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
70 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
71 related applications. Previously an "H" was used where available info
72 says that "M" should be, so change to match.
74 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
75 as arguments, so an implementation trying to copy these into a local
76 buffer was taking a taint-enforcement trap. Fix by using dynamically
77 created buffers. Similar fix for radius expansion condition.
79 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
80 reasonable, eg. to count headers. Fix by using dynamically created
81 buffers rather than a local. Do similar fixes for ACL actions "dcc",
82 "log_reject_target", "malware" and "spam"; the arguments are expanded
83 so could be handling tainted values.
85 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
86 broken the (no-op) support for this sendmail command. Restore it
87 to doing nothing, silently, and returning good status.
89 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
90 record path was given (or the default used) without a leading directory
91 path, an error occurred on trying to open it. Use the transport's working
94 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
95 transport. Previously it was the name on the DNS A-record; use instead
96 the head of the CNAME chain leading there (if there is one). This seems
97 to align better with RFC 6125.
99 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
100 smtp_accept_max_per_host allocated resources which were not released
101 when the limit was exceeded. This eventually crashed the daemon. Fix
102 by adding a release action in that path.
104 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
105 expanded; previously using tainted values was rejected. Fix by using
106 dynamically-created buffers.
108 JH/09 Relax restrictions on ACL verify condition needing access to message
109 headers. Previously they were only permitted in data and non-smtp ACLs;
110 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
111 not_blind, header_sender and header_names_ascii verification.
113 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
114 Previously a macro used one argument twice; when called with the
115 argument as an expression having side-effects, incorrect operation
116 resulted. Use an inlineable function.
118 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
119 held open for a verify callout. Previously this wan not accounted for
120 and a corrupt onward SMTP conversation resulted.
122 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
123 Previously they were embedded naively in the constructed address; when
124 needed, strip the quoting and quote the entire local_part.
125 Also make the inbound_srs expansion condition handle quoting.
127 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
128 excluded, not matching the documentation.
130 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
131 was given for the sqlite_dbfile a trap resulted.
133 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
134 "name" argument resulted in a trap. There is no reason to disallow such;
135 this was a coding error.
137 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
138 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
139 spent suspended, ignoring the POSIX definition. Previously we assumed
140 it did and a constant offset from real time could be used as a correction.
141 Change to using the same clock source for the start-of-message and the
142 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
143 exists, just to get a clock slightly more aligned to reality.
145 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
146 RFC says it is optional some validators care. The missing char was not
147 intended but triggered by a line-wrap alignment. Discovery and fix by
148 Guillaume Outters, hacked on by JH.
150 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
151 name being quoted was tainted a trap would be taken. Fix by using
152 dynamically created buffers. The routine could have been called by a
153 rewrite with the "h" flag, by using the "-F" command-line option, or
154 by using a "name=" option on a control=submission ACL modifier.
156 JH/19 SPF: change the Authentication-Results expansion component to give
157 smtp.helo when the sender domain is empty. Previously it gave
160 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
161 not in 127.0.0.0/8 to help in spotting list domains taken over by a
162 domain-parking registrar.
164 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
165 Previously when a whitespace character was specified it was not inserted
166 after removing the newline.
168 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
169 the domain part of the recipient address. This overrides any tls_sni
170 option set, which was previously used.
172 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
175 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
176 is_tainted() had an off-by-one error in the overenthusiastic direction.
177 Find and fix by Gavan. Although NetBSD is not a supported platform for
178 4.94 this bug could affect other platforms.
180 PP/01 Fix default prime selection to be consistent.
181 One path used ike23 still, instead of exim.dev.20160529.3; now both
182 execution flows will use the same DH primes (currently
183 exim.dev.20160529.3).
185 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
186 option in smtp transport, to match the documentation. Previously
187 verification was not being done in some cases where it should have been.
189 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
190 than one server was defined and depending on the platform memory layout
191 details, an internal consistency trap could be hit while walking the list
194 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
195 files. The value is used by the authresults expansion item. Previously
196 if this was used in a router or transport, a crash could result.
198 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
199 have a sufficient incidence of EINTR returns from select that an
200 interaction with other operations done by the main daemon loop exposed
201 a bug in the error-handling. This was benign apart from the log
204 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
205 for consistency with delivered "=>" and failed "**" lines. While we're
206 there, handle PRX and TFO.
208 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
209 applied. This resulted, if any header-line rewrite rules were configured,
210 in a panic-log triggerable by sending a message with a long address in
211 a header. Fix by increasing the arbitrary limit to larger than a single
212 (dewrapped) 5322 header line maximum size.
214 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
215 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
216 RFC 6648 which deprecates X- options in protocols as a general practice.
217 Changeover between the implementations is handled by the mechanisms
220 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
221 a remote which uses callout/hold. Previously the local was queued.
223 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
226 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
227 log line. The asterisk is now consistently appended to the remote IP
228 (and port, if given), and will also be provided on defer and fail log
229 lines. Previously it could be placed on the local IP if that was being
230 logged, and was only provided on delivery lines.
232 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
234 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
235 authenticator client_send option. Previously the next char, after a pair
236 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
237 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
238 documentation. There is still no way to get a leading ^ immediately
239 after a NUL (ie. for the password of a PLAIN method authenticator.
241 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
242 files. For bad sizes read, delete the record and whine to paniclog.
244 JH/38 When logging an AUTH failure, as server, do not include sensitive
245 information. Previously, the credentials would be included if given
246 as part of the AUTH command line and an ACL denied authentication.
248 JH/39 Bug 2691: fix $local_part_data. When the matching list element
249 referred to a file, bad data was returned. This likely also affected
252 JH/40 The gsasl authenticator now supports caching of the salted password
253 generated by the client-side implementation. This required the addition
254 of a new variable: $auth4.
256 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
257 left undeleted; the attempt to re-create it then failed - resulting in
258 the usual "SIGHUP tp have daemon reload configuration" to not work.
259 This affected any platform not supporting "abstract" Unix-domain
260 sockets (i.e. not Linux).
262 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
263 recipients" response to RCPT in a later response, with a 250. The
264 previous coding assumed this would not happen, and under PIPELINING
265 would result in both lost and duplicate recipients for a message.
267 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
268 Previously the weighting was incorrectly applied. Similar fix for socks
269 proxies. Found and fixed by Heiko Schlichting.
271 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
272 not handle sub-lists included using the +namedlist syntax. While
273 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
274 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
277 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
278 Previously the permanent pool was used, so the sore could not be freed.
279 This meant a connection with many messages would use continually-growing
282 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
283 per-pool so as not to waste too much space. Previously a constant size
284 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
285 DOS attacks harder. The cost is wasted memory use in the larger blocks.
287 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
288 same space repeatedly, and a smaller process footprint.
290 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
291 for GnuTLS. Previously the name started "%s" which, while not a bug,
292 looked as if if might be one.
294 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
295 "continued connection") make the $proxy_* variables available. Previously
296 the information was not passed across the exec() call for subsequent
297 transport executions. This also mean that the log lines for the
298 messages can show the proxy information.
300 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
301 receive time. With modern systems the difference is significant.
302 The historical behaviour can be restored by disabling (a new) log_selector
303 "queue_time_exclusive".
305 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
306 content) were specifically tested for. Now, also cover expansions
307 resulting in ACL names and inline ACL content.
309 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
310 making it unusable in complex expressions.
312 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
313 fail with a taint-check on the spoolfile name, and leave the message
316 HS/01 Enforce absolute PID file path name.
318 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
320 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
322 PP/02 Bug 2643: Correct TLS DH constants.
323 A missing NUL termination in our code-generation tool had led to some
324 incorrect Diffie-Hellman constants in the Exim source.
325 Reported by kylon94, code-gen tool fix by Simon Arlott.
327 PP/03 Impose security length checks on various command-line options.
328 Fixes CVE-2020-SPRSS reported by Qualys.
330 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
331 better. Reported by Qualys.
333 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
334 providing a particularly obnoxious sender full name.
337 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
339 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
342 PP/08 Change default for recipients_max from unlimited to 50,000.
344 PP/09 Fix security issue with too many recipients on a message (to remove a
345 known security problem if someone does set recipients_max to unlimited,
346 or if local additions add to the recipient list).
347 Fixes CVE-2020-RCPTL reported by Qualys.
349 PP/10 Fix security issue in SMTP verb option parsing
350 Fixes CVE-2020-EXOPT reported by Qualys.
352 PP/11 Fix security issue in BDAT state confusion.
353 Ensure we reset known-good where we know we need to not be reading BDAT
354 data, as a general case fix, and move the places where we switch to BDAT
355 mode until after various protocol state checks.
356 Fixes CVE-2020-BDATA reported by Qualys.
358 HS/03 Die on "/../" in msglog file names
360 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
361 the Exim runtime user are allowed to create files.
363 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
366 QS/03 When reading the output from interpreted forward files we do not
367 pass the pipe between the parent and the interpreting process to
368 executed child processes (if any).
370 QS/04 Always die if requested from internal logging, even is logging is
373 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
374 the API; compilation noo longer completes with DMARC support included.
375 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
376 been reported on other platforms.
378 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
380 JH/56 Make the majority of info read from config files readonly, for defence-in-
381 depth against exploits. Suggestion by Qualys.
382 Not supported on Solaris 10.
384 JH/57 Fix control=fakreject for a custom message containing tainted data.
385 Previously this resulted in a log complaint, due to a re-expansion present
386 since fakereject was originally introduced.
388 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
389 modifier was given, a loop resulted.
391 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
392 pipelined SMTP command followed the BDAT LAST then it would be
393 incorrectly treated as part of the message body, causing a verification
396 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
397 require looser validation rules than those for 821-level addresses,
398 which only permit IP addresses.
404 JH/01 Avoid costly startup code when not strictly needed. This reduces time
405 for some exim process initialisations. It does mean that the logging
406 of TLS configuration problems is only done for the daemon startup.
408 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
410 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
411 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
414 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
415 DKIM signing is being done. Previously a transport_filter would always
416 disable CHUNKING, falling back to traditional DATA.
418 JH/05 Regard command-line recipients as tainted.
420 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
422 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
423 PAM library frees one of the arguments given to it, despite the
424 documentation. Therefore a plain malloc must be used.
426 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
427 on-stack buffers were used, resulting in a taint trap when DSN information
428 copied from a received message was written into the buffer.
430 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
431 the ordering of its ARC headers. This caused a crash.
433 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
434 a new record was being constructed with information from the peer, a trap
437 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
438 installation would get error messages from DMARC verify, when it hit the
439 nonexistent file indicated by the default. Distros wanting DMARC enabled
440 should both provide the file and set the option.
441 Also enforce no DMARC verification for command-line sourced messages.
443 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
444 could, depending on the platform, hang at the STARTTLS response.
446 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
447 message on a connection. Previously if one message had ARC headers and
448 the following one did not, a crash could result when adding an
449 Authentication-Results: header.
451 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
452 Exim main code and Exim-related utities. The introduction of taint
453 tracking also did many adjustments to string handling. Since then, eximon
454 frequently terminated with an assert failure.
456 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
457 check for 452 responses. This slightly helps the inefficieny of doing
458 a large alias-expansion into a recipient-limited target. The max_rcpt
459 transport option still applies (and at the current default, will override
460 the new feature). The check is done for either cause of synch, and forces
461 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
462 connection. The new facility is not tunable at this time.
464 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
465 library live data was being used, so the results became garbage. Make
466 copies while it is still usable.
468 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
469 on delivery deferred (==) and failed (**) lines (if a delivery was
470 attemtped). Previously it was only on completion (=>) lines.
472 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
473 for the expansion of the server_scram_iter and server_scram_salt options.
475 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
476 are now specifically given a NO_DATA response without hitting the system
477 resolver. The library goes on to do the now-standard TXT lookup.
478 Use of dnsdb lookups is not affected.
480 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
481 only retrieve the errormessage once. Previously two calls to dlerror()
482 were used, and the second one (for mainlog/paniclog) retrieved null
485 JH/20 Taint checking: disallow use of tainted data for
486 - the appendfile transport file and directory options
487 - the pipe transport command
488 - the autoreply transport file, log and once options
489 - file names used by the redirect router (including filter files)
491 - paths used by single-key lookups
492 Previously this was permitted.
494 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
495 adjusted the size of a major service buffer; this failed because the
496 buffer was in use at the time. Change to a compile-time increase in the
497 buffer size, when this authenticator is compiled into exim.
499 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
500 previous fast-mode was untenable in the face of glibs using mmap to
501 support larger malloc requests.
503 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
504 New values supported, if defined on system where compiled:
505 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
506 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
508 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
509 running a limited number of proceses in parallel, a benefit is gained. The
510 amount varies with the platform hardware and load. The use of the option
511 queue_run_in_order means we cannot do this, as ordering becomes
514 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
515 had introduced a string-copy (for ensuring NUL-termination) which was not
516 appropriate for that case, which can include embedded NUL bytes in the
517 block of data. Investigation showed the copy to actually be needless, the
518 data being length-specified.
520 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
521 done during a receiving connection, and both used TLS, global info was
522 used rather than per-connection info for tracking the state of data
523 queued for transmission. This could result in a connection hang.
525 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
526 Previously, when delivering serveral messages down a single connection
527 only the first would provide a SIZE. This was due to the size information
528 not being properly tracked.
530 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
531 TAI (at 37 seconds currently), pretend to be in UTC for time-related
532 expansion and logging. Previously, spurious values such as a future
533 minute could be seen.
535 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
536 it could crash from a null-deref. This could also affect the
537 ${addresses: } operator and ${readsock } item.
539 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
540 message following a mime one, the variable was not reset.
542 JH/30 When an pipelined-connect fails at the first response, assume incorrect
543 cached capability (perhaps the peer reneged?) and immediately retry in
546 JH/31 Fix spurious detection of timeout while writing to transport filter.
548 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
549 an attempt to copy the string was made before checking it.
551 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
552 taint of the lookup key was maintained; we now regard the presence in the
553 filesystem as sufficient validation.
555 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
556 argument is supplied.
558 JH/35 The dsearch lookup now requires that the directory is an absolute path.
559 Previously this was not checked, and nonempty relative paths made an
560 access under Exim's current working directory.
562 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
563 Previously no event was raised.
565 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
566 parameter supplied by the sender MAIL FROM command. Previously it was
567 ignored, and only the check_spool_space option value for the required
570 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
571 the size of the signing public-key. Previously it was instead giving
572 the size of the signature hash.
574 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
575 the default. See the (new) dkim_verify_min_keysizes option.
577 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
578 and an ACL use a lookup for checking either the local_part or domain,
579 stale data could be accessed. Ensure that variable references are
580 dropped between messages.
582 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
583 by the client was not checked as pointing within response data before
584 being used. A malicious client could thus cause an out-of-bounds read and
585 possibly gain authentication. Fix by adding the check.
587 JH/42 Internationalisation: change the default for downconversion in the smtp
588 transport to be "if needed". Previously it was "as previously set" for
589 the message, which usually meant "if needed" for message-submission but
590 "no" for everything else. However, MTAs have been seen using SMTPUTF8
591 even when the envelope addresses did not need it, resulting in forwarding
592 failures to non-supporting MTAs. A downconvert in such cases will be
593 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
594 transport. The change does mean that addresses needing conversion will
595 be converted when previously a delivery failure would occur.
597 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
598 response was received it would be used unchecked in a fail-DSN, violating
599 standards on line-length limits. Truncate if needed.
601 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
602 doesn't work. (Additionally add a new main config option to configure the
609 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
610 side, to decode a TLS 1.3 packet capture.
612 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
613 Previously the default library behaviour applied, sending two, each in
616 JH/03 Debug output for ACL now gives the config file name and line number for
619 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
621 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
623 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
624 buffer overrun for (non-chunking) other transports.
626 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
627 TLS1.3, means that a server rejecting a client certificate is not visible
628 to the client until the first read of encrypted data (typically the
629 response to EHLO). Add detection for that case and treat it as a failed
630 TLS connection attempt, so that the normal retry-in-clear can work (if
631 suitably configured).
633 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
634 and/or domain. Found and fixed by Jason Betts.
636 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
637 configuration). If a CNAME target was not a wellformed name pattern, a
640 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
641 the OS reports them interleaved with other addresses.
643 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
644 used both for input and for a verify callout, both encrypted, SMTP
645 responses being sent by the server could be lost. This resulted in
646 dropped connections and sometimes bounces generated by a peer sending
649 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
650 string. Previously it was possible to cause undefined behaviour in a
651 library routine (usually a crash). Found by "zerons".
653 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
656 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
657 API was removed, so update to use the newer ones.
659 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
660 any timeout set, is taking a long time. Previously we would hang on to a
661 rotated logfile "forever" if the input was arriving with long gaps
662 (a previous attempt to fix addressed lack, for a long time, of initial
665 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
666 shared (NFS) environment. The length of the tempfile name is now
667 4 + 16 ("hdr.$message_exim_id") which might break on file
668 systems which restrict the file name length to lower values.
671 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
672 shared (NFS) environment.
674 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
675 did for all versions <4.90). Notably -M, -m, --invert, -I may be
678 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
679 on some platforms for bit 31.
681 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
682 to changes apparently associated with TLS1.3 handling some of the APIs
683 previously used were either nonfunctional or inappropriate. Strings
684 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
685 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
686 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
687 This affects log line X= elements, the $tls_{in,out}_cipher variables,
688 and the use of specific cipher names in the encrypted= ACL condition.
690 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
692 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
693 verification result was not updated unless hosts_require_ocsp applied.
695 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
696 queue_list_requires_admin set to false, non-admin users were denied the
699 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
700 directory-of-certs mode. Previously they were advertised despite the
703 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
704 A single TCP connection by a client will now hold a TLS connection open
705 for multiple message deliveries, by default. Previously the default was to
708 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
709 default. If built with the facility, DANE will be used. The facility
710 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
712 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
713 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
714 must be defined and you must still, unless you define DISABLE_TLS, manage
715 the the include-dir and library-file requirements that go with that
716 choice. Non-TLS builds are still supported.
718 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
721 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
722 default. If the platform supports and has the facility enabled, it will
723 be requested on all coneections.
725 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
726 controlled by the build-time option SUPPORT_PIPE_CONNECT.
728 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
730 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
731 success-DSN messages. Previously the From: header was always the default
732 one for these; the option was ignored.
734 JH/28 Fix the timeout on smtp response to apply to the whole response.
735 Previously it was reset for every read, so a teergrubing peer sending
736 single bytes within the time limit could extend the connection for a
737 long time. Credit to Qualsys Security Advisory Team for the discovery.
739 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
740 delivery address, which leaked information of the results of local
741 forwarding. Change to the original envelope recipient address, per
744 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
745 requested. Previously not bounce was generated and a log entry of
746 error ignored was made.
748 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
750 JH/32 Introduce a general tainting mechanism for values read from the input
751 channel, and values derived from them. Refuse to expand any tainted
752 values, to catch one form of exploit.
754 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
755 was unused and the unexpanded text used for the test. Found and
756 fixed by Ruben Jenster.
758 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
759 an attempt to use a TLS library read routine dereffed a nul pointer,
762 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
763 them in our smtp response.
765 JH/36 Have the general router option retry_use_local_part default to true when
766 any of the restrictive preconditions are set (to anything). Previously it
767 was only for check_local user. The change removes one item of manual
768 configuration which is required for proper retries when a remote router
769 handles a subset of addresses for a domain.
771 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
772 link count into consideration.
774 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
775 caused the extension of big_buffer, the following lines were ignored.
777 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
778 accordance with RFC 2308. Previously there was no expiry, so a longlived
779 receive process (eg. due to ACL delays) versus a short SOA value could
782 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
784 JH/39 Promote DMARC support to mainline.
786 JH/40 Bug 2452: Add a References: header to DSNs.
788 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
789 parameters. The relevant library call is documented as "Deprecated: This
790 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
791 3.6.0, DH parameters are negotiated following RFC7919."
793 HS/06 Change the default of dnssec_request_domains to "*"
795 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
796 carried on and emitted a BDAT command, even when PIPELINING was not
799 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
800 buffer was used for the filename, resulting in a trap when tainted
801 arguments (eg. $domain) were used.
803 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
804 recommended to avoid a possible server-load attack. The feature can be
805 re-enabled via the openssl_options main cofiguration option.
807 JH/45 local_scan API: documented the current smtp_printf() call. This changed
808 for version 4.90 - adding a "more data" boolean to the arguments.
809 Bumped the ABI version number also, this having been missed previously;
810 release versions 4.90 to 4.92.3 inclusive were effectively broken in
811 respect of usage of smtp_printf() by either local_scan code or libraries
812 accessed via the ${dlfunc } expansion item. Both will need coding
813 adjustment for any calls to smtp_printf() to match the new function
814 signature; a FALSE value for the new argument is always safe.
816 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
817 the file-offset (which the Linux syscall does, and exim expects); this
818 resulted in an indefinite loop.
820 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
821 to do ARC verification. The Authentication-Results: header line added
822 by the configuration then had no ARC item.
824 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
825 an incorrect mode was used for reading the result, resulting in it being
832 JH/01 Remove code calling the customisable local_scan function, unless a new
833 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
835 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
836 non-signal-safe functions being used.
838 JH/03 Bug 2269: When presented with a received message having a stupidly large
839 number of DKIM-Signature headers, disable DKIM verification to avoid
840 a resource-consumption attack. The limit is set at twenty.
842 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
843 report of oldest_pass in ${authres } in consequence, and separate out
844 some descriptions of reasons for verification fail.
846 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
847 files in the spool were present and unlocked. A queue-runner could spot
848 them, resulting in a duplicate delivery. Fix that by doing the unlock
849 after the unlink. Investigation by Tim Stewart. Take the opportunity to
850 add more error-checking on spoolfile handling while that code is being
853 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
854 No known attacks, no CVE, this is defensive hardening.
856 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
857 a queue-runner could start a delivery while other operations were ongoing.
858 Cutthrough delivery was a common victim, resulting in duplicate delivery.
859 Found and investigated by Tim Stewart. Fix by using the open message data
860 file handle rather than opening another, and not locally closing it (which
861 releases a lock) for that case, while creating the temporary .eml format
862 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
864 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
865 $sender_verify_failure/$recipient_verify_failure to "random".
867 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
870 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
871 Previously this would segfault.
873 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
876 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
877 like zero, since the resolver should be doing this for us, But we need one
878 as a CNAME but no MX presence gets the CNAME returned; we need to check
879 that doesn't point to an MX to declare it "no MX returned" rather than
880 "error, loop". A new main option is added so the older capability of
881 following some limited number of chain links is maintained.
883 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
885 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
886 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
887 now-deprecated earlier definitions used only specified the range up to TLS
888 1.2 (in the older-version library docs).
890 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
892 JH/15 Rework TLS client-side context management. Stop using a global, and
893 explicitly pass a context around. This enables future use of TLS for
894 connections to service-daemons (eg. malware scanning) while a client smtp
895 connection is using TLS; with cutthrough connections this is quite likely.
897 JH/16 Fix ARC verification to do AS checks in reverse order.
899 JH/17 Support a "tls" option on the ${readsocket } expansion item.
901 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
902 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
903 Previously the "utf8" would be re-prepended for every additional message.
905 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
906 Previously thery were accepted, resulting in issues when attempting to
907 forward messages to a non-supporting MTA.
909 PP/02 Let -n work with printing macros too, not just options.
911 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
912 one parent address was copied, and bogus data was used at delivery-logging
913 time. Either a crash (after delivery) or bogus log data could result.
914 Discovery and analysis by Tim Stewart.
916 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
917 Previously if the string ended mid-character, we did not insert the
918 promised '?' replacement.
920 PP/04 Documentation: current string operators work on bytes, not codepoints.
922 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
923 should pack well giving a smaller memory footprint so better caching and
924 therefore performance. Group the declarations where this can't be done so
925 that the byte-sized flag variables are not interspersed among pointer
926 variables, giving a better chance of good packing by the compiler.
928 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
929 non-null, to avoid issues with sites running BATV. Previously reports were
930 sent with an empty envelope sender so looked like bounces.
932 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
933 The ignore_error flag wasn't being returned from the filter subprocess so
934 was not set for later routers. Investigation and fix by Matthias Kurz.
936 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
937 and a msg:complete for the whole, when a message is manually removed using
938 -Mrm. Developement by Matthias Kurz, hacked on by JH.
940 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
941 a "Gnu special" function, asprintf() in the DB utility binary builds; I
942 hope that is portable enough.
944 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
945 requiring a known-CA anchor certificate; make it now rely entirely on the
946 TLSA as an anchor. Checking the name on the leaf cert against the name
947 on the A-record for the host is still done for TA (but not for EE mode).
949 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
950 included in delivery lines for non-proxied connections, when compiled with
951 SUPPORT_SOCKS and running with proxy logging enabled.
953 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
954 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
955 move the existing event to fire before the normal logging of message
956 failure so that custom logging is bracketed by normal logging.
958 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
959 msg:fail:internal event. Developement by Matthias Kurz.
961 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
962 far too small for todays use of crypto signatures stored there. Go all
963 the way to the max DNS message size of 64kB, even though this might be
964 overmuch for IOT constrained device use.
966 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
967 copy a string over itself. The library routine is documented as not
968 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
970 JH/32 For main options check_spool_space and check_inode_space, where the
971 platform supports 64b integers, support more than the previous 2^31 kB
972 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
973 the previous G, M, k.
975 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
976 $authenticated_fail_id variable on authentication failure. Previously
979 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
980 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
981 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
982 GNUTLS_SEC_PARAM_MEDIUM.
984 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
985 side. Previously we would continue as if no SNI had been received.
987 JH/36 Harden the handling of string-lists. When a list consisted of a sole
988 "<" character, which should be a list-separator specification, we walked
989 off past the nul-terimation.
991 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
992 causes) even when the retry time is not yet met. Previously they were
993 not, meaning that when (say) an account was over-quota and temp-rejecting,
994 and multiple senders' messages were queued, only one sender would get
995 notified on each configured delay_warning cycle.
997 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
999 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1000 cipher-suites, an error can be left on the stack even for a succeeding
1001 accept; this results in impossible error messages when a later operation
1004 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1005 return error codes indicating retry. Under TLS1.3 this becomes required.
1007 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1008 it only wrote the new authenticators, resulting in a lack of tracking of
1009 peer changes of ESMTP extensions until the next cache flush.
1011 JH/41 Fix the loop reading a message header line to check for integer overflow,
1012 and more-often against header_maxsize. Previously a crafted message could
1013 induce a crash of the recive process; now the message is cleanly rejected.
1015 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1016 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1022 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1023 When redis_servers is set to a list of > 1 element, and the Redis servers
1024 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1025 case of MOVED into a DEFER case instead, thus moving the query onto the
1026 next server in the list. For a cluster of N elements, all N servers must
1027 be defined in redis_servers.
1029 GF/02 Catch and remove uninitialized value warning in exiqsumm
1030 Check for existence of @ARGV before looking at $ARGV[0]
1032 JH/01 Replace the store_release() internal interface with store_newblock(),
1033 which internalises the check required to safely use the old one, plus
1034 the allocate and data copy operations duplicated in both (!) of the
1035 extant use locations.
1037 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1038 modifier. This matches the restriction on the commandline.
1040 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1041 Previously only the last row was returned.
1043 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1044 we assumed that tags in the header were well-formed, and parsed the
1045 element content after inspecting only the first char of the tag.
1046 Assumptions at that stage could crash the receive process on malformed
1049 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1050 While running the DKIM ACL we operate on the Permanent memory pool so that
1051 variables created with "set" persist to the DATA ACL. Also (at any time)
1052 DNS lookups that fail create cache records using the Permanent pool. But
1053 expansions release any allocations made on the current pool - so a dnsdb
1054 lookup expansion done in the DKIM ACL releases the memory used for the
1055 DNS negative-cache, and bad things result. Solution is to switch to the
1056 Main pool for expansions.
1057 While we're in that code, add checks on the DNS cache during store_reset,
1058 active in the testsuite.
1059 Problem spotted, and debugging aided, by Wolfgang Breyha.
1061 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1062 When none of the hosts presented to a transport match an already-open
1063 connection, close it and proceed with the list. Previously we would
1064 queue the message. Spotted by Lena with Yahoo, probably involving
1067 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1068 Previously a spurious "250 OK id=" response was appended to the proper
1071 JH/08 The "support for" informational output now, which built with Content
1072 Scanning support, has a line for the malware scanner interfaces compiled
1073 in. Interface can be individually included or not at build time.
1075 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1076 by the template makefile "src/EDITME". The "STREAM" support for an older
1077 ClamAV interface method is removed.
1079 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1080 rows affected is given instead).
1082 JH/11 The runtime Berkeley DB library version is now additionally output by
1083 "exim -d -bV". Previously only the compile-time version was shown.
1085 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1086 SMTP connection. Previously, when one had more recipients than the
1087 first, an abortive onward connection was made. Move to full support for
1088 multiple onward connections in sequence, handling cutthrough connection
1089 for all multi-message initiating connections.
1091 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1092 routers. Previously, a multi-recipient message would fail to match the
1093 onward-connection opened for the first recipient, and cause its closure.
1095 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1096 a timeout on read on a GnuTLS initiating connection, resulting in the
1097 initiating connection being dropped. This mattered most when the callout
1098 was marked defer_ok. Fix to keep the two timeout-detection methods
1101 JH/15 Relax results from ACL control request to enable cutthrough, in
1102 unsupported situations, from error to silently (except under debug)
1103 ignoring. This covers use with PRDR, frozen messages, queue-only and
1106 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1108 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1109 metadata, resulting in a crash in free().
1111 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1112 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1113 Broken also in d185889f4, with init system revamp.
1115 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1116 Previously we abruptly closed the connection after reading a malware-
1117 found indication; now we go on to read the "scan ok" response line,
1120 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1121 ACL. Previously, a crash would result.
1123 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1124 macro text after a replacement (previously it was only once per line) and
1125 by skipping builtin macros when searching for an uppercase lead character.
1127 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1128 for the build is renamed.
1130 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1131 was allocated for every new TLS startup, meaning one per message. Fix
1132 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1134 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1135 reported the original. Fix to report (as far as possible) the ACL
1136 result replacing the original.
1138 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1139 OpenSSL. Certificate information is loaded for every new TLS startup,
1140 and the resources needed to be freed.
1142 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1144 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1147 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1148 DATA response info to the (existing) per-recipient response info for
1149 the "C=" log element. It can have useful tracking info from the
1150 destination system. Patch from Simon Arlott.
1152 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1153 length value. Previously this would segfault.
1155 HS/02 Support Avast multiline protoocol, this allows passing flags to
1156 newer versions of the scanner.
1158 JH/28 Ensure that variables possibly set during message acceptance are marked
1159 dead before release of memory in the daemon loop. This stops complaints
1160 about them when the debug_store option is enabled. Discovered specifically
1161 for sender_rate_period, but applies to a whole set of variables.
1162 Do the same for the queue-runner and queue-list loops, for variables set
1163 from spool message files. Do the same for the SMTP per-message loop, for
1164 certain variables indirectly set in ACL operations.
1166 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1167 as a multi-recipient message from a mailinglist manager). The coding had
1168 an arbitrary cutoff number of characters while checking for more input;
1169 enforced by writing a NUL into the buffer. This corrupted long / fast
1170 input. The problem was exposed more widely when more pipelineing of SMTP
1171 responses was introduced, and one Exim system was feeding another.
1172 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1173 receiving system, and refused recipients seen by the sending system
1174 (propating to people being dropped from mailing lists).
1175 Discovered and pinpointed by David Carter.
1177 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1178 replaced by the ${authresults } expansion.
1180 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1182 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1183 allows proper process termination in container environments.
1185 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1186 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1188 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1189 and "err_perm", deprecated since 4.83 when the RFC-defined words
1190 "temperror" and "permerror" were introduced.
1192 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1193 transport-filters or DKIM-signing. The restriction was lost in the
1194 consolidation of verify-callout and delivery SMTP handling.
1195 Extend the restriction to also cover ARC-signing.
1197 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1198 in defer=pass mode supply a 450 to the initiator. Previously the message
1201 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1202 tls_require_ciphers is used as before.
1204 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1205 "pass_unscanned". Only tmpfails from the scanner are written to
1206 the paniclog, as they may require admin intervention (permission
1207 denied, license issues). Other scanner errors (like decompression
1208 bombs) do not cause a paniclog entry.
1210 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1211 Previously it was possible to log spurious information in receive log
1214 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1215 triggered odd behaviour from Outlook Express clients.
1217 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1220 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1221 since the IETF WG has not yet settled on that versus the original
1222 "bare" representation.
1224 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1225 Previously the millisecond value corrupted the output.
1226 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1227 corrupted the output.
1233 JH/01 Rework error string handling in TLS interface so that the caller in
1234 more cases is responsible for logging. This permits library-sourced
1235 string to be attached to addresses during delivery, and collapses
1236 pairs of long lines into single ones.
1238 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1239 during configuration. Wildcards are allowed and expanded.
1241 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1242 This permits better logging.
1244 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1245 do not close it down and have the child transport start it up again on
1246 the passed-on TCP connection. Instead, proxy the child (and any
1247 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1248 affected: the continued delivery log lines do not have any DNSSEC, TLS
1249 Certificate or OCSP information. TLS cipher information is still logged.
1251 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1252 identical IP addresses on different listening ports. Will also affect
1255 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1256 add noisy ifdef guards to special-case this sillyness.
1257 Patch from Bernd Kuhls.
1259 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1260 than 255 are no longer allowed.
1262 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1263 Disable the session-cache too, which might reduce our load. Since we
1264 currrectly use a new context for every connection, both as server and
1265 client, there is no benefit for these.
1266 GnuTLS appears to not support tickets server-side by default (we don't
1267 call gnutls_session_ticket_enable_server()) but client side is enabled
1268 by default on recent versions (3.1.3 +) unless the PFS priority string
1271 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1272 <https://reproducible-builds.org/specs/source-date-epoch/>.
1274 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1275 the check for any unsuccessful recipients did not notice the limit, and
1276 erroneously found still-pending ones.
1278 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1279 MSG_MORE. Only in-clear (not on TLS connections).
1281 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1282 transport-filter is involved we can buffer the headers in memory for
1283 creating the signature, and read the spool data file once for the
1284 signature and again for transmission.
1286 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1287 4.77 as the kernel support then wasn't solid, having issues in 64bit
1288 mode. Now, it's been long enough. Add support for FreeBSD also.
1290 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1291 case where the routing stage had gathered several addresses to send to
1292 a host before calling the transport for the first, we previously failed
1293 to close down TLS in the old transport process before passing the TCP
1294 connection to the new process. The new one sent a STARTTLS command
1295 which naturally failed, giving a failed delivery and bloating the retry
1296 database. Investigation and fix prototype from Wolfgang Breyha.
1298 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1299 false-negatives in the check that the sender had not preempted a response
1300 or prompt from Exim (running as a server), due to that code's lack of
1301 awareness of the SMTP input buffering.
1303 PP/04 Add commandline_checks_require_admin option.
1304 Exim drops privileges sanely, various checks such as -be aren't a
1305 security problem, as long as you trust local users with access to their
1306 own account. When invoked by services which pass untrusted data to
1307 Exim, this might be an issue. Set this option in main configuration
1308 AND make fixes to the calling application, such as using `--` to stop
1311 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1312 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1313 responses to those, into a single TLS record each way (this usually means
1314 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1317 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1318 affects you only if you're dancing at the edge of the param size limits.
1319 If you are, and this message makes sense to you, then: raise the
1320 configured limit or use OpenSSL 1.1. Nothing we can do for older
1323 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1324 cmdline element to get the documented default one. Previously it was
1327 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1328 are unsuable for non-message-specific reasons - eg. connection timeout,
1329 banner-time rejection.
1331 JH/16 Fix logging of delivery remote port, when specified by router, under
1334 PP/06 Repair manualroute's ability to take options in any order, even if one
1335 is the name of a transport.
1338 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1340 JH/17 Change the list-building routines interface to use the expanding-string
1341 triplet model, for better allocation and copying behaviour.
1343 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1344 Previously it was constructed the first time a possibly-matching string
1345 was met in the configuration file input during startup; now it is done
1348 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1349 compatible one, to avoid the (poorly documented) possibility of a config
1350 file in the working directory redirecting the DB files, possibly correpting
1351 some existing file. CVE-2017-10140 assigned for BDB.
1353 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1354 cache-hot. Previously, although the result was properly cached, the
1355 initial verify call returned a defer.
1357 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1358 the main verify for receipient in uncached-mode.
1360 JH/22 Retire historical build files to an "unsupported" subdir. These are
1361 defined as "ones for which we have no current evidence of testing".
1363 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1364 if present. Previously it was ignored.
1366 JH/24 Start using specified-initialisers in C structure init coding. This is
1367 a C99 feature (it's 2017, so now considered safe).
1369 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1370 if was a fixed-sized field and bitmask ops via macros; it is now more
1373 PP/07 GitHub PR 56: Apply MariaDB build fix.
1374 Patch provided by Jaroslav Škarvada.
1376 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1377 during Coverity cleanups [4.87 JH/47]
1378 Diagnosis and fix provided by Michael Fischer v. Mollard.
1380 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1381 the right size to place the terminating semicolon on its own folded
1382 line, the header hash was calculated to an incorrect value thanks to
1383 the (relaxed) space the fold became.
1385 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1386 and confused the parent.
1388 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1389 which could crash as a result. This could lead to undeliverable messages.
1391 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1394 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1395 matters on fast-turnover and PID-randomising systems, which were getting
1396 out-of-order delivery.
1398 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1399 a possibly-overlapping copy. The symptom was that "Remote host closed
1400 connection in response to HELO" was logged instead of the actual 4xx
1403 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1404 Previously only that bufferd was discarded, resulting in SYMTP command
1407 JH/32 DKIM: when a message has multiple signatures matching an identity given
1408 in dkim_verify_signers, run the dkim acl once for each. Previously only
1409 one run was done. Bug 2189.
1411 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1412 "panic the current process" to "deliberately defer". The panic log is
1413 still written with the problem list name; the mail and reject logs now
1414 get a temp-reject line for the message that was being handled, saying
1415 something like "domains check lookup or other defer". The SMTP 451
1416 message is still "Temporary local problem".
1418 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1419 A crafted sequence of BDAT commands could result in in-use memory beeing
1420 freed. CVE-2017-16943.
1422 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1423 from SMTP input. Previously it was always done; now only done for DATA
1424 and not BDAT commands. CVE-2017-16944.
1426 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1427 to the message (such as an overlong header line). Previously this was
1428 not done and we did not exit BDAT mode. Followon from the previous item
1429 though a different problem.
1435 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1436 than -2003 did; needs libidn2 in addition to libidn.
1438 JH/02 The path option on a pipe transport is now expanded before use.
1440 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1441 Patch provided by "Björn", documentation fix added too.
1443 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1444 missing a wire-to-host endian conversion.
1446 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1447 close after a BDAT command line could be taken as a following command,
1448 giving a synch failure. Fix by only checking for synch immediately
1449 before acknowledging the chunk.
1451 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1452 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1453 macro. Patches provided by Josh Soref.
1455 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1456 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1457 by IANA as a service extension. However, John Klensin suggests that we
1460 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1461 the dkim code may be unix-mode line endings rather than smtp wire-format
1462 CRLF, so prepend a CR to any bare LF.
1464 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1465 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1467 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1468 the first were themselves being wrongly included in the feed into dkim
1469 processing; with most chunk sizes in use this resulted in an incorrect
1470 body hash calculated value.
1472 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1473 DKIM signature block, for verification. Although advised against by
1474 standards it is specifically not ruled illegal.
1476 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1478 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1479 missing a body hash (the bh= tag).
1481 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1482 It seems that HAProxy sends the Proxy Protocol information in clear and
1483 only then does a TLS startup, so do the same.
1485 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1486 TCP connections (such as for Spamd) unless the daemon successfully set
1487 Fast Open mode on its listening sockets. This fixes breakage seen on
1488 too-old kernels or those not configured for Fast Open, at the cost of
1489 requiring both directions being enabled for TFO, and TFO never being used
1490 by non-daemon-related Exim processes.
1492 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1493 endings, at least on the first header line. Try to canonify any that get
1494 past that check, despite the cost.
1496 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1497 now limited to an arbitrary five deep, while parsing addresses with the
1498 strip_excess_angle_brackets option enabled.
1500 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1501 instead leave the unprompted TLS handshake in socket buffer for the
1502 TLS library to consume.
1504 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1506 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1508 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1509 region in one operation, for speed, and this leaves assigned pointers
1510 dangling. Add checks run only under the testsuite which checks all
1511 variables at a store-reset and panics on a dangling pointer; add code
1512 explicitly nulling out all the variables discovered. Fixes one known
1513 bug: a transport crash, where a dangling pointer for $sending_ip_address
1514 originally assigned in a verify callout, is re-used.
1516 PP/06 Drop '.' from @INC in various Perl scripts.
1518 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1520 PP/08 Reduce a number of compilation warnings under clang; building with
1521 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1522 should be warning-free.
1524 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1526 HS/01 Fix portability problems introduced by PP/08 for platforms where
1527 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1529 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1530 chunk. This allows us to accept broken chunked messages. We need a more
1531 general solution here.
1533 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1534 already-broken messages in the queue.
1536 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1538 JH/19 Fix reference counting bug in routing-generated-address tracking.
1544 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1545 supports it and a size is available (ie. the sending peer gave us one).
1547 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1548 years of being deprecated). The replacements are the ACLs
1549 acl_smtp_mime and acl_not_smtp_mime.
1551 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1552 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1553 or even an in-clear connection were permitted. Now, if the host lookup
1554 was dnssec and dane was requested then the host is only used if the
1555 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1556 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1557 if one fails this test.
1558 This means that a poorly-configured remote DNS will make it incommunicado;
1559 but it protects against a DNS-interception attack on it.
1561 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1562 non-noisy when a race steals the message being considered.
1564 JH/05 If main configuration option tls_certificate is unset, generate a
1565 self-signed certificate for inbound TLS connections.
1567 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1568 in rewrites and routers.
1570 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1571 and logged a warning sing 4.83; now they are a configuration file error.
1573 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1574 (lacking @domain). Apply the same qualification processing as RCPT.
1576 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1578 JH/10 Support ${sha256:} applied to a string (as well as the previous
1581 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1582 a cutthrough deliver is pending, as we always want to make a connection.
1583 This also avoids re-routing the message when later placing the cutthrough
1584 connection after a verify cache hit.
1585 Do not update it with the verify result either.
1587 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1588 when routing results in more than one destination address.
1590 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1591 signing (which inhibits the cutthrough capability). Previously only
1592 the presence of an option was tested; now an expansion evaluating as
1593 empty is permissible (obviously it should depend only on data available
1594 when the cutthrough connection is made).
1596 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1597 the relevant preceding SMTP command did not note the pipelining mode.
1599 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1600 Previously they were not counted.
1602 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1603 as one having no matching records. Previously we deferred the message
1604 that needed the lookup.
1606 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1607 distinguished as "(=".
1609 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1610 for missing MX records. Previously it only worked for missing A records.
1612 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1614 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1615 after the data-go-ahead and data-ack. Patch from Jason Betts.
1617 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1618 even for a "none" policy. Patch from Tony Meyer.
1620 JH/22 Fix continued use of a connection for further deliveries. If a port was
1621 specified by a router, it must also match for the delivery to be
1624 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1625 When one of the recipients of a message was unsuitable for the connection
1626 (has no matching addresses), we lost track of needing to mark it
1627 deferred. As a result mail would be lost.
1629 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1631 JH/25 Decoding ACL controls is now done using a binary search; the source code
1632 takes up less space and should be simpler to maintain. Merge the ACL
1633 condition decode tables also, with similar effect.
1635 JH/26 Fix problem with one_time used on a redirect router which returned the
1636 parent address unchanged. A retry would see the parent address marked as
1637 delivered, so not attempt the (identical) child. As a result mail would
1640 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1641 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1642 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1645 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1646 Main config options check_{log,spool}_{inodes,space} are now
1647 100 inodes, 10MB unless set otherwise in the configuration.
1649 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1650 Previously it only applied to the main-section connection policy
1653 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1655 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1656 by me. Added RFC7919 DH primes as an alternative.
1658 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1659 are not in the system include path.
1661 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1662 GnuTLS, when a session startup failed (eg because the client disconnected)
1663 Exim did stdio operations after fclose. This was exposed by a recent
1664 change which nulled out the file handle after the fclose.
1666 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1667 signed directly by the cert-signing cert, rather than an intermediate
1668 OCSP-signing cert. This is the model used by LetsEncrypt.
1670 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1672 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1673 an incoming connection.
1675 HS/02 Bug 1802: Do not half-close the connection after sending a request
1678 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1679 fallback to "prime256v1".
1681 JH/34 SECURITY: Use proper copy of DATA command in error message.
1682 Could leak key material. Remotely exploitable. CVE-2016-9963.
1688 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1689 and 3.4.4 - once the server is enabled to respond to an OCSP request
1690 it does even when not requested, resulting in a stapling non-aware
1691 client dropping the TLS connection.
1693 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1694 support variable-length bit vectors. No functional change.
1696 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1697 The I= interface field on outgoing lines is now after the H= remote
1698 host field, same as incoming lines. There is a separate
1699 outgoing_interface log selector which allows you to disable the
1702 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1703 If not running log_selector +smtp_connection the mainlog would be held
1704 open indefinitely after a "too many connections" event, including to a
1705 deleted file after a log rotate. Leave the per net connection logging
1706 leaving it open for efficiency as that will be quickly detected by the
1707 check on the next write.
1709 HS/01 Bug 1671: Fix post transport crash.
1710 Processing the wait-<transport> messages could crash the delivery
1711 process if the message IDs didn't exist for some reason. When
1712 using 'split_spool_directory=yes' the construction of the spool
1713 file name failed already, exposing the same netto behaviour.
1715 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1716 mime_regex ACL conditions.
1718 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1719 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1720 to HELO, local diagnostic string.
1722 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1723 log line to debug. Even when configured with a tls authenticator many
1724 client connections are expected to not authenticate in this way, so
1725 an authenticate fail is not an error.
1727 HS/02 Add the Exim version string to the process info. This way exiwhat
1728 gives some more detail about the running daemon.
1730 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1731 matter for fast-change records such as DNSBLs.
1733 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1734 retry record, even if constant. There may be multiple transports with
1735 different interface settings and the retry behaviour needs to be kept
1738 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1740 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1742 JH/10 Bug 840: fix log_defer_output option of pipe transport
1744 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1745 hosts_randomize. This matters a lot when combined with mua_wrapper.
1747 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1748 ${quote_pgsql:<string>} operator.
1750 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1752 JH/14 Change header file name for freeradius-client. Relevant if compiling
1753 with Radius support; from the Gentoo tree and checked under Fedora.
1755 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1757 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1758 option transport_filter, meaning no filtering.
1760 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1762 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1763 defaults to "*" (all hosts). The variable is now available when not built
1764 with TLS, default unset, mainly to enable keeping the testsuite sane.
1765 If a server certificate is not supplied (via tls_certificate) an error is
1766 logged, and clients will find TLS connections fail on startup. Presumably
1767 they will retry in-clear.
1768 Packagers of Exim are strongly encouraged to create a server certificate
1769 at installation time.
1771 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1772 with the $config_file variable.
1774 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1775 in transport context, after the attempt, and per-recipient. The latter type
1776 is per host attempted. The event data is the error message, and the errno
1777 information encodes the lookup type (A vs. MX) used for the (first) host,
1778 and the trailing two digits of the smtp 4xx response.
1780 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1781 to write to mainlog (or rejectlog, paniclog) in the window between file
1782 creation and permissions/ownership being changed. Particularly affects
1783 installations where exicyclog is run as root, rather than exim user;
1784 result is that the running daemon panics and dies.
1786 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1788 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1789 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1790 "pri" and "weight". Note that the previous implicit priority given by the
1791 list order is no longer honoured.
1793 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1794 for DKIM processing.
1796 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1797 by defining SUPPORT_SOCKS.
1799 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1800 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1801 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1802 variables are renamed to proxy_{local,external}_{address,port}.
1804 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1805 for a build by defining SUPPORT_I18N
1807 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1808 of the query string, and make ${quote_redis:} do that quoting.
1810 JH/29 Move Events support from Experimental to mainline, enabled by default
1811 and removable for a build by defining DISABLE_EVENT.
1813 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1815 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1816 cached by the daemon.
1818 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1819 by defining LOOKUP_REDIS. The libhiredis library is required.
1821 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1822 keys are given for lookup.
1824 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1825 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1826 only supported when built with TLS support. The PolarSSL SHA routines
1827 are still used when the TLS library is too old for convenient support.
1829 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1830 openssl_options), for security. OpenSSL forces this from version 1.1.0
1831 server-side so match that on older versions.
1833 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1834 allocation for $value could be released as the expansion processing
1835 concluded, but leaving the global pointer active for it.
1837 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1838 and to use the domains and local_parts ACL conditions.
1840 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1841 incorrectly not doubled on cutthrough transmission, hence seen as a
1842 body-termination at the receiving system - resulting in truncated mails.
1843 Commonly the sender saw a TCP-level error, and retransmitted the message
1844 via the normal store-and-forward channel. This could result in duplicates
1845 received - but deduplicating mailstores were liable to retain only the
1846 initial truncated version.
1848 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1850 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1852 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1853 we're in there, support oversigning also; bug 1309.
1855 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1857 HS/04 Add support for keep_environment and add_environment options.
1859 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1860 either intentional arithmetic overflow during PRNG, or testing config-
1863 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1864 delivery resulted in actual delivery. Cancel cutthrough before DATA
1867 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1868 rejects a recipient: pass the reject to the originator.
1870 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1871 Many were false-positives and ignorable, but it's worth fixing the
1874 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1875 for the new environment-manipulation done at startup. Move the routines
1876 from being local to tls.c to being global via the os.c file.
1878 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1879 an extract embedded as result-arg for a map, the first arg for extract
1880 is unavailable so we cannot tell if this is a numbered or keyed
1881 extraction. Accept either.
1887 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1890 JH/02 The smtp transport option "multi_domain" is now expanded.
1892 JH/03 The smtp transport now requests PRDR by default, if the server offers
1895 JH/04 Certificate name checking on server certificates, when exim is a client,
1896 is now done by default. The transport option tls_verify_cert_hostnames
1897 can be used to disable this per-host. The build option
1898 EXPERIMENTAL_CERTNAMES is withdrawn.
1900 JH/05 The value of the tls_verify_certificates smtp transport and main options
1901 default to the word "system" to access the system default CA bundle.
1902 For GnuTLS, only version 3.0.20 or later.
1904 JH/06 Verification of the server certificate for a TLS connection is now tried
1905 (but not required) by default. The verification status is now logged by
1906 default, for both outbound TLS and client-certificate supplying inbound
1909 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1912 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1913 Status Notification (bounce) messages are now MIME format per RFC 3464.
1914 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1915 under the control of the dsn_advertise_hosts option, and routers may
1916 have a dsn_lasthop option.
1918 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1919 default, modifiable by a malware= option. The list separator for
1920 the options can now be changed in the usual way. Bug 68.
1922 JH/10 The smtp_receive_timeout main option is now expanded before use.
1924 JH/11 The incoming_interface log option now also enables logging of the
1925 local interface on delivery outgoing connections.
1927 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1928 if the interface and destination host and port all match.
1930 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1933 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1934 Patch from Andrew Lewis.
1936 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1937 now supports optional time-restrictions, weighting, and priority
1938 modifiers per server. Patch originally by <rommer@active.by>.
1940 JH/16 The spamd_address main option now supports a mixed list of local
1941 and remote servers. Remote servers can be IPv6 addresses, and
1942 specify a port-range.
1944 JH/17 Bug 68: The spamd_address main option now supports an optional
1945 timeout value per server.
1947 JH/18 Bug 1581: Router and transport options headers_add/remove can
1948 now have the list separator specified.
1950 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1953 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1956 JH/21 Support for the A6 type of dns record is withdrawn.
1958 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1959 rather than the verbs used.
1961 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1962 from 255 to 1024 chars.
1964 JH/24 Verification callouts now attempt to use TLS by default.
1966 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1967 are generic router options now. The defaults didn't change.
1969 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1970 Original patch from Alexander Shikoff, worked over by JH.
1972 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1973 files if ZCAT_COMMAND is not executable.
1975 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1977 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1979 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1980 Normally benign, it bites when the pair was led to by a CNAME;
1981 modern usage is to not canonicalize the domain to a CNAME target
1982 (and we were inconsistent anyway for A-only vs AAAA+A).
1984 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1986 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1987 when evaluating $sender_host_dnssec.
1989 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1990 $sender_helo_dnssec variable.
1992 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1994 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1996 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1998 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1999 documented as working, but never had. Support all but $spam_report.
2001 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2002 added for tls authenticator.
2004 HS/03 Add perl_taintmode main config option
2010 TL/01 When running the test suite, the README says that variables such as
2011 no_msglog_check are global and can be placed anywhere in a specific
2012 test's script, however it was observed that placement needed to be near
2013 the beginning for it to behave that way. Changed the runtest perl
2014 script to read through the entire script once to detect and set these
2015 variables, reset to the beginning of the script, and then run through
2016 the script parsing/test process like normal.
2018 TL/02 The BSD's have an arc4random API. One of the functions to induce
2019 adding randomness was arc4random_stir(), but it has been removed in
2020 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2021 function when detected.
2023 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2024 cause callback expansion.
2026 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2027 syntax errors in an expansion can be treated as a string instead of
2028 logging or causing an error, due to the internal use of bool_lax
2029 instead of bool when processing it.
2031 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2032 server certificates when making smtp deliveries.
2034 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2036 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2038 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2040 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2041 Merged patch from Sebastian Wiedenroth.
2043 JH/05 Fix results-pipe from transport process. Several recipients, combined
2044 with certificate use, exposed issues where response data items split
2045 over buffer boundaries were not parsed properly. This eventually
2046 resulted in duplicates being sent. This issue only became common enough
2047 to notice due to the introduction of connection certificate information,
2048 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2050 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2051 size buffer was used, resulting in syntax errors when an expansion
2054 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2055 version 3.3.6 or later.
2057 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2058 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2059 both become "event_action", the variables become $event_name, $event_data
2060 and $event_defer_errno. There is a new variable $verify_mode, usable in
2061 routers, transports and related events. The tls:cert event is now also
2062 raised for inbound connections, if the main configuration event_action
2065 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2066 early OCSP support, but no stapling (appears to be less than 1.0.0).
2068 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2069 server certificate names available under the smtp transport option
2070 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2073 JH/10 Time-related extraction expansions from certificates now use the main
2074 option "timezone" setting for output formatting, and are consistent
2075 between OpenSSL and GnuTLS compilations. Bug 1541.
2077 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2078 encoded parameter in the incoming message. Bug 1558.
2080 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2081 include certificate info, eximon was claiming there were spoolfile
2084 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2086 JH/14 Log delivery-related information more consistently, using the sequence
2087 "H=<name> [<ip>]" wherever possible.
2089 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2090 are problematic for Debian distribution, omit them from the release
2093 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2095 JH/16 Fix string representation of time values on 64bit time_t architectures.
2098 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2099 output list separator was used.
2104 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2105 checkers that were complaining about end of non-void function with no
2108 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2109 This was a regression introduced in 4.83 by another bugfix.
2111 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2113 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2114 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2120 TF/01 Correctly close the server side of TLS when forking for delivery.
2122 When a message was received over SMTP with TLS, Exim failed to clear up
2123 the incoming connection properly after forking off the child process to
2124 deliver the message. In some situations the subsequent outgoing
2125 delivery connection happened to have the same fd number as the incoming
2126 connection previously had. Exim would try to use TLS and fail, logging
2127 a "Bad file descriptor" error.
2129 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2130 utilities have not been installed.
2132 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2133 temporary space as the ACL may create new global variables.
2135 TL/01 LDAP support uses per connection or global context settings, depending
2136 upon the detected version of the libraries at build time.
2138 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2139 to extract and use the src ip:port in logging and expansions as if it
2140 were a direct connection from the outside internet. PPv2 support was
2141 updated based on HAProxy spec change in May 2014.
2143 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2145 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2146 Properly escape header and check for NULL return.
2148 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2151 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2153 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2154 characters in header names, implemented as a verify condition.
2155 Contributed by Michael Fischer v. Mollard.
2157 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2158 results permerror and temperror. Previous values are deprecated but
2159 still accepted. In a future release, err_perm and err_temp will be
2160 completely removed, which will be a backward incompatibility if the
2161 ACL tests for either of these two old results. Patch contributed by
2162 user bes-internal on the mailing list.
2164 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2166 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2167 selectors, in both main and reject logs.
2169 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2172 JH/07 Add malware type "sock" for talking to simple daemon.
2174 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2176 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2177 routers/transports under cutthrough routing.
2179 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2180 numbers. Touch up "bool" conditional to keep the same definition.
2182 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2184 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2186 JH/12 Expand items in router/transport headers_add or headers_remove lists
2187 individually rather than the list as a whole. Bug 1452.
2189 Required for reasonable handling of multiple headers_ options when
2190 they may be empty; requires that headers_remove items with embedded
2191 colons must have them doubled (or the list-separator changed).
2193 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2194 view the policy declared in the DMARC record. Currently, $dmarc_status
2195 is a combined value of both the record presence and the result of the
2198 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2200 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2201 dnslookup router and the smtp transport (applying to the forward
2204 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2205 of ldap servers used for a specific lookup. Patch provided by Heiko
2208 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2209 New variable $lookup_dnssec_authenticated for observability.
2211 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2212 Patch submitted by Lars Timman.
2214 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2216 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2217 Requires trusted mode and valid format message id, aborts otherwise.
2218 Patch contributed by Heiko Schlichting.
2220 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2221 certextract with support for various fields. Bug 1358.
2223 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2224 is requested by default, modifiable by smtp transport option
2227 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2228 operate on certificate variables to give certificate fingerprints
2229 Also new ${sha256:cert_variable}.
2231 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2233 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2234 Christian Aistleitner.
2236 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2238 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2239 file. Patch from Wolfgang Breyha.
2241 JH/25 Expand the coverage of the delivery $host and $host_address to
2242 client authenticators run in verify callout. Bug 1476.
2244 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2245 align with daemon_smtp_ports. Bug 72.
2247 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2248 support and error reporting did not work properly.
2250 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2251 and is readable. Patch from Andrew Colin Kissa.
2253 TL/14 Enhance documentation of ${run expansion and how it parses the
2254 commandline after expansion, particularly in the case when an
2255 unquoted variable expansion results in an empty value.
2257 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2259 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2260 by renaming away from it. Addresses GH issue 15, reported by
2263 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2265 TL/15 SECURITY: prevent double expansion in math comparison functions
2266 (can expand unsanitized data). Not remotely exploitable.
2273 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2275 PP/02 Make -n do something, by making it not do something.
2276 When combined with -bP, the name of an option is not output.
2278 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2281 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2282 $sender_host_name and config options to manage this, and basic check
2285 PP/05 DSCP support for outbound connections and control modifier for inbound.
2287 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2288 (Only plugin which currently uses this is kerberos4, which nobody should
2289 be using, but we should make it available and other future plugins might
2290 conceivably use it, even though it would break NAT; stuff *should* be
2291 using channel bindings instead).
2293 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2294 name; added for Sendmail compatibility; requires admin caller.
2295 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2296 just ignore it); requires trusted caller.
2297 Also parse but ignore: -Ac -Am -X<logfile>
2300 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2302 TL/02 Add +smtp_confirmation as a default logging option.
2304 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2305 Patch by Magnus Holmgren from 2007-02-20.
2307 TL/04 Bugzilla 1281 - Spec typo.
2308 Bugzilla 1283 - Spec typo.
2309 Bugzilla 1290 - Spec grammar fixes.
2311 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2313 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2315 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2316 mailing list by by Dmitry Isaikin.
2318 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2320 JH/02 Support "G" suffix to numbers in ${if comparisons.
2322 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2324 NM/01 Bugzilla 1197 - Spec typo
2325 Bugzilla 1196 - Spec examples corrections
2327 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2329 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2330 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2333 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2334 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2336 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2337 "acl {{name}{arg}...}", and optional args on acl condition
2340 JH/05 Permit multiple router/transport headers_add/remove lines.
2342 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2344 JH/07 Avoid using a waiting database for a single-message-only transport.
2345 Performance patch from Paul Fisher. Bugzilla 1262.
2347 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2350 JH/09 Add $headers_added variable, with content from use of ACL modifier
2351 add_header (but not yet added to the message). Bugzilla 199.
2353 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2354 Pulled from Bugzilla 817 by Wolfgang Breyha.
2356 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2358 (nb: this is the same fix as in Exim 4.80.1)
2360 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2363 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2366 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2368 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2369 advertise SMTP AUTH mechanism to us, instead of a generic
2370 protocol violation error. Also, make Exim more robust to bad
2371 data from the Dovecot auth socket.
2373 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2375 When a queue runner is handling a message, Exim first routes the
2376 recipient addresses, during which it prunes them based on the retry
2377 hints database. After that it attempts to deliver the message to
2378 any remaining recipients. It then updates the hints database using
2381 So if a recipient address works intermittently, it can get repeatedly
2382 deferred at routing time. The retry hints record remains fresh so the
2383 address never reaches the final cutoff time.
2385 This is a fairly common occurrence when a user is bumping up against
2386 their storage quota. Exim had some logic in its local delivery code
2387 to deal with this. However it did not apply to per-recipient defers
2388 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2390 This change adds a proper retry rule check during routing so that the
2391 final cutoff time is checked against the message's age. We only do
2392 this check if there is an address retry record and there is not a
2393 domain retry record; this implies that previous attempts to handle
2394 the address had the retry_use_local_parts option turned on. We use
2395 this as an approximation for the destination being like a local
2396 delivery, as in LMTP.
2398 I suspect this new check makes the old local delivery cutoff check
2399 redundant, but I have not verified this so I left the code in place.
2401 TF/02 Correct gecos expansion when From: is a prefix of the username.
2403 Test 0254 submits a message to Exim with the header
2407 When I ran the test suite under the user fanf2, Exim expanded
2408 the header to contain my full name, whereas it should have added
2409 a Resent-Sender: header. It erroneously treats any prefix of the
2410 username as equal to the username.
2412 This change corrects that bug.
2414 GF/01 DCC debug and logging tidyup
2415 Error conditions log to paniclog rather than rejectlog.
2416 Debug lines prefixed by "DCC: " to remove any ambiguity.
2418 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2420 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2421 Bug spotted by Jeremy Harris; was flawed since initial commit.
2422 Would have resulted in OCSP responses post-SNI triggering an Exim
2423 NULL dereference and crash.
2425 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2427 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2428 Bug detection, analysis and fix by Samuel Thibault.
2429 Bugzilla 1331, Debian bug #698092.
2431 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2433 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2434 Server implementation by Todd Lyons, client by JH.
2435 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2436 config variable "prdr_enable" controls whether the server
2437 advertises the facility. If the client requests PRDR a new
2438 acl_data_smtp_prdr ACL is called once for each recipient, after
2439 the body content is received and before the acl_smtp_data ACL.
2440 The client is controlled by both of: a hosts_try_prdr option
2441 on the smtp transport, and the server advertisement.
2442 Default client logging of deliveries and rejections involving
2443 PRDR are flagged with the string "PRDR".
2445 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2446 fclose(). Diagnosis by Todd Lyons.
2448 PP/17 Update configure.default to handle IPv6 localhost better.
2449 Patch by Alain Williams (plus minor tweaks).
2452 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2453 This is now consistent with GnuTLS, and is now documented: the
2454 previous undocumented portable approach to treating the option as
2455 unset was to force an expansion failure. That still works, and
2456 an empty string is now equivalent.
2458 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2459 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2460 not performing validation itself.
2462 PP/20 Added force_command boolean option to pipe transport.
2463 Patch from Nick Koston, of cPanel Inc.
2465 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2468 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2470 PP/21 Fix eximon continuous updating with timestamped log-files.
2471 Broken in a format-string cleanup in 4.80, missed when I repaired the
2472 other false fix of the same issue.
2473 Report and fix from Heiko Schlichting.
2476 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2477 Report from Prashanth Katuri.
2479 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2480 It's SecureTransport, so affects any MacOS clients which use the
2481 system-integrated TLS libraries, including email clients.
2483 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2484 using a MIME ACL for non-SMTP local injection.
2485 Report and assistance in diagnosis by Warren Baker.
2487 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2489 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2491 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2492 last id that failed so it may be referenced in subsequent ACL's.
2494 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2497 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2498 ldap library initialization, allowing self-signed CA's to be
2499 used. Also properly sets require_cert option later in code by
2500 using NULL (global ldap config) instead of ldap handle (per
2501 session). Bug diagnosis and testing by alxgomz.
2503 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2504 the src/util/ subdirectory.
2506 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2507 renamed to Transport Post Delivery Action by Jeremy Harris, as
2510 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2511 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2512 redis_servers = needs to be configured which will be used by the redis
2513 lookup. Patch from Warren Baker, of The Packet Hub.
2515 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2517 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2518 hostname or reverse DNS when processing a host list. Used suggestions
2519 from multiple comments on this bug.
2521 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2523 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2524 Missed a few lines, added it to make the runtest require no keyboard
2527 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2528 contains upper case chars. Make router use caseful_local_part.
2530 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2531 support when GnuTLS has been built with p11-kit.
2537 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2539 This, or similar/improved, will also be change PP/11 of 4.82.
2545 PP/01 Handle short writes when writing local log-files.
2546 In practice, only affects FreeBSD (8 onwards).
2547 Bugzilla 1053, with thanks to Dmitry Isaikin.
2549 NM/01 Bugzilla 949 - Documentation tweak
2551 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2554 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2556 PP/02 Implemented gsasl authenticator.
2558 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2560 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2561 `pkg-config foo` for cflags/libs.
2563 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2564 with rest of GSASL and with heimdal_gssapi.
2566 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2567 `pkg-config foo` for cflags/libs for the TLS implementation.
2569 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2570 properties get this fed in as external SSF. A number of robustness
2571 and debugging improvements to the cyrus_sasl authenticator.
2573 PP/08 cyrus_sasl server now expands the server_realm option.
2575 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2576 Patch by Jeremy Harris.
2578 PP/10 Added dbmjz lookup type.
2580 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2582 PP/12 MAIL args handles TAB as well as SP, for better interop with
2583 non-compliant senders.
2584 Analysis and variant patch by Todd Lyons.
2586 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2587 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2588 Patch from Dirk Mueller <dmueller@suse.com>
2590 PP/13 tls_peerdn now print-escaped for spool files.
2591 Observed some $tls_peerdn in wild which contained \n, which resulted
2592 in spool file corruption.
2594 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2595 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2596 or write after TLS renegotiation, which otherwise led to messages
2599 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2600 as a tracking header (ie: a signed header comes before the signature).
2601 Patch from Wolfgang Breyha.
2603 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2604 comma-sep list; embedded commas doubled.
2606 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2608 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2610 Report and patch from Dmitry Banschikov.
2612 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2613 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2614 failures. We appear to now support TLS1.1+ with Exim.
2616 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2617 lets Exim select keys and certificates based upon TLS SNI from client.
2618 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2619 before an outbound SMTP session. New log_selector, +tls_sni.
2621 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2622 NULL dereference. Report and patch from Alun Jones.
2624 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2625 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2626 Not seeing resolver debug output on NetBSD, but suspect this is a
2627 resolver implementation change.
2629 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2630 Left warnings. Added "eximon gdb" invocation mode.
2632 PP/21 Defaulting "accept_8bitmime" to true, not false.
2634 PP/22 Added -bw for inetd wait mode support.
2636 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2637 locate the relevant includes and libraries. Made this the default.
2639 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2640 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2642 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2643 This may cause build issues on older platforms.
2645 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2646 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2647 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2648 Added SNI support via GnuTLS too.
2649 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2651 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2653 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2654 Applied second patch from Janne, implementing suggestion to default
2655 multiple-strings-in-record handling to match SPF spec.
2657 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2659 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2660 read-only, out of scope).
2661 Patch from Wolfgang Breyha, report from Stuart Northfield.
2663 PP/29 Fix three issues highlighted by clang analyser static analysis.
2664 Only crash-plausible issue would require the Cambridge-specific
2665 iplookup router and a misconfiguration.
2666 Report from Marcin Mirosław.
2668 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2670 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2671 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2672 As part of this, removing so much warning spew let me fix some minor
2673 real issues in debug logging.
2675 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2676 assignment on my part. Fixed.
2678 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2679 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2680 Janne Snabb (who went above and beyond: thank you).
2682 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2683 string otherwise requires a connection and a bunch more work and it's
2684 relatively easy to get wrong. Should also expose TLS library linkage
2687 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2688 64-bit ${eval} (JH/03).
2690 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2691 GNU libc to support some of the 64-bit stuff, should not lead to
2692 conflicts. Defined before os.h is pulled in, so if a given platform
2693 needs to override this, it can.
2695 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2696 protection layer was required, which is not implemented.
2697 Bugzilla 1254, patch from Wolfgang Breyha.
2699 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2700 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2701 tls_dhparam take prime identifiers. Also unbreak combination of
2702 OpenSSL+DH_params+TLSSNI.
2704 PP/39 Disable SSLv2 by default in OpenSSL support.
2710 PP/01 Solaris build fix for Oracle's LDAP libraries.
2711 Bugzilla 1109, patch from Stephen Usher.
2713 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2715 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2718 TF/02 Fix a couple more cases where we did not log the error message
2719 when unlink() failed. See also change 4.74-TF/03.
2721 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2722 lock up or crash if it happened to be inside a call to libc when it
2723 got a SIGUSR1 from exiwhat.
2725 The SIGUSR1 handler appends the current process status to the process
2726 log which is later printed by exiwhat. It used to use the general
2727 purpose logging code to do this, but several functions it calls are
2728 not safe for signals.
2730 The new output code in the SIGUSR1 handler is specific to the process
2731 log, and simple enough that it's easy to inspect for signal safety.
2732 Removing some special cases also simplifies the general logging code.
2733 Removing the spurious timestamps from the process log simplifies
2736 TF/04 Improved ratelimit ACL condition.
2738 The /noupdate option has been deprecated in favour of /readonly which
2739 has clearer semantics. The /leaky, /strict, and /readonly update modes
2740 are mutually exclusive. The update mode is no longer included in the
2741 database key; it just determines when the database is updated. (This
2742 means that when you upgrade Exim will forget old rate measurements.)
2744 Exim now checks that the per_* options are used with an update mode that
2745 makes sense for the current ACL. For example, when Exim is processing a
2746 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2747 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2748 must specify per_mail/readonly. If you omit the update mode it defaults to
2749 /leaky where that makes sense (as before) or /readonly where required.
2751 The /noupdate option is now undocumented but still supported for
2752 backwards compatibility. It is equivalent to /readonly except that in
2753 ACLs where /readonly is required you may specify /leaky/noupdate or
2754 /strict/noupdate which are treated the same as /readonly.
2756 A useful new feature is the /count= option. This is a generalization
2757 of the per_byte option, so that you can measure the throughput of other
2758 aggregate values. For example, the per_byte option is now equivalent
2759 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2761 The per_rcpt option has been generalized using the /count= mechanism
2762 (though it's more complicated than the per_byte equivalence). When it is
2763 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2764 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2765 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2766 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2767 Note that using per_rcpt with a non-readonly update mode in more than
2768 one ACL will cause the recipients to be double-counted. (The per_mail
2769 and per_byte options don't have this problem.)
2771 The handling of very low rates has changed slightly. If the computed rate
2772 is less than the event's count (usually one) then this event is the first
2773 after a long gap. In this case the rate is set to the same as this event's
2774 count, so that the first message of a spam run is counted properly.
2776 The major new feature is a mechanism for counting the rate of unique
2777 events. The new per_addr option counts the number of different
2778 recipients that someone has sent messages to in the last time period. It
2779 behaves like per_rcpt if all the recipient addresses are different, but
2780 duplicate recipient addresses do not increase the measured rate. Like
2781 the /count= option this is a general mechanism, so the per_addr option
2782 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2783 example, measure the rate that a client uses different sender addresses
2784 with the options per_mail/unique=$sender_address. There are further
2785 details in the main documentation.
2787 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2789 TF/06 Removed a few PCRE remnants.
2791 TF/07 Automatically extract Exim's version number from tags in the git
2792 repository when doing development or release builds.
2794 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2795 Bugzilla 879. Patch from Paul Fisher.
2797 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2798 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2801 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2803 PP/05 Variable $av_failed, true if the AV scanner deferred.
2804 Bugzilla 1078. Patch from John Horne.
2806 PP/06 Stop make process more reliably on build failure.
2807 Bugzilla 1087. Patch from Heiko Schlittermann.
2809 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2810 Bugzilla 1089. Patch from Heiko Schlittermann.
2812 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2813 Bugzilla 1131. Patch from Holger Weiß.
2815 PP/09 Handle IPv6 addresses with SPF.
2816 Bugzilla 860. Patch from Wolfgang Breyha.
2818 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2820 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2823 PP/11 match_* no longer expand right-hand-side by default.
2824 New compile-time build option, EXPAND_LISTMATCH_RHS.
2825 New expansion conditions, "inlist", "inlisti".
2827 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2829 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2831 PP/14 fix log_write() format string regression from TF/03.
2832 Bugzilla 1152. Patch from Dmitry Isaikin.
2838 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2840 PP/02 Harmonised TLS library version reporting; only show if debugging.
2841 Layout now matches that introduced for other libraries in 4.74 PP/03.
2843 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2845 PP/04 New "dns_use_edns0" global option.
2847 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2850 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2851 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2853 TK/01 Updated PolarSSL code to 0.14.2.
2854 Bugzilla 1097. Patch from Andreas Metzler.
2856 PP/07 Catch divide-by-zero in ${eval:...}.
2857 Fixes bugzilla 1102.
2859 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2862 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2863 format-string attack -- SECURITY: remote arbitrary code execution.
2865 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2866 time unintentionally subject to list matching rules, letting the header
2867 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2868 arbitrary string expansion). This allowed for information disclosure.
2870 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2871 INT_MIN/-1 -- value coerced to INT_MAX.
2877 NM/01 Workaround for PCRE version dependency in version reporting
2880 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2881 This fixes portability to compilers other than gcc, notably
2882 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2884 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2885 makefiles for portability to HP-UX and POSIX correctness.
2887 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2888 Also via indented variable definition in the Makefile.
2889 (Debugging by Oliver Heesakkers).
2891 PP/02 Restore caching of spamd results with expanded spamd_address.
2892 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2894 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2895 Improves build reliability. Fix from: Frank Elsner
2897 NM/02 Fix wide character breakage in the rfc2047 coding
2898 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2900 NM/03 Allow underscore in dnslist lookups
2901 Fixes bug 1026. Patch from Graeme Fowler
2903 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2904 Code patches from Adam Ciarcinski of NetBSD.
2906 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2909 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2910 is logged, to avoid truncation. Patch from John Horne.
2912 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2913 Patch from Jakob Hirsch.
2915 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2916 SQL string expansion failure details.
2917 Patch from Andrey Oktyabrski.
2919 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2920 Patch from Simon Arlott.
2922 PP/09 New lookups functionality failed to compile on old gcc which rejects
2923 extern declarations in function scope.
2924 Patch from Oliver Fleischmann
2926 PP/10 Use sig_atomic_t for flags set from signal handlers.
2927 Check getgroups() return and improve debugging.
2928 Fixed developed for diagnosis in bug 927 (which turned out to be
2931 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2932 Patch from Mark Zealey.
2934 PP/12 Bugzilla 1056: Improved spamd server selection.
2935 Patch from Mark Zealey.
2937 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2938 Based on patch from Heiko Schlittermann.
2940 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2941 Patch from Uwe Doering, sign-off by Michael Haardt.
2943 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2944 variable declaration deep within a block. Bug and patch from
2947 PP/15 lookups-Makefile IRIX compatibility coercion.
2949 PP/16 Make DISABLE_DKIM build knob functional.
2951 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2952 Patch by Simon Arlott
2954 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2955 variable argument macros. Our copy now differs from upstream.
2961 TF/01 Failure to get a lock on a hints database can have serious
2962 consequences so log it to the panic log.
2964 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2965 controlled using the smtp_confirmation log selector.
2967 TF/03 Include the error message when we fail to unlink a spool file.
2969 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2970 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2971 for maintaining out-of-tree patches for some time.
2973 PP/01 Bugzilla 139: Documentation and portability issues.
2974 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2975 Handle per-OS dynamic-module compilation flags.
2977 PP/02 Let /dev/null have normal permissions.
2978 The 4.73 fixes were a little too stringent and complained about the
2979 permissions on /dev/null. Exempt it from some checks.
2980 Reported by Andreas M. Kirchwitz.
2982 PP/03 Report version information for many libraries, including
2983 Exim version information for dynamically loaded libraries. Created
2984 version.h, now support a version extension string for distributors
2985 who patch heavily. Dynamic module ABI change.
2987 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2988 privilege escalation vulnerability whereby the Exim run-time user
2989 can cause root to append content of the attacker's choosing to
2992 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2995 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2996 If dropping privileges for untrusted macros, we disabled normal logging
2997 on the basis that it would fail; for the Exim run-time user, this is not
2998 the case, and it resulted in successful deliveries going unlogged.
2999 Fixed. Reported by Andreas Metzler.
3005 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3006 only prepend for the Resent-* case. Fixes regression introduced in
3007 Exim 4.70 by NM/22 for Bugzilla 607.
3009 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3010 increasing numbers of administrators be bitten by this.
3012 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3014 PP/03 Bugzilla 994: added openssl_options main configuration option.
3016 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3018 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3020 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3022 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3023 without a peer certificate, leading to a segfault because of an
3024 assumption that peers always have certificates. Be a little more
3025 paranoid. Problem reported by Martin Tscholak.
3027 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3028 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3029 NB: ClamAV planning to remove STREAM in "middle of 2010".
3030 CL also introduces -bmalware, various -d+acl logging additions and
3031 more caution in buffer sizes.
3033 PP/09 Implemented reverse_ip expansion operator.
3035 PP/10 Bugzilla 937: provide a "debug" ACL control.
3037 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3039 PP/12 Bugzilla 973: Implement --version.
3041 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3043 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3045 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3047 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3048 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3049 ignore trailing whitespace.
3051 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3053 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3056 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3057 Notification from Dr Andrew Aitchison.
3059 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3060 ExtendedDetectionInfo response format.
3061 Notification from John Horne.
3063 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3066 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3067 XSL and documented dependency on system catalogs, with examples of how
3070 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3073 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3074 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3075 configuration file which is writeable by the Exim user or group.
3077 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3078 of configuration files to cover files specified with the -C option if
3079 they are going to be used with root privileges, not just the default
3082 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3083 option (effectively making it always true).
3085 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3086 files to be used while preserving root privileges.
3088 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3089 that rogue child processes cannot use them.
3091 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3092 run-time user, instead of root.
3094 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3095 Exim run-time user without dropping privileges.
3097 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3098 result string, instead of calling string_vformat() twice with the same
3101 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3102 for other users. Others should always drop root privileges if they use
3103 -C on the command line, even for a whitelisted configure file.
3105 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3107 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3113 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3114 $data_path, and $header_path variables; fixed documentation bugs and
3117 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3118 exipick to access non-standard spools, including the "frozen" queue
3121 NM/01 Bugzilla 965: Support mysql stored procedures.
3122 Patch from Alain Williams
3124 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3126 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3127 Patch from Andreas Metzler
3129 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3130 Patch from Kirill Miazine
3132 NM/05 Bugzilla 671: Added umask to procmail example.
3134 JJ/03 installed exipick 20100323.0, fixing doc bug
3136 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3137 directory. Notification and patch from Dan Rosenberg.
3139 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3141 TK/02 Improve log output when DKIM signing operation fails.
3143 MH/01 Treat the transport option dkim_domain as a colon separated
3144 list, not as a single string, and sign the message with each element,
3145 omitting multiple occurences of the same signer.
3147 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3148 Bugzilla 985, 986. Patch by Simon Arlott
3150 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3151 Patch by Simon Arlott
3153 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3154 MBX locking. Notification from Dan Rosenberg.
3160 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3162 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3164 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3166 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3168 NM/04 Bugzilla 915: Flag broken perl installation during build.
3174 TK/01 Added patch by Johannes Berg that expands the main option
3175 "spamd_address" if it starts with a dollar sign.
3177 TK/02 Write list of recipients to X-Envelope-Sender header when building
3178 the mbox-format spool file for content scanning (suggested by Jakob
3181 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3182 (http://www.dcc-servers.net/) support via dccifd. Activated by
3183 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3185 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3186 by Mark Daniel Reidel <mr@df.eu>.
3188 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3189 When building exim an external PCRE library is now needed -
3190 PCRE is a system library on the majority of modern systems.
3191 See entry on PCRE_LIBS in EDITME file.
3193 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3194 conversation. Added nologin parameter to request.
3195 Patch contributed by Kirill Miazine.
3197 TF/01 Do not log submission mode rewrites if they do not change the address.
3199 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3201 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3202 log files in place. Contributed by Roberto Lima.
3204 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3206 TF/03 Bugzilla 615: When checking the local_parts router precondition
3207 after a local_part_suffix or local_part_prefix option, Exim now
3208 does not use the address's named list lookup cache, since this
3209 contains cached lookups for the whole local part.
3211 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3212 Robert Millan. Documentation is in experimental-spec.txt.
3214 TF/04 Bugzilla 668: Fix parallel build (make -j).
3216 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3218 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3219 Patch provided by Jan Srzednicki.
3221 TF/05 Leading white space used to be stripped from $spam_report which
3222 wrecked the formatting. Now it is preserved.
3224 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3225 that they are available at delivery time.
3227 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3229 TF/08 TLS error reporting now respects the incoming_interface and
3230 incoming_port log selectors.
3232 TF/09 Produce a more useful error message if an SMTP transport's hosts
3233 setting expands to an empty string.
3235 NM/06 Bugzilla 744: EXPN did not work under TLS.
3236 Patch provided by Phil Pennock.
3238 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3239 Patch provided by Richard Godbee.
3241 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3242 acl_smtp_notquit, added index entry.
3244 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3245 Patch provided by Eugene Bujak.
3247 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3248 accept(). Patch provided by Maxim Dounin.
3250 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3251 Patch provided by Phil Pennock.
3253 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3255 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3256 Patch provided by Brad "anomie" Jorsch.
3258 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3259 Patch provided by Dean Brooks.
3261 TK/05 Add native DKIM support (does not depend on external libraries).
3263 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3264 Patch provided by Graeme Fowler.
3266 NM/16 Bugzilla 851: Documentation example syntax fix.
3268 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3270 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3273 NM/19 Bugzilla 745: TLS version reporting.
3274 Patch provided by Phil Pennock.
3276 NM/20 Bugzilla 167: bool: condition support.
3277 Patch provided by Phil Pennock.
3279 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3280 clients. Patch provided by Phil Pennock.
3282 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3283 Patch provided by Brad "anomie" Jorsch.
3285 NM/23 Bugzilla 687: Fix misparses in eximstats.
3286 Patch provided by Heiko Schlittermann.
3288 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3289 Patch provided by Heiko Schlittermann.
3291 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3292 plus update to original patch.
3294 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3296 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3297 Patch provided by David Brownlee.
3299 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3301 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3303 NM/30 Bugzilla 888: TLS documentation bugfixes.
3305 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3307 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3308 Unlike the original bugzilla I have changed all shell scripts in src tree.
3310 NM/33 Bugzilla 898: Transport filter timeout fix.
3311 Patch by Todd Rinaldo.
3313 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3314 Patch by Serge Demonchaux.
3316 NM/35 Bugzilla 39: Base64 decode bug fixes.
3317 Patch by Jakob Hirsch.
3319 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3321 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3323 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3325 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3331 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3332 ALT-N's libdkim that I have put here:
3333 http://duncanthrax.net/exim-experimental/
3335 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3336 were called 'true' and it seems that C99 defines that as a reserved
3337 keyword to be used with 'bool' variable types. That means you could
3338 not include C99-style headers which use bools without triggering
3339 build errors in sieve.c.
3341 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3342 as mailq or other aliases. Changed the --help handling significantly
3343 to do whats expected. exim_usage() emits usage/help information.
3345 SC/01 Added the -bylocaldomain option to eximstats.
3347 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3349 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3351 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3357 PH/01 Another patch from the Sieve maintainer.
3359 PH/02 When an IPv6 address is converted to a string for single-key lookup
3360 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3361 dots are used instead of colons so that keys in lsearch files need not
3362 contain colons. This was done some time before quoting was made available
3363 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3364 (notated using the quote facility) so as to distinguish them from IPv4
3365 keys. This meant that lookups for IP addresses in host lists did not work
3366 for iplsearch lookups.
3368 This has been fixed by arranging for IPv6 addresses to be expressed with
3369 colons if the lookup type is iplsearch. This is not incompatible, because
3370 previously such lookups could never work.
3372 The situation is now rather anomalous, since one *can* have colons in
3373 ordinary lsearch keys. However, making the change in all cases is
3374 incompatible and would probably break a number of configurations.
3376 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3379 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3380 conversion specification without a maximum field width, thereby enabling
3381 a rogue spamd server to cause a buffer overflow. While nobody in their
3382 right mind would setup Exim to query an untrusted spamd server, an
3383 attacker that gains access to a server running spamd could potentially
3384 exploit this vulnerability to run arbitrary code as the Exim user.
3386 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3387 $primary_hostname instead of what libspf2 thinks the hosts name is.
3389 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3390 a directory entry by the name of the lookup key. Previously, if a
3391 symlink pointed to a non-existing file or a file in a directory that
3392 Exim lacked permissions to read, a lookup for a key matching that
3393 symlink would fail. Now it is enough that a matching directory entry
3394 exists, symlink or not. (Bugzilla 503.)
3396 PH/03 The body_linecount and body_zerocount variables are now exported in the
3399 PH/04 Added the $dnslist_matched variable.
3401 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3402 This means they are set thereafter only if the connection becomes
3405 PH/06 Added the client_condition to authenticators so that some can be skipped
3406 by clients under certain conditions.
3408 PH/07 The error message for a badly-placed control=no_multiline_responses left
3409 "_responses" off the end of the name.
3411 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3413 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3414 (without spaces) instead of just copying the configuration text.
3416 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3418 PH/11 Added $max_received_linelength.
3420 PH/12 Added +ignore_defer and +include_defer to host lists.
3422 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3423 way in which PCRE > 7.0 is built.
3425 PH/14 Implemented queue_only_load_latch.
3427 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3428 MAIL command. The effect was to mangle the value on 64-bit systems.
3430 PH/16 Another patch from the Sieve maintainer.
3432 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3434 PH/18 If a system quota error occurred while trying to create the file for
3435 a maildir delivery, the message "Mailbox is full" was not appended to the
3436 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3437 only to a quota excession during the actual writing of the file.
3439 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3440 characters?) which causes problems in log lines. The DN values are now
3441 passed through string_printing() before being added to log lines.
3443 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3444 and InterBase are left for another time.)
3446 PH/21 Added message_body_newlines option.
3448 PH/22 Guard against possible overflow in moan_check_errorcopy().
3450 PH/23 POSIX allows open() to be a macro; guard against that.
3452 PH/24 If the recipient of an error message contained an @ in the local part
3453 (suitably quoted, of course), incorrect values were put in $domain and
3454 $local_part during the evaluation of errors_copy.
3460 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3461 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3464 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3465 issue a MAIL command.
3467 PH/02 In an ACL statement such as
3469 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3471 if a client was not listed at all, or was listed with a value other than
3472 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3473 the condition was not true (as it should be), so access was not denied.
3474 The bug was that the ! inversion was incorrectly passed on to the second
3475 item. This has been fixed.
3477 PH/03 Added additional dnslists conditions == and =& which are different from
3478 = and & when the dns lookup returns more than one IP address.
3480 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3481 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3483 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3484 FSYNC, which compiles an option called disable_fsync that allows for
3485 bypassing fsync(). The documentation is heavily laced with warnings.
3487 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3489 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3490 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3491 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3492 including adding "make clean"; (3) Added -fPIC when compiling the test
3493 dynamically loaded module, to get rid of a warning.
3495 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3496 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3497 = 0s. The bug is otherwise harmless.
3499 PH/07 There was a bug in the dovecot authenticator such that the value of
3500 $auth1 could be overwritten, and so not correctly preserved, after a
3501 successful authentication. This usually meant that the value preserved by
3502 the server_setid option was incorrect.
3504 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3506 PH/09 Installed PCRE release 7.0.
3508 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3509 run for batched SMTP input. It is now run at the start of every message
3510 in the batch. While fixing this I discovered that the process information
3511 (output by running exiwhat) was not always getting set for -bs and -bS
3512 input. This is fixed, and it now also says "batched" for BSMTP.
3514 PH/11 Added control=no_pipelining.
3516 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3517 patch, slightly modified), and move the expansion of helo_data till after
3518 the connection is made in the smtp transport (so it can use these
3521 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3523 PH/14 Added log_selector = +pid.
3525 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3527 PH/16 Add ${if forany and ${if forall.
3529 PH/17 Added dsn_from option to vary the From: line in DSNs.
3531 PH/18 Flush SMTP output before performing a callout, unless control =
3532 no_callout_flush is set.
3534 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3535 was true (the default) a successful delivery failed to delete the retry
3536 item, thus causing premature timeout of the address. The bug is now
3539 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3541 PH/21 Long custom messages for fakedefer and fakereject are now split up
3542 into multiline responses in the same way that messages for "deny" and
3543 other ACL rejections are.
3545 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3546 with slight modification.
3548 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3549 draft, changing the syntax and factoring some duplicate code".
3551 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3552 for deliveries of the second and subsequent messages over the same SMTP
3555 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3556 ${reduce, with only minor "tidies".
3558 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3560 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3561 expansion side effects.
3563 PH/27 When a message times out after an over-quota error from an Exim-imposed
3564 quota, the bounce message says "mailbox is full". This message was not
3565 being given when it was a system quota that was exceeded. It now should
3568 MH/03 Made $recipients available in local_scan(). local_scan() already has
3569 better access to the recipient list through recipients_list[], but
3570 $recipients can be useful in postmaster-provided expansion strings.
3572 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3573 in the case of a MAIL command with additional options following the
3574 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3575 were accidentally chopped off.
3577 PH/29 SMTP synchronization checks are implemented when a command is read -
3578 there is a check that no more input is waiting when there shouldn't be
3579 any. However, for some commands, a delay in an ACL can mean that it is
3580 some time before the response is written. In this time, more input might
3581 arrive, invalidly. So now there are extra checks after an ACL has run for
3582 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3583 pipelining has not been advertised.
3585 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3587 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3588 returned in the SMTP response when smtp_return_error_details was false.
3589 This has been fixed.
3591 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3592 instead of the C I/O that was originally supplied, because problems were
3593 reported on Solaris.
3595 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3596 Exim which did not show up earlier: it was assuming that a call to
3597 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3598 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3599 was a macro that became an assignment, so it seemed to work. This has
3600 changed to a proper function call with a void return, hence the compile
3601 error. Exim's code has been fixed.
3603 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3606 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3608 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3610 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3611 because it turns out to be too big or there is a timeout) there is no
3612 "Completed" line in the log. When some messages of this type were
3613 selected by exigrep, they were listed as "not completed". Others were
3614 picked up by some special patterns. I have improved the selection
3615 criteria to be more general.
3617 PH/38 The host_find_failed option in the manualroute router can now be set
3618 to "ignore", to completely ignore a host whose IP address cannot be
3619 found. If all hosts are ignored, the behaviour is controlled by the new
3620 host_all_ignored option.
3622 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3623 homing or because of multiple MX records with /mx) generated more than
3624 one IP address, and the following item turned out to be the local host,
3625 all the secondary addresses of the first item were incorrectly removed
3626 from the list, along with the local host and any following hosts (which
3627 is what is supposed to happen).
3629 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3630 whoever called Exim into the -H file. In the case of the daemon it was
3631 behaving confusingly. When first started, it used values for whoever
3632 started the daemon, but after a SIGHUP it used the Exim user (because it
3633 calls itself on a restart). I have changed the code so that it now always
3636 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3637 message are rejected with the same error (e.g. no authentication or bad
3638 sender address), and a DATA command is nevertheless sent (as can happen
3639 with PIPELINING or a stupid MUA), the error message that was given to the
3640 RCPT commands is included in the rejection of the DATA command. This is
3641 intended to be helpful for MUAs that show only the final error to their
3644 PH/42 Another patch from the Sieve maintainer.
3646 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3647 Eximstats - Fixed some broken HTML links and added missing column headers
3649 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3650 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3652 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3654 PH/43 Yet another patch from the Sieve maintainer.
3656 PH/44 I found a way to check for a TCP/IP connection going away before sending
3657 the response to the final '.' that terminates a message, but only in the
3658 case where the client has not sent further data following the '.'
3659 (unfortunately, this is allowed). However, in many cases there won't be
3660 any further data because there won't be any more messages to send. A call
3661 to select() can be used: if it shows that the input is "ready", there is
3662 either input waiting, or the socket has been closed. An attempt to read
3663 the next input character can distinguish the two cases. Previously, Exim
3664 would have sent an OK response which the client would never have see.
3665 This could lead to message repetition. This fix should cure that, at
3666 least in a lot of common cases.
3668 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3669 advertised in response to EHLO.
3675 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3676 fixed by 4.65/MH/01 (is this a record?) are fixed:
3678 (i) An empty string was always treated as zero by the numeric comparison
3679 operators. This behaviour has been restored.
3681 (ii) It is documented that the numeric comparison operators always treat
3682 their arguments as decimal numbers. This was broken in that numbers
3683 starting with 0 were being interpreted as octal.
3685 While fixing these problems I realized that there was another issue that
3686 hadn't been noticed. Values of message_size_limit (both the global option
3687 and the transport option) were treated as octal if they started with 0.
3688 The documentation was vague. These values are now always treated as
3689 decimal, and I will make that clear in the documentation.
3695 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3696 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3699 MH/01 Don't check that the operands of numeric comparison operators are
3700 integers when their expansion is in "skipping" mode (fixes bug
3701 introduced by 4.64-PH/07).
3703 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3704 child addresses, Exim now panics and dies. Previously, because the count
3705 is held in a short int, deliveries were likely to be lost. As such a
3706 large number of recipients for a single message is ridiculous
3707 (performance will be very, very poor), I have chosen to impose a limit
3708 rather than extend the field.
3714 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3715 leftover -K file (the existence of which was triggered by #402).
3716 While we were at it, introduced process PID as part of the -K
3717 filename. This should rule out race conditions when creating
3720 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3721 processing considerably. Previous code took too long for large mails,
3722 triggering a timeout which in turn triggers #401.
3724 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3725 in the DK code in transports.c. sendfile() is not really portable,
3726 hence the _LINUX specificness.
3728 TF/01 In the add_headers option to the mail command in an Exim filter,
3729 there was a bug that Exim would claim a syntax error in any
3730 header after the first one which had an odd number of characters
3733 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3734 callout verification, Exim cached a "reject" for the entire domain. This
3735 is correct for most verifications, but it is not correct for a recipient
3736 verification with use_sender or use_postmaster set, because in that case
3737 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3738 case of MAIL FROM:<> rejection from other early rejections (e.g.
3739 rejection of HELO). When verifying a recipient using a non-null MAIL
3740 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3741 Whatever the result of the callout, the value of the domain cache is
3742 left unchanged (for any other kind of callout, getting as far as trying
3743 RCPT means that the domain itself is ok).
3745 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3748 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3749 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3750 session, the daemon ignores ECONNECT errors and logs others; it now
3751 ignores EPIPE as well.
3753 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3754 (quoted-printable decoding).
3756 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3757 later the small subsequent patch to fix an introduced bug.
3759 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3761 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3763 PH/08 An error is now given if message_size_limit is specified negative.
3765 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3766 to be given (somewhat) arbitrary names.
3768 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3771 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3772 miscellaneous code fixes
3774 PH/10 Added the log_reject_target ACL modifier to specify where to log
3777 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3778 hostname. This is wrong, because it relates to the incoming message (and
3779 probably the interface on which it is arriving) and not to the outgoing
3780 callout (which could be using a different interface). This has been
3781 changed to use the value of the helo_data option from the smtp transport
3782 instead - this is what is used when a message is actually being sent. If
3783 there is no remote transport (possible with a router that sets up host
3784 addresses), $smtp_active_hostname is used.
3786 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3787 tweaks were necessary in order to get it to work (see also 21 below):
3788 (a) The code assumed that strncpy() returns a negative number on buffer
3789 overflow, which isn't the case. Replaced with Exim's string_format()
3791 (b) There were several signed/unsigned issues. I just did the minimum
3792 hacking in of casts. There is scope for a larger refactoring.
3793 (c) The code used strcasecmp() which is not a standard C function.
3794 Replaced with Exim's strcmpic() function.
3795 (d) The code set only $1; it now sets $auth1 as well.
3796 (e) A simple test gave the error "authentication client didn't specify
3797 service in request". It would seem that Dovecot has changed its
3798 interface. Fortunately there's a specification; I followed it and
3799 changed what the client sends and it appears to be working now.
3801 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3804 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3805 address A is aliased to B and C, where B exists and C does not. Without
3806 -v the output is "A verified" because verification stops after a
3807 successful redirection if more than one address is generated. However,
3808 with -v the child addresses are also verified. Exim was outputting "A
3809 failed to verify" and then showing the successful verification for C,
3810 with its parentage. It now outputs "B failed to verify", showing B's
3811 parentage before showing the successful verification of C.
3813 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3814 look up a TXT record in a specific list after matching in a combined
3817 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3818 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3819 they consult the DNS. I had assumed they would set it the way they
3820 wanted; and indeed my experiments on Linux seem to show that in some
3821 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3822 To be on the safe side, however, I have now made the interface to
3823 host_find_byname() similar to host_find_bydns(), with an argument
3824 containing the DNS resolver options. The host_find_byname() function now
3825 sets these options at its start, just as host_find_bydns() does. The smtp
3826 transport options dns_qualify_single and dns_search_parents are passed to
3827 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3828 of host_find_byname() use the default settings of RES_DEFNAMES
3829 (qualify_single) but not RES_DNSRCH (search_parents).
3831 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3832 spool_read_header() do less string testing, by means of a preliminary
3833 switch on the second character of optional "-foo" lines. (This is
3834 overdue, caused by the large number of possibilities that now exist.
3835 Originally there were few.) While I was there, I also converted the
3836 str(n)cmp tests so they don't re-test the leading "-" and the first
3837 character, in the hope this might squeeze out yet more improvement.
3839 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3840 flag allowing group syntax was set by the header_syntax check but not
3841 turned off, possible causing trouble later; (2) The flag was not being
3842 set at all for the header_verify test, causing "group"-style headers to
3843 be rejected. I have now set it in this case, and also caused header_
3844 verify to ignore an empty address taken from a group. While doing this, I
3845 came across some other cases where the code for allowing group syntax
3846 while scanning a header line wasn't quite right (mostly, not resetting
3847 the flag correctly in the right place). These bugs could have caused
3848 trouble for malformed header lines. I hope it is now all correct.
3850 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3851 with the "reply" argument non-NULL. The code, however (which originally
3852 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3853 but it didn't always do it. This confused somebody who was copying the
3854 code for some other use. I have removed all the tests.
3856 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3857 feature that was used to support insecure browsers during the U.S. crypto
3858 embargo. It requires special client support, and Exim is probably the
3859 only MTA that supported it -- and would never use it because real RSA is
3860 always available. This code has been removed, because it had the bad
3861 effect of slowing Exim down by computing (never used) parameters for the
3862 RSA_EXPORT functionality.
3864 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3865 authenticator to fail if there's a tab character in the incoming data
3866 (there should never be unless someone is messing about, as it's supposed
3867 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3868 if the connection is using TLS or if the remote IP is the same as the
3869 local IP, and the "valid-client-cert option" if a client certificate has
3872 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3873 authenticators. This can be used for authorization after authentication
3874 succeeds. (In the case of plaintext, it servers for both authentication
3877 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3878 if any retry times were supplied.
3880 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3881 connection, where there is no client IP address to check. In this
3882 situation, the verify now always succeeds.
3884 PH/25 Applied John Jetmore's -Mset patch.
3886 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3888 PH/27 In a string expansion for a processed (not raw) header when multiple
3889 headers of the same name were present, leading whitespace was being
3890 removed from all of them, but trailing whitespace was being removed only
3891 from the last one. Now trailing whitespace is removed from each header
3892 before concatenation. Completely empty headers in a concatenation (as
3893 before) are ignored.
3895 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3896 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3898 PH/29 [Removed. This was a change that I later backed out, and forgot to
3899 correct the ChangeLog entry (that I had efficiently created) before
3900 committing the later change.]
3902 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3903 address errors (4xx response to RCPT) over the same connection as other
3904 messages routed to the same hosts. Such deliveries are always "forced",
3905 so retry times are not inspected. This resulted in far too many retries
3906 for the affected addresses. The effect occurred only when there were more
3907 hosts than the hosts_max_try setting in the smtp transport when it had
3908 the 4xx errors. Those hosts that it had tried were not added to the list
3909 of hosts for which the message was waiting, so if all were tried, there
3910 was no problem. Two fixes have been applied:
3912 (i) If there are any address or message errors in an SMTP delivery, none
3913 of the hosts (tried or untried) are now added to the list of hosts
3914 for which the message is waiting, so the message should not be a
3915 candidate for sending over the same connection that was used for a
3916 successful delivery of some other message. This seems entirely
3917 reasonable: after all the message is NOT "waiting for some host".
3918 This is so "obvious" that I'm not sure why it wasn't done
3919 previously. Hope I haven't missed anything, but it can't do any
3920 harm, as the worst effect is to miss an optimization.
3922 (ii) If, despite (i), such a delivery is accidentally attempted, the
3923 routing retry time is respected, so at least it doesn't keep
3924 hammering the server.
3926 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3927 in ${readsocket because some servers need this prod.
3929 PH/32 Added some extra debug output when updating a wait-xxx database.
3931 PH/33 The hint "could be header name not terminated by colon", which has been
3932 given for certain expansion errors for a long time, was not being given
3933 for the ${if def:h_colon_omitted{... case.
3935 PH/34 The spec says: "With one important exception, whenever a domain list is
3936 being scanned, $domain contains the subject domain." There was at least
3937 one case where this was not true.
3939 PH/35 The error "getsockname() failed: connection reset by peer" was being
3940 written to the panic log as well as the main log, but it isn't really
3941 panic-worthy as it just means the connection died rather early on. I have
3942 removed the panic log writing for the ECONNRESET error when getsockname()
3945 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3946 runs only) independently of the message's sender address. This meant
3947 that, if the 4xx error was in fact related to the sender, a different
3948 message to the same recipient with a different sender could confuse
3949 things. In particular, this can happen when sending to a greylisting
3950 server, but other circumstances could also provoke similar problems.
3951 I have changed the default so that the retry time for these errors is now
3952 based a combination of the sender and recipient addresses. This change
3953 can be overridden by setting address_retry_include_sender=false in the
3956 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3957 remote server are returned as part of bounce messages. This was not
3958 happening for LMTP over a pipe (the lmtp transport), but now it is the
3959 same for both kinds of LMTP.
3961 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3962 in header lines that were in fact CNAMEs. This is no longer the case.
3964 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3965 and queue runs started by the daemon processed all messages. This has
3966 been fixed so that -R and -S can now usefully be given with -q<time>.
3968 PH/40 Import PCRE release 6.7 (fixes some bugs).
3970 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3972 PH/42 Give an error if -q is specified more than once.
3974 PH/43 Renamed the variables $interface_address and $interface_port as
3975 $received_ip_address and $received_port, to make it clear that these
3976 values apply to message reception, and not to the outgoing interface when
3977 a message is delivered. (The old names remain recognized, of course.)
3979 PH/44 There was no timeout on the connect() call when using a Unix domain
3980 socket in the ${readsocket expansion. There now is.
3982 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3983 be meaningful with "accept".
3985 SC/01 Eximstats V1.43
3986 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3988 SC/02 Eximstats V1.44
3989 Use a glob alias rather than an array ref in the generated
3990 parser. This improves both readability and performance.
3992 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3993 Collect SpamAssassin and rejection statistics.
3994 Don't display local sender or destination tables unless
3995 there is data to show.
3996 Added average volumes into the top table text output.
3998 SC/04 Eximstats V1.46
3999 Collect data on the number of addresses (recipients)
4000 as well as the number of messages.
4002 SC/05 Eximstats V1.47
4003 Added 'Message too big' to the list of mail rejection
4004 reasons (thanks to Marco Gaiarin).
4006 SC/06 Eximstats V1.48
4007 Mainlog lines which have GMT offsets and are too short to
4008 have a flag are now skipped.
4010 SC/07 Eximstats V1.49 (Alain Williams)
4011 Added the -emptyok flag.
4013 SC/08 Eximstats V1.50
4014 Fixes for obtaining the IP address from reject messages.
4016 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4017 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4018 whitespace changes from 4.64-PH/27
4020 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4023 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4024 are found, allow negative numbers in numeric criteria)
4026 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4028 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4029 to match changes made in 4.64-PH/43
4031 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4033 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4034 "domain" is actually an IP address, and force a failure. This locks out
4035 those revolvers/nameservers that support "A-for-A" lookups, in
4036 contravention of the specifications.
4038 PH/48 When a host name was looked up from an IP address, and the subsequent
4039 forward lookup of the name timed out, the host name was left in
4040 $sender_host_name, contrary to the specification.
4042 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4043 restricted to single-key lookups, Exim was not diagnosing an error if
4044 * or *@ was used with a query-style lookup.
4046 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4048 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4049 long ago, but noone interested enough thought of it. Let's just say that
4050 the "1.1" means that there are some new functions that weren't there at
4051 some point in the past.
4053 PH/51 Error processing for expansion failure of helo_data from an smtp
4054 transport during callout processing was broken.
4056 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4057 tested/used via the -bh/-bhc/-bs options.
4059 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4060 bug, fixed in subsequent PCRE releases).
4062 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4063 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4065 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4071 SC/01 Use a glob alias rather than an array ref in eximstats generated
4072 parser. This improves both readability and performance.
4074 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4075 Don't display local sender or destination tables in eximstats unless
4076 there is data to show.
4077 Added average volumes into the eximstats top table text output.
4079 SC/03 Collect data on the number of addresses (recipients) as well
4080 as the number of messages in eximstats.
4082 TF/01 Correct an error in the documentation for the redirect router. Exim
4083 does (usually) call initgroups() when daemonizing.
4085 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4086 with consistent privilege compared to when running as a daemon.
4088 TF/03 Note in the spec that $authenticated_id is not set for local
4089 submissions from trusted users.
4091 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4092 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4094 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4095 by adding some example configuration directives to the default
4096 configuration file. A little bit of work is required to uncomment the
4097 directives and define how usernames and passwords are checked, but
4098 there is now a framework to start from.
4100 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4101 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4102 without this. I don't know how relevant this is to other LDAP libraries.
4104 PH/02 Add the verb name to the "unknown ACL verb" error.
4106 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4108 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4110 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4111 directory not expanded when it should be if an expanded home directory
4112 was set for the address (which is overridden by the transport).
4114 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4117 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4118 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4119 because it is too late at that time, and has no effect.
4121 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4122 security issue with \' (bugzilla #107). I could not use the
4123 PQescapeStringConn() function, because it needs a PGconn value as one of
4126 PH/08 When testing addresses using -bt, indicate those final addresses that
4127 are duplicates that would not cause an additional delivery. At least one
4128 person was confused, thinking that -bt output corresponded to deliveries.
4129 (Suppressing duplicates isn't a good idea as you lose the information
4130 about possibly different redirections that led to the duplicates.)
4132 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4133 systems where poll() doesn't work, in particular OS X.
4135 PH/10 Added more information to debugging output for retry time not reached.
4137 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4138 operations in malware.c.
4140 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4143 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4144 syslog_duplication set false, log lines that would normally be written
4145 both the the main log and to the reject log were not written to syslog at
4148 PH/14 In the default configuration, change the use of "message" in ACL warn
4149 statements to "add_header".
4151 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4152 not followed by a command (e.g. "seen endif").
4154 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4155 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4158 PH/17 Added extra conditions to the default value of delay_warning_condition
4162 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4163 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4164 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4167 The Auto-Submitted: and various List- headers are standardised, whereas I
4168 don't think Precedence: ever was.
4170 PH/18 Refactored debugging code in route_finduser() to show more information,
4171 in particular, the error code if getpwnam() issues one.
4173 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4174 This is apparently needed in addition to the PH/07 change above to avoid
4175 any possible encoding problems.
4177 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4178 but not after initializing Perl.
4180 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4181 output them only if debugging. By default they are written stderr,
4182 apparently, which is not desirable.
4184 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4187 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4190 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4192 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4193 authenticated or an ident call has been made. Suppress the default
4194 values for $authenticated_id and $authenticated_sender (but permit -oMai
4195 and -oMas) when testing with -bh.
4197 PH/24 Re-jigged the order of the tests in the default configuration so that the
4198 tests for valid domains and recipients precede the DNS black list and CSA
4199 tests, on the grounds that those ones are more expensive.
4201 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4202 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4203 This bug exists in every version of Exim that I still have, right back to
4206 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4207 However, an attempt to turn on case-sensitivity in a regex key by
4208 including (?-i) didn't work because the subject string was already
4209 lowercased, and the effects were non-intuitive. It turns out that a
4210 one-line patch can be used to allow (?-i) to work as expected.
4216 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4217 other effects) broke the use of negated acl sub-conditions.
4219 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4222 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4223 "Deny" causes Exim to reject the incoming connection with a 554 error.
4224 Unfortunately, if there is a major crisis, such as a disk failure,
4225 tcp-wrappers gives "deny", whereas what one would like would be some
4226 kind of temporary error. A kludge has been added to help with this.
4227 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4228 554 error is used if errno is still zero or contains ENOENT (which occurs
4229 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4232 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4234 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4235 errors. Otherwise a message that provokes a temporary error (when other
4236 messages do not) can cause a whole host to time out.
4238 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4239 addresses were routed directly to files or pipes from a redirect router.
4240 File deliveries just didn't batch; pipe deliveries might have suffered
4243 PH/06 A failure to get a lock for a hints database would erroneously always say
4244 "Failed to get write lock", even when it was really a read lock.
4246 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4247 of 0600. This has been changed to use the value of the lockfile_mode
4248 option (which defaults to 0600).
4250 PH/08 Applied small patch from the Sieve maintainer.
4252 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4253 folder from quota calculations, a direct delivery into this folder messed
4254 up the contents of the maildirsize file. This was because the regex was
4255 used only to exclude .Trash (or whatever) when the size of the mailbox
4256 was calculated. There was no check that a delivery was happening into an
4257 excluded directory. This bug has been fixed by ignoring all quota
4258 processing for deliveries into excluded directories.
4260 PH/10 Added the maildirfolder_create_regex option to appendfile.
4266 PH/01 The code for finding all the local interface addresses on a FreeBSD
4267 system running IPv6 was broken. This may well have applied to all BSD
4268 systems, as well as to others that have similar system calls. The broken
4269 code found IPv4 interfaces correctly, but gave incorrect values for the
4270 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4271 that it would not match correctly against @[] and not recognize the IPv6
4274 PH/02 The ipliteral router was not recognizing addresses of the form user@
4275 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4277 PH/03 Added disable_ipv6.
4279 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4280 decoded form, because it is most often used to construct To: headers
4281 lines in autoreplies, and the decoded form may well be syntactically
4282 invalid. However, $reply_address has leading white space removed, and all
4283 newlines turned into spaces so that the autoreply transport does not
4286 PH/05 If group was specified without a user on a router, and no group or user
4287 was specified on a transport, the group from the router was ignored.
4289 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4290 for visible compile-time settings that can be used to change these
4291 numbers, for those that want even more. Backwards compatibility with old
4292 spool files has been maintained. However, going back to a previous Exim
4293 release will lost any variables that are in spool files.
4295 PH/07 Two small changes when running in the test harness: increase delay when
4296 passing a TCP/IP connection to a new process, in case the original
4297 process has to generate a bounce, and remove special handling of
4298 127.0.0.2 (sic), which is no longer necessary.
4300 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4301 be the same on different OS.
4303 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4306 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4307 whether --show-vars was specified or not
4309 JJ/02 exipick: Added support for new ACL variable spool format introduced
4312 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4313 syntactically invalid From: or Reply-to: line, and a filter used this to
4314 generate an autoreply, and therefore failed to obtain an address for the
4315 autoreply, Exim could try to deliver to a non-existent relative file
4316 name, causing unrelated and misleading errors. What now happens is that
4317 it logs this as a hard delivery error, but does not attempt to create a
4320 PH/11 The exinext utility has a -C option for testing purposes, but although
4321 the given file was scanned by exinext itself; it wasn't being passed on
4322 when Exim was called.
4324 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4325 an end-of-file indication when reading a command response.
4327 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4328 compiled. In many other places in Exim, IPv6 addresses are always
4329 recognized, so I have changed this. It also means that IPv4 domain
4330 literals of the form [IPV4:n.n.n.n] are now always recognized.
4332 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4333 used if the router is not running as root, for example, when verifying at
4334 ACL time, or when using -bh. The debugging output from this situation was
4335 non-existent - all you got was a failure to exec. I have made two
4338 (a) Failures to set uid/gid, the current directory, or a process leader
4339 in a subprocess such as that created by queryprogram now generate
4340 suitable debugging output when -d is set.
4342 (b) The queryprogram router detects when it is not running as root,
4343 outputs suitable debugging information if -d is set, and then runs
4344 the subprocess without attempting to change uid/gid.
4346 PH/15 Minor change to Makefile for building test_host (undocumented testing
4349 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4350 additional section of a DNS packet that returns MX or SRV records.
4351 Instead, it always explicitly searches for A/AAAA records. This avoids
4352 major problems that occur when a DNS server includes only records of one
4353 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4354 fixed another bug: if SRV records were looked up and the corresponding
4355 address records were *not* found in the additional section, the port
4356 values from the SRV records were lost.
4358 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4359 using the correct key (the original address) when searching the retry
4360 rules in order to find which one to use for generating the retry hint.
4362 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4363 adding the default one. Similarly, if it contains a Reply-To: header, the
4364 errors_reply_to option, if set, is not used.
4366 PH/19 When calculating a retry time, Exim used to measure the "time since
4367 failure" by looking at the "first failed" field in the retry record. Now
4368 it does not use this if it is later than than the arrival time of the
4369 message. Instead it uses the arrival time. This makes for better
4370 behaviour in cases where some deliveries succeed, thus re-setting the
4371 "first failed" field. An example is a quota failure for a huge message
4372 when small messages continue to be delivered. Without this change, the
4373 "time since failure" will always be short, possible causing more frequent
4374 delivery attempts for the huge message than are intended.
4375 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4377 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4378 $1, $2, $3) because the numerical variables can be reset during some
4379 expansion items (e.g. "match"), thereby losing the authentication data.
4381 PH/21 Make -bV show the size of off_t variables so that the test suite can
4382 decide whether to run tests for quotas > 2G.
4384 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4385 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4386 filecount value is greater than 2G or if a quota value is greater than 2G
4387 on a system where the size of off_t is not greater than 4, a panic error
4390 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4391 never match. The debug and -bh output now contains an explicit error
4392 message indicating a malformed IPv4 address or mask.
4394 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4395 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4396 PH/23 above applies.
4398 PH/25 Do not write to syslog when running in the test harness. The only
4399 occasion when this arises is a failure to open the main or panic logs
4400 (for which there is an explicit test).
4402 PH/26 Added the /no_tell option to "control=freeze".
4404 PH/27 If a host name lookup failed very early in a connection, for example, if
4405 the IP address matched host_lookup and the reverse lookup yielded a name
4406 that did not have a forward lookup, an error message of the form "no IP
4407 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4408 could be logged. Now it outputs the IP address instead of "NULL".
4410 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4411 allows the sender and the authenticated sender to be set when
4412 submitting a message from within Exim. Since child_open_exim() is
4413 documented for local_scan(), the new function should be too.
4415 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4416 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4417 results in an empty string is now treated as unset.
4419 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4421 PH/31 Added sender_verify_fail logging option.
4423 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4424 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4425 tidied the source and removed it altogether.
4427 PH/33 When a queue run was abandoned because the load average was too high, a
4428 log line was always written; now it is written only if the queue_run log
4429 selector is set. In addition, the log line for abandonment now contains
4430 information about the queue run such as the pid. This is always present
4431 in "start" and "stop" lines but was omitted from the "abandon" line.
4433 PH/34 Omit spaces between a header name and the colon in the error message that
4434 is given when verify = headers_syntax fails (if there are lots of them,
4435 the message gets confusing).
4437 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4438 names, as there are now some PTR records that contain slashes. This check
4439 is only to protect against broken name servers that fall over on strange
4440 characters, so the fact that it applies to all lookups doesn't matter.
4442 PH/36 Now that the new test suite is complete, we can remove some of the
4443 special code in Exim that was needed for the old test suite. For example,
4444 sorting DNS records because real resolvers return them in an arbitrary
4445 order. The new test suite's fake resolver always returns records in the
4448 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4449 bounces) except when queue_only is set, to avoid logging races between
4450 the different processes.
4452 PH/38 Panic-die if .include specifies a non-absolute path.
4454 PH/39 A tweak to the "H" retry rule from its user.
4456 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4457 a label. They prevented compilation on older perls.
4459 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4460 a warning to be raised on newish perls.
4462 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4463 on queue. Changes to match documented behaviour of showing count of
4464 messages matching specified criteria.
4466 PH/40 Changed the default ident timeout from 30s to 5s.
4468 PH/41 Added support for the use of login_cap features, on those BSD systems
4469 that have them, for controlling the resources used by pipe deliveries.
4471 PH/42 The content-scanning code uses fopen() to create files in which to put
4472 message data. Previously it was not paying any attention to the mode of
4473 the files. Exim runs with umask(0) because the rest of the code creates
4474 files with open(), and sets the required mode explicitly. Thus, these
4475 files were ending up world-writeable. This was not a big issue, because,
4476 being within the spool directory, they were not world-accessible. I have
4477 created a function called modefopen, which takes an additional mode
4478 argument. It sets umask(777), creates the file, chmods it to the required
4479 mode, then resets the umask. All the relevant calls to fopen() in the
4480 content scanning code have been changed to use this function.
4482 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4483 to 24 hours. This avoids potential overflow problems when processing G
4484 and H retry rules. I suspect nobody ever tinkers with this value.
4486 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4488 PH/45 When the plaintext authenticator is running as a client, the server's
4489 challenges are checked to ensure they are valid base64 strings. By
4490 default, the authentication attempt is cancelled if an invalid string is
4491 received. Setting client_ignore_invalid_base64 true ignores these errors.
4492 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4493 they are received. Thus, the responses can be made to depend on the
4494 challenges. If an invalid string is ignored, an empty string is placed in
4497 PH/46 Messages that are created by the autoreply transport now contains a
4498 References: header, in accordance with RFCs 2822 and 3834.
4500 PH/47 Added authenticated_sender_force to the smtp transport.
4502 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4504 PH/49 Installed latest patch from the Sieve maintainer.
4506 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4507 was also set, the appendfile transport was unnecessarily scanning a
4508 directory of message files (e.g. for maildir delivery) to find the count
4509 of files (along with the size), even though it did not need this
4510 information. It now does the scan only if it needs to find either the
4511 size of the count of files.
4513 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4515 PH/52 Two bugs concerned with error handling when the smtp transport is
4518 (i) Exim was not creating retry information for temporary errors given
4519 for individual recipients after the DATA command when the smtp transport
4520 was used in LMTP mode. This meant that they could be retried too
4521 frequently, and not timed out correctly.
4523 (ii) Exim was setting the flag that allows error details to be returned
4524 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4525 recipients that were returned after the DATA command.
4527 PH/53 This is related to PH/52, but is more general: for any failing address,
4528 when detailed error information was permitted to be returned to the
4529 sender, but the error was temporary, then after the final timeout, only
4530 "retry timeout exceeded" was returned. Now it returns the full error as
4531 well as "retry timeout exceeded".
4533 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4534 do this, and (what is worse) MTAs that accept it.
4536 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4537 will now be deprecated.
4539 PH/56 New os.c-cygwin from the Cygwin maintainer.
4541 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4542 formats (previously only available in exim formats via -bpr, -bpru,
4543 and -bpra. Now also available in native and exiqgrep formats)
4545 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4546 with very large, slow to parse queues
4548 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4550 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4552 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4553 responses to authentication challenges, though it was showing the
4554 challenges; (ii) I've removed the CR characters from the debug output for
4557 PH/58 Allow for the insertion of a newline as well as a space when a string
4558 is turned into more than one encoded-word during RFC 2047 encoding. The
4559 Sieve code now uses this.
4561 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4562 data_4xx, lost_connection, tls_required.
4564 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4565 message was being sent as an SMTP response.
4567 PH/61 Add -l and -k options to exicyclog.
4569 PH/62 When verifying, if an address was redirected to one new address, so that
4570 verification continued, and the new address failed or deferred after
4571 having set something in $address_data, the value of $address_data was not
4572 passed back to the ACL. This was different to the case when no
4573 redirection occurred. The value is now passed back in both cases.
4575 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4576 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4577 use login_cap.h, so on its own it isn't the distinguishing feature. The
4578 new name refers directly to the setclassresources() function.
4580 PH/65 Added configuration files for NetBSD3.
4582 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4584 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4585 is preferred over IPv4.
4587 PH/68 The bounce_return_message and bounce_return_body options were not being
4588 honoured for bounces generated during the reception of non-SMTP messages.
4589 In particular, this applied to messages rejected by the ACL. This bug has
4590 been fixed. However, if bounce_return_message is true and bounce_return_
4591 body is false, the headers that are returned for a non-SMTP message
4592 include only those that have been read before the error was detected.
4593 (In the case of an ACL rejection, they have all been read.)
4595 PH/69 The HTML version of the specification is now built in a directory called
4596 spec_html instead of spec.html, because the latter looks like a path with
4597 a MIME-type, and this confuses some software.
4599 PH/70 Catch two compiler warnings in sieve.c.
4601 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4602 function verify_get_ident() calls ip_connect() to connect a socket, but
4603 if the "connect()" function timed out, ip_connect() used to close the
4604 socket. However, verify_get_ident() also closes the socket later, and in
4605 between Exim writes to the log, which may get opened at this point. When
4606 the socket was closed in ip_connect(), the log could get the same file
4607 descriptor number as the socket. This naturally causes chaos. The fix is
4608 not to close the socket in ip_connect(); the socket should be closed by
4609 the function that creates it. There was only one place in the code where
4610 this was missing, in the iplookup router, which I don't think anybody now
4611 uses, but I've fixed it anyway.
4613 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4614 well as to direct DNS lookups. Otherwise the handling of names in host
4615 lists is inconsistent and therefore confusing.
4621 PH/01 Two changes to the default runtime configuration:
4623 (1) Move the checks for relay_from_hosts and authenticated clients from
4624 after to before the (commented out) DNS black list checks.
4626 (2) Add control=submission to the relay_from_hosts and authenticated
4627 clients checks, on the grounds that messages accepted by these
4628 statements are most likely to be submissions.
4630 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4632 (1) Generate an error if the third argument for the ${prvs expansion is
4635 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4638 (3) Reset the variables that are obtained from the first argument of
4639 ${prvscheck and used in the second argument before leaving the code,
4640 because their memory is reclaimed, so using them afterwards may do
4643 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4644 one (it's much easier than Tom thought :-).
4646 (5) Because of (4), we can now allow for the use of $prvscheck_result
4647 inside the third argument.
4649 PH/03 For some reason, the default setting of PATH when running a command from
4650 a pipe transport was just "/usr/bin". I have changed it to
4653 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4654 anything to be listed in the output from -bV.
4656 PH/05 When a filter generated an autoreply, the entire To: header line was
4657 quoted in the delivery log line, like this:
4659 => >A.N.Other <ano@some.domain> <original@ddress> ...
4661 This has been changed so that it extracts the operative address. There
4662 may be more than one such address. If so, they are comma-separated, like
4665 => >ano@some.domain,ona@other.domain <original@ddress> ...
4667 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4668 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4669 not being looked up in the rDNS to get a host name, Exim was showing the
4670 IP address twice in Received: lines, even though the IP addresses were
4671 identical. For example:
4673 Received: from [1.2.3.4] (helo=[1.2.3.4])
4675 However, if the real host name was known, it was omitting the HELO data
4676 if it matched the actual IP address. This has been tidied up so that it
4677 doesn't show the same IP address twice.
4679 PH/07 When both +timestamp and +memory debugging was on, the value given by
4680 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4681 called by the debug printing, thereby overwriting the timestamp buffer.
4682 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4684 PH/08 When the original message was included in an autoreply transport, it
4685 always said "this is a copy of the message, including all the headers",
4686 even if body_only or headers_only was set. It now gives an appropriate
4689 PH/09 Applied a patch from the Sieve maintainer which:
4691 o fixes some comments
4692 o adds the (disabled) notify extension core
4693 o adds some debug output for the result of if/elsif tests
4694 o points to the current vacation draft in the documentation
4695 and documents the missing references header update
4699 o fixes a bug in processing the envelope test (when testing
4700 multiple envelope elements, the last element determined the
4703 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4704 Electronic Mail") by including:
4706 Auto-submitted: auto-generated
4708 in the messages that it generates (bounce messages and others, such as
4709 warnings). In the case of bounce messages for non-SMTP messages, there was
4710 also a typo: it was using "Auto_submitted" (underscore instead of
4711 hyphen). Since every message generated by Exim is necessarily in response
4712 to another message, thes have all been changed to:
4714 Auto-Submitted: auto-replied
4716 in accordance with these statements in the RFC:
4718 The auto-replied keyword:
4720 - SHOULD be used on messages sent in direct response to another
4721 message by an automatic process,
4723 - MUST NOT be used on manually-generated messages,
4725 - MAY be used on Delivery Status Notifications (DSNs) and Message
4726 Disposition Notifications (MDNs),
4728 - MUST NOT be used on messages generated by automatic or periodic
4729 processes, except for messages which are automatic responses to
4732 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4733 to the default Received: header definition.
4735 PH/12 Added log selector acl_warn_skipped (default on).
4737 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4738 variables because (a) they are in the wrong storage pool and (b) even if
4739 they were copied, it wouldn't work properly because of the caching.
4741 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4742 checking when decoding. Apparently there are clients that generate
4743 overlong encoded strings. Why am I not surprised?
4745 PH/15 If the first argument of "${if match_address" was not empty, but did not
4746 contain an "@" character, Exim crashed. Now it writes a panic log message
4747 and treats the condition as false.
4749 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4751 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4752 extension "envelope-auth". The code is finished and in agreement with
4753 other implementations, but there is no documentation so far and in fact,
4754 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4755 not changing the active code.
4757 Print executed "if" and "elsif" statements when debugging is used. This
4758 helps a great deal to understand what a filter does.
4760 Document more things not specified clearly in RFC3028. I had all this
4761 sorted out, when out of a sudden new issues came to my mind. Oops."
4763 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4766 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4767 canonical form (as documented). However, after a host name lookup from
4768 the IP address, check_host() was doing a simple string comparison with
4769 addresses acquired from the DNS when checking that the found name did
4770 have the original IP as one of its addresses. Since any found IPv6
4771 addresses are likely to be in abbreviated form, the comparison could
4772 fail. Luckily, there already exists a function for doing the comparison
4773 by converting both addresses to binary, so now that is used instead of
4774 the text comparison.
4776 PH/20 There was another similar case to PH/19, when a complete host name was
4777 given in a host list; looking up its IP address could give an abbreviated
4778 form, whereas the current host's name might or might not be abbreviated.
4779 The same fix has been applied.
4785 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4786 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4789 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4790 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4792 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4794 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4795 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4796 Also, for "Auto-Submitted", treat anything other than "no" as
4797 non-personal, in accordance with RFC 3834. (Previously it treated
4798 anything starting "auto-" as non-personal.)
4800 TF/01 The control=submission/name=... option had a problem with syntax
4801 errors if the name included a slash character. The /name= option
4802 now slurps the rest of the string, so it can include any characters
4803 but it must come last in the list of options (after /sender_retain
4806 PH/05 Some modifications to the interface to the fake nameserver for the new
4814 TK/01 Added the "success_on_redirect" address verification option. See
4815 NewStuff for rationale and an example.
4817 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4819 PH/02 Patch to exigrep to allow it to work on syslog lines.
4821 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4822 fread() to skip over the body file's header line, because in Cygwin the
4823 header line is locked and is inaccessible.
4825 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4826 co-exist for some time) to make it clear that it is the Exim ID that is
4827 referenced, not the Message-ID: header line.
4829 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4830 string_format() function, because snprintf() does not exist on all
4833 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4834 ${stat: expansion item.
4836 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4837 protocol synchronization error", to keep the pedants happy.
4839 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4840 well as for IRIX systems, when gcc is being used. See the host.c source
4843 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4845 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4848 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4849 transport and to the smtp transport in LMTP mode.
4851 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4853 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4854 run a filter in a subprocess. This could lead to confusion in subsequent
4855 lookups in the parent process. There should also be a search_tidyup() at
4856 the end of the subprocess.
4858 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4859 only if the host matched helo_try_verify_hosts, which caused the
4860 verification to occur when the EHLO/HELO command was issued. The ACL just
4861 tested the remembered result. Now, if a previous verification attempt has
4862 not happened, "verify = helo" does it there and then.
4864 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4866 TK/03 Fix log output including CR from clamd.
4868 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4869 exist provoked a memory error which could cause a segfault.
4871 PH/15 Installed PCRE 6.2
4873 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4875 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4876 of the problem. Specifically, suggested +O2 rather than +O1 for the
4879 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4881 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4882 record, the port was not correctly logged when the outgoing_port log
4883 selector was set (it logged the transort's default port).
4885 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4886 fallback_hosts, and "hosts" in the smtp transport.
4888 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4889 host errors such as "Connection refused".
4891 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4892 authentication with radiusclient 0.4.9:
4894 - Error returned from rc_read_config was caught wrongly
4895 - Username/password not passed on to radius server due to wrong length.
4897 The presumption is that some radiusclient API changes for 4.51/PH/17
4898 were not taken care of correctly. The code is still untested by me (my
4899 Linux distribution still has 0.3.2 of radiusclient), but it was
4900 contributed by a Radius user.
4902 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4903 expanding the "port" option of the smtp transport.
4905 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4906 while reading a MIME header. Thanks to Tom Hughes for a patch.
4908 PH/24 Include config.h inside local_scan.h so that configuration settings are
4911 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4912 that in an ACL for RCPT (for example), you can examine exactly what was
4915 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4916 commands, but it was not correctly comparing the address with the actual
4917 client host address. Thus, it would show the EHLO address in Received:
4918 header lines when this was not necessary.
4920 PH/27 Added the % operator to ${eval:}.
4922 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4923 it should be ignoring failures (because with -C, for example, it has lost
4924 privilege). It wasn't ignoring creation failures other than "already
4927 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4930 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4931 because an input line was too long, either on its own, or by virtue of
4932 too many continuations, the temporary file was not being removed, and the
4933 return code was incorrect.
4935 PH/31 Missing "BOOL" in function definition in filtertest.c.
4937 PH/32 Applied Sieve patches from the maintainer.
4939 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4941 PH/33 Added "verify = not_blind".
4943 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4944 Local/Makefile (with some defaults set). These are used in built scripts
4945 such as exicyclog, but they have never been used in the exim_install
4946 script (though there are many overriding facilities there). I have
4947 arranged that the exim_install script now takes note of these two
4950 PH/35 Installed configuration files for Dragonfly.
4952 PH/36 When a locally submitted message by a trusted user did not contain a
4953 From: header, and the sender address was obtained from -f or from an SMTP
4954 MAIL command, and the trusted user did not use -F to supply a sender
4955 name, $originator_name was incorrectly used when constructing a From:
4956 header. Furthermore, $originator_name was used for submission mode
4957 messages from external hosts without From: headers in a similar way,
4958 which is clearly wrong.
4960 PH/37 Added control=suppress_local_fixups.
4962 PH/38 When log_selector = +received_sender was set, and the addition of the
4963 sender made the log line's construction buffer exactly full, or one byte
4964 less than full, an overflow happened when the terminating "\n" was
4967 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4968 when the result of a list match is failure because a DNS lookup failed.
4970 PH/40 RM_COMMAND is now used in the building process.
4972 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4973 the "build-* directories that it finds.
4975 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4976 address in a domain literal was a prefix of an interface address.
4978 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4979 when verifying a sender address, unless rewrite_headers is false.
4981 PH/44 Wrote a long comment about why errors_to addresses are verified as
4982 recipients, not senders.
4984 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4985 the ratelimit ACL was added.
4987 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4989 PH/46 Added extra information about PostgreSQL errors to the error string.
4991 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4992 avoiding the need to install special zones in a real server. This is
4993 backwards compatible; if it can't find the fake resolver, it drops back.
4994 Thus, both old and new test suites can be run.
4996 TF/02 Added util/ratelimit.pl
4998 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4999 clock is set back in time.
5001 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5002 Candler <B.Candler@pobox.com>.
5004 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5005 OK for addresses that are the result of redirections.
5007 PH/48 A number of further additions for the benefit of the new test suite,
5008 including a fake gethostbyname() that interfaces to the fake DNS resolver
5011 TF/06 The fix for widen_domains has also been applied to qualify_single and
5012 search_parents which are the other dnslookup options that can cause
5015 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5018 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5020 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5021 service" if it encountered an MX record with an empty target hostname.
5022 The message is now "an MX or SRV record indicated no SMTP service".
5024 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5025 if the DNS of the sending site is misconfigured. This is quite a
5026 common situation. This change restores the behaviour of treating a
5027 helo verification defer as a failure.
5029 PH/51 If self=fail was set on a router, the bounce message did not include the
5030 actual error message.
5036 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5038 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5039 command itself ended in error, the underlying message about the transport
5040 filter timeout was being overwritten with the pipe command error. Now the
5041 underlying error message should be appended to the second error message.
5043 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5045 PH/02 Reduce the amount of output that "make" produces by default. Full output
5046 can still be requested.
5048 PH/03 The warning log line about a condition test deferring for a "warn" verb
5049 was being output only once per connection, rather than after each
5050 occurrence (because it was using the same function as for successful
5051 "warn" verbs). This seems wrong, so I have changed it.
5053 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5054 it should not have, which might have caused a crash in the right
5055 circumstances, but probably never did.
5057 PH/04 Installed a modified version of Tony Finch's patch to make submission
5058 mode fix the return path as well as the Sender: header line, and to
5059 add a /name= option so that you can make the user's friendly name appear
5062 TF/03 Added the control = fakedefer ACL modifier.
5064 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5065 Mark Lowes for thorough testing.
5067 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5069 TK/03 Merged latest SRS patch from Miles Wilton.
5071 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5072 with the definition in sysexits.h (which is #included earlier).
5073 Fortunately, Exim does not actually use EX_OK. The code used to try to
5074 preserve the sysexits.h value, by assuming that macro definitions were
5075 scanned for macro replacements. I have been disabused of this notion,
5076 so now the code just undefines EX_OK before #including unistd.h.
5078 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5079 in the smtp transport. When a block could not be written in a single
5080 write() function, the timeout was being re-applied to each part-write.
5081 This seems wrong - if the receiver was accepting one byte at a time it
5082 would take for ever. The timeout is now adjusted when this happens. It
5083 doesn't have to be particularly precise.
5085 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5086 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5088 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5089 without a domain if the check to <postmaster@domain> fails.
5091 SC/01 Eximstats: added -xls and the ability to specify output files
5092 (patch written by Frank Heydlauf).
5094 SC/02 Eximstats: use FileHandles for outputting results.
5096 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5098 SC/04 Eximstats: fixed display of large numbers with -nvr option
5100 SC/05 Eximstats: fixed merging of reports with empty tables.
5102 SC/06 Eximstats: added the -include_original_destination flag
5104 SC/07 Eximstats: removed tabs and trailing whitespace.
5106 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5108 TK/06 MBOX spool code: Add real "From " MBOX separator line
5109 so the .eml file is really in mbox format (even though
5110 most programs do not really care). Patch from Alex Miller.
5112 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5113 The latter is generated from $received_to and is only set if the
5114 message has one envelope recipient. SA can use these headers,
5115 obviously out-of-the-box. Patch from Alex Miller.
5117 PH/08 The ${def test on a variable was returning false if the variable's
5118 value was "0", contrary to what the specification has always said!
5119 The result should be true unless the variable is empty.
5121 PH/09 The syntax error of a character other than { following "${if
5122 def:variable_name" (after optional whitespace) was not being diagnosed.
5123 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5124 accidental colon was present, for example, could give incorrect results.
5126 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5127 result is used (not including appendfile, where other changes are about
5130 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5131 This involved changing a lot of size variables from int to off_t. It
5132 should work with maildirs and everything.
5134 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5135 spamd dying while we are connected to it.
5137 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5140 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5141 function for BDB 4.3.
5143 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5145 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5146 the sources for PCRE have been re-arranged and more files are now
5149 PH/15 The code I had for printing potentially long long variables in PH/11
5150 above was not the best (it lost precision). The length of off_t variables
5151 is now inspected at build time, and an appropriate printing format (%ld
5152 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5153 to be "long long int" or "long int". This is needed for the internal
5154 formatting function string_vformat().
5156 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5157 the configuration file to be ":syslog", then the script "guesses" where
5158 the logs files are, rather than using the compiled in default. In our
5159 case the guess is not the same as the compiled default, so the script
5160 suddenly stopped working when I started to use syslog. The patch checks
5161 to see if log_file_path is "". If so, it attempts to read it from exim
5162 with no configuration file to get the compiled in version, before it
5163 falls back to the previous guessing code."
5165 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5166 implementing BATV in an Exim configuration. See NewStuff for the gory
5169 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5170 Makefile that are specific to HP-UX.
5172 PH/18 If the "use_postmaster" option was set for a recipient callout together
5173 with the "random" option, the postmaster address was used as the MAIL
5174 FROM address for the random test, but not for the subsequent recipient
5175 test. It is now used for both.
5177 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5178 patch removes a few documentation additions to RFC 3028, because the
5179 latest draft now contains them. It adds the new en;ascii-case comparator
5180 and a new error check for 8bit text in MIME parts. Comparator and
5181 require names are now matched exactly. I enabled the subaddress
5182 extension, but it is not well tested yet (read: it works for me)."
5184 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5185 rework some of the code of TK/09 above to avoid the hardwired use of
5186 "%lld" and "long long". Replaced the call to snprintf() with a call to
5189 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5190 records point to non-existent hosts", "retry timeout exceeded", and
5191 "retry time not reached for any host after a long failure period".
5193 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5194 experimental DomainKeys support:
5196 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5197 (2) On an error such as an illegally used "control", the wrong name for
5198 the control was given.
5200 These problems did NOT occur unless DomainKeys support was compiled.
5202 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5204 PH/24 Added ${if match_ip condition.
5206 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5207 databases so that it will be absolutely obvious if a crash occurs in the
5208 DB library. This is a regular occurrence (often caused by mis-matched
5211 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5212 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5213 picked up on a user's system that detects such things. There doesn't seem
5214 to be a gcc warning option for this - only an attribute that has to be
5215 put on the function's prototype. It seems that in Fedora Core 4 they have
5216 set this on a number of new functions. No doubt there will be more in due
5219 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5220 specify a transport. However, if an address that was verified by such a
5221 router was the subject of a callout, Exim crashed because it tried to
5222 read the rcpt_include_affixes from the non-existent transport. Now it
5223 just assumes that the setting of that option is false. This bug was
5224 introduced by 4.51/PH/31.
5226 PH/28 Changed -d+all to exclude +memory, because that information is very
5227 rarely of interest, but it makes the output a lot bigger. People tend to
5228 do -d+all out of habit.
5230 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5231 code in os-type was giving problems when libc.so lives in lib64, like on
5234 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5235 aren't the modern standard, and it seems that some systems' include files
5236 don't always have them. Exim was already checking for some of the newer
5237 ones like T_AAAA, and defining it itself. I've added checks for all the
5238 record types that Exim uses.
5240 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5241 not automatically generating a new one, as it is supposed to. This
5242 prevented TLS from working. If the file did exist, but contained invalid
5243 data, a new version was generated, as expected. It was only the case of a
5244 non-existent file that was broken.
5246 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5247 with a change in libdomainkeys > 0.64.
5249 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5250 from DNS. If the selector record carries the flag, it now has
5251 precedence over the domain-wide flag.
5253 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5255 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5256 the use of a port name that isn't defined in /etc/services) occurred, the
5257 message was deferred as in a normal delivery, and thus remained on the
5258 spool, instead of being failed because of the mua_wrapper setting. This
5259 is now fixed, and I tidied up some of the mua_wrapper messages at the
5262 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5263 the messages in a hash of arrays rather than using individual hashes.
5264 This is a bit cleaner and results in dramatic memory savings, albeit
5265 at a slight CPU cost.
5267 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5268 as requested by Marc Sherman.
5270 SC/10 Eximstats: added histograms for user specified patterns as requested
5273 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5275 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5276 fopen() in the content-scanning modules that did not already have it.
5282 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5283 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5285 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5287 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5289 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5290 to test Sieve filters that use "vacation".
5292 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5293 that changes the way the GnuTLS parameters are stored in the cache file.
5294 The new format can be generated externally. For backward compatibility,
5295 if the data in the cache doesn't make sense, Exim assumes it has read an
5296 old-format file, and it generates new data and writes a new file. This
5297 means that you can't go back to an older release without removing the
5300 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5301 work if there are any delivery delays because "one_time" forces the
5302 parent to be marked "delivered", so its unseen clone is never tried
5303 again. For this reason, Exim now forbids the simultaneous setting of
5306 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5307 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5308 entry if you want to know the details. The fix, however, neglected to
5309 consider the case when local delivery batching is involved. The test for
5310 "previously delivered" was not happening when checking to see if an
5311 address could be batched with a previous (undelivered) one; under
5312 certain circumstances this could lead to multiple deliveries to the same
5315 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5316 in its include files, and this causes problems building Exim.
5318 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5319 header_syntax) but Exim was just ignoring anything given after a slash.
5320 In particular, this caused confusion with an attempt to use "verify =
5321 reverse_host_lookup/defer_ok". An error is now given when options are
5322 supplied for verify items that do not have them. (Maybe reverse_host_
5323 lookup should have a defer_ok option, but that's a different point.)
5325 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5326 defined by RFC 821) to 2048, because there were problems with some AUTH
5327 commands, and RFC 1869 says the size should be increased for extended
5328 SMTP commands that take arguments.
5330 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5333 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5334 "unknown" error; now it says that the functionality isn't in the binary.
5336 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5337 an address' error message when a string expansion fails (syntax or
5338 whatever). Otherwise the password may appear in the log. Following change
5339 PH/42 below, there is no longer a chance of it appearing in a bounce
5342 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5344 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5345 of its addresses was ever tried. (Bugzilla bug #2.)
5347 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5348 the result incorrectly in the debug output. (It correctly added a newline
5349 to what was transported.)
5351 TF/01 Added $received_time.
5353 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5354 commented out examples of how to interface to a virus scanner and to
5355 SpamAssassin. Also added commented examples of av_scanner and
5356 spamd_address settings.
5358 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5359 and controls are allowed in which ACLs. There were a couple of minor
5360 errors. Some of the entries in the conditions table (which is a table of
5361 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5362 negation of where the condition IS allowed.
5364 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5366 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5367 header file does not have a version number, so I've had to invent a new
5368 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5369 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5370 radiusclient), but it was contributed by a Radius user.
5372 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5373 files or directories, for OpenSSL.
5375 PH/19 When an Exim process that is running as root has to create an Exim log
5376 file, it does so in a subprocess that runs as exim:exim so as to get the
5377 ownership right at creation (otherwise, other Exim processes might see
5378 the file with the wrong ownership). There was no test for failure of this
5379 fork() call, which would lead to the process getting stuck as it waited
5380 for a non-existent subprocess. Forks do occasionally fail when resources
5381 run out. I reviewed all the other calls to fork(); they all seem to check
5384 PH/20 When checking for unexpected SMTP input at connect time (before writing
5385 the banner), Exim was not dealing correctly with a non-positive return
5386 from the read() function. If the client had disconnected by this time,
5387 the result was a log entry for a synchronization error with an empty
5388 string after "input=" when read() returned zero. If read() returned -1
5389 (an event I could not check), uninitialized data bytes were printed.
5390 There were reports of junk text (parts of files, etc) appearing after
5393 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5395 PH/22 Added support for macro redefinition, and (re)definition in between
5396 driver and ACL definitions.
5398 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5399 forgetting to use the resulting value; it was using the unexpanded value.
5401 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5402 hadn't been configured. The fix is from Juergen Kreileder, who
5403 understands it better than I do:
5405 "Here's what I see happening with three configured cyrus_sasl
5406 authenticators configured (plain, login, cram-md5):
5408 On startup auth_cyrus_sasl_init() gets called for each of these.
5409 This means three calls to sasl_listmech() without a specified mech_list.
5410 => SASL tests which mechs of all available mechs actually work
5411 => three warnings about OTP not working
5412 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5414 With the patch, sasl_listmech() also gets called three times. But now
5415 SASL's mech_list option is set to the server_mech specified in the the
5416 authenticator. Or in other words, the answer from sasl_listmech()
5417 gets limited to just the mech you're testing for (which is different
5419 => the return list contains just 'plain' or 'login', 'cram-md5' or
5420 nothing depending on the value of ob->server_mech.
5422 I've just tested the patch: Authentication still works fine,
5423 unavailable mechs specified in the exim configuration are still
5424 caught, and the auth.log warnings about OTP are gone."
5426 PH/25 When debugging is enabled, the contents of the command line are added
5427 to the debugging output, even when log_selector=+arguments is not
5430 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5431 answer is "GNU", and only if the return is "GNU/something" is the answer
5434 PH/27 $acl_verify_message is now set immediately after the failure of a
5435 verification in an ACL, and so is available in subsequent modifiers. In
5436 particular, the message can be preserved by coding like this:
5438 warn !verify = sender
5439 set acl_m0 = $acl_verify_message
5441 Previously, $acl_verify_message was set only while expanding "message"
5442 and "log_message" when a very denied access.
5444 PH/28 Modified OS/os.c-Linux with
5446 -#ifndef OS_LOAD_AVERAGE
5447 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5449 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5450 nomenclature these days.)
5452 PH/29 Installed patch from the Sieve maintainer that adds the options
5453 sieve_useraddress and sieve_subaddress to the redirect router.
5455 PH/30 In these circumstances:
5456 . Two addresses routed to the same list of hosts;
5457 . First host does not offer TLS;
5458 . First host accepts first address;
5459 . First host gives temporary error to second address;
5460 . Second host offers TLS and a TLS session is established;
5461 . Second host accepts second address.
5462 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5463 and peerdn, if requested) that were in fact used only for the second
5466 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5467 attention to any local part prefix or suffix that was matched by the
5468 router that accepted the address. It now behaves in the same way as it
5469 does for delivery: the affixes are removed from the local part unless
5470 rcpt_include_affixes is set on the transport.
5472 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5473 timeout during the DATA phase of an incoming message.
5475 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5476 applied a patch sanctioned by the Sieve maintainer.
5478 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5479 the uid or gid is negative. A case of a negative gid caused this to be
5480 noticed. The fix allows for either to be negative.
5482 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5483 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5486 PH/36 The change PH/12 above was broken. Fixed it.
5488 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5489 the grounds that routing the same address twice would always produce the
5490 same answer. This might have been true once, but it is certainly no
5491 longer true now. Routing a child address may depend on the previous
5492 routing that produced that child. Some complicated redirection strategies
5493 went wrong when messages had multiple recipients, and made Exim's
5494 behaviour dependent on the order in which the addresses were given.
5496 I have moved the duplicate checking until after the routing is complete.
5497 Exim scans the addresses that are assigned to local and remote
5498 transports, and removes any duplicates. This means that more work will be
5499 done, as duplicates will always all be routed, but duplicates are
5500 presumably rare, so I don't expect this is of any significance.
5502 For deliveries to pipes, files, and autoreplies, the duplicate checking
5503 still happens during the routing process, since they are not going to be
5506 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5507 It corrects a timeout issue with spamd. This is Ian's comment: "The
5508 background is that sometimes spamd either never reads data from a
5509 connection it has accepted, or it never writes response data. The exiscan
5510 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5511 blindly assumes that writes won't block so it may never time out."
5513 PH/39 Allow G after quota size as well as K and M.
5515 PH/40 The value set for $authenticated_id in an authenticator may not contain
5516 binary zeroes or newlines because the value is written to log lines and
5517 to spool files. There was no check on this. Now the value is run through
5518 the string_printing() function so that such characters are converted to
5519 printable escape sequences.
5521 PH/41 $message_linecount is a new variable that contains the total number of
5522 lines in the message. Compare $body_linecount, which is the count for the
5525 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5526 bounce and delay warning messages, except in certain special cases, which
5529 (a) An SMTP error message from a remote host;
5530 (b) A message specified in a :fail: redirection;
5531 (c) A message specified in a "fail" command in a system filter;
5532 (d) A message specified in a FAIL return from the queryprogram router;
5533 (e) A message specified by the cannot_route_message router option.
5535 In these cases only, Exim does include the error details in bounce and
5536 warning messages. There are also a few cases where bland messages such
5537 as "unrouteable address" or "local delivery error" are given.
5539 PH/43 $value is now also set for the "else" part of a ${run expansion.
5541 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5542 being worked on, but at least Exim now implements the latest version to
5545 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5546 process to complete was treated as a delivery failure, a timeout while
5547 writing the message to the pipe was logged, but erroneously treated as a
5548 successful delivery. Such timeouts include transport filter timeouts. For
5549 consistency with the overall process timeout, these timeouts are now
5550 treated as errors, giving rise to delivery failures by default. However,
5551 there is now a new Boolean option for the pipe transport called
5552 timeout_defer, which, if set TRUE, converts the failures into defers for
5553 both kinds of timeout. A transport filter timeout is now identified in
5556 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5557 systems where "make" and "gmake" are different, calling "gmake" at top
5558 level broke things. I've arranged for the value of $(MAKE) to be passed
5559 from the Makefile to this script so that it can call the same version of
5563 A note about Exim versions 4.44 and 4.50
5564 ----------------------------------------
5566 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5567 changes of various kinds. As a consequence, a big documentation update was
5568 needed. This delayed the release for rather longer than seemed good, especially
5569 in the light of a couple of (minor) security issues. Therefore, the changes
5570 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5571 release. So 4.44 and 4.50 are in effect two different branches that both start
5574 I have left the 4.50 change log unchanged; it contains all the changes since
5575 4.43. The change log for 4.44 is below; many of its items are identical to
5576 those for 4.50. This seems to be the most sensible way to preserve the
5577 historical information.
5583 1. Minor wording change to the doc/README.SIEVE file.
5585 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5586 computation of the current number of files was incorrect.
5588 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5589 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5592 4. Give more explanation in the error message when the command for a transport
5593 filter fails to execute.
5595 5. There are several places where Exim runs a non-Exim command in a
5596 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5597 was being done only for the command run by the queryprogram router. It is
5598 now done for all such subprocesses. The other cases are: ${run, transport
5599 filters, and the commands run by the lmtp and pipe transports.
5601 6. Added CONFIGURE_GROUP build-time option.
5603 7. Some older OS have a limit of 256 on the maximum number of file
5604 descriptors. Exim was using setrlimit() to set 1000 as a large value
5605 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5606 systems. I've change it so that if it can't get 1000, it tries for 256.
5608 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5609 was an oversight, and furthermore, ever since the addition of extra
5610 controls (e.g. 4.43/32), the checks on when to allow different forms of
5611 "control" were broken. There should now be diagnostics for all cases when a
5612 control that does not make sense is encountered.
5614 9. Added the /retain_sender option to "control=submission".
5616 10. $recipients is now available in the predata ACL (oversight).
5618 11. Tidy the search cache before the fork to do a delivery from a message
5619 received from the command line. Otherwise the child will trigger a lookup
5620 failure and thereby defer the delivery if it tries to use (for example) a
5621 cached ldap connection that the parent has called unbind on.
5623 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5624 of $address_data from the recipient verification was clobbered by the
5625 sender verification.
5627 13. The value of address_data from a sender verification is now available in
5628 $sender_address_data in subsequent conditions in the ACL statement.
5630 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5632 15. Added a new option "connect=<time>" to callout options, to set a different
5635 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5636 was its contents. (It was OK if the option was not defined at all.)
5638 17. A "Completed" log line is now written for messages that are removed from
5639 the spool by the -Mrm option.
5641 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5642 information about exactly what failed.
5644 19. Added -dd to debug only the daemon process.
5646 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5647 handles timeouts, both on the server side and network timeouts. Renamed the
5648 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5650 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5651 It is now set to "smtps".
5653 22. $host_address is now set to the target address during the checking of
5654 ignore_target_hosts.
5656 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5657 being passed; this would have caused $sender_host_name to have been used if
5658 matching the list had actually called for a host name (not very likely,
5659 since this list is usually IP addresses). A host name is now passed as
5662 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5663 code that specifies a non-restarting handler (typically sigaction() in
5664 modern systems) in an attempt to fix a rare and obscure crash bug.
5666 25. Narrowed the window for a race in the daemon that could cause it to ignore
5667 SIGCHLD signals. This is not a major problem, because they are used only to
5668 wake it up if nothing else does.
5670 26. A malformed maildirsize file could cause Exim to calculate negative values
5671 for the mailbox size or file count. Odd effects could occur as a result.
5672 The maildirsize information is now recalculated if the size or filecount
5675 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5676 support for a long time. Removed HAVE_SYS_VFS_H.
5678 28. Installed the latest version of exipick from John Jetmore.
5680 29. In an address list, if the pattern was not a regular expression, an empty
5681 subject address (from a bounce message) matched only if the pattern was an
5682 empty string. Non-empty patterns were not even tested. This was the wrong
5683 because it is perfectly reasonable to use an empty address as part of a
5684 database query. An empty address is now tested by patterns that are
5685 lookups. However, all the other forms of pattern expect the subject to
5686 contain a local part and a domain, and therefore, for them, an empty
5687 address still always fails if the pattern is not itself empty.
5689 30. Exim went into a mad DNS loop when attempting to do a callout where the
5690 host was specified on an smtp transport, and looking it up yielded more
5691 than one IP address.
5693 31. Re-factored the code for checking spool and log partition space into a
5694 function that finds that data and another that does the check. The former
5695 is then used to implement four new variables: $spool_space, $log_space,
5696 $spool_inodes, and $log_inodes.
5698 32. The RFC2047 encoding function was originally intended for short strings
5699 such as real names; it was not keeping to the 75-character limit for
5700 encoded words that the RFC imposes. It now respects the limit, and
5701 generates multiple encoded words if necessary. To be on the safe side, I
5702 have increased the buffer size for the ${rfc2047: expansion operator from
5705 33. It is now permitted to omit both strings after an "if" condition; if the
5706 condition is true, the result is "true". As before, when the second string
5707 is omitted, a false condition yields an empty string. This makes it less
5708 cumbersome to write custom ACL and router conditions.
5710 34. Failure to deliver a bounce message always caused it to be frozen, even if
5711 there was an errors_to setting on the router. The errors_to setting is now
5714 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5715 canonical form (fully expanded) before being placed in
5716 $sender_host_address.
5718 36. The table in the code that translates DNS record types into text (T_A to
5719 "A" for instance) was missing entries for NS and CNAME. It is just possible
5720 that this could have caused confusion if both these types were looked up
5721 for the same domain, because the text type is used as part of Exim's
5722 per-process caching. But the chance of anyone hitting this buglet seems
5725 37. The dnsdb lookup has been extended in a number of ways.
5727 (1) There is a new type, "zns", which walks up the domain tree until it
5728 finds some nameserver records. It should be used with care.
5730 (2) There is a new type, "mxh", which is like "mx" except that it returns
5731 just the host names, not the priorities.
5733 (3) It is now possible to give a list of domains (or IP addresses) to be
5734 looked up. The behaviour when one of the lookups defers can be
5735 controlled by a keyword.
5737 (4) It is now possible to specify the separator character for use when
5738 multiple records are returned.
5740 38. The dnslists ACL condition has been extended: it is now possible to supply
5741 a list of IP addresses and/or domains to be looked up in a particular DNS
5744 39. Added log_selector=+queue_time_overall.
5746 40. When running the queue in the test harness, wait just a tad after forking a
5747 delivery process, to get repeatability of debugging output.
5749 41. Include certificate and key file names in error message when GnuTLS fails
5750 to set them up, because the GnuTLS error message doesn't include the name
5751 of the failing file when there is a problem reading it.
5753 42. Allow both -bf and -bF in the same test run.
5755 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5757 44. The "Exiscan patch" is now merged into the mainline Exim source.
5759 45. Sometimes the final signoff response after QUIT could fail to get
5760 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5761 < 0 before doing a fflush(). This bug looks as though it goes back to the
5762 introduction of TLS in release 3.20, but "sometimes" must have been rare
5763 because the tests only now provoked it.
5765 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5766 (this can affect the format of dates).
5768 47. exim_tidydb, when checking for the continued existence of a message for
5769 which it has found a message-specific retry record, was not finding
5770 messages that were in split spool directories. Consequently, it was
5771 deleting retry records that should have stayed in existence.
5773 48. Steve fixed some bugs in eximstats.
5775 49. The SPA authentication driver was not abandoning authentication and moving
5776 on to the next authenticator when an expansion was forced to fail,
5777 contradicting the general specification for all authenticators. Instead it
5778 was generating a temporary error. It now behaves as specified.
5780 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5781 (the order specifies the preference for clients). The order is now AES256,
5782 AES128, 3DES, ARCFOUR128.
5784 51. Small patch to Sieve code - explicitly set From: when generating an
5787 52. Exim crashed if a remote delivery caused a very long error message to be
5788 recorded - for instance if somebody sent an entire SpamAssassin report back
5789 as a large number of 550 error lines. This bug was coincidentally fixed by
5790 increasing the size of one of Exim's internal buffers (big_buffer) that
5791 happened as part of the Exiscan merge. However, to be on the safe side, I
5792 have made the code more robust (and fixed the comments that describe what
5795 53. Now that there can be additional text after "Completed" in log lines (if
5796 the queue_time_overall log selector is set), a one-byte patch to exigrep
5797 was needed to allow it to recognize "Completed" as not the last thing in
5800 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5801 patch that reportedly fixes this has been added. I am not expert enough to
5802 create a test for it. This is what the patch creator wrote:
5804 "I found a little strange behaviour of ldap code when working with
5805 Windows 2003 AD Domain, where users was placed in more than one
5806 Organization Units. When I tried to give exim partial DN, the exit code
5807 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5808 But simultaneously result of request was absolutely normal ldap result,
5809 so I produce this patch..."
5811 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5812 REFERENCE, so I have modified the code to exclude the patch when that macro
5815 55. Some experimental protocols are using DNS PTR records for new purposes. The
5816 keys for these records are domain names, not reversed IP addresses. The
5817 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5818 leaves it alone. Component reversal etc. now happens only for IP addresses.
5821 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5823 57. Double the size of the debug message buffer (to 2048) so that more of very
5824 long debug lines gets shown.
5826 58. The exicyclog utility now does better if the number of log files to keep
5827 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5829 59. Two changes related to the smtp_active_hostname option:
5831 (1) $smtp_active_hostname is now available as a variable.
5832 (2) The default for smtp_banner uses $smtp_active_hostname instead
5833 of $primary_hostname.
5835 60. The host_aton() function is supposed to be passed a string that is known
5836 to be a valid IP address. However, in the case of IPv6 addresses, it was
5837 not checking this. This is a hostage to fortune. Exim now panics and dies
5838 if the condition is not met. A case was found where this could be provoked
5839 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5840 components; fortuitously, this particular loophole had already been fixed
5841 by change 4.50/55 above.
5843 If there are any other similar loopholes, the new check in host_aton()
5844 itself should stop them being exploited. The report I received stated that
5845 data on the command line could provoke the exploit when Exim was running as
5846 exim, but did not say which command line option was involved. All I could
5847 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5848 running as the user.
5851 61. There was a buffer overflow vulnerability in the SPA authentication code
5852 (which came originally from the Samba project). I have added a test to the
5853 spa_base64_to_bits() function which I hope fixes it.
5856 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5857 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5859 63. The daemon start-up calls getloadavg() while still root for those OS that
5860 need the first call to be done as root, but it missed one case: when
5861 deliver_queue_load_max is set with deliver_drop_privilege. This is
5862 necessary for the benefit of the queue runner, because there is no re-exec
5863 when deliver_drop_privilege is set.
5865 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5866 This has been fixed.
5868 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5869 was in use, was not putting the data itself into the right store pool;
5870 consequently, it could be overwritten for a subsequent message in the same
5871 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5874 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5876 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5877 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5878 the function were treating the return as a boolean value, which happened to
5879 work because 0=false and not-0=true, but is not correct code.
5881 68. The host_aton() function was not handling scoped IPv6 addresses (those
5882 with, for example, "%eth0" on the end) correctly.
5884 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5885 negated items (that is, ~something) in unsigned ints. Some compilers
5886 apparently mutter when there is no cast.
5888 70. If an address verification called from an ACL failed, and did not produce a
5889 user-specific message (i.e. there was only a "system" message), nothing was
5890 put in $acl_verify_message. In this situation, it now puts the system
5893 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5894 session; change 4.31/43 added the unwanted input to the log line - except
5895 that it did not do this in the start of session case. It now does.
5897 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5898 This is wrong and can cause the other end to generate a synchronization
5899 error if it is another Exim or anything else that does the synchronization
5900 check. A QUIT command is no longer sent after a timeout.
5902 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5903 during host lookups.
5905 74. The defer_ok option of callout verification was not working if it was used
5906 when verifying addresses in header lines, that is, for this case:
5908 verify = header_sender/callout=defer_ok
5910 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5911 those file descriptors could be used for SMTP connections. If anything
5912 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5913 could be sent to the SMTP client, causing chaos. The daemon now opens
5914 stdin, stdout, and stderr to /dev/null when it puts itself into the
5917 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5918 log by default. The user can override this with suitable Perl magic.
5920 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5921 the log message when discard triggers, was not working for the DATA ACL or
5922 for the non-SMTP ACL.
5924 78. Error message wording change in sieve.c.
5926 79. If smtp_accept_max_per_host was set, the number of connections could be
5927 restricted to fewer than expected, because the daemon was trying to set up
5928 a new connection before checking whether the processes handling previous
5929 connections had finished. The check for completed processes is now done
5930 earlier. On busy systems, this bug wouldn't be noticed because something
5931 else would have woken the daemon, and it would have reaped the completed
5934 80. If a message was submitted locally by a user whose login name contained one
5935 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5936 It caused a spool format error. I have fixed the spool reading code. A
5937 related problem was that the "from" clause in the Received: line became
5938 illegal because of the space(s). It is now covered by ${quote_local_part.
5940 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5943 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5945 83. Patch from Sieve maintainer to fix unterminated string problem in
5946 "vacation" handling.
5948 84. Some minor changes to the Linux configuration files to help with other
5949 OS variants using glibc.
5951 85. One more patch for Sieve to update vacation handling to latest spec.
5954 ----------------------------------------------------
5955 See the note above about the 4.44 and 4.50 releases.
5956 ----------------------------------------------------
5962 1. Change 4.43/35 introduced a bug that caused file counts to be
5963 incorrectly computed when quota_filecount was set in an appendfile
5966 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5967 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5970 3. Give more explanation in the error message when the command for a transport
5971 filter fails to execute.
5973 4. There are several places where Exim runs a non-Exim command in a
5974 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5975 was being done only for the command run by the queryprogram router. It is
5976 now done for all such subprocesses. The other cases are: ${run, transport
5977 filters, and the commands run by the lmtp and pipe transports.
5979 5. Some older OS have a limit of 256 on the maximum number of file
5980 descriptors. Exim was using setrlimit() to set 1000 as a large value
5981 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5982 systems. I've change it so that if it can't get 1000, it tries for 256.
5984 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5985 was an oversight, and furthermore, ever since the addition of extra
5986 controls (e.g. 4.43/32), the checks on when to allow different forms of
5987 "control" were broken. There should now be diagnostics for all cases when a
5988 control that does not make sense is encountered.
5990 7. $recipients is now available in the predata ACL (oversight).
5992 8. Tidy the search cache before the fork to do a delivery from a message
5993 received from the command line. Otherwise the child will trigger a lookup
5994 failure and thereby defer the delivery if it tries to use (for example) a
5995 cached ldap connection that the parent has called unbind on.
5997 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5998 of $address_data from the recipient verification was clobbered by the
5999 sender verification.
6001 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6002 was its contents. (It was OK if the option was not defined at all.)
6004 11. A "Completed" log line is now written for messages that are removed from
6005 the spool by the -Mrm option.
6007 12. $host_address is now set to the target address during the checking of
6008 ignore_target_hosts.
6010 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6011 being passed; this would have caused $sender_host_name to have been used if
6012 matching the list had actually called for a host name (not very likely,
6013 since this list is usually IP addresses). A host name is now passed as
6016 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6017 code that specifies a non-restarting handler (typically sigaction() in
6018 modern systems) in an attempt to fix a rare and obscure crash bug.
6020 15. Narrowed the window for a race in the daemon that could cause it to ignore
6021 SIGCHLD signals. This is not a major problem, because they are used only to
6022 wake it up if nothing else does.
6024 16. A malformed maildirsize file could cause Exim to calculate negative values
6025 for the mailbox size or file count. Odd effects could occur as a result.
6026 The maildirsize information is now recalculated if the size or filecount
6029 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6030 support for a long time. Removed HAVE_SYS_VFS_H.
6032 18. Updated exipick to current release from John Jetmore.
6034 19. Allow an empty sender to be matched against a lookup in an address list.
6035 Previously the only cases considered were a regular expression, or an
6038 20. Exim went into a mad DNS lookup loop when doing a callout where the
6039 host was specified on the transport, if the DNS lookup yielded more than
6042 21. The RFC2047 encoding function was originally intended for short strings
6043 such as real names; it was not keeping to the 75-character limit for
6044 encoded words that the RFC imposes. It now respects the limit, and
6045 generates multiple encoded words if necessary. To be on the safe side, I
6046 have increased the buffer size for the ${rfc2047: expansion operator from
6049 22. Failure to deliver a bounce message always caused it to be frozen, even if
6050 there was an errors_to setting on the router. The errors_to setting is now
6053 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6054 canonical form (fully expanded) before being placed in
6055 $sender_host_address.
6057 24. Updated eximstats to version 1.33
6059 25. Include certificate and key file names in error message when GnuTLS fails
6060 to set them up, because the GnuTLS error message doesn't include the name
6061 of the failing file when there is a problem reading it.
6063 26. Expand error message when OpenSSL has problems setting up cert/key files.
6066 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6067 (this can affect the format of dates).
6069 28. exim_tidydb, when checking for the continued existence of a message for
6070 which it has found a message-specific retry record, was not finding
6071 messages that were in split spool directories. Consequently, it was
6072 deleting retry records that should have stayed in existence.
6074 29. eximstats updated to version 1.35
6075 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6076 1.35 - bugfix such that pie charts by volume are generated correctly
6078 30. The SPA authentication driver was not abandoning authentication and moving
6079 on to the next authenticator when an expansion was forced to fail,
6080 contradicting the general specification for all authenticators. Instead it
6081 was generating a temporary error. It now behaves as specified.
6083 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6084 (the order specifies the preference for clients). The order is now AES256,
6085 AES128, 3DES, ARCFOUR128.
6087 31. Small patch to Sieve code - explicitly set From: when generating an
6090 32. Exim crashed if a remote delivery caused a very long error message to be
6091 recorded - for instance if somebody sent an entire SpamAssassin report back
6092 as a large number of 550 error lines. This bug was coincidentally fixed by
6093 increasing the size of one of Exim's internal buffers (big_buffer) that
6094 happened as part of the Exiscan merge. However, to be on the safe side, I
6095 have made the code more robust (and fixed the comments that describe what
6098 33. Some experimental protocols are using DNS PTR records for new purposes. The
6099 keys for these records are domain names, not reversed IP addresses. The
6100 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6101 leaves it alone. Component reversal etc. now happens only for IP addresses.
6104 34. The host_aton() function is supposed to be passed a string that is known
6105 to be a valid IP address. However, in the case of IPv6 addresses, it was
6106 not checking this. This is a hostage to fortune. Exim now panics and dies
6107 if the condition is not met. A case was found where this could be provoked
6108 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6109 components; fortuitously, this particular loophole had already been fixed
6110 by change 4.50/55 or 4.44/33 above.
6112 If there are any other similar loopholes, the new check in host_aton()
6113 itself should stop them being exploited. The report I received stated that
6114 data on the command line could provoke the exploit when Exim was running as
6115 exim, but did not say which command line option was involved. All I could
6116 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6117 running as the user.
6120 35. There was a buffer overflow vulnerability in the SPA authentication code
6121 (which came originally from the Samba project). I have added a test to the
6122 spa_base64_to_bits() function which I hope fixes it.
6125 36. The daemon start-up calls getloadavg() while still root for those OS that
6126 need the first call to be done as root, but it missed one case: when
6127 deliver_queue_load_max is set with deliver_drop_privilege. This is
6128 necessary for the benefit of the queue runner, because there is no re-exec
6129 when deliver_drop_privilege is set.
6131 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6132 was in use, was not putting the data itself into the right store pool;
6133 consequently, it could be overwritten for a subsequent message in the same
6134 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6137 38. Sometimes the final signoff response after QUIT could fail to get
6138 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6139 < 0 before doing a fflush(). This bug looks as though it goes back to the
6140 introduction of TLS in release 3.20, but "sometimes" must have been rare
6141 because the tests only now provoked it.
6147 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6148 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6149 Now it returns an integer. A number of calls were still expecting a T/F
6150 return. Fortuitously, in all cases, the tests worked in OK situations,
6151 which is the norm. However, things would have gone wrong on any write
6152 failures on the smtp file descriptor. This function is used when sending
6153 messages over SMTP and also when doing verify callouts.
6155 2. When Exim is called to do synchronous delivery of a locally submitted
6156 message (the -odf or -odi options), it no longer closes stderr before doing
6159 3. Implemented the mua_wrapper option.
6161 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6163 5. Implemented the functions header_remove(), header_testname(),
6164 header_add_at_position(), and receive_remove_recipient(), and exported them
6167 6. If an ACL "warn" statement specified the addition of headers, Exim already
6168 inserted X-ACL-Warn: at the start if there was no header name. However, it
6169 was not making this test for the second and subsequent header lines if
6170 there were newlines in the string. This meant that an invalid header could
6171 be inserted if Exim was badly configured.
6173 7. Allow an ACL "warn" statement to add header lines at the start or after all
6174 the Received: headers, as well as at the end.
6176 8. Added the rcpt_4xx retry error code.
6178 9. Added postmaster_mailfrom=xxx to callout verification option.
6180 10. Added mailfrom=xxxx to the callout verification option, for verify=
6183 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6184 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6186 12. Inserted some casts to stop certain compilers warning when using pointer
6187 differences as field lengths or precisions in printf-type calls (mostly
6188 affecting debugging statements).
6190 13. Added optional readline() support for -be (dynamically loaded).
6192 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6193 same clock tick as a message's arrival, so that its received time was the
6194 same as the "first fail" time on the retry record, and that message
6195 remained on the queue past the ultimate address timeout, every queue runner
6196 would try a delivery (because it was past the ultimate address timeout) but
6197 after another failure, the ultimate address timeout, which should have then
6198 bounced the address, did not kick in. This was a "< instead of <=" error;
6199 in most cases the first failure would have been in the next clock tick
6200 after the received time, and all would be well.
6202 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6203 being recognized when the domain list was tested by the match_domain
6204 condition in an expansion string.
6206 16. Added the ${str2b64: operator.
6208 17. Exim was always calling setrlimit() to set a large limit for the number of
6209 processes, without checking whether the existing limit was already
6210 adequate. (It did check for the limit on file descriptors.) Furthermore,
6211 errors from getrlimit() and setrlimit() were being ignored. Now they are
6212 logged to the main and panic logs, but Exim does carry on, to try to do its
6213 job under whatever limits there are.
6215 18. Imported PCRE 5.0.
6217 19. Trivial typo in log message " temporarily refused connection" (the leading
6220 20. If the log selector return_path_on_delivery was set and an address was
6221 redirected to /dev/null, the delivery process crashed because it assumed
6222 that a return path would always be set for a "successful" delivery. In this
6223 case, the whole delivery is bypassed as an optimization, and therefore no
6226 21. Internal re-arrangement: the function for sending a challenge and reading
6227 a response while authentication was assuming a zero-terminated challenge
6228 string. It's now changed to take a pointer and a length, to allow for
6229 binary data in such strings.
6231 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6233 23. Exim was not respecting finduser_retries when seeking the login of the
6234 uid under which it was called; it was always trying 10 times. (The default
6235 setting of finduser_retries is zero.) Also, it was sleeping after the final
6236 failure, which is pointless.
6238 24. Implemented tls_on_connect_ports.
6240 25. Implemented acl_smtp_predata.
6242 26. If the domain in control=submission is set empty, Exim assumes that the
6243 authenticated id is a complete email address when it generates From: or
6244 Sender: header lines.
6246 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6247 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6248 chown and chgrp in /bin and hostname in /usr/bin.
6250 28. Exim was keeping the "process log" file open after each use, just as it
6251 does for the main log. This opens the possibility of it remaining open for
6252 long periods when the USR1 signal hits a daemon. Occasional processlog
6253 errors were reported, that could have been caused by this. Anyway, it seems
6254 much more sensible not to leave this file open at all, so that is what now
6257 29. The long-running daemon process does not normally write to the log once it
6258 has entered its main loop, and it closes the log before doing so. This is
6259 so that log files can straightforwardly be renamed and moved. However,
6260 there are a couple of unusual error situations where the daemon does write
6261 log entries, and I had neglected to close the log afterwards.
6263 30. The text of an SMTP error response that was received during a remote
6264 delivery was being truncated at 512 bytes. This is too short for some of
6265 the long messages that one sometimes sees. I've increased the limit to
6268 31. It is now possible to make retry rules that apply only when a message has a
6269 specific sender, in particular, an empty sender.
6271 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6272 it possible to be selective about when SMTP synchronization is enforced.
6274 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6276 32. Implemented hosts_connection_nolog.
6278 33. Added an ACL for QUIT.
6280 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6283 35. Added mailbox_size and mailbox_filecount to appendfile.
6285 36. Added control = no_multiline_responses to ACLs.
6287 37. There was a bug in the logic of the code that waits for the clock to tick
6288 in the case where the clock went backwards by a substantial amount such
6289 that the microsecond fraction of "now" was more than the microsecond
6290 fraction of "then" (but the whole seconds number was less).
6292 38. Added support for the libradius Radius client library this is found on
6293 FreeBSD (previously only the radiusclient library was supported).
6299 1. When certain lookups returned multiple values in the form name=value, the
6300 quoting of the values was not always being done properly. Specifically:
6301 (a) If the value started with a double quote, but contained no whitespace,
6303 (b) If the value contained whitespace other than a space character (i.e.
6304 tabs or newlines or carriage returns) it was not quoted.
6305 This fix has been applied to the mysql and pgsql lookups by writing a
6306 separate quoting function and calling it from the lookup code. The fix
6307 should probably also be applied to nisplus, ibase and oracle lookups, but
6308 since I cannot test any of those, I have not disturbed their existing code.
6310 2. A hit in the callout cache for a specific address caused a log line with no
6311 reason for rejecting RCPT. Now it says "Previous (cached) callout
6312 verification failure".
6314 3. There was an off-by-one bug in the queryprogram router. An over-long
6315 return line was truncated at 256 instead of 255 characters, thereby
6316 overflowing its buffer with the terminating zero. As well as fixing this, I
6317 have increased the buffer size to 1024 (and made a note to document this).
6319 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6320 when Exim is waiting for an SMTP response from a remote server, Exim
6321 restarts its select() call on the socket, thereby resetting its timeout.
6322 This is not a problem when such interrupts are rare. Somebody set up a cron
6323 job to run exiwhat every 2 minutes, which is less than the normal select()
6324 timeout (5 or 10 minutes). This meant that the select() timeout never
6325 kicked in because it was always reset. I have fixed this by comparing the
6326 time when an interrupt arrives with the time at the start of the first call
6327 to select(). If more time than the timeout has elapsed, the interrupt is
6328 treated as a timeout.
6330 5. Some internal re-factoring in preparation for the addition of Sieve
6331 extensions (by MH). In particular, the "personal" test is moved to a
6332 separate function, and given an option for scanning Cc: and Bcc: (which is
6333 not set for Exim filters).
6335 6. When Exim created an email address using the login of the caller as the
6336 local part (e.g. when creating a From: or Sender: header line), it was not
6337 quoting the local part when it contained special characters such as @.
6339 7. Installed new OpenBSD configuration files.
6341 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6342 try to make them clearer.
6344 9. Callout options, other than the timeout value, were being ignored when
6345 verifying sender addresses in header lines. For example, when using
6347 verify = header_sender/callout=no_cache
6349 the cache was (incorrectly) being used.
6351 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6352 only the Cygwin environment.
6354 11. When return_path_on_delivery was set as a log selector, if different remote
6355 addresses in the same message used different return paths and parallel
6356 remote delivery occurred, the wrong values would sometimes be logged.
6357 (Whenever a remote delivery process finished, the return path value from
6358 the most recently started remote delivery process was logged.)
6360 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6361 lines when AUTH and/or TLS are in use. This is the "received protocol"
6362 field. Exim used to use "asmtp" for authenticated SMTP, without any
6363 indication (in the protocol name) for TLS use. Now it follows the RFC and
6364 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6365 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6366 names appear in log lines as well as in Received: header lines.
6368 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6369 extensions, and comparison tests, and to fix some bugs.
6371 14. Changes to the "personal" filter test:
6373 (1) The test was buggy in that it was just doing the equivalent of
6374 "contains" tests on header lines. For example, if a user's address was
6375 anne@some.where, the "personal" test would incorrectly be true for
6377 To: susanne@some.where
6379 This test is now done by extracting each address from the header in turn,
6380 and checking the entire address. Other tests that are part of "personal"
6381 are now done using regular expressions (for example, to check local parts
6382 of addresses in From: header lines).
6384 (2) The list of non-personal local parts in From: addresses has been
6385 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6386 taken from the Sieve specification recommendations.
6388 (3) If the message contains any header line starting with "List-" it is
6389 treated as non-personal.
6391 (4) The test for "circular" in the Subject: header line has been removed
6392 because it now seems ill-conceived.
6394 15. Minor typos in src/EDITME comments corrected.
6396 16. Installed latest exipick from John Jetmore.
6398 17. If headers_add on a router specified a text string that was too long for
6399 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6400 of string_sprintf() is now avoided.
6402 18. $message_body_size was not set (it was always zero) when running the DATA
6403 ACL and the local_scan() function.
6405 19. For the "mail" command in an Exim filter, no default was being set for
6406 the once_repeat time, causing a random time value to be used if "once" was
6407 specified. (If the value happened to be <= 0, no repeat happened.) The
6408 default is now 0s, meaning "never repeat". The "vacation" command was OK
6409 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6410 (I found it when inspecting the code).
6412 20. There is now an overall timeout for performing a callout verification. It
6413 defaults to 4 times the callout timeout, which applies to individual SMTP
6414 commands during the callout. The overall timeout applies when there is more
6415 than one host that can be tried. The timeout is checked before trying the
6416 next host. This prevents very long delays if there are a large number of
6417 hosts and all are timing out (e.g. when the network connections are timing
6418 out). The value of the overall timeout can be changed by specifying an
6419 additional sub-option for "callout", called "maxwait". For example:
6421 verify = sender/callout=5s,maxwait=20s
6423 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6424 to the end before writing, but this should make it even safer).
6426 22. Exim was forgetting that it had advertised PIPELINING for the second and
6427 subsequent messages on an SMTP connection. It was also not resetting its
6428 memory on STARTTLS and an internal HELO.
6430 23. When Exim logs an SMTP synchronization error within a session, it now
6431 records whether PIPELINING has been advertised or not.
6433 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6434 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6435 rather than long int.
6437 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6439 26. Added the never_mail option to autoreply.
6445 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6446 crash if the getsockname() call failed; this can happen if a connection is
6447 closed very soon after it is established. The problem was simply in the
6448 order in which certain operations were done, causing Exim to try to write
6449 to the SMTP stream before it had set up the file descriptor. The bug has
6450 been fixed by making things happen in the correct order.
6456 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6457 before the connection was closed, thus losing the rejection response.
6459 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6460 some early Solaris releases, but causes trouble in current releases where
6461 socklen_t is defined.
6463 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6466 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6469 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6470 that releasing the top of it at the end releases what was used for sub-
6471 expansions (unless the block got too big). However, discard this block if
6472 the first thing is a variable or header, so that we can use its block when
6473 it is dynamic (useful for very large $message_headers, for example).
6475 6. Lookups now cache *every* query, not just the most recent. A new, separate
6476 store pool is used for this. It can be recovered when all lookup caches are
6477 flushed. Lookups now release memory at the end of their result strings.
6478 This has involved some general refactoring of the lookup sources.
6480 7. Some code has been added to the store_xxx() functions to reduce the amount
6481 of flapping under certain conditions.
6483 8. log_incoming_interface used to affect only the <= reception log lines. Now
6484 it causes the local interface and port to be added to several more SMTP log
6485 lines, for example "SMTP connection from", and rejection lines.
6487 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6489 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6491 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6492 could be overwritten at the end of the current message (or the start of a
6493 new message if it was set in a HELO ACL). The value is now preserved for
6494 the duration of the SMTP connection.
6496 12. If a transport had a headers_rewrite setting, and a matching header line
6497 contained an unqualified address, that address was qualified, even if it
6498 did not match any rewriting rules. The underlying bug was that the values
6499 of the flags that permit the existence of unqualified sender and recipient
6500 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6501 non-local messages, and by -bnq for local messages) were not being
6502 preserved with the message after it was received.
6504 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6505 "next input=" as part of the text comprising the host identity instead of
6506 the correct text. The code was using the same buffer for two different
6507 strings. However, depending on which order the printing function evaluated
6508 its arguments, the bug did not always show up. Under Linux, for example, my
6509 test suite worked just fine.
6511 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6512 doesn't work with some older versions of Perl. It has been changed to "my",
6513 which in any case is probably the better facility to use.
6515 15. A really picky compiler found some instances of statements for creating
6516 error messages that either had too many or two few arguments for the format
6519 16. The size of the buffer for calls to the DNS resolver has been increased
6520 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6521 for addresses that have a lot of PTR records. This alleviates a problem; it
6522 does not fully solve it.
6524 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6525 buffer now truncates the list and logs the incident, which is the same
6526 action as happens when Exim is looking up a host name and its aliases.
6527 Previously in this situation something unpredictable would happen;
6528 sometimes it was "internal error: store_reset failed".
6530 18. If a server dropped the connection unexpectedly when an Exim client was
6531 using GnuTLS and trying to read a response, the client delivery process
6532 crashed while trying to generate an error log message.
6534 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6535 string, for example:
6537 warn message = H1: something\nH2: something
6539 the text was added as a single header line from Exim's point of view
6540 though it ended up OK in the delivered message. However, searching for the
6541 second and subsequent header lines using $h_h2: did not work. This has been
6542 fixed. Similarly, if a system filter added multiple headers in this way,
6543 the routers could not see them.
6545 20. Expanded the error message when iplsearch is called with an invalid key to
6546 suggest using net-iplsearch in a host list.
6548 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6549 ACLs are no longer actually imposed (and a message to that effect is
6552 22. If a "gecos" field in a passwd entry contained escaped characters, in
6553 particular, if it contained a \" sequence, Exim got it wrong when building
6554 a From: or a Sender: header from that name. A second bug also caused
6555 incorrect handling when an unquoted " was present following a character
6556 that needed quoting.
6558 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6559 was not being matched caselessly.
6561 24. Arranged for all hyphens in the exim.8 source to be escaped with
6564 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6565 properly. Recipient callout cache records were still being keyed to include
6566 the sender, even when use_sender was set false. This led to far more
6567 callouts that were necessary. The sender is no longer included in the key
6568 when use_sender is false.
6570 26. Added "control = submission" modifier to ACLs.
6572 27. Added the ${base62d: operator to decode base 62 numbers.
6574 28. dnsdb lookups can now access SRV records.
6576 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6577 the configuration file.
6579 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6580 (-v) mode. This makes the output for a verbose queue run more intelligible.
6582 31. Added a use_postmaster feature to recipient callouts.
6584 32. Added the $body_zerocount variable, containing the number of binary zero
6585 bytes in the message body.
6587 33. The time of last modification of the "new" subdirectory is now used as the
6588 "mailbox time last read" when there is a quota error for a maildir
6591 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6593 35. Added +ignore_unknown as a special item in host lists.
6595 36. Code for decoding IPv6 addresses in host lists is now included, even if
6596 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6597 address was recognized as an IP address, but was then not correctly decoded
6598 into binary, causing unexpected and incorrect effects when compared with
6605 1. Very minor rewording of debugging text in manualroute to say "list of
6606 hosts" instead of "hostlist".
6608 2. If verify=header_syntax was set, and a header line with an unqualified
6609 address (no domain) and a large number of spaces between the end of the
6610 name and the colon was received, the reception process suffered a buffer
6611 overflow, and (when I tested it) crashed. This was caused by some obsolete
6612 code that should have been removed. The fix is to remove it!
6614 3. When running in the test harness, delay a bit after writing a bounce
6615 message to get a bit more predictability in the log output.
6617 4. Added a call to search_tidyup() just before forking a reception process. In
6618 theory, someone could use a lookup in the expansion of smtp_accept_max_
6619 per_host which, without the tidyup, could leave open a database connection.
6621 5. Added the variables $recipient_data and $sender_data which get set from a
6622 lookup success in an ACL "recipients" or "senders" condition, or a router
6623 "senders" option, similar to $domain_data and $local_part_data.
6625 6. Moved the writing of debug_print from before to after the "senders" test
6628 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6629 problems for message scanning, either using a data ACL, or using
6630 local_scan() because the Received: header was not generated till after they
6631 were called (in order to set the time as the time of reception completion).
6632 I have revised the way this works. The header is now generated after the
6633 body is received, but before the ACL or local_scan() are called. After they
6634 are run, the timestamp in the header is updated.
6640 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6641 before starting a queue runner without re-exec. This happened only when
6642 deliver_drop_privilege was set or when the Exim user was set to root. The
6643 effect of the bug was that timeouts during subsequent deliveries caused
6644 crashes instead of being properly handled. The handler is now left at its
6645 default (and expected) setting.
6647 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6648 message, again when deliver_drop_privilege is set or Exim is run as root.
6649 The bug described in (1) was not present in this case, but the tidying up
6650 of the other signals was missing. I have made the two cases consistent.
6652 3. The ignore_target_hosts setting on a manualroute router was being ignored
6653 for hosts that were looked up using the /MX notation.
6655 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6658 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6659 operated on the sender address. After changing the $sender_address to <>
6660 for the sender address verify, Exim was re-instated it as the original
6661 (before rewriting) address, but remembering that it had rewritten it, so it
6662 wasn't rewriting it again. This bug also had the effect of breaking the
6663 sender address verification caching when the sender address was rewritten.
6665 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6666 This has been changed so that if the ip literal address matches
6667 ignore_target_hosts, the router declines.
6669 7. Added expansion conditions match_domain, match_address, and match_local_
6670 part (NOT match_host).
6672 8. The placeholder for the Received: header didn't have a length field set.
6674 9. Added code to Exim itself and to exim_lock to test for a specific race
6675 condition that could lead to file corruption when using MBX delivery. The
6676 issue is with the lockfile that is created in /tmp. If this file is removed
6677 after a process has opened it but before that process has acquired a lock,
6678 there is the potential for a second process to recreate the file and also
6679 acquire a lock. This could lead to two Exim processes writing to the file
6680 at the same time. The added code performs the same test as UW imapd; it
6681 checks after acquiring the lock that its file descriptor still refers to
6682 the same named file.
6684 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6685 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6688 11. Added the smtp_active_hostname option. If used, this will typically be made
6689 to depend on the incoming interface address. Because $interface_address is
6690 not set up until the daemon has forked a reception process, error responses
6691 that can happen earlier (such as "too many connections") no longer contain
6694 12. If an expansion in a condition on a "warn" statement fails because a lookup
6695 defers, the "warn" statement is abandoned, and the next ACL statement is
6696 processed. Previously this caused the whole ACL to be aborted.
6698 13. Added the iplsearch lookup type.
6700 14. Added ident_timeout as a log selector.
6702 15. Added tls_certificate_verified as a log selector.
6704 16. Added a global option tls_require_ciphers (compare the smtp transport
6705 option of the same name). This controls incoming TLS connections.
6707 17. I finally figured out how to make tls_require_ciphers do a similar thing
6708 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6709 before starting the TLS session.
6711 18. Tabs are now shown as \t in -bP output.
6713 19. If the log selector return_path_on_delivery was set, Exim crashed when
6714 bouncing a message because it had too many Received: header lines.
6716 20. If two routers both had headers_remove settings, and the first one included
6717 a superfluous trailing colon, the final name in the first list and the
6718 first name in the second list were incorrectly joined into one item (with a
6719 colon in the middle).
6725 1. Added -C and -D options to the exinext utility, mainly to make it easier
6726 to include in the automated testing, but these could be helpful when
6727 multiple configurations are in use.
6729 2. The exinext utility was not formatting the output nicely when there was
6730 an alternate port involved in the retry record key, nor when there was a
6731 message id as well (for retries that were specific to a specific message
6732 and a specific host). It was also confused by IPv6 addresses, because of
6733 the additional colons they contain. I have fixed the IPv4 problem, and
6734 patched it up to do a reasonable job for IPv6.
6736 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6737 delivery, the log line now contains "pipelined" if PIPELINING was used.
6739 4. An SMTP transport process used to panic and die if the bind() call to set
6740 an explicit outgoing interface failed. This has been changed; it is now
6741 treated in the same way as a connect() failure.
6743 5. A reference to $sender_host_name in the part of a conditional expansion
6744 that was being skipped was still causing a DNS lookup. This no longer
6747 6. The def: expansion condition was not recognizing references to header lines
6748 that used bh_ and bheader_.
6750 7. Added the _cache feature to named lists.
6752 8. The code for checking quota_filecount in the appendfile transport was
6753 allowing one more file than it should have been.
6755 9. For compatibility with Sendmail, the command line option
6763 and sets the incoming protocol and host name (for trusted callers). The
6764 host name and its colon can be omitted when only the protocol is to be set.
6765 Note the Exim already has two private options, -pd and -ps, that refer to
6766 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6767 "s", but I don't think that's a major issue.
6769 10. A number of refactoring changes to the code, none of which should affect
6772 (a) The number of logging options was getting close to filling up the
6773 32-bit word that was used as a bit map. I have split them into two classes:
6774 those that are passed in the argument to log_write(), and those that are
6775 only ever tested independently outside of that function. These are now in
6776 separate 32-bit words, so there is plenty of room for expansion again.
6777 There is no change in the user interface or the logging behaviour.
6779 (b) When building, for example, log lines, the code previously used a
6780 macro that called string_cat() twice, in order to add two strings. This is
6781 not really sufficiently general. Furthermore, there was one instance where
6782 it was actually wrong because one of the argument was used twice, and in
6783 one call a function was used. (As it happened, calling the function twice
6784 did not affect the overall behaviour.) The macro has been replaced by a
6785 function that can join an arbitrary number of extra strings onto a growing
6788 (c) The code for expansion conditions now uses a table and a binary chop
6789 instead of a serial search (which was left over from when there were very
6790 few conditions). Also, it now recognizes conditions like "pam" even when
6791 the relevant support is not compiled in: a suitably worded error message is
6792 given if an attempt is made to use such a condition.
6794 11. Added ${time_interval:xxxxx}.
6796 12. A bug was causing one of the ddress fields not to be passed back correctly
6797 from remote delivery subprocesses. The field in question was not being
6798 subsequently used, so this caused to problems in practice.
6800 13. Added new log selectors queue_time and deliver_time.
6802 14. Might have fixed a bug in maildirsizefile handling that threw up
6803 "unexpected character" debug warnings, and recalculated the data
6804 unnecessarily. In any case, I expanded the warning message to give more
6807 15. Added the message "Restricted characters in address" to the statements in
6808 the default ACL that block characters like @ and % in local parts.
6810 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6811 Three changes have been made:
6813 (a) There was a serious bug; a negative response to MAIL caused the whole
6814 recipient domain to be cached as invalid, thereby blocking all messages
6815 to all local parts at the same domain, from all senders. This bug has
6816 been fixed. The domain is no longer cached after a negative response to
6817 MAIL if the sender used is not empty.
6819 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6822 (c) A new callout option, "use_sender" has been added for people who want
6823 the modified behaviour.
6829 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6832 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6833 indeed breaks things for older releases.
6835 3. Added additional logging to the case where there is a problem reading data
6836 from a filter that is running in a subprocess using a pipe, in order to
6837 try to track down a specific problem.
6839 4. Testing facility fudge: when running in the test harness and attempting
6840 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6841 getting "No route to host". Convert this to a timeout.
6843 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6846 6. Some OS don't have socklen_t but use size_t instead. This affects the
6847 fifth argument of getsockopt() amongst other things. This is now
6848 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6849 can be set for individual OS. I have set it for SunOS5, OSF1, and
6850 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6851 some earlier ones do not.
6853 7. Change 4.30/15 was not doing the test caselessly.
6855 8. The standard form for an IPv6 address literal was being rejected by address
6856 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6857 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6858 this, as well as the form without the "IPv6" on the front (but only when
6859 address literals are enabled, of course).
6861 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6863 10. Exim crashed if a message with an empty sender address specified by -f
6864 encountered a router with an errors_to setting. This could be provoked only
6865 by a command such as
6869 where an empty string was supplied; "<>" did not hit this bug.
6871 11. Installed PCRE release 4.5.
6873 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6874 remained set. It is now erased.
6876 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6877 times from message ids (which are base 36 rather than the normal 62).
6879 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6880 were being counted as actual protocol errors, and logged if the log
6881 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6882 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6883 and DATA following a set of rejected RCPTs do not count as protocol errors.
6884 In other words, Exim assumes they were pipelined, though this may not
6885 actually be the case. Of course, in all cases the client gets an
6886 appropriate error code.
6888 15. If a lookup fails in an ACL condition, a message about the failure may
6889 be available; it is used if testing the ACL cannot continue, because most
6890 such messages specify what the cause of the deferral is. However, some
6891 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6892 that caused an old message to be retained and used if a later statement
6893 caused a defer, replacing the real cause of the deferral.
6895 16. If an IP address had so many PTR records that the DNS lookup buffer
6896 was not large enough to hold them, Exim could crash while trying to process
6897 the truncated data. It now detects and logs this case.
6899 17. Further to 4.21/58, another change has been made: if (and only if) the
6900 first line of a message (the first header line) ends with CRLF, a bare LF
6901 in a subsequent header line has a space inserted after it, so as not to
6902 terminate the header.
6904 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6905 unnecessarily, used atoi() instead of strtol(), and didn't check the
6906 termination when getting file sizes from file names by regex.
6908 19. Completely re-implemented the support for maildirsize files, in the light
6909 of a number of problems with the previous contributed implementation
6910 (4.30/29). In particular:
6912 . If the quota is zero, the maildirsize file is maintained, but no quota is
6915 . If the maildir directory does not exist, it is created before any attempt
6916 to write a maildirsize file.
6918 . The quota value in the file is just a cache; if the quota is changed in
6919 the transport, the new value overrides.
6921 . A regular expression is available for excluding directories from the
6924 20. The autoreply transport checks the characters in options that define the
6925 message's headers; it allows continued headers, but it was checking with
6926 isspace() after an embedded newline instead of explicitly looking for a
6929 21. If all the "regular" hosts to which an address was routed had passed their
6930 expiry times, and had not reached their retry times, the address was
6931 bounced, even if fallback hosts were defined. Now Exim should go on to try
6934 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6935 equivalent code in the SMTP transport. Some hosts send humungous responses
6936 to HELO/EHLO, more than 1024 it seems.
6938 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6939 gives compiler warnings in some environments. I've now done it "properly",
6942 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6943 (because of problems with the built-in one) was declared to return uschar *
6944 instead of char *, causing compiler failure.
6946 25. Fixed a file descriptor leak when processing alias/forward files.
6948 26. Fixed a minor format string issue in dbfn.c.
6950 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6952 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6953 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6956 29. When a sender address is verified, it is cached, to save repeating the test
6957 when there is more than one recipient in a message. However, when the
6958 verification involves a callout, it is possible for different callout
6959 options to be set for different recipients. It is too complicated to keep
6960 track of this in the cache, so now Exim always runs a verification when a
6961 callout is required, relying on the callout cache for the optimization.
6962 The overhead is duplication of the address routing, but this should not be
6965 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6966 to be verified with callout=postmaster, and the main callout worked but the
6967 postmaster check failed, the verification correctly failed. However, if a
6968 subsequent RCPT command asked for sender verification *without* the
6969 postmaster check, incorrect caching caused this verification also to fail,
6972 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6973 it was not caching the DNS options (qualify_single, search_parents) that
6974 were used when the lookup failed. A subsequent lookup with different
6975 options therefore always gave the same answer, though there were cases
6976 where it should not have. (Example: a "domains = !$mx_any" option on a
6977 dnslookup router: the "domains" option is always processed without any
6978 widening, but the router might have qualify_single set.) Now Exim uses the
6979 cached value only when the same options are set.
6981 32. Added John Jetmore's "exipick" utility to the distribution.
6983 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6984 than a timeout (e.g. a certificate is required, and is not provided), an
6985 Exim server now closes the connection immediately. Previously it waited for
6986 the client to close - but if the client is SSL, it seems that they each
6987 wait for each other, leading to a delay before one of them times out.
6989 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6990 maintained 0.8.x compatibility because I don't think many are using it, and
6991 it is clearly obsolete.
6993 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6996 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6997 client certificate was expired. A simple patch fixes this, though I don't
6998 understand the full logic of why the verify callback is called multiple
7001 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7002 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7003 which causes problems with some clients (such as the Certicom SSL Plus
7004 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7005 disables the coutermeasure allowing Eudora to connect."
7007 38. Exim was not checking that a write() to a log file succeeded. This could
7008 lead to Bad Things if a log got too big, in particular if it hit a file
7009 size limit. Exim now panics and dies if it cannot write to a log file, just
7010 as it does if it cannot open a log file.
7012 39. Modified OS/Makefile-Linux so that it now contains
7014 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7016 The two -D definitions ensure that Exim is compiled with large file
7017 support, which makes it possible to handle log files that are bigger than
7020 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7021 instance) a domain was checked against a named list that involved a lookup,
7022 causing $domain_data to be set, then another domain was checked against the
7023 same list, then the first domain was re-checked, the value of $domain_data
7024 after the final check could be wrong. In particular, if the second check
7025 failed, it could be set empty. This bug probably also applied to
7028 41. The strip_trailing_dot option was not being applied to the address given
7029 with the -f command-line option.
7031 42. The code for reading a message's header from the spool was incrementing
7032 $received_count, but never initializing it. This meant that the value was
7033 incorrect (doubled) while delivering a message in the same process in which
7034 it was received. In the most common configuration of Exim, this never
7035 happens - a fresh exec is done - but it can happen when
7036 deliver_drop_privilege is set.
7038 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7039 it now includes up to 150 characters of the unexpected data in the log
7042 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7043 and building data strings. The size of both of these buffers was 10 000
7044 bytes - far larger than anybody would *ever* want, thought I. Needless to
7045 say, somebody hit the limit. I have increased the maximum line length to
7046 20 000 and the maximum data length of concatenated lines to 100 000. I have
7047 also fixed two bugs, because there was no checking on these buffers. Tsk,
7048 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7049 buffer is too small.
7051 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7052 lsearch lookups. Now it does.
7054 46. When parsing a route_list item in a manualroute router, a fixed-length
7055 buffer was used for the list of hosts. I made this 1024 bytes long,
7056 thinking that nobody would ever have a list of hosts that long. Wrong.
7057 Somebody had a whole pile of complicated expansion conditions, and the
7058 string was silently truncated, leading to an expansion error. It turns out
7059 that it is easier to change to an unlimited length (owing to other changes
7060 that have happened since this code was originally written) than to build
7061 structure for giving a limitation error. The length of the item that
7062 expands into the list of hosts is now unlimited.
7064 47. The lsearch lookup could not handle data where the length of text line was
7065 more than 4095 characters. Such lines were truncated, leading to shortened
7066 data being returned. It should now handle lines of any length.
7068 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7069 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7072 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7073 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7074 no longer happens in comment lines. A list of replacements is now placed
7075 at the head of all of the source files, except those whose only change is
7076 to replace PERL_COMMAND in the very first #! line.
7078 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7079 messages on the queue, with a bottom-up merge sort, using code contributed
7080 by Michael Haardt. This should make operations like -bp somewhat faster on
7081 large queues. It won't affect queue runners, except when queue_run_in_order
7084 51. Installed eximstats 1.31 in the distribution.
7086 52. Added support for SRV lookups to the dnslookup router.
7088 53. If an ACL referred to $message_body or $message_body_end, the value was not
7089 reset for any messages that followed in the same SMTP session.
7091 54. The store-handling optimization for building very long strings was not
7092 differentiating between the different store pools. I don't think this
7093 actually made any difference in practice, but I've tidied it.
7095 55. While running the routers to verify a sender address, $sender_address
7096 was still set to the sender address. This is wrong, because when routing to
7097 send a bounce to the sender, it would be empty. Therefore, I have changed
7098 it so that, while verifying a sender address, $sender_address is set to <>.
7099 (There is no change to what happens when verifying a recipient address.)
7101 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7102 target A or AAAA records (if not already returned) without resetting the
7103 qualify_single or search_parents options of the DNS resolver. These are
7104 inappropriate in this case because the targets of MX and SRV records must
7105 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7106 target that, when qualified, matched something in the local domain. These
7107 two options are now turned off when doing these lookups.
7109 57. It seems that at least some releases of Reiserfs (which does not have the
7110 concept of a fixed number of inodes) returns zero and not -1 for the
7111 number of available inodes. This interacted badly with check_spool_inodes,
7112 which assumed that -1 was the "no such thing" setting. What I have done is
7113 to check that the total number of inodes is greater than zero before doing
7114 the test of how many are available.
7116 58. When a "warn" ACL statement has a log_message modifier, the message is
7117 remembered, and not repeated. This is to avoid a lot of repetition when a
7118 message has many recipients that cause the same warning to be written.
7119 However, Exim was preserving the list of already written lines for an
7120 entire SMTP session, which doesn't seem right. The memory is now reset if a
7121 new message is started.
7123 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7124 debugging output unless log_rewrite was also set.
7126 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7127 of (int)(handle) when we know that handle contains (void *)(-1).
7129 61. The Exim daemon panic-logs an error return when it closes the incoming
7130 connection. However "connection reset by peer" seems to be common, and
7131 isn't really an error worthy of noting specially, so that particular error
7134 62. When Exim is trying to find all the local interfaces, it used to panic and
7135 die if the ioctl to get the interface flags failed. However, it seems that
7136 on at least one OS (Solaris 9) it is possible to have an interface that is
7137 included in the list of interfaces, but for which you get a failure error
7138 for this call. This happens when the interface is not "plumbed" into a
7139 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7140 failure of the "get flags" call assumes that the interface is down.
7142 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7143 makes life easier for people who are doing arithmetic on fields extracted
7144 from dates, where you often get leading zeros that should not be
7145 interpreted as octal.
7147 64. Added qualify_domain to the redirect router, to override the global
7150 65. If a pathologically long header line contained very many addresses (the
7151 report of this problem mentioned 10 000) and each of them was rewritten,
7152 Exim could use up a very large amount of memory. (It kept on making new
7153 copies of the header line as it rewrote, and never released the old ones.)
7154 At the expense of a bit more processing, the header rewriting function has
7155 been changed so that it no longer eats memory in this way.
7157 66. The generation of the Received: header has been moved from the time that a
7158 message starts to be received, to the time that it finishes. The timestamp
7159 in the Received: header should now be very close to that of the <= log
7160 line. There are two side-effects of this change:
7162 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7163 logged header lines no longer include the local Received: line, because
7164 it has not yet been created. The same applies to a copy of the message
7165 that is returned to a non-SMTP sender when a message is rejected.
7167 (b) When a filter file is tested using -bf, no additional Received: header
7168 is added to the test message. After some thought, I decided that this
7171 This change does not affect the value of $received_for. It is still set
7172 after address rewriting, but before local_scan() is called.
7174 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7176 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7177 gave an unhelpful panic error message, and a defer error. I have managed to
7178 change this behaviour so that it now rejects any supplied certificate,
7179 which seems right, as the list of acceptable certificates is empty.
7181 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7182 gave an unhelpful defer error. I have not managed to make this reject any
7183 supplied certificates, but the error message it gives is "no certificate
7184 supplied", which is not helpful.
7186 70. exigrep's output now also includes lines that are not associated with any
7187 message, but which match the given pattern. Implemented by a patch from
7188 Martin Sluka, which also tidied up the Perl a bit.
7190 71. Recipient callout verification, like sender verification, was using <> in
7191 the MAIL FROM command. This isn't really the right thing, since the actual
7192 sender may affect whether the remote host accepts the recipient or not. I
7193 have changed it to use the actual sender in the callout; this means that
7194 the cache record is now keyed on a recipient/sender pair, not just the
7195 recipient address. There doesn't seem to be a real danger of callout loops,
7196 since a callout by the remote host to check the sender would use <>.
7197 [SEE ABOVE: changed after hitting problems.]
7199 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7200 temporary errors. However, in the case of such a code being given after
7201 the end of a data transmission (i.e. after ".") Exim was failing to write
7202 a retry record for the message. (Yes, there was some broken host that was
7203 actually sending 8xx at this point.)
7205 73. An unknown lookup type in a host list could cause Exim to panic-die when
7206 the list was checked. (An example that provoked this was putting <; in the
7207 middle of a list instead of at the start.) If this happened during a DATA
7208 ACL check, a -D file could be left lying around. This kind of configuration
7209 error no longer causes Exim to die; instead it causes a defer error. The
7210 incident is still logged to the main and panic logs.
7212 74. Buglet left over from Exim 3 conversion. The message "too many messages
7213 in one connection" was written to the rejectlog but not the mainlog, except
7214 when address rewriting (yes!) was being logged.
7216 75. Added write_rejectlog option.
7218 76. When a system filter was run not as root (that is, when system_filter_user
7219 was set), the values of the $n variables were not being returned to the
7220 main process; thus, they were not subsequently available in the $sn
7223 77. Added +return_path_on_delivery log selector.
7225 78. A connection timeout was being treated differently from recipients deferred
7226 when testing hosts_max_try with a message that was older than the host's
7227 retry timeout. (The host should not be counted, thus allowing all hosts to
7228 be tried at least once before bouncing.) This may have been the cause of an
7229 occasionally reported bug whereby a message would remain on the queue
7230 longer than the retry timeout, but would be bounced if a delivery was
7231 forced. I say "may" because I never totally pinned down the problem;
7232 setting up timeout/retry tests is difficult. See also the next item.
7234 79. The ultimate address timeout was not being applied to errors that involved
7235 a combination of host plus message (for example, a timeout on a MAIL
7236 command). When an address resolved to a number of possible hosts, and they
7237 were not all tried for each delivery (e.g. because of hosts_max_try), a
7238 message could remain on the queue longer than the retry timeout.
7240 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7243 81. Fixed an obscure SMTP outgoing bug which required at least the following
7244 conditions: (a) there was another message waiting for the same server;
7245 (b) the server returned 5xx to all RCPT commands in the first message so
7246 that the message was not completed; (c) the server dropped the connection
7247 or gave a negative response to the RSET that Exim sends to abort the
7248 transaction. The observed case was a dropped connection after DATA that had
7249 been sent in pipelining mode. That is, the server had advertised PIPELINING
7250 but was not implementing it correctly. The effect of the bug was incorrect
7251 behaviour, such as trying another host, and this could lead to a crash.
7257 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7258 and daemon.c were passed as pointers to ints; they should have been
7259 pointers to socklen_t variables (which are typically unsigned ints).
7261 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7264 3. Fixed a really odd bug that affected only the testing scheme; patching a
7265 certain fixed string in the binary changed the value of another string that
7266 happened to be identical to the end of the original first string.
7268 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7269 name", it returns that address as the IP address. On some operating
7270 systems (e.g. Solaris), it also passes back the IP address string as the
7271 "host name". However, on others (e.g. Linux), it passes back an empty
7272 string. Exim wasn't checking for this, and was changing the host name to an
7273 empty string, assuming it had been canonicalized.
7275 5. Although rare, it is permitted to have more than one PTR record for a given
7276 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7277 all the names associated with an address, because they do in Solaris.
7278 However, it seems that they do not in Linux for data that comes from the
7279 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7280 I found this out when I moved to a new Linux workstation and tried to run
7281 the Exim test suite.
7283 To get round this problem I have changed the code so that it now does its
7284 own call to the DNS to look up PTR records when searching for a host name.
7285 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7286 addresses that are only in /etc/hosts are still found.
7288 This behaviour is, however, controlled by an option called host_lookup_
7289 order, which defaults to "bydns:byaddr". If people want to use the other
7290 order, or indeed, just use one or the other means of lookup, they can
7291 specify it in this variable.
7293 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7294 some operating systems, this comes back from gethostbyaddr() as an empty
7295 string, and this is what Exim used to test for. However, it seems that in
7296 other systems, "." is yielded. Exim now tests for this case too.
7298 7. The values of check_spool_space and check_log_space are now held internally
7299 as a number of kilobytes instead of an absolute number of bytes. If a
7300 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7301 kilobyte. This means that much larger values can be stored.
7303 8. Exim monitor: an attempt to get the action menu when not actually pointing
7304 at a message produces an empty menu entitled "No message selected". This
7305 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7306 no entries in it ("Shell widget menu has zero width and/or height"). So I
7307 have added a single, blank menu entry in this case.
7309 9. Added ${quote_local_part.
7311 10. MIME decoding is now applied to the contents of Subject: header lines when
7314 11. Now that a reference to $sender_host_address automatically causes a reverse
7315 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7316 host lookup before query-style lookups in lists that might use this
7317 variable. This has therefore been abolished, and the "net-" prefix is no
7318 longer necessary for query-style lookups.
7320 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7321 have been a typo for LFLAGS, so it has been changed.
7323 13. The install script calls Exim with "-C /dev/null" in order to find the
7324 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7325 to be output. However, since Exim outputs its version number before the
7326 error, it didn't break the script. It just looked ugly. I fixed this by
7327 always allowing "-C /dev/null" if the caller is root.
7329 14. Ignore overlarge ACL variable number when reading spool file - insurance
7330 against a later release with more variables having written the file.
7332 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7333 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7334 this, as well as the form without the "IPv6" on the front.
7336 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7337 OS/Makefile-Darwin file.
7339 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7340 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7342 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7343 variable. However, some OS use ints for inodes. I've added cast to long int
7344 to get rid of the compiler warning.
7346 19. I had forgotten to lock out "/../" in configuration file names when
7347 ALT_CONFIG_PREFIX was set.
7349 20. Routers used for verification do not need to specify transports. However,
7350 if such a router generated a host list, and callout was configured, Exim
7351 crashed, because it could not find a port number from the (non-existent)
7352 transport. It now assumes port 25 in this circumstance.
7354 21. Added the -t option to exigrep.
7356 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7357 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7358 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7359 them is set, LOOKUP_LSEARCH is forced.
7361 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7363 24. Added sender and host information to the "rejected by local_scan()" log
7364 line; previously there was no indication of these.
7366 25. Added .include_if_exists.
7368 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7369 Linux. It turns out that not all file systems support this. Apparently some
7370 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7371 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7372 not supported on the file descriptor, is now ignored when Exim is trying to
7373 sync a directory. This applies only to Linux.
7375 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7377 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7378 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7381 29. Added support for maildirsize files from supplied patch (modified a bit).
7383 30. The use of :fail: followed by an empty string could lead Exim to respond to
7384 sender verification failures with (e.g.):
7386 550 Verification failed for <xxx>
7387 550 Sender verify failed
7389 where the first response line was missing the '-' that indicates it is not
7390 the final line of the response.
7392 31. The loop for finding the name of the user that called Exim had a hardwired
7393 limit of 10; it now uses the value of finduser_retries, which is used for
7394 all other user lookups.
7396 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7399 33. Exim was neglecting to zero errno before one call of strtol() when
7400 expanding a string and expecting an integer value. On some systems this
7401 resulted in spurious "integer overflow" errors. Also, it was casting the
7402 result into an int without checking.
7404 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7405 did not work. The code looks as if it has *never* worked, though it appears
7406 to have been documented since at least release 1.62. I have made it work.
7408 35. The "timeout_DNS" error in retry rules, also documented since at least
7409 1.62, also never worked. As it isn't clear exactly what this means, and
7410 clearly it isn't a major issue, I have abolished the feature by treating it
7411 as "timeout", and writing a warning to the main and panic logs.
7413 36. The display of retry rules for -brt wasn't always showing the error code
7416 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7417 timeout_connect_A, timeout_connect_MX.
7419 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7420 to the empty sender.
7422 39. The daemon was not analysing the content of -oX till after it had closed
7423 stderr and disconnected from the controlling terminal. This meant that any
7424 syntax errors were only noted on the panic log, and the return code from
7425 the command was 0. By re-arranging the code a little, I've made the
7426 decoding happen first, so such errors now appear on stderr, and the return
7427 code is 1. However, the actual setting up of the sockets still happens in
7428 the disconnected process, so errors there are still only recorded on the
7431 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7432 connections (as happens on some IP stacks) was logged at start up time as
7433 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7434 it from "IPv6 and IPv4", which means that two separate sockets are being
7437 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7438 says whether AF_INET or AF_INET6 was passed as an argument.
7440 42. Exiwhat output was messed up when time zones were included in log
7443 43. Exiwhat now gives more information about the daemon's listening ports,
7444 and whether -tls-on-connect was used.
7446 44. The "port" option of the smtp transport is now expanded.
7448 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7449 silently ignored. Now an error message is written to the main and panic
7452 46. There's a new ACL modifier called "logwrite" which writes to a log file
7453 as soon as it is encountered.
7455 47. Added $local_user_uid and $local_user_gid at routing time.
7457 48. Exim crashed when trying to verify a sender address that was being
7460 49. Exim was recognizing only a space character after ".include". It now also
7461 recognizes a tab character.
7463 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7464 extracting the relevant information from the specification. The man page no
7465 longer contains scrambled data for the -d option, and I've added a section
7466 at the front about calling Exim under different names.
7468 51. Added "extra_headers" argument to the "mail" command in filter files.
7470 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7473 53. Installed eximstats 1.29.
7475 54. Added transport_filter_timeout as a generic transport option.
7477 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7478 Cc: header lines. This was required by RFC 822, but it not required by RFC
7481 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7482 incoming messages that did not have them. Now it does so only if the
7483 message originates locally, that is, if there is no associated remote host
7484 address. When Resent- header lines are present, this applies to the Resent-
7485 lines rather than the non-Resent- lines.
7487 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7488 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7490 58. Messages for configuration errors now include the name of the main
7491 configuration file - useful now that there may be more than one file in a
7492 list (.included file names were always shown).
7494 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7495 for those rare installations that do not start the daemon as root or run it
7496 setuid root. I've cut out the call to initgroups() if the daemon is not
7499 60. The Exim user and group can now be bound into the binary as text strings
7500 that are looked up at the start of Exim's processing.
7502 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7504 62. Added $mailstore_basename variable.
7506 63. Installed patch to sieve.c from Michael Haardt.
7508 64. When Exim failed to open the panic log after failing to open the main log,
7509 the original message it was trying to log was written to stderr and debug
7510 output, but if they were not available (the usual case in production), it
7511 was lost. Now it is written to syslog before the two lines that record the
7512 failures to open the logs.
7514 65. Users' Exim filters run in subprocesses under the user's uid. It is
7515 possible for a "deliver" command or an alias in a "personal" command to
7516 provoke an address rewrite. If logging of address rewriting is configured,
7517 this fails because the process is not running as root or exim. There may be
7518 a better way of dealing with this, but for the moment (because 4.30 needs
7519 to be released), I have disabled address rewrite logging when running a
7520 filter in a non-root, non-exim process.
7526 1. The buildconfig auxiliary program wasn't quoting the value set for
7527 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7528 not defined. This bug was masked in 4.22 by the effect that was fixed in
7531 2. Some messages that were rejected after a message id was allocated were
7532 shown as "incomplete" by exigrep. It no longer does this for messages that
7533 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7535 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7536 have allow_domain_literals set, the ID did not get logged in the <= line.
7537 Domain literals are now always recognized in Message-ID: header lines.
7539 4. The first argument for a ${extract expansion item is the key name or field
7540 number. Leading and trailing spaces in this item were not being ignored,
7541 causing some misleading effects.
7543 5. When deliver_drop_privilege was set, single queue runner processes started
7544 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7545 same command in the process it spins off) were not dropping privilege.
7547 6. When the daemon running as "exim" started a queue runner, it always
7548 re-executed Exim in the spun-off process. This is a waste of effort when
7549 deliver_drop_privilege is set. The new process now just calls the
7550 queue-runner function directly.
7556 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7559 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7560 ignored. Though the use of -oP was forcing the writing of a pid file, it
7561 was always written to the default place.
7563 3. If the message "no IP address found for host xxxx" is generated during
7564 incoming verification, it is now followed by identification of the incoming
7565 connection (so you can more easily find what provoked it).
7567 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7569 5. Added some features to "harden" Exim a bit more against certain attacks:
7571 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7572 be put in Local/Makefile. This is like the never_users runtime option,
7573 but it cannot be overridden. The default setting is "root".
7575 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7576 prefix string with which any file named in a -C command line option
7579 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7580 is retained for -C and -D only if the caller of Exim is root. Without
7581 it, the exim user may also use -C and -D and retain privilege.
7583 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7584 command line option is disabled.
7586 6. Macro names set by the -D option must start with an upper case letter, just
7587 like macro names defined in the configuration file.
7589 7. Added "dereference=" facility to LDAP.
7591 8. Two instances of the typo "uknown" in the source files are fixed.
7593 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7594 the Configure-Makefile script screwed up while processing it.
7596 10. Incorporated PCRE 4.4.
7598 11. The SMTP synchronization check was not operating right at the start of an
7599 SMTP session. For example, it could not catch a HELO sent before the client
7600 waited for the greeting. There is now a check for outstanding input at the
7601 point when the greeting is written. Because of the duplex, asynchronous
7602 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7603 way, but not yet received, when the check is performed.
7605 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7606 on TCP/IP sockets, because this apparently causes some broken clients to
7609 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7610 unchanged) from the Cygwin maintainer.
7612 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7613 was supported instead of testing for mbx. Effectively a typo.
7615 15. The spa authenticator server code was not checking that the input it
7616 received was valid base64.
7618 16. The debug output line for the "set" modifier in ACLs was not showing the
7619 name of the variable that was being set.
7621 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7623 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7624 lookup on its own. Something else was needed to trigger the lookup. For
7625 example, a match in host_lookup or the need for a host name in a host list.
7626 Now, if $sender_host_name is referenced and the host name has not yet been
7627 looked up, a lookup is performed. If the lookup fails, the variable remains
7628 empty, and $host_lookup_failed is set to "1".
7630 19. Added "eqi" as a case-independent comparison operator.
7632 20. The saslauthd authentication condition could segfault if neither service
7633 nor realm was specified.
7635 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7636 error message that was logged was misleading, and incoming SMTP
7637 connections were dropped. The message is now more accurate, and temporary
7638 errors are given to SMTP connections.
7640 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7641 (except RSET) with a 421 error, until QUIT is received. However, it was
7642 failing to send a response to QUIT.
7644 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7645 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7646 ACL is now run after the helo_try_verify_hosts code.
7648 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7649 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7650 case-independent, so other case variants are also recognized). Apparently
7651 some systems use these upper case variants.
7653 25. If more than two messages were waiting for the same host, and a transport
7654 filter was specified for the transport, Exim sent two messages over the
7655 same TCP/IP connection, and then failed with "socket operation on non-
7656 socket" when it tried to send the third.
7658 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7660 27. The extern definition of crypt16() in expand.c was not being excluded when
7661 the OS had its own crypt16() function.
7663 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7664 as a preferred synonym for return_size_limit, both as an option and as an
7667 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7669 30. Changed the default configuration ACL to relax the local part checking rule
7670 for addresses that are not in any local domains. For these addresses,
7671 slashes and pipe symbols are allowed within local parts, but the sequence
7672 /../ is explicitly forbidden.
7674 31. SPA server authentication was not clearing the challenge buffer before
7677 32. log_message in a "warn" ACL statement was writing to the reject log as
7678 well as to the main log, which contradicts the documentation and doesn't
7679 seem right (because no rejection is happening). So I have stopped it.
7681 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7682 However, I am unable to do any testing of this.
7684 34. Fixed an infelicity in the appendfile transport. When checking directories
7685 for a mailbox, to see if any needed to be created, it was accidentally
7686 using path names with one or more superfluous leading slashes; tracing
7687 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7689 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7690 contents are added to the log line that is written for every discarded
7691 recipient. (Previously a log_message setting was ignored.)
7693 36. The ${quote: operator now quotes the string if it is empty.
7695 37. The install script runs exim in order to find its version number. If for
7696 some reason other than non-existence or emptiness, which it checks, it
7697 could not run './exim', it was installing it with an empty version number,
7698 i.e. as "exim-". This error state is now caught, and the installation is
7701 38. An argument was missing from the function that creates an error message
7702 when Exim fails to connect to the socket for saslauthd authentication.
7703 This could cause Exim to crash, or give a corrupted message.
7705 39. Added isip, isip4, and isip6 to ${if conditions.
7707 40. The ACL variables $acl_xx are now saved with the message, and can be
7708 accessed later in routers, transports, and filters.
7710 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7711 strings in the file are not string-expanded.
7713 42. If a MAIL command specified a SIZE value that was too large to fit into an
7714 int variable, the check against message_size_limit failed. Such values are
7715 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7716 day this will have to be increased, but I don't think I want to be around
7717 when emails are that large.
7724 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7725 iconv() is not standard in FreeBSD.
7727 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7728 IPv6 enabled. The observed symptom was a segmentation fault on return from
7729 the function os_common_find_running_interfaces() in src/os.c.
7731 3. In the check_special_case() function in daemon.c I had used "errno" as an
7732 argument name, which causes warnings on some systems. This was basically a
7733 typo, since it was named "eno" in the comments!
7735 4. The code that waits for the clock to tick (at a resolution of some fraction
7736 of a second) so as to ensure message-id uniqueness was always waiting for
7737 at least one whole tick, when it could have waited for less. [This is
7738 almost certainly not relevant at current processor speeds, where it is
7739 unlikely to ever wait at all. But we try to future-proof.]
7741 5. The function that sleeps for a time interval that includes fractions of a
7742 second contained a race. It did not block SIGALRM between setting the
7743 timer, and suspending (a couple of lines later). If the interval was short
7744 and the sigsuspend() was delayed until after it had expired, the suspension
7745 never ended. On busy systems this could lead to processes getting stuck for
7748 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7749 process, before it forks any delivery processes. The open lookup caching
7750 mechanism meant that the open file or database connection was passed into
7751 the delivery process. The problem was that delivery processes always tidy
7752 up cached lookup data. This could cause a problem for the next delivery
7753 process started by the queue runner, because the external queue runner
7754 process does not know about the closure. So the next delivery process
7755 still has data in the lookup cache. In the case of a file lookup, there was
7756 no problem because closing a file descriptor in a subprocess doesn't affect
7757 the parent. However, if the lookup was caching a connection to a database,
7758 the connection was closed, and the second delivery process was likely to
7759 see errors such as "PGSQL: query failed: server closed the connection
7760 unexpectedly". The problem has been fixed by closing all cached lookups
7761 in a queue runner before running a delivery process.
7763 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7764 seem to have the "const" qualifier which it has on other OS. I've
7767 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7768 *of the same type* (client or server) with the same public name that an
7769 error should be diagnosed.
7771 9. When Exim looked up a host name for an IP address, but failed to find the
7772 original IP address when looking up the host name (a safety check), it
7773 output the message "<ip address> does not match any IP for NULL", which was
7774 confusing, to say the least. The bug was that the host name should have
7775 appeared instead of "NULL".
7777 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7778 user that is built into the binary, and the -C or -D options is used, root
7779 privilege is dropped before the configuration file is read. In addition,
7780 logging is switched to stderr instead of the normal log files. If the
7781 configuration then re-defines the Exim user, the unprivileged environment
7782 is probably not what is expected, so Exim logs a panic warning message (but
7785 However, if deliver_drop_privilege is set, the unprivileged state may well
7786 be exactly what is intended, so the warning has been cut out in that case,
7787 and Exim is allowed to try to write to its normal log files.
7793 1. smtp_return_error_details was not giving details for temporary sender
7794 or receiver verification errors.
7796 2. Diagnose a configuration error if two authenticators have the same public
7799 3. Exim used not to create the message log file for a message until the first
7800 delivery attempt. This could be confusing when incoming messages were held
7801 for policy or load reasons. The message log file is now created at the time
7802 the message is received, and an initial "Received" line is written to it.
7804 4. The automatically generated man page for command line options had a minor
7805 bug that caused no ill effects; however, a more serious problem was that
7806 the procedure for building the man page automatically didn't always
7807 operate. Consequently, release 4.20 contains an out-of-date version. This
7808 shouldn't happen again.
7810 5. When building Exim with embedded Perl support, the script that builds the
7811 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7812 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7814 6. The freeze_tell option was not being used for messages that were frozen on
7815 arrival, either by an ACL or by local_scan().
7817 7. Added the smtp_incomplete_transaction log selector.
7819 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7820 was accepting AUTH without a new EHLO.
7822 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7823 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7826 10. Logging of TCP/IP connections (when configured) now happens in the main
7827 daemon process instead of the child process, so that the TCP/IP connection
7828 count is more accurate (but it can never be perfect).
7830 11. The use of "drop" in a nested ACL was not being handled correctly in the
7831 outer ACL. Now, if condition failure induced by the nested "drop" causes
7832 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7833 or "require"), the connection is dropped.
7835 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7836 that yield "discard" can now be used with an "accept" or a "discard" verb,
7837 but an error is generated for any others (because I can't see a useful way
7838 to define what should happen).
7840 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7841 are now processed in the same way as lines in the Exim configuration file.
7842 In particular, continuation lines are supported.
7844 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7846 15. Added -ti meaning -t -i.
7848 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7849 domains, and warn by logging if others are found.
7851 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7852 returned from GIFCONF when Exim is trying to find the list of interfaces on
7853 a host. The code in os.c has been modified to copy each ifreq to an aligned
7854 structure in all cases.
7856 Also, in some cases, the returned ifreq's were being copied to a 'struct
7857 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7858 means the last couple of bytes of an IPv6 address could be chopped if the
7859 ifreq contained only a normal sockaddr (14 bytes storage).
7861 18. Named domain lists were not supported in the hosts_treat_as_local option.
7862 An entry such as +xxxx was not recognized, and was treated as a literal
7865 19. Ensure that header lines added by a DATA ACL are included in the reject log
7866 if the ACL subsequently rejects the message.
7868 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7869 MD5 (which is deprecated).
7871 21. When testing a filter file using -bf, Exim was writing a message when it
7872 took the sender from a "From " line in the message, but it was not doing so
7873 when it took $return_path from a Return-Path: header line. It now does.
7875 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7876 with a valid header line field name (a series of printing characters
7877 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7879 23. Changed "disc" in the source to "disk" to conform to the documentation and
7880 the book and for uniformity.
7882 24. Ignore Sendmail's -Ooption=value command line item.
7884 25. When execve() failed while trying to run a command in a pipe transport,
7885 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7886 could be confused with a return value of 69 from the command itself. This
7887 has been changed to 127, the value the shell returns if it is asked to run
7888 a non-existent command. The wording for the related log line suggests a
7889 non-existent command as the problem.
7891 26. If received_header_text expands to an empty string, do not add a Received:
7892 header line to the message. (Well, it adds a token one on the spool, but
7893 marks it "old" so that it doesn't get used or transmitted.)
7895 27. Installed eximstats 1.28 (addition of -nt option).
7897 28. There was no check for failure on the call to getsockname() in the daemon
7898 code. This can fail if there is a shortage of resources on the system, with
7899 ENOMEM, for example. A temporary error is now given on failure.
7901 29. Contrary to the C standard, it seems that in some environments, the
7902 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7903 program. Exim now does this explicitly; it affects the formatting of
7904 timestamps using strftime().
7906 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7907 complaints. I've now initialized all the variables, to avoid this.
7909 32. Header lines added by a system filter were not being "seen" during
7910 transport-time rewrites.
7912 33. The info_callback() function passed to OpenSSL is set up with type void
7913 (*)(SSL *, int, int), as described somewhere. However, when calling the
7914 function (actually a macro) that sets it up, the type void(*)() is
7915 expected. I've put in a cast to prevent warnings from picky compilers.
7917 34. If a DNS black list lookup found a CNAME record, but there were no A
7918 records associated with the domain it pointed at, Exim crashed.
7920 35. If a DNS black list lookup returned more than one A record, Exim ignored
7921 all but the first. It now scans all returned addresses if a particular IP
7922 value is being sought. In this situation, the contents of the
7923 $dnslist_value variable are a list of all the addresses, separated by a
7926 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7927 to accept a host name and all its aliases if the forward lookup for any of
7928 them yielded the IP address of the incoming connection. Now it accepts only
7929 those names whose forward lookup yields the correct IP address. Any other
7930 names are discarded. This closes a loophole whereby a rogue DNS
7931 administrator could create reverse DNS records to break through a
7932 wildcarded host restriction in an ACL.
7934 37. If a user filter or a system filter that ran in a subprocess used any of
7935 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7936 the wrong values were passed to the pipe command ($thisaddress had the
7937 value of $0, $0 had the value of $1, etc). This bug was introduced by
7938 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7940 38. Improved the line breaking for long SMTP error messages from ACLs.
7941 Previously, if there was no break point between 40 and 75 characters, Exim
7942 left the rest of the message alone. Two changes have been made: (a) I've
7943 reduced the minimum length to 35 characters; (b) if it can't find a break
7944 point between 35 and 75 characters, it looks ahead and uses the first one
7945 that it finds. This may give the occasional overlong line, but at least the
7946 remaining text gets split now.
7948 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7949 file descriptors might be low, and that setting 1000 would always raise it.
7950 It turns out that in some environments, the limit is already over 1000 and
7951 that lowering it causes trouble. So now Exim takes care not to decrease it.
7953 40. When delivering a message, the value of $return_path is set to $sender_
7954 address at the start of routing (routers may change the value). By an
7955 oversight, this default was not being set up when an address was tested by
7956 -bt or -bv, which affected the outcome if any router or filter referred to
7959 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7960 log only when "message" or "log_message" are set. However, if one of the
7961 conditions was an address verification, or a call to a nested ACL, the
7962 messages generated by the underlying test were being passed through. This
7963 no longer happens. The underlying message is available in $acl_verify_
7964 message for both "message" and "log_message" expansions, so it can be
7965 passed through if needed.
7967 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7968 new expansion $bh_ to give the encoded byte string without charset
7969 translation. Translation happens only if iconv() is available; HAVE_ICONV
7970 indicates this at build time. HEADERS_CHARSET gives the charset to
7971 translate to; headers_charset can change it in the configuration, and
7972 "headers charset" can change it in an individual filter file.
7974 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7975 that creates RFC 2047 encoded "words" labels them as that charset instead
7976 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7977 expansion operator; (ii) when Exim creates a From: line for a local
7978 message; (iii) when a header line is rewritten to include a "phrase" part.
7980 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7981 buggy, causing it to skip the first lines of messages whose message ID
7982 ended in 'D'. This would not have bitten before Exim release 4.14, because
7983 message IDs were unlikely to end in 'D' before then. The effect was to have
7984 incorrect size information for certain domains.
7986 45. #include "config.h" was missing at the start of the crypt16.c module. This
7987 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7990 46. If there was a timeout during a "random" callout check, Exim treated it as
7991 a failure of the random address, and carried on sending RSET and the real
7992 address. If the delay was just some slowness somewhere, the response to the
7993 original RCPT would be taken as a response to RSET and so on, causing
7994 mayhem of various kinds.
7996 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7997 when I implemented it. It didn't allow for the fact that some option values
7998 may legitimately be negative (e.g. size_addition), and it didn't even do
7999 the right test for positive values.
8001 48. Domain names in DNS records are case-independent. Exim always looks them up
8002 in lower case. Some resolvers return domain names in exactly the case they
8003 appear in the zone file, that is, they may contain uppercase letters. Not
8004 all resolvers do this - some return always lower case. Exim was treating a
8005 change of case by a resolver as a change of domain, similar to a widening
8006 of a domain abbreviation. This triggered its re-routing code and so it was
8007 trying to route what was effectively the same domain again. This normally
8008 caused routing to fail (because the router wouldn't handle the domain
8009 twice). Now Exim checks for this case specially, and just changes the
8010 casing of the domain that it ultimately uses when it transmits the message
8013 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8016 50. If a filter generated a file delivery with a non-absolute name (possible if
8017 no home directory exists for the router), the forbid_file option was not
8020 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8021 the existing equality matching.
8023 52. Exim was using ints instead of ino_t variables in some places where it was
8024 dealing with inode numbers.
8026 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8027 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8028 called TMPDIR, and if it finds it is different, it changes its value.
8030 54. The smtp_printf() function is now made available to local_scan() so
8031 additional output lines can be written before returning. There is also an
8032 smtp_fflush() function to enable the detection of a dropped connection.
8033 The variables smtp_input and smtp_batched_input are exported to
8036 55. Changed the default runtime configuration: the message "Unknown user"
8037 has been removed from the ACL, and instead placed on the localuser router,
8038 using the cannot_route_message feature. This means that any verification
8039 failures that generate their own messages won't get overridden. Similarly,
8040 the "Unrouteable address" message that was in the ACL for unverifiable
8041 relay addresses has also been removed.
8043 56. Added hosts_avoid_esmtp to the smtp transport.
8045 57. The exicyclog script was not checking for the esoteric option
8046 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8047 will work only if exicyclog is run under the appropriate euid.
8049 58. Following a discussion on the list, the rules by which Exim recognises line
8050 endings on incoming messages have been changed. The -dropcr and drop_cr
8051 options are now no-ops, retained only for backwards compatibility. The
8052 following line terminators are recognized: LF CRLF CR. However, special
8053 processing applies to CR:
8055 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8056 nor a local message in the state where . is a terminator.
8058 (ii) If a bare CR is encountered in a header line, an extra space is added
8059 after the line terminator so as not to end the header. The reasoning
8060 behind this is that bare CRs in header lines are most likely either
8061 to be mistakes, or people trying to play silly games.
8063 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8064 was being incorrectly given as 18 bytes larger than it should have been.
8065 This is a VOB (very old bug).
8067 60. This may never have affected anything current, but just in case it has:
8068 When the local host is found other than at the start of a list of hosts,
8069 the local host, those with the same MX, and any that follow, are discarded.
8070 When the list in question was part of a longer list of hosts, the following
8071 hosts (not currently being processed) were also being discarded. This no
8072 longer happens. I'm not sure if this situation could ever has previously
8075 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8078 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8079 header. This is something that is recommended in a new Internet Draft, and
8080 is something that is documented as being done by Sendmail. There are two
8081 possible values. For messages generated by the autoreply transport, Exim
8084 Auto-Submitted: auto-replied
8086 whereas for all other generated messages (e.g. bounces) it adds
8088 Auto-Submitted: auto-generated
8090 63. The "personal" condition in filters now includes a test for the
8091 Auto-Submitted: header. If it contains the string "auto-" the message it
8092 not considered personal.
8094 64. Added rcpt_include_affixes as a generic transport option.
8096 65. Added queue_only_override (default true).
8098 66. Added the syslog_duplication option.
8100 67. If what should have been the first header line of a message consisted of
8101 a space followed by a colon, Exim was mis-interpreting it as a header line.
8102 It isn't of course - it is syntactically invalid and should therefore be
8103 treated as the start of the message body. The misbehaviour could have
8104 caused a number of strange effects, including loss of data in subsequent
8105 header lines, and spool format errors.
8107 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8108 client host had authenticated. This control can now be exercised by an ACL
8109 for more flexibility.
8111 69. By default, callouts do not happen when testing with -bh. There is now a
8112 variant, -bhc, which does actually run the callout code, including
8113 consulting and updating the callout cache.
8115 70. Added support for saslauthd authentication, courtesy of Alexander
8118 71. If statvfs() failed on the spool or log directories while checking their
8119 size for availability, Exim confusingly gave the error "space shortage".
8120 Furthermore, in debugging mode it crashed with a floating point exception.
8121 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8122 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8123 serious problem, Exim now writes to the main and panic logs when this
8124 happens, with details of the failure. It then refuses to accept the
8125 incoming message, giving the message "spool directory problem" or "log
8126 directory problem" with a 421 code for SMTP messages.
8128 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8129 0, 1, and 2 exist, because some OS complain for execs without them (see
8130 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8131 descriptors. However, the code omitted to check that the open succeeded,
8132 causing mysterious errors if for some reason the permissions on /dev/null
8133 got screwed. Now Exim writes a message to the main and panic logs, and
8134 bombs out if it can't open /dev/null.
8136 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8137 interact so that it is all more flexible. It is supposed to remain
8138 backwards compatible. Also added extra_local_interfaces.
8140 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8141 to bomb out with an assertion failure - to the client this appears as a
8142 connection drop. This problem occurs in the part of the code that was taken
8143 from the Samba project. Fortunately, the assertion is in a very simple
8144 function, so I have fixed this by reproducing the function inline in the
8145 one place where it is called, and arranging for authentication to fail
8146 instead of killing the process with assert().
8148 75. The SPA client code was not working when the server requested OEM rather
8149 than Unicode encoding.
8151 76. Added code to make require_files with a specific uid setting more usable in
8152 the case where statting the file as root fails - usually a non-root-mounted
8153 NFS file system. When this happens and the failure is EACCES, Exim now
8154 forks a subprocess and does the per-uid checking as the relevant uid.
8156 77. Added process_log_path.
8158 78. If log_file_path was not explicitly set, a setting of check_log_space or
8159 check_log_inodes was ignored.
8161 79. If a space check for the spool or log partitions fails, the incident is now
8162 logged. Of course, in the latter case the data may get lost...
8164 80. Added the %p formatting code to string_format() so that it can be used to
8165 print addresses in debug_print(). Adjusted all the address printing in the
8166 debugging in store.c to use %p rather than %d.
8168 81. There was a concern that a line of code in smtp_in.c could overflow a
8169 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8170 initially expressed, the concern was not well-founded, because trailing
8171 spaces are removed early. However, if the trailing spaces were followed by
8172 a NULL, they did not get removed, so the overflow was possible. Two fixes
8175 (a) I re-wrote the offending code in a cleaner fashion.
8176 (b) If an incoming SMTP command contains a NULL character, it is rejected
8179 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8180 runs initgroups(), so that if the Exim user is in any additional groups,
8181 they will be used during message reception.
8187 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob