTP synchronization check was not operating right at the start of an
7605 SMTP session. For example, it could not catch a HELO sent before the client
7606 waited for the greeting. There is now a check for outstanding input at the
7607 point when the greeting is written. Because of the duplex, asynchronous
7608 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7609 way, but not yet received, when the check is performed.
7611 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7612 on TCP/IP sockets, because this apparently causes some broken clients to
7615 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7616 unchanged) from the Cygwin maintainer.
7618 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7619 was supported instead of testing for mbx. Effectively a typo.
7621 15. The spa authenticator server code was not checking that the input it
7622 received was valid base64.
7624 16. The debug output line for the "set" modifier in ACLs was not showing the
7625 name of the variable that was being set.
7627 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7629 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7630 lookup on its own. Something else was needed to trigger the lookup. For
7631 example, a match in host_lookup or the need for a host name in a host list.
7632 Now, if $sender_host_name is referenced and the host name has not yet been
7633 looked up, a lookup is performed. If the lookup fails, the variable remains
7634 empty, and $host_lookup_failed is set to "1".
7636 19. Added "eqi" as a case-independent comparison operator.
7638 20. The saslauthd authentication condition could segfault if neither service
7639 nor realm was specified.
7641 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7642 error message that was logged was misleading, and incoming SMTP
7643 connections were dropped. The message is now more accurate, and temporary
7644 errors are given to SMTP connections.
7646 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7647 (except RSET) with a 421 error, until QUIT is received. However, it was
7648 failing to send a response to QUIT.
7650 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7651 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7652 ACL is now run after the helo_try_verify_hosts code.
7654 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7655 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7656 case-independent, so other case variants are also recognized). Apparently
7657 some systems use these upper case variants.
7659 25. If more than two messages were waiting for the same host, and a transport
7660 filter was specified for the transport, Exim sent two messages over the
7661 same TCP/IP connection, and then failed with "socket operation on non-
7662 socket" when it tried to send the third.
7664 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7666 27. The extern definition of crypt16() in expand.c was not being excluded when
7667 the OS had its own crypt16() function.
7669 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7670 as a preferred synonym for return_size_limit, both as an option and as an
7673 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7675 30. Changed the default configuration ACL to relax the local part checking rule
7676 for addresses that are not in any local domains. For these addresses,
7677 slashes and pipe symbols are allowed within local parts, but the sequence
7678 /../ is explicitly forbidden.
7680 31. SPA server authentication was not clearing the challenge buffer before
7683 32. log_message in a "warn" ACL statement was writing to the reject log as
7684 well as to the main log, which contradicts the documentation and doesn't
7685 seem right (because no rejection is happening). So I have stopped it.
7687 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7688 However, I am unable to do any testing of this.
7690 34. Fixed an infelicity in the appendfile transport. When checking directories
7691 for a mailbox, to see if any needed to be created, it was accidentally
7692 using path names with one or more superfluous leading slashes; tracing
7693 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7695 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7696 contents are added to the log line that is written for every discarded
7697 recipient. (Previously a log_message setting was ignored.)
7699 36. The ${quote: operator now quotes the string if it is empty.
7701 37. The install script runs exim in order to find its version number. If for
7702 some reason other than non-existence or emptiness, which it checks, it
7703 could not run './exim', it was installing it with an empty version number,
7704 i.e. as "exim-". This error state is now caught, and the installation is
7707 38. An argument was missing from the function that creates an error message
7708 when Exim fails to connect to the socket for saslauthd authentication.
7709 This could cause Exim to crash, or give a corrupted message.
7711 39. Added isip, isip4, and isip6 to ${if conditions.
7713 40. The ACL variables $acl_xx are now saved with the message, and can be
7714 accessed later in routers, transports, and filters.
7716 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7717 strings in the file are not string-expanded.
7719 42. If a MAIL command specified a SIZE value that was too large to fit into an
7720 int variable, the check against message_size_limit failed. Such values are
7721 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7722 day this will have to be increased, but I don't think I want to be around
7723 when emails are that large.
7730 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7731 iconv() is not standard in FreeBSD.
7733 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7734 IPv6 enabled. The observed symptom was a segmentation fault on return from
7735 the function os_common_find_running_interfaces() in src/os.c.
7737 3. In the check_special_case() function in daemon.c I had used "errno" as an
7738 argument name, which causes warnings on some systems. This was basically a
7739 typo, since it was named "eno" in the comments!
7741 4. The code that waits for the clock to tick (at a resolution of some fraction
7742 of a second) so as to ensure message-id uniqueness was always waiting for
7743 at least one whole tick, when it could have waited for less. [This is
7744 almost certainly not relevant at current processor speeds, where it is
7745 unlikely to ever wait at all. But we try to future-proof.]
7747 5. The function that sleeps for a time interval that includes fractions of a
7748 second contained a race. It did not block SIGALRM between setting the
7749 timer, and suspending (a couple of lines later). If the interval was short
7750 and the sigsuspend() was delayed until after it had expired, the suspension
7751 never ended. On busy systems this could lead to processes getting stuck for
7754 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7755 process, before it forks any delivery processes. The open lookup caching
7756 mechanism meant that the open file or database connection was passed into
7757 the delivery process. The problem was that delivery processes always tidy
7758 up cached lookup data. This could cause a problem for the next delivery
7759 process started by the queue runner, because the external queue runner
7760 process does not know about the closure. So the next delivery process
7761 still has data in the lookup cache. In the case of a file lookup, there was
7762 no problem because closing a file descriptor in a subprocess doesn't affect
7763 the parent. However, if the lookup was caching a connection to a database,
7764 the connection was closed, and the second delivery process was likely to
7765 see errors such as "PGSQL: query failed: server closed the connection
7766 unexpectedly". The problem has been fixed by closing all cached lookups
7767 in a queue runner before running a delivery process.
7769 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7770 seem to have the "const" qualifier which it has on other OS. I've
7773 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7774 *of the same type* (client or server) with the same public name that an
7775 error should be diagnosed.
7777 9. When Exim looked up a host name for an IP address, but failed to find the
7778 original IP address when looking up the host name (a safety check), it
7779 output the message "<ip address> does not match any IP for NULL", which was
7780 confusing, to say the least. The bug was that the host name should have
7781 appeared instead of "NULL".
7783 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7784 user that is built into the binary, and the -C or -D options is used, root
7785 privilege is dropped before the configuration file is read. In addition,
7786 logging is switched to stderr instead of the normal log files. If the
7787 configuration then re-defines the Exim user, the unprivileged environment
7788 is probably not what is expected, so Exim logs a panic warning message (but
7791 However, if deliver_drop_privilege is set, the unprivileged state may well
7792 be exactly what is intended, so the warning has been cut out in that case,
7793 and Exim is allowed to try to write to its normal log files.
7799 1. smtp_return_error_details was not giving details for temporary sender
7800 or receiver verification errors.
7802 2. Diagnose a configuration error if two authenticators have the same public
7805 3. Exim used not to create the message log file for a message until the first
7806 delivery attempt. This could be confusing when incoming messages were held
7807 for policy or load reasons. The message log file is now created at the time
7808 the message is received, and an initial "Received" line is written to it.
7810 4. The automatically generated man page for command line options had a minor
7811 bug that caused no ill effects; however, a more serious problem was that
7812 the procedure for building the man page automatically didn't always
7813 operate. Consequently, release 4.20 contains an out-of-date version. This
7814 shouldn't happen again.
7816 5. When building Exim with embedded Perl support, the script that builds the
7817 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7818 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7820 6. The freeze_tell option was not being used for messages that were frozen on
7821 arrival, either by an ACL or by local_scan().
7823 7. Added the smtp_incomplete_transaction log selector.
7825 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7826 was accepting AUTH without a new EHLO.
7828 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7829 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7832 10. Logging of TCP/IP connections (when configured) now happens in the main
7833 daemon process instead of the child process, so that the TCP/IP connection
7834 count is more accurate (but it can never be perfect).
7836 11. The use of "drop" in a nested ACL was not being handled correctly in the
7837 outer ACL. Now, if condition failure induced by the nested "drop" causes
7838 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7839 or "require"), the connection is dropped.
7841 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7842 that yield "discard" can now be used with an "accept" or a "discard" verb,
7843 but an error is generated for any others (because I can't see a useful way
7844 to define what should happen).
7846 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7847 are now processed in the same way as lines in the Exim configuration file.
7848 In particular, continuation lines are supported.
7850 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7852 15. Added -ti meaning -t -i.
7854 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7855 domains, and warn by logging if others are found.
7857 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7858 returned from GIFCONF when Exim is trying to find the list of interfaces on
7859 a host. The code in os.c has been modified to copy each ifreq to an aligned
7860 structure in all cases.
7862 Also, in some cases, the returned ifreq's were being copied to a 'struct
7863 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7864 means the last couple of bytes of an IPv6 address could be chopped if the
7865 ifreq contained only a normal sockaddr (14 bytes storage).
7867 18. Named domain lists were not supported in the hosts_treat_as_local option.
7868 An entry such as +xxxx was not recognized, and was treated as a literal
7871 19. Ensure that header lines added by a DATA ACL are included in the reject log
7872 if the ACL subsequently rejects the message.
7874 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7875 MD5 (which is deprecated).
7877 21. When testing a filter file using -bf, Exim was writing a message when it
7878 took the sender from a "From " line in the message, but it was not doing so
7879 when it took $return_path from a Return-Path: header line. It now does.
7881 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7882 with a valid header line field name (a series of printing characters
7883 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7885 23. Changed "disc" in the source to "disk" to conform to the documentation and
7886 the book and for uniformity.
7888 24. Ignore Sendmail's -Ooption=value command line item.
7890 25. When execve() failed while trying to run a command in a pipe transport,
7891 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7892 could be confused with a return value of 69 from the command itself. This
7893 has been changed to 127, the value the shell returns if it is asked to run
7894 a non-existent command. The wording for the related log line suggests a
7895 non-existent command as the problem.
7897 26. If received_header_text expands to an empty string, do not add a Received:
7898 header line to the message. (Well, it adds a token one on the spool, but
7899 marks it "old" so that it doesn't get used or transmitted.)
7901 27. Installed eximstats 1.28 (addition of -nt option).
7903 28. There was no check for failure on the call to getsockname() in the daemon
7904 code. This can fail if there is a shortage of resources on the system, with
7905 ENOMEM, for example. A temporary error is now given on failure.
7907 29. Contrary to the C standard, it seems that in some environments, the
7908 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7909 program. Exim now does this explicitly; it affects the formatting of
7910 timestamps using strftime().
7912 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7913 complaints. I've now initialized all the variables, to avoid this.
7915 32. Header lines added by a system filter were not being "seen" during
7916 transport-time rewrites.
7918 33. The info_callback() function passed to OpenSSL is set up with type void
7919 (*)(SSL *, int, int), as described somewhere. However, when calling the
7920 function (actually a macro) that sets it up, the type void(*)() is
7921 expected. I've put in a cast to prevent warnings from picky compilers.
7923 34. If a DNS black list lookup found a CNAME record, but there were no A
7924 records associated with the domain it pointed at, Exim crashed.
7926 35. If a DNS black list lookup returned more than one A record, Exim ignored
7927 all but the first. It now scans all returned addresses if a particular IP
7928 value is being sought. In this situation, the contents of the
7929 $dnslist_value variable are a list of all the addresses, separated by a
7932 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7933 to accept a host name and all its aliases if the forward lookup for any of
7934 them yielded the IP address of the incoming connection. Now it accepts only
7935 those names whose forward lookup yields the correct IP address. Any other
7936 names are discarded. This closes a loophole whereby a rogue DNS
7937 administrator could create reverse DNS records to break through a
7938 wildcarded host restriction in an ACL.
7940 37. If a user filter or a system filter that ran in a subprocess used any of
7941 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7942 the wrong values were passed to the pipe command ($thisaddress had the
7943 value of $0, $0 had the value of $1, etc). This bug was introduced by
7944 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7946 38. Improved the line breaking for long SMTP error messages from ACLs.
7947 Previously, if there was no break point between 40 and 75 characters, Exim
7948 left the rest of the message alone. Two changes have been made: (a) I've
7949 reduced the minimum length to 35 characters; (b) if it can't find a break
7950 point between 35 and 75 characters, it looks ahead and uses the first one
7951 that it finds. This may give the occasional overlong line, but at least the
7952 remaining text gets split now.
7954 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7955 file descriptors might be low, and that setting 1000 would always raise it.
7956 It turns out that in some environments, the limit is already over 1000 and
7957 that lowering it causes trouble. So now Exim takes care not to decrease it.
7959 40. When delivering a message, the value of $return_path is set to $sender_
7960 address at the start of routing (routers may change the value). By an
7961 oversight, this default was not being set up when an address was tested by
7962 -bt or -bv, which affected the outcome if any router or filter referred to
7965 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7966 log only when "message" or "log_message" are set. However, if one of the
7967 conditions was an address verification, or a call to a nested ACL, the
7968 messages generated by the underlying test were being passed through. This
7969 no longer happens. The underlying message is available in $acl_verify_
7970 message for both "message" and "log_message" expansions, so it can be
7971 passed through if needed.
7973 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7974 new expansion $bh_ to give the encoded byte string without charset
7975 translation. Translation happens only if iconv() is available; HAVE_ICONV
7976 indicates this at build time. HEADERS_CHARSET gives the charset to
7977 translate to; headers_charset can change it in the configuration, and
7978 "headers charset" can change it in an individual filter file.
7980 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7981 that creates RFC 2047 encoded "words" labels them as that charset instead
7982 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7983 expansion operator; (ii) when Exim creates a From: line for a local
7984 message; (iii) when a header line is rewritten to include a "phrase" part.
7986 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7987 buggy, causing it to skip the first lines of messages whose message ID
7988 ended in 'D'. This would not have bitten before Exim release 4.14, because
7989 message IDs were unlikely to end in 'D' before then. The effect was to have
7990 incorrect size information for certain domains.
7992 45. #include "config.h" was missing at the start of the crypt16.c module. This
7993 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7996 46. If there was a timeout during a "random" callout check, Exim treated it as
7997 a failure of the random address, and carried on sending RSET and the real
7998 address. If the delay was just some slowness somewhere, the response to the
7999 original RCPT would be taken as a response to RSET and so on, causing
8000 mayhem of various kinds.
8002 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8003 when I implemented it. It didn't allow for the fact that some option values
8004 may legitimately be negative (e.g. size_addition), and it didn't even do
8005 the right test for positive values.
8007 48. Domain names in DNS records are case-independent. Exim always looks them up
8008 in lower case. Some resolvers return domain names in exactly the case they
8009 appear in the zone file, that is, they may contain uppercase letters. Not
8010 all resolvers do this - some return always lower case. Exim was treating a
8011 change of case by a resolver as a change of domain, similar to a widening
8012 of a domain abbreviation. This triggered its re-routing code and so it was
8013 trying to route what was effectively the same domain again. This normally
8014 caused routing to fail (because the router wouldn't handle the domain
8015 twice). Now Exim checks for this case specially, and just changes the
8016 casing of the domain that it ultimately uses when it transmits the message
8019 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8022 50. If a filter generated a file delivery with a non-absolute name (possible if
8023 no home directory exists for the router), the forbid_file option was not
8026 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8027 the existing equality matching.
8029 52. Exim was using ints instead of ino_t variables in some places where it was
8030 dealing with inode numbers.
8032 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8033 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8034 called TMPDIR, and if it finds it is different, it changes its value.
8036 54. The smtp_printf() function is now made available to local_scan() so
8037 additional output lines can be written before returning. There is also an
8038 smtp_fflush() function to enable the detection of a dropped connection.
8039 The variables smtp_input and smtp_batched_input are exported to
8042 55. Changed the default runtime configuration: the message "Unknown user"
8043 has been removed from the ACL, and instead placed on the localuser router,
8044 using the cannot_route_message feature. This means that any verification
8045 failures that generate their own messages won't get overridden. Similarly,
8046 the "Unrouteable address" message that was in the ACL for unverifiable
8047 relay addresses has also been removed.
8049 56. Added hosts_avoid_esmtp to the smtp transport.
8051 57. The exicyclog script was not checking for the esoteric option
8052 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8053 will work only if exicyclog is run under the appropriate euid.
8055 58. Following a discussion on the list, the rules by which Exim recognises line
8056 endings on incoming messages have been changed. The -dropcr and drop_cr
8057 options are now no-ops, retained only for backwards compatibility. The
8058 following line terminators are recognized: LF CRLF CR. However, special
8059 processing applies to CR:
8061 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8062 nor a local message in the state where . is a terminator.
8064 (ii) If a bare CR is encountered in a header line, an extra space is added
8065 after the line terminator so as not to end the header. The reasoning
8066 behind this is that bare CRs in header lines are most likely either
8067 to be mistakes, or people trying to play silly games.
8069 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8070 was being incorrectly given as 18 bytes larger than it should have been.
8071 This is a VOB (very old bug).
8073 60. This may never have affected anything current, but just in case it has:
8074 When the local host is found other than at the start of a list of hosts,
8075 the local host, those with the same MX, and any that follow, are discarded.
8076 When the list in question was part of a longer list of hosts, the following
8077 hosts (not currently being processed) were also being discarded. This no
8078 longer happens. I'm not sure if this situation could ever has previously
8081 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8084 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8085 header. This is something that is recommended in a new Internet Draft, and
8086 is something that is documented as being done by Sendmail. There are two
8087 possible values. For messages generated by the autoreply transport, Exim
8090 Auto-Submitted: auto-replied
8092 whereas for all other generated messages (e.g. bounces) it adds
8094 Auto-Submitted: auto-generated
8096 63. The "personal" condition in filters now includes a test for the
8097 Auto-Submitted: header. If it contains the string "auto-" the message it
8098 not considered personal.
8100 64. Added rcpt_include_affixes as a generic transport option.
8102 65. Added queue_only_override (default true).
8104 66. Added the syslog_duplication option.
8106 67. If what should have been the first header line of a message consisted of
8107 a space followed by a colon, Exim was mis-interpreting it as a header line.
8108 It isn't of course - it is syntactically invalid and should therefore be
8109 treated as the start of the message body. The misbehaviour could have
8110 caused a number of strange effects, including loss of data in subsequent
8111 header lines, and spool format errors.
8113 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8114 client host had authenticated. This control can now be exercised by an ACL
8115 for more flexibility.
8117 69. By default, callouts do not happen when testing with -bh. There is now a
8118 variant, -bhc, which does actually run the callout code, including
8119 consulting and updating the callout cache.
8121 70. Added support for saslauthd authentication, courtesy of Alexander
8124 71. If statvfs() failed on the spool or log directories while checking their
8125 size for availability, Exim confusingly gave the error "space shortage".
8126 Furthermore, in debugging mode it crashed with a floating point exception.
8127 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8128 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8129 serious problem, Exim now writes to the main and panic logs when this
8130 happens, with details of the failure. It then refuses to accept the
8131 incoming message, giving the message "spool directory problem" or "log
8132 directory problem" with a 421 code for SMTP messages.
8134 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8135 0, 1, and 2 exist, because some OS complain for execs without them (see
8136 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8137 descriptors. However, the code omitted to check that the open succeeded,
8138 causing mysterious errors if for some reason the permissions on /dev/null
8139 got screwed. Now Exim writes a message to the main and panic logs, and
8140 bombs out if it can't open /dev/null.
8142 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8143 interact so that it is all more flexible. It is supposed to remain
8144 backwards compatible. Also added extra_local_interfaces.
8146 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8147 to bomb out with an assertion failure - to the client this appears as a
8148 connection drop. This problem occurs in the part of the code that was taken
8149 from the Samba project. Fortunately, the assertion is in a very simple
8150 function, so I have fixed this by reproducing the function inline in the
8151 one place where it is called, and arranging for authentication to fail
8152 instead of killing the process with assert().
8154 75. The SPA client code was not working when the server requested OEM rather
8155 than Unicode encoding.
8157 76. Added code to make require_files with a specific uid setting more usable in
8158 the case where statting the file as root fails - usually a non-root-mounted
8159 NFS file system. When this happens and the failure is EACCES, Exim now
8160 forks a subprocess and does the per-uid checking as the relevant uid.
8162 77. Added process_log_path.
8164 78. If log_file_path was not explicitly set, a setting of check_log_space or
8165 check_log_inodes was ignored.
8167 79. If a space check for the spool or log partitions fails, the incident is now
8168 logged. Of course, in the latter case the data may get lost...
8170 80. Added the %p formatting code to string_format() so that it can be used to
8171 print addresses in debug_print(). Adjusted all the address printing in the
8172 debugging in store.c to use %p rather than %d.
8174 81. There was a concern that a line of code in smtp_in.c could overflow a
8175 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8176 initially expressed, the concern was not well-founded, because trailing
8177 spaces are removed early. However, if the trailing spaces were followed by
8178 a NULL, they did not get removed, so the overflow was possible. Two fixes
8181 (a) I re-wrote the offending code in a cleaner fashion.
8182 (b) If an incoming SMTP command contains a NULL character, it is rejected
8185 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8186 runs initgroups(), so that if the Exim user is in any additional groups,
8187 they will be used during message reception.
8193 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob