1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes, when delivery is immediate. Previously debugging
74 stopped any time Exim re-execs.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
84 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
86 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
87 supplied as the db file-pair basename. Previously, if a directory
88 path was given, for example via the autoreply "once" option, the DB
89 file.pag and file.dir files would be created in that directory's
92 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
93 "taint" log_selector. These were previously deprecated.
99 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
100 related applications. Previously an "H" was used where available info
101 says that "M" should be, so change to match.
103 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
104 as arguments, so an implementation trying to copy these into a local
105 buffer was taking a taint-enforcement trap. Fix by using dynamically
106 created buffers. Similar fix for radius expansion condition.
108 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
109 reasonable, eg. to count headers. Fix by using dynamically created
110 buffers rather than a local. Do similar fixes for ACL actions "dcc",
111 "log_reject_target", "malware" and "spam"; the arguments are expanded
112 so could be handling tainted values.
114 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
115 broken the (no-op) support for this sendmail command. Restore it
116 to doing nothing, silently, and returning good status.
118 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
119 record path was given (or the default used) without a leading directory
120 path, an error occurred on trying to open it. Use the transport's working
123 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
124 transport. Previously it was the name on the DNS A-record; use instead
125 the head of the CNAME chain leading there (if there is one). This seems
126 to align better with RFC 6125.
128 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
129 smtp_accept_max_per_host allocated resources which were not released
130 when the limit was exceeded. This eventually crashed the daemon. Fix
131 by adding a release action in that path.
133 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
134 expanded; previously using tainted values was rejected. Fix by using
135 dynamically-created buffers.
137 JH/09 Relax restrictions on ACL verify condition needing access to message
138 headers. Previously they were only permitted in data and non-smtp ACLs;
139 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
140 not_blind, header_sender and header_names_ascii verification.
142 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
143 Previously a macro used one argument twice; when called with the
144 argument as an expression having side-effects, incorrect operation
145 resulted. Use an inlineable function.
147 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
148 held open for a verify callout. Previously this wan not accounted for
149 and a corrupt onward SMTP conversation resulted.
151 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
152 Previously they were embedded naively in the constructed address; when
153 needed, strip the quoting and quote the entire local_part.
154 Also make the inbound_srs expansion condition handle quoting.
156 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
157 excluded, not matching the documentation.
159 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
160 was given for the sqlite_dbfile a trap resulted.
162 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
163 "name" argument resulted in a trap. There is no reason to disallow such;
164 this was a coding error.
166 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
167 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
168 spent suspended, ignoring the POSIX definition. Previously we assumed
169 it did and a constant offset from real time could be used as a correction.
170 Change to using the same clock source for the start-of-message and the
171 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
172 exists, just to get a clock slightly more aligned to reality.
174 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
175 RFC says it is optional some validators care. The missing char was not
176 intended but triggered by a line-wrap alignment. Discovery and fix by
177 Guillaume Outters, hacked on by JH.
179 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
180 name being quoted was tainted a trap would be taken. Fix by using
181 dynamically created buffers. The routine could have been called by a
182 rewrite with the "h" flag, by using the "-F" command-line option, or
183 by using a "name=" option on a control=submission ACL modifier.
185 JH/19 SPF: change the Authentication-Results expansion component to give
186 smtp.helo when the sender domain is empty. Previously it gave
189 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
190 not in 127.0.0.0/8 to help in spotting list domains taken over by a
191 domain-parking registrar.
193 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
194 Previously when a whitespace character was specified it was not inserted
195 after removing the newline.
197 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
198 the domain part of the recipient address. This overrides any tls_sni
199 option set, which was previously used.
201 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
204 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
205 is_tainted() had an off-by-one error in the overenthusiastic direction.
206 Find and fix by Gavan. Although NetBSD is not a supported platform for
207 4.94 this bug could affect other platforms.
209 PP/01 Fix default prime selection to be consistent.
210 One path used ike23 still, instead of exim.dev.20160529.3; now both
211 execution flows will use the same DH primes (currently
212 exim.dev.20160529.3).
214 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
215 option in smtp transport, to match the documentation. Previously
216 verification was not being done in some cases where it should have been.
218 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
219 than one server was defined and depending on the platform memory layout
220 details, an internal consistency trap could be hit while walking the list
223 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
224 files. The value is used by the authresults expansion item. Previously
225 if this was used in a router or transport, a crash could result.
227 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
228 have a sufficient incidence of EINTR returns from select that an
229 interaction with other operations done by the main daemon loop exposed
230 a bug in the error-handling. This was benign apart from the log
233 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
234 for consistency with delivered "=>" and failed "**" lines. While we're
235 there, handle PRX and TFO.
237 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
238 applied. This resulted, if any header-line rewrite rules were configured,
239 in a panic-log triggerable by sending a message with a long address in
240 a header. Fix by increasing the arbitrary limit to larger than a single
241 (dewrapped) 5322 header line maximum size.
243 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
244 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
245 RFC 6648 which deprecates X- options in protocols as a general practice.
246 Changeover between the implementations is handled by the mechanisms
249 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
250 a remote which uses callout/hold. Previously the local was queued.
252 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
255 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
256 log line. The asterisk is now consistently appended to the remote IP
257 (and port, if given), and will also be provided on defer and fail log
258 lines. Previously it could be placed on the local IP if that was being
259 logged, and was only provided on delivery lines.
261 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
263 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
264 authenticator client_send option. Previously the next char, after a pair
265 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
266 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
267 documentation. There is still no way to get a leading ^ immediately
268 after a NUL (ie. for the password of a PLAIN method authenticator.
270 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
271 files. For bad sizes read, delete the record and whine to paniclog.
273 JH/38 When logging an AUTH failure, as server, do not include sensitive
274 information. Previously, the credentials would be included if given
275 as part of the AUTH command line and an ACL denied authentication.
277 JH/39 Bug 2691: fix $local_part_data. When the matching list element
278 referred to a file, bad data was returned. This likely also affected
281 JH/40 The gsasl authenticator now supports caching of the salted password
282 generated by the client-side implementation. This required the addition
283 of a new variable: $auth4.
285 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
286 left undeleted; the attempt to re-create it then failed - resulting in
287 the usual "SIGHUP tp have daemon reload configuration" to not work.
288 This affected any platform not supporting "abstract" Unix-domain
289 sockets (i.e. not Linux).
291 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
292 recipients" response to RCPT in a later response, with a 250. The
293 previous coding assumed this would not happen, and under PIPELINING
294 would result in both lost and duplicate recipients for a message.
296 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
297 Previously the weighting was incorrectly applied. Similar fix for socks
298 proxies. Found and fixed by Heiko Schlichting.
300 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
301 not handle sub-lists included using the +namedlist syntax. While
302 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
303 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
306 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
307 Previously the permanent pool was used, so the sore could not be freed.
308 This meant a connection with many messages would use continually-growing
311 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
312 per-pool so as not to waste too much space. Previously a constant size
313 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
314 DOS attacks harder. The cost is wasted memory use in the larger blocks.
316 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
317 same space repeatedly, and a smaller process footprint.
319 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
320 for GnuTLS. Previously the name started "%s" which, while not a bug,
321 looked as if if might be one.
323 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
324 "continued connection") make the $proxy_* variables available. Previously
325 the information was not passed across the exec() call for subsequent
326 transport executions. This also mean that the log lines for the
327 messages can show the proxy information.
329 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
330 receive time. With modern systems the difference is significant.
331 The historical behaviour can be restored by disabling (a new) log_selector
332 "queue_time_exclusive".
334 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
335 content) were specifically tested for. Now, also cover expansions
336 resulting in ACL names and inline ACL content.
338 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
339 making it unusable in complex expressions.
341 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
342 fail with a taint-check on the spoolfile name, and leave the message
345 HS/01 Enforce absolute PID file path name.
347 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
349 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
351 PP/02 Bug 2643: Correct TLS DH constants.
352 A missing NUL termination in our code-generation tool had led to some
353 incorrect Diffie-Hellman constants in the Exim source.
354 Reported by kylon94, code-gen tool fix by Simon Arlott.
356 PP/03 Impose security length checks on various command-line options.
357 Fixes CVE-2020-SPRSS reported by Qualys.
359 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
360 better. Reported by Qualys.
362 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
363 providing a particularly obnoxious sender full name.
366 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
368 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
371 PP/08 Change default for recipients_max from unlimited to 50,000.
373 PP/09 Fix security issue with too many recipients on a message (to remove a
374 known security problem if someone does set recipients_max to unlimited,
375 or if local additions add to the recipient list).
376 Fixes CVE-2020-RCPTL reported by Qualys.
378 PP/10 Fix security issue in SMTP verb option parsing
379 Fixes CVE-2020-EXOPT reported by Qualys.
381 PP/11 Fix security issue in BDAT state confusion.
382 Ensure we reset known-good where we know we need to not be reading BDAT
383 data, as a general case fix, and move the places where we switch to BDAT
384 mode until after various protocol state checks.
385 Fixes CVE-2020-BDATA reported by Qualys.
387 HS/03 Die on "/../" in msglog file names
389 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
390 the Exim runtime user are allowed to create files.
392 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
395 QS/03 When reading the output from interpreted forward files we do not
396 pass the pipe between the parent and the interpreting process to
397 executed child processes (if any).
399 QS/04 Always die if requested from internal logging, even is logging is
402 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
403 the API; compilation noo longer completes with DMARC support included.
404 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
405 been reported on other platforms.
407 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
409 JH/56 Make the majority of info read from config files readonly, for defence-in-
410 depth against exploits. Suggestion by Qualys.
411 Not supported on Solaris 10.
413 JH/57 Fix control=fakreject for a custom message containing tainted data.
414 Previously this resulted in a log complaint, due to a re-expansion present
415 since fakereject was originally introduced.
417 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
418 modifier was given, a loop resulted.
420 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
421 pipelined SMTP command followed the BDAT LAST then it would be
422 incorrectly treated as part of the message body, causing a verification
425 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
426 require looser validation rules than those for 821-level addresses,
427 which only permit IP addresses.
433 JH/01 Avoid costly startup code when not strictly needed. This reduces time
434 for some exim process initialisations. It does mean that the logging
435 of TLS configuration problems is only done for the daemon startup.
437 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
439 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
440 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
443 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
444 DKIM signing is being done. Previously a transport_filter would always
445 disable CHUNKING, falling back to traditional DATA.
447 JH/05 Regard command-line recipients as tainted.
449 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
451 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
452 PAM library frees one of the arguments given to it, despite the
453 documentation. Therefore a plain malloc must be used.
455 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
456 on-stack buffers were used, resulting in a taint trap when DSN information
457 copied from a received message was written into the buffer.
459 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
460 the ordering of its ARC headers. This caused a crash.
462 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
463 a new record was being constructed with information from the peer, a trap
466 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
467 installation would get error messages from DMARC verify, when it hit the
468 nonexistent file indicated by the default. Distros wanting DMARC enabled
469 should both provide the file and set the option.
470 Also enforce no DMARC verification for command-line sourced messages.
472 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
473 could, depending on the platform, hang at the STARTTLS response.
475 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
476 message on a connection. Previously if one message had ARC headers and
477 the following one did not, a crash could result when adding an
478 Authentication-Results: header.
480 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
481 Exim main code and Exim-related utities. The introduction of taint
482 tracking also did many adjustments to string handling. Since then, eximon
483 frequently terminated with an assert failure.
485 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
486 check for 452 responses. This slightly helps the inefficieny of doing
487 a large alias-expansion into a recipient-limited target. The max_rcpt
488 transport option still applies (and at the current default, will override
489 the new feature). The check is done for either cause of synch, and forces
490 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
491 connection. The new facility is not tunable at this time.
493 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
494 library live data was being used, so the results became garbage. Make
495 copies while it is still usable.
497 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
498 on delivery deferred (==) and failed (**) lines (if a delivery was
499 attemtped). Previously it was only on completion (=>) lines.
501 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
502 for the expansion of the server_scram_iter and server_scram_salt options.
504 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
505 are now specifically given a NO_DATA response without hitting the system
506 resolver. The library goes on to do the now-standard TXT lookup.
507 Use of dnsdb lookups is not affected.
509 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
510 only retrieve the errormessage once. Previously two calls to dlerror()
511 were used, and the second one (for mainlog/paniclog) retrieved null
514 JH/20 Taint checking: disallow use of tainted data for
515 - the appendfile transport file and directory options
516 - the pipe transport command
517 - the autoreply transport file, log and once options
518 - file names used by the redirect router (including filter files)
520 - paths used by single-key lookups
521 Previously this was permitted.
523 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
524 adjusted the size of a major service buffer; this failed because the
525 buffer was in use at the time. Change to a compile-time increase in the
526 buffer size, when this authenticator is compiled into exim.
528 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
529 previous fast-mode was untenable in the face of glibs using mmap to
530 support larger malloc requests.
532 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
533 New values supported, if defined on system where compiled:
534 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
535 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
537 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
538 running a limited number of proceses in parallel, a benefit is gained. The
539 amount varies with the platform hardware and load. The use of the option
540 queue_run_in_order means we cannot do this, as ordering becomes
543 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
544 had introduced a string-copy (for ensuring NUL-termination) which was not
545 appropriate for that case, which can include embedded NUL bytes in the
546 block of data. Investigation showed the copy to actually be needless, the
547 data being length-specified.
549 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
550 done during a receiving connection, and both used TLS, global info was
551 used rather than per-connection info for tracking the state of data
552 queued for transmission. This could result in a connection hang.
554 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
555 Previously, when delivering serveral messages down a single connection
556 only the first would provide a SIZE. This was due to the size information
557 not being properly tracked.
559 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
560 TAI (at 37 seconds currently), pretend to be in UTC for time-related
561 expansion and logging. Previously, spurious values such as a future
562 minute could be seen.
564 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
565 it could crash from a null-deref. This could also affect the
566 ${addresses: } operator and ${readsock } item.
568 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
569 message following a mime one, the variable was not reset.
571 JH/30 When an pipelined-connect fails at the first response, assume incorrect
572 cached capability (perhaps the peer reneged?) and immediately retry in
575 JH/31 Fix spurious detection of timeout while writing to transport filter.
577 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
578 an attempt to copy the string was made before checking it.
580 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
581 taint of the lookup key was maintained; we now regard the presence in the
582 filesystem as sufficient validation.
584 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
585 argument is supplied.
587 JH/35 The dsearch lookup now requires that the directory is an absolute path.
588 Previously this was not checked, and nonempty relative paths made an
589 access under Exim's current working directory.
591 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
592 Previously no event was raised.
594 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
595 parameter supplied by the sender MAIL FROM command. Previously it was
596 ignored, and only the check_spool_space option value for the required
599 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
600 the size of the signing public-key. Previously it was instead giving
601 the size of the signature hash.
603 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
604 the default. See the (new) dkim_verify_min_keysizes option.
606 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
607 and an ACL use a lookup for checking either the local_part or domain,
608 stale data could be accessed. Ensure that variable references are
609 dropped between messages.
611 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
612 by the client was not checked as pointing within response data before
613 being used. A malicious client could thus cause an out-of-bounds read and
614 possibly gain authentication. Fix by adding the check.
616 JH/42 Internationalisation: change the default for downconversion in the smtp
617 transport to be "if needed". Previously it was "as previously set" for
618 the message, which usually meant "if needed" for message-submission but
619 "no" for everything else. However, MTAs have been seen using SMTPUTF8
620 even when the envelope addresses did not need it, resulting in forwarding
621 failures to non-supporting MTAs. A downconvert in such cases will be
622 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
623 transport. The change does mean that addresses needing conversion will
624 be converted when previously a delivery failure would occur.
626 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
627 response was received it would be used unchecked in a fail-DSN, violating
628 standards on line-length limits. Truncate if needed.
630 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
631 doesn't work. (Additionally add a new main config option to configure the
638 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
639 side, to decode a TLS 1.3 packet capture.
641 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
642 Previously the default library behaviour applied, sending two, each in
645 JH/03 Debug output for ACL now gives the config file name and line number for
648 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
650 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
652 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
653 buffer overrun for (non-chunking) other transports.
655 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
656 TLS1.3, means that a server rejecting a client certificate is not visible
657 to the client until the first read of encrypted data (typically the
658 response to EHLO). Add detection for that case and treat it as a failed
659 TLS connection attempt, so that the normal retry-in-clear can work (if
660 suitably configured).
662 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
663 and/or domain. Found and fixed by Jason Betts.
665 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
666 configuration). If a CNAME target was not a wellformed name pattern, a
669 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
670 the OS reports them interleaved with other addresses.
672 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
673 used both for input and for a verify callout, both encrypted, SMTP
674 responses being sent by the server could be lost. This resulted in
675 dropped connections and sometimes bounces generated by a peer sending
678 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
679 string. Previously it was possible to cause undefined behaviour in a
680 library routine (usually a crash). Found by "zerons".
682 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
685 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
686 API was removed, so update to use the newer ones.
688 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
689 any timeout set, is taking a long time. Previously we would hang on to a
690 rotated logfile "forever" if the input was arriving with long gaps
691 (a previous attempt to fix addressed lack, for a long time, of initial
694 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
695 shared (NFS) environment. The length of the tempfile name is now
696 4 + 16 ("hdr.$message_exim_id") which might break on file
697 systems which restrict the file name length to lower values.
700 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
701 shared (NFS) environment.
703 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
704 did for all versions <4.90). Notably -M, -m, --invert, -I may be
707 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
708 on some platforms for bit 31.
710 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
711 to changes apparently associated with TLS1.3 handling some of the APIs
712 previously used were either nonfunctional or inappropriate. Strings
713 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
714 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
715 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
716 This affects log line X= elements, the $tls_{in,out}_cipher variables,
717 and the use of specific cipher names in the encrypted= ACL condition.
719 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
721 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
722 verification result was not updated unless hosts_require_ocsp applied.
724 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
725 queue_list_requires_admin set to false, non-admin users were denied the
728 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
729 directory-of-certs mode. Previously they were advertised despite the
732 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
733 A single TCP connection by a client will now hold a TLS connection open
734 for multiple message deliveries, by default. Previously the default was to
737 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
738 default. If built with the facility, DANE will be used. The facility
739 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
741 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
742 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
743 must be defined and you must still, unless you define DISABLE_TLS, manage
744 the the include-dir and library-file requirements that go with that
745 choice. Non-TLS builds are still supported.
747 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
750 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
751 default. If the platform supports and has the facility enabled, it will
752 be requested on all coneections.
754 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
755 controlled by the build-time option SUPPORT_PIPE_CONNECT.
757 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
759 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
760 success-DSN messages. Previously the From: header was always the default
761 one for these; the option was ignored.
763 JH/28 Fix the timeout on smtp response to apply to the whole response.
764 Previously it was reset for every read, so a teergrubing peer sending
765 single bytes within the time limit could extend the connection for a
766 long time. Credit to Qualsys Security Advisory Team for the discovery.
768 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
769 delivery address, which leaked information of the results of local
770 forwarding. Change to the original envelope recipient address, per
773 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
774 requested. Previously not bounce was generated and a log entry of
775 error ignored was made.
777 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
779 JH/32 Introduce a general tainting mechanism for values read from the input
780 channel, and values derived from them. Refuse to expand any tainted
781 values, to catch one form of exploit.
783 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
784 was unused and the unexpanded text used for the test. Found and
785 fixed by Ruben Jenster.
787 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
788 an attempt to use a TLS library read routine dereffed a nul pointer,
791 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
792 them in our smtp response.
794 JH/36 Have the general router option retry_use_local_part default to true when
795 any of the restrictive preconditions are set (to anything). Previously it
796 was only for check_local user. The change removes one item of manual
797 configuration which is required for proper retries when a remote router
798 handles a subset of addresses for a domain.
800 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
801 link count into consideration.
803 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
804 caused the extension of big_buffer, the following lines were ignored.
806 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
807 accordance with RFC 2308. Previously there was no expiry, so a longlived
808 receive process (eg. due to ACL delays) versus a short SOA value could
811 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
813 JH/39 Promote DMARC support to mainline.
815 JH/40 Bug 2452: Add a References: header to DSNs.
817 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
818 parameters. The relevant library call is documented as "Deprecated: This
819 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
820 3.6.0, DH parameters are negotiated following RFC7919."
822 HS/06 Change the default of dnssec_request_domains to "*"
824 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
825 carried on and emitted a BDAT command, even when PIPELINING was not
828 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
829 buffer was used for the filename, resulting in a trap when tainted
830 arguments (eg. $domain) were used.
832 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
833 recommended to avoid a possible server-load attack. The feature can be
834 re-enabled via the openssl_options main cofiguration option.
836 JH/45 local_scan API: documented the current smtp_printf() call. This changed
837 for version 4.90 - adding a "more data" boolean to the arguments.
838 Bumped the ABI version number also, this having been missed previously;
839 release versions 4.90 to 4.92.3 inclusive were effectively broken in
840 respect of usage of smtp_printf() by either local_scan code or libraries
841 accessed via the ${dlfunc } expansion item. Both will need coding
842 adjustment for any calls to smtp_printf() to match the new function
843 signature; a FALSE value for the new argument is always safe.
845 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
846 the file-offset (which the Linux syscall does, and exim expects); this
847 resulted in an indefinite loop.
849 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
850 to do ARC verification. The Authentication-Results: header line added
851 by the configuration then had no ARC item.
853 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
854 an incorrect mode was used for reading the result, resulting in it being
861 JH/01 Remove code calling the customisable local_scan function, unless a new
862 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
864 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
865 non-signal-safe functions being used.
867 JH/03 Bug 2269: When presented with a received message having a stupidly large
868 number of DKIM-Signature headers, disable DKIM verification to avoid
869 a resource-consumption attack. The limit is set at twenty.
871 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
872 report of oldest_pass in ${authres } in consequence, and separate out
873 some descriptions of reasons for verification fail.
875 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
876 files in the spool were present and unlocked. A queue-runner could spot
877 them, resulting in a duplicate delivery. Fix that by doing the unlock
878 after the unlink. Investigation by Tim Stewart. Take the opportunity to
879 add more error-checking on spoolfile handling while that code is being
882 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
883 No known attacks, no CVE, this is defensive hardening.
885 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
886 a queue-runner could start a delivery while other operations were ongoing.
887 Cutthrough delivery was a common victim, resulting in duplicate delivery.
888 Found and investigated by Tim Stewart. Fix by using the open message data
889 file handle rather than opening another, and not locally closing it (which
890 releases a lock) for that case, while creating the temporary .eml format
891 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
893 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
894 $sender_verify_failure/$recipient_verify_failure to "random".
896 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
899 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
900 Previously this would segfault.
902 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
905 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
906 like zero, since the resolver should be doing this for us, But we need one
907 as a CNAME but no MX presence gets the CNAME returned; we need to check
908 that doesn't point to an MX to declare it "no MX returned" rather than
909 "error, loop". A new main option is added so the older capability of
910 following some limited number of chain links is maintained.
912 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
914 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
915 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
916 now-deprecated earlier definitions used only specified the range up to TLS
917 1.2 (in the older-version library docs).
919 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
921 JH/15 Rework TLS client-side context management. Stop using a global, and
922 explicitly pass a context around. This enables future use of TLS for
923 connections to service-daemons (eg. malware scanning) while a client smtp
924 connection is using TLS; with cutthrough connections this is quite likely.
926 JH/16 Fix ARC verification to do AS checks in reverse order.
928 JH/17 Support a "tls" option on the ${readsocket } expansion item.
930 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
931 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
932 Previously the "utf8" would be re-prepended for every additional message.
934 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
935 Previously thery were accepted, resulting in issues when attempting to
936 forward messages to a non-supporting MTA.
938 PP/02 Let -n work with printing macros too, not just options.
940 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
941 one parent address was copied, and bogus data was used at delivery-logging
942 time. Either a crash (after delivery) or bogus log data could result.
943 Discovery and analysis by Tim Stewart.
945 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
946 Previously if the string ended mid-character, we did not insert the
947 promised '?' replacement.
949 PP/04 Documentation: current string operators work on bytes, not codepoints.
951 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
952 should pack well giving a smaller memory footprint so better caching and
953 therefore performance. Group the declarations where this can't be done so
954 that the byte-sized flag variables are not interspersed among pointer
955 variables, giving a better chance of good packing by the compiler.
957 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
958 non-null, to avoid issues with sites running BATV. Previously reports were
959 sent with an empty envelope sender so looked like bounces.
961 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
962 The ignore_error flag wasn't being returned from the filter subprocess so
963 was not set for later routers. Investigation and fix by Matthias Kurz.
965 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
966 and a msg:complete for the whole, when a message is manually removed using
967 -Mrm. Developement by Matthias Kurz, hacked on by JH.
969 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
970 a "Gnu special" function, asprintf() in the DB utility binary builds; I
971 hope that is portable enough.
973 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
974 requiring a known-CA anchor certificate; make it now rely entirely on the
975 TLSA as an anchor. Checking the name on the leaf cert against the name
976 on the A-record for the host is still done for TA (but not for EE mode).
978 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
979 included in delivery lines for non-proxied connections, when compiled with
980 SUPPORT_SOCKS and running with proxy logging enabled.
982 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
983 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
984 move the existing event to fire before the normal logging of message
985 failure so that custom logging is bracketed by normal logging.
987 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
988 msg:fail:internal event. Developement by Matthias Kurz.
990 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
991 far too small for todays use of crypto signatures stored there. Go all
992 the way to the max DNS message size of 64kB, even though this might be
993 overmuch for IOT constrained device use.
995 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
996 copy a string over itself. The library routine is documented as not
997 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
999 JH/32 For main options check_spool_space and check_inode_space, where the
1000 platform supports 64b integers, support more than the previous 2^31 kB
1001 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1002 the previous G, M, k.
1004 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1005 $authenticated_fail_id variable on authentication failure. Previously
1008 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1009 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1010 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1011 GNUTLS_SEC_PARAM_MEDIUM.
1013 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1014 side. Previously we would continue as if no SNI had been received.
1016 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1017 "<" character, which should be a list-separator specification, we walked
1018 off past the nul-terimation.
1020 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1021 causes) even when the retry time is not yet met. Previously they were
1022 not, meaning that when (say) an account was over-quota and temp-rejecting,
1023 and multiple senders' messages were queued, only one sender would get
1024 notified on each configured delay_warning cycle.
1026 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1028 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1029 cipher-suites, an error can be left on the stack even for a succeeding
1030 accept; this results in impossible error messages when a later operation
1033 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1034 return error codes indicating retry. Under TLS1.3 this becomes required.
1036 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1037 it only wrote the new authenticators, resulting in a lack of tracking of
1038 peer changes of ESMTP extensions until the next cache flush.
1040 JH/41 Fix the loop reading a message header line to check for integer overflow,
1041 and more-often against header_maxsize. Previously a crafted message could
1042 induce a crash of the recive process; now the message is cleanly rejected.
1044 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1045 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1051 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1052 When redis_servers is set to a list of > 1 element, and the Redis servers
1053 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1054 case of MOVED into a DEFER case instead, thus moving the query onto the
1055 next server in the list. For a cluster of N elements, all N servers must
1056 be defined in redis_servers.
1058 GF/02 Catch and remove uninitialized value warning in exiqsumm
1059 Check for existence of @ARGV before looking at $ARGV[0]
1061 JH/01 Replace the store_release() internal interface with store_newblock(),
1062 which internalises the check required to safely use the old one, plus
1063 the allocate and data copy operations duplicated in both (!) of the
1064 extant use locations.
1066 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1067 modifier. This matches the restriction on the commandline.
1069 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1070 Previously only the last row was returned.
1072 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1073 we assumed that tags in the header were well-formed, and parsed the
1074 element content after inspecting only the first char of the tag.
1075 Assumptions at that stage could crash the receive process on malformed
1078 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1079 While running the DKIM ACL we operate on the Permanent memory pool so that
1080 variables created with "set" persist to the DATA ACL. Also (at any time)
1081 DNS lookups that fail create cache records using the Permanent pool. But
1082 expansions release any allocations made on the current pool - so a dnsdb
1083 lookup expansion done in the DKIM ACL releases the memory used for the
1084 DNS negative-cache, and bad things result. Solution is to switch to the
1085 Main pool for expansions.
1086 While we're in that code, add checks on the DNS cache during store_reset,
1087 active in the testsuite.
1088 Problem spotted, and debugging aided, by Wolfgang Breyha.
1090 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1091 When none of the hosts presented to a transport match an already-open
1092 connection, close it and proceed with the list. Previously we would
1093 queue the message. Spotted by Lena with Yahoo, probably involving
1096 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1097 Previously a spurious "250 OK id=" response was appended to the proper
1100 JH/08 The "support for" informational output now, which built with Content
1101 Scanning support, has a line for the malware scanner interfaces compiled
1102 in. Interface can be individually included or not at build time.
1104 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1105 by the template makefile "src/EDITME". The "STREAM" support for an older
1106 ClamAV interface method is removed.
1108 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1109 rows affected is given instead).
1111 JH/11 The runtime Berkeley DB library version is now additionally output by
1112 "exim -d -bV". Previously only the compile-time version was shown.
1114 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1115 SMTP connection. Previously, when one had more recipients than the
1116 first, an abortive onward connection was made. Move to full support for
1117 multiple onward connections in sequence, handling cutthrough connection
1118 for all multi-message initiating connections.
1120 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1121 routers. Previously, a multi-recipient message would fail to match the
1122 onward-connection opened for the first recipient, and cause its closure.
1124 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1125 a timeout on read on a GnuTLS initiating connection, resulting in the
1126 initiating connection being dropped. This mattered most when the callout
1127 was marked defer_ok. Fix to keep the two timeout-detection methods
1130 JH/15 Relax results from ACL control request to enable cutthrough, in
1131 unsupported situations, from error to silently (except under debug)
1132 ignoring. This covers use with PRDR, frozen messages, queue-only and
1135 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1137 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1138 metadata, resulting in a crash in free().
1140 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1141 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1142 Broken also in d185889f4, with init system revamp.
1144 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1145 Previously we abruptly closed the connection after reading a malware-
1146 found indication; now we go on to read the "scan ok" response line,
1149 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1150 ACL. Previously, a crash would result.
1152 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1153 macro text after a replacement (previously it was only once per line) and
1154 by skipping builtin macros when searching for an uppercase lead character.
1156 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1157 for the build is renamed.
1159 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1160 was allocated for every new TLS startup, meaning one per message. Fix
1161 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1163 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1164 reported the original. Fix to report (as far as possible) the ACL
1165 result replacing the original.
1167 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1168 OpenSSL. Certificate information is loaded for every new TLS startup,
1169 and the resources needed to be freed.
1171 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1173 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1176 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1177 DATA response info to the (existing) per-recipient response info for
1178 the "C=" log element. It can have useful tracking info from the
1179 destination system. Patch from Simon Arlott.
1181 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1182 length value. Previously this would segfault.
1184 HS/02 Support Avast multiline protoocol, this allows passing flags to
1185 newer versions of the scanner.
1187 JH/28 Ensure that variables possibly set during message acceptance are marked
1188 dead before release of memory in the daemon loop. This stops complaints
1189 about them when the debug_store option is enabled. Discovered specifically
1190 for sender_rate_period, but applies to a whole set of variables.
1191 Do the same for the queue-runner and queue-list loops, for variables set
1192 from spool message files. Do the same for the SMTP per-message loop, for
1193 certain variables indirectly set in ACL operations.
1195 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1196 as a multi-recipient message from a mailinglist manager). The coding had
1197 an arbitrary cutoff number of characters while checking for more input;
1198 enforced by writing a NUL into the buffer. This corrupted long / fast
1199 input. The problem was exposed more widely when more pipelineing of SMTP
1200 responses was introduced, and one Exim system was feeding another.
1201 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1202 receiving system, and refused recipients seen by the sending system
1203 (propating to people being dropped from mailing lists).
1204 Discovered and pinpointed by David Carter.
1206 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1207 replaced by the ${authresults } expansion.
1209 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1211 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1212 allows proper process termination in container environments.
1214 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1215 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1217 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1218 and "err_perm", deprecated since 4.83 when the RFC-defined words
1219 "temperror" and "permerror" were introduced.
1221 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1222 transport-filters or DKIM-signing. The restriction was lost in the
1223 consolidation of verify-callout and delivery SMTP handling.
1224 Extend the restriction to also cover ARC-signing.
1226 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1227 in defer=pass mode supply a 450 to the initiator. Previously the message
1230 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1231 tls_require_ciphers is used as before.
1233 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1234 "pass_unscanned". Only tmpfails from the scanner are written to
1235 the paniclog, as they may require admin intervention (permission
1236 denied, license issues). Other scanner errors (like decompression
1237 bombs) do not cause a paniclog entry.
1239 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1240 Previously it was possible to log spurious information in receive log
1243 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1244 triggered odd behaviour from Outlook Express clients.
1246 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1249 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1250 since the IETF WG has not yet settled on that versus the original
1251 "bare" representation.
1253 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1254 Previously the millisecond value corrupted the output.
1255 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1256 corrupted the output.
1262 JH/01 Rework error string handling in TLS interface so that the caller in
1263 more cases is responsible for logging. This permits library-sourced
1264 string to be attached to addresses during delivery, and collapses
1265 pairs of long lines into single ones.
1267 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1268 during configuration. Wildcards are allowed and expanded.
1270 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1271 This permits better logging.
1273 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1274 do not close it down and have the child transport start it up again on
1275 the passed-on TCP connection. Instead, proxy the child (and any
1276 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1277 affected: the continued delivery log lines do not have any DNSSEC, TLS
1278 Certificate or OCSP information. TLS cipher information is still logged.
1280 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1281 identical IP addresses on different listening ports. Will also affect
1284 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1285 add noisy ifdef guards to special-case this sillyness.
1286 Patch from Bernd Kuhls.
1288 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1289 than 255 are no longer allowed.
1291 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1292 Disable the session-cache too, which might reduce our load. Since we
1293 currrectly use a new context for every connection, both as server and
1294 client, there is no benefit for these.
1295 GnuTLS appears to not support tickets server-side by default (we don't
1296 call gnutls_session_ticket_enable_server()) but client side is enabled
1297 by default on recent versions (3.1.3 +) unless the PFS priority string
1300 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1301 <https://reproducible-builds.org/specs/source-date-epoch/>.
1303 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1304 the check for any unsuccessful recipients did not notice the limit, and
1305 erroneously found still-pending ones.
1307 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1308 MSG_MORE. Only in-clear (not on TLS connections).
1310 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1311 transport-filter is involved we can buffer the headers in memory for
1312 creating the signature, and read the spool data file once for the
1313 signature and again for transmission.
1315 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1316 4.77 as the kernel support then wasn't solid, having issues in 64bit
1317 mode. Now, it's been long enough. Add support for FreeBSD also.
1319 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1320 case where the routing stage had gathered several addresses to send to
1321 a host before calling the transport for the first, we previously failed
1322 to close down TLS in the old transport process before passing the TCP
1323 connection to the new process. The new one sent a STARTTLS command
1324 which naturally failed, giving a failed delivery and bloating the retry
1325 database. Investigation and fix prototype from Wolfgang Breyha.
1327 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1328 false-negatives in the check that the sender had not preempted a response
1329 or prompt from Exim (running as a server), due to that code's lack of
1330 awareness of the SMTP input buffering.
1332 PP/04 Add commandline_checks_require_admin option.
1333 Exim drops privileges sanely, various checks such as -be aren't a
1334 security problem, as long as you trust local users with access to their
1335 own account. When invoked by services which pass untrusted data to
1336 Exim, this might be an issue. Set this option in main configuration
1337 AND make fixes to the calling application, such as using `--` to stop
1340 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1341 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1342 responses to those, into a single TLS record each way (this usually means
1343 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1346 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1347 affects you only if you're dancing at the edge of the param size limits.
1348 If you are, and this message makes sense to you, then: raise the
1349 configured limit or use OpenSSL 1.1. Nothing we can do for older
1352 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1353 cmdline element to get the documented default one. Previously it was
1356 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1357 are unsuable for non-message-specific reasons - eg. connection timeout,
1358 banner-time rejection.
1360 JH/16 Fix logging of delivery remote port, when specified by router, under
1363 PP/06 Repair manualroute's ability to take options in any order, even if one
1364 is the name of a transport.
1367 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1369 JH/17 Change the list-building routines interface to use the expanding-string
1370 triplet model, for better allocation and copying behaviour.
1372 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1373 Previously it was constructed the first time a possibly-matching string
1374 was met in the configuration file input during startup; now it is done
1377 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1378 compatible one, to avoid the (poorly documented) possibility of a config
1379 file in the working directory redirecting the DB files, possibly correpting
1380 some existing file. CVE-2017-10140 assigned for BDB.
1382 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1383 cache-hot. Previously, although the result was properly cached, the
1384 initial verify call returned a defer.
1386 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1387 the main verify for receipient in uncached-mode.
1389 JH/22 Retire historical build files to an "unsupported" subdir. These are
1390 defined as "ones for which we have no current evidence of testing".
1392 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1393 if present. Previously it was ignored.
1395 JH/24 Start using specified-initialisers in C structure init coding. This is
1396 a C99 feature (it's 2017, so now considered safe).
1398 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1399 if was a fixed-sized field and bitmask ops via macros; it is now more
1402 PP/07 GitHub PR 56: Apply MariaDB build fix.
1403 Patch provided by Jaroslav Škarvada.
1405 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1406 during Coverity cleanups [4.87 JH/47]
1407 Diagnosis and fix provided by Michael Fischer v. Mollard.
1409 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1410 the right size to place the terminating semicolon on its own folded
1411 line, the header hash was calculated to an incorrect value thanks to
1412 the (relaxed) space the fold became.
1414 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1415 and confused the parent.
1417 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1418 which could crash as a result. This could lead to undeliverable messages.
1420 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1423 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1424 matters on fast-turnover and PID-randomising systems, which were getting
1425 out-of-order delivery.
1427 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1428 a possibly-overlapping copy. The symptom was that "Remote host closed
1429 connection in response to HELO" was logged instead of the actual 4xx
1432 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1433 Previously only that bufferd was discarded, resulting in SYMTP command
1436 JH/32 DKIM: when a message has multiple signatures matching an identity given
1437 in dkim_verify_signers, run the dkim acl once for each. Previously only
1438 one run was done. Bug 2189.
1440 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1441 "panic the current process" to "deliberately defer". The panic log is
1442 still written with the problem list name; the mail and reject logs now
1443 get a temp-reject line for the message that was being handled, saying
1444 something like "domains check lookup or other defer". The SMTP 451
1445 message is still "Temporary local problem".
1447 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1448 A crafted sequence of BDAT commands could result in in-use memory beeing
1449 freed. CVE-2017-16943.
1451 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1452 from SMTP input. Previously it was always done; now only done for DATA
1453 and not BDAT commands. CVE-2017-16944.
1455 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1456 to the message (such as an overlong header line). Previously this was
1457 not done and we did not exit BDAT mode. Followon from the previous item
1458 though a different problem.
1464 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1465 than -2003 did; needs libidn2 in addition to libidn.
1467 JH/02 The path option on a pipe transport is now expanded before use.
1469 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1470 Patch provided by "Björn", documentation fix added too.
1472 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1473 missing a wire-to-host endian conversion.
1475 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1476 close after a BDAT command line could be taken as a following command,
1477 giving a synch failure. Fix by only checking for synch immediately
1478 before acknowledging the chunk.
1480 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1481 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1482 macro. Patches provided by Josh Soref.
1484 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1485 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1486 by IANA as a service extension. However, John Klensin suggests that we
1489 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1490 the dkim code may be unix-mode line endings rather than smtp wire-format
1491 CRLF, so prepend a CR to any bare LF.
1493 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1494 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1496 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1497 the first were themselves being wrongly included in the feed into dkim
1498 processing; with most chunk sizes in use this resulted in an incorrect
1499 body hash calculated value.
1501 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1502 DKIM signature block, for verification. Although advised against by
1503 standards it is specifically not ruled illegal.
1505 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1507 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1508 missing a body hash (the bh= tag).
1510 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1511 It seems that HAProxy sends the Proxy Protocol information in clear and
1512 only then does a TLS startup, so do the same.
1514 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1515 TCP connections (such as for Spamd) unless the daemon successfully set
1516 Fast Open mode on its listening sockets. This fixes breakage seen on
1517 too-old kernels or those not configured for Fast Open, at the cost of
1518 requiring both directions being enabled for TFO, and TFO never being used
1519 by non-daemon-related Exim processes.
1521 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1522 endings, at least on the first header line. Try to canonify any that get
1523 past that check, despite the cost.
1525 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1526 now limited to an arbitrary five deep, while parsing addresses with the
1527 strip_excess_angle_brackets option enabled.
1529 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1530 instead leave the unprompted TLS handshake in socket buffer for the
1531 TLS library to consume.
1533 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1535 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1537 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1538 region in one operation, for speed, and this leaves assigned pointers
1539 dangling. Add checks run only under the testsuite which checks all
1540 variables at a store-reset and panics on a dangling pointer; add code
1541 explicitly nulling out all the variables discovered. Fixes one known
1542 bug: a transport crash, where a dangling pointer for $sending_ip_address
1543 originally assigned in a verify callout, is re-used.
1545 PP/06 Drop '.' from @INC in various Perl scripts.
1547 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1549 PP/08 Reduce a number of compilation warnings under clang; building with
1550 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1551 should be warning-free.
1553 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1555 HS/01 Fix portability problems introduced by PP/08 for platforms where
1556 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1558 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1559 chunk. This allows us to accept broken chunked messages. We need a more
1560 general solution here.
1562 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1563 already-broken messages in the queue.
1565 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1567 JH/19 Fix reference counting bug in routing-generated-address tracking.
1573 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1574 supports it and a size is available (ie. the sending peer gave us one).
1576 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1577 years of being deprecated). The replacements are the ACLs
1578 acl_smtp_mime and acl_not_smtp_mime.
1580 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1581 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1582 or even an in-clear connection were permitted. Now, if the host lookup
1583 was dnssec and dane was requested then the host is only used if the
1584 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1585 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1586 if one fails this test.
1587 This means that a poorly-configured remote DNS will make it incommunicado;
1588 but it protects against a DNS-interception attack on it.
1590 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1591 non-noisy when a race steals the message being considered.
1593 JH/05 If main configuration option tls_certificate is unset, generate a
1594 self-signed certificate for inbound TLS connections.
1596 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1597 in rewrites and routers.
1599 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1600 and logged a warning sing 4.83; now they are a configuration file error.
1602 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1603 (lacking @domain). Apply the same qualification processing as RCPT.
1605 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1607 JH/10 Support ${sha256:} applied to a string (as well as the previous
1610 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1611 a cutthrough deliver is pending, as we always want to make a connection.
1612 This also avoids re-routing the message when later placing the cutthrough
1613 connection after a verify cache hit.
1614 Do not update it with the verify result either.
1616 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1617 when routing results in more than one destination address.
1619 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1620 signing (which inhibits the cutthrough capability). Previously only
1621 the presence of an option was tested; now an expansion evaluating as
1622 empty is permissible (obviously it should depend only on data available
1623 when the cutthrough connection is made).
1625 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1626 the relevant preceding SMTP command did not note the pipelining mode.
1628 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1629 Previously they were not counted.
1631 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1632 as one having no matching records. Previously we deferred the message
1633 that needed the lookup.
1635 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1636 distinguished as "(=".
1638 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1639 for missing MX records. Previously it only worked for missing A records.
1641 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1643 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1644 after the data-go-ahead and data-ack. Patch from Jason Betts.
1646 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1647 even for a "none" policy. Patch from Tony Meyer.
1649 JH/22 Fix continued use of a connection for further deliveries. If a port was
1650 specified by a router, it must also match for the delivery to be
1653 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1654 When one of the recipients of a message was unsuitable for the connection
1655 (has no matching addresses), we lost track of needing to mark it
1656 deferred. As a result mail would be lost.
1658 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1660 JH/25 Decoding ACL controls is now done using a binary search; the source code
1661 takes up less space and should be simpler to maintain. Merge the ACL
1662 condition decode tables also, with similar effect.
1664 JH/26 Fix problem with one_time used on a redirect router which returned the
1665 parent address unchanged. A retry would see the parent address marked as
1666 delivered, so not attempt the (identical) child. As a result mail would
1669 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1670 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1671 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1674 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1675 Main config options check_{log,spool}_{inodes,space} are now
1676 100 inodes, 10MB unless set otherwise in the configuration.
1678 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1679 Previously it only applied to the main-section connection policy
1682 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1684 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1685 by me. Added RFC7919 DH primes as an alternative.
1687 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1688 are not in the system include path.
1690 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1691 GnuTLS, when a session startup failed (eg because the client disconnected)
1692 Exim did stdio operations after fclose. This was exposed by a recent
1693 change which nulled out the file handle after the fclose.
1695 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1696 signed directly by the cert-signing cert, rather than an intermediate
1697 OCSP-signing cert. This is the model used by LetsEncrypt.
1699 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1701 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1702 an incoming connection.
1704 HS/02 Bug 1802: Do not half-close the connection after sending a request
1707 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1708 fallback to "prime256v1".
1710 JH/34 SECURITY: Use proper copy of DATA command in error message.
1711 Could leak key material. Remotely exploitable. CVE-2016-9963.
1717 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1718 and 3.4.4 - once the server is enabled to respond to an OCSP request
1719 it does even when not requested, resulting in a stapling non-aware
1720 client dropping the TLS connection.
1722 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1723 support variable-length bit vectors. No functional change.
1725 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1726 The I= interface field on outgoing lines is now after the H= remote
1727 host field, same as incoming lines. There is a separate
1728 outgoing_interface log selector which allows you to disable the
1731 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1732 If not running log_selector +smtp_connection the mainlog would be held
1733 open indefinitely after a "too many connections" event, including to a
1734 deleted file after a log rotate. Leave the per net connection logging
1735 leaving it open for efficiency as that will be quickly detected by the
1736 check on the next write.
1738 HS/01 Bug 1671: Fix post transport crash.
1739 Processing the wait-<transport> messages could crash the delivery
1740 process if the message IDs didn't exist for some reason. When
1741 using 'split_spool_directory=yes' the construction of the spool
1742 file name failed already, exposing the same netto behaviour.
1744 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1745 mime_regex ACL conditions.
1747 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1748 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1749 to HELO, local diagnostic string.
1751 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1752 log line to debug. Even when configured with a tls authenticator many
1753 client connections are expected to not authenticate in this way, so
1754 an authenticate fail is not an error.
1756 HS/02 Add the Exim version string to the process info. This way exiwhat
1757 gives some more detail about the running daemon.
1759 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1760 matter for fast-change records such as DNSBLs.
1762 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1763 retry record, even if constant. There may be multiple transports with
1764 different interface settings and the retry behaviour needs to be kept
1767 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1769 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1771 JH/10 Bug 840: fix log_defer_output option of pipe transport
1773 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1774 hosts_randomize. This matters a lot when combined with mua_wrapper.
1776 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1777 ${quote_pgsql:<string>} operator.
1779 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1781 JH/14 Change header file name for freeradius-client. Relevant if compiling
1782 with Radius support; from the Gentoo tree and checked under Fedora.
1784 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1786 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1787 option transport_filter, meaning no filtering.
1789 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1791 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1792 defaults to "*" (all hosts). The variable is now available when not built
1793 with TLS, default unset, mainly to enable keeping the testsuite sane.
1794 If a server certificate is not supplied (via tls_certificate) an error is
1795 logged, and clients will find TLS connections fail on startup. Presumably
1796 they will retry in-clear.
1797 Packagers of Exim are strongly encouraged to create a server certificate
1798 at installation time.
1800 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1801 with the $config_file variable.
1803 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1804 in transport context, after the attempt, and per-recipient. The latter type
1805 is per host attempted. The event data is the error message, and the errno
1806 information encodes the lookup type (A vs. MX) used for the (first) host,
1807 and the trailing two digits of the smtp 4xx response.
1809 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1810 to write to mainlog (or rejectlog, paniclog) in the window between file
1811 creation and permissions/ownership being changed. Particularly affects
1812 installations where exicyclog is run as root, rather than exim user;
1813 result is that the running daemon panics and dies.
1815 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1817 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1818 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1819 "pri" and "weight". Note that the previous implicit priority given by the
1820 list order is no longer honoured.
1822 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1823 for DKIM processing.
1825 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1826 by defining SUPPORT_SOCKS.
1828 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1829 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1830 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1831 variables are renamed to proxy_{local,external}_{address,port}.
1833 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1834 for a build by defining SUPPORT_I18N
1836 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1837 of the query string, and make ${quote_redis:} do that quoting.
1839 JH/29 Move Events support from Experimental to mainline, enabled by default
1840 and removable for a build by defining DISABLE_EVENT.
1842 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1844 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1845 cached by the daemon.
1847 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1848 by defining LOOKUP_REDIS. The libhiredis library is required.
1850 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1851 keys are given for lookup.
1853 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1854 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1855 only supported when built with TLS support. The PolarSSL SHA routines
1856 are still used when the TLS library is too old for convenient support.
1858 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1859 openssl_options), for security. OpenSSL forces this from version 1.1.0
1860 server-side so match that on older versions.
1862 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1863 allocation for $value could be released as the expansion processing
1864 concluded, but leaving the global pointer active for it.
1866 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1867 and to use the domains and local_parts ACL conditions.
1869 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1870 incorrectly not doubled on cutthrough transmission, hence seen as a
1871 body-termination at the receiving system - resulting in truncated mails.
1872 Commonly the sender saw a TCP-level error, and retransmitted the message
1873 via the normal store-and-forward channel. This could result in duplicates
1874 received - but deduplicating mailstores were liable to retain only the
1875 initial truncated version.
1877 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1879 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1881 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1882 we're in there, support oversigning also; bug 1309.
1884 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1886 HS/04 Add support for keep_environment and add_environment options.
1888 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1889 either intentional arithmetic overflow during PRNG, or testing config-
1892 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1893 delivery resulted in actual delivery. Cancel cutthrough before DATA
1896 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1897 rejects a recipient: pass the reject to the originator.
1899 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1900 Many were false-positives and ignorable, but it's worth fixing the
1903 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1904 for the new environment-manipulation done at startup. Move the routines
1905 from being local to tls.c to being global via the os.c file.
1907 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1908 an extract embedded as result-arg for a map, the first arg for extract
1909 is unavailable so we cannot tell if this is a numbered or keyed
1910 extraction. Accept either.
1916 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1919 JH/02 The smtp transport option "multi_domain" is now expanded.
1921 JH/03 The smtp transport now requests PRDR by default, if the server offers
1924 JH/04 Certificate name checking on server certificates, when exim is a client,
1925 is now done by default. The transport option tls_verify_cert_hostnames
1926 can be used to disable this per-host. The build option
1927 EXPERIMENTAL_CERTNAMES is withdrawn.
1929 JH/05 The value of the tls_verify_certificates smtp transport and main options
1930 default to the word "system" to access the system default CA bundle.
1931 For GnuTLS, only version 3.0.20 or later.
1933 JH/06 Verification of the server certificate for a TLS connection is now tried
1934 (but not required) by default. The verification status is now logged by
1935 default, for both outbound TLS and client-certificate supplying inbound
1938 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1941 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1942 Status Notification (bounce) messages are now MIME format per RFC 3464.
1943 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1944 under the control of the dsn_advertise_hosts option, and routers may
1945 have a dsn_lasthop option.
1947 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1948 default, modifiable by a malware= option. The list separator for
1949 the options can now be changed in the usual way. Bug 68.
1951 JH/10 The smtp_receive_timeout main option is now expanded before use.
1953 JH/11 The incoming_interface log option now also enables logging of the
1954 local interface on delivery outgoing connections.
1956 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1957 if the interface and destination host and port all match.
1959 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1962 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1963 Patch from Andrew Lewis.
1965 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1966 now supports optional time-restrictions, weighting, and priority
1967 modifiers per server. Patch originally by <rommer@active.by>.
1969 JH/16 The spamd_address main option now supports a mixed list of local
1970 and remote servers. Remote servers can be IPv6 addresses, and
1971 specify a port-range.
1973 JH/17 Bug 68: The spamd_address main option now supports an optional
1974 timeout value per server.
1976 JH/18 Bug 1581: Router and transport options headers_add/remove can
1977 now have the list separator specified.
1979 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1982 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1985 JH/21 Support for the A6 type of dns record is withdrawn.
1987 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1988 rather than the verbs used.
1990 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1991 from 255 to 1024 chars.
1993 JH/24 Verification callouts now attempt to use TLS by default.
1995 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1996 are generic router options now. The defaults didn't change.
1998 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1999 Original patch from Alexander Shikoff, worked over by JH.
2001 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2002 files if ZCAT_COMMAND is not executable.
2004 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2006 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2008 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2009 Normally benign, it bites when the pair was led to by a CNAME;
2010 modern usage is to not canonicalize the domain to a CNAME target
2011 (and we were inconsistent anyway for A-only vs AAAA+A).
2013 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2015 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2016 when evaluating $sender_host_dnssec.
2018 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2019 $sender_helo_dnssec variable.
2021 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2023 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2025 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2027 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2028 documented as working, but never had. Support all but $spam_report.
2030 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2031 added for tls authenticator.
2033 HS/03 Add perl_taintmode main config option
2039 TL/01 When running the test suite, the README says that variables such as
2040 no_msglog_check are global and can be placed anywhere in a specific
2041 test's script, however it was observed that placement needed to be near
2042 the beginning for it to behave that way. Changed the runtest perl
2043 script to read through the entire script once to detect and set these
2044 variables, reset to the beginning of the script, and then run through
2045 the script parsing/test process like normal.
2047 TL/02 The BSD's have an arc4random API. One of the functions to induce
2048 adding randomness was arc4random_stir(), but it has been removed in
2049 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2050 function when detected.
2052 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2053 cause callback expansion.
2055 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2056 syntax errors in an expansion can be treated as a string instead of
2057 logging or causing an error, due to the internal use of bool_lax
2058 instead of bool when processing it.
2060 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2061 server certificates when making smtp deliveries.
2063 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2065 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2067 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2069 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2070 Merged patch from Sebastian Wiedenroth.
2072 JH/05 Fix results-pipe from transport process. Several recipients, combined
2073 with certificate use, exposed issues where response data items split
2074 over buffer boundaries were not parsed properly. This eventually
2075 resulted in duplicates being sent. This issue only became common enough
2076 to notice due to the introduction of connection certificate information,
2077 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2079 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2080 size buffer was used, resulting in syntax errors when an expansion
2083 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2084 version 3.3.6 or later.
2086 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2087 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2088 both become "event_action", the variables become $event_name, $event_data
2089 and $event_defer_errno. There is a new variable $verify_mode, usable in
2090 routers, transports and related events. The tls:cert event is now also
2091 raised for inbound connections, if the main configuration event_action
2094 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2095 early OCSP support, but no stapling (appears to be less than 1.0.0).
2097 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2098 server certificate names available under the smtp transport option
2099 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2102 JH/10 Time-related extraction expansions from certificates now use the main
2103 option "timezone" setting for output formatting, and are consistent
2104 between OpenSSL and GnuTLS compilations. Bug 1541.
2106 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2107 encoded parameter in the incoming message. Bug 1558.
2109 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2110 include certificate info, eximon was claiming there were spoolfile
2113 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2115 JH/14 Log delivery-related information more consistently, using the sequence
2116 "H=<name> [<ip>]" wherever possible.
2118 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2119 are problematic for Debian distribution, omit them from the release
2122 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2124 JH/16 Fix string representation of time values on 64bit time_t architectures.
2127 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2128 output list separator was used.
2133 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2134 checkers that were complaining about end of non-void function with no
2137 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2138 This was a regression introduced in 4.83 by another bugfix.
2140 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2142 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2143 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2149 TF/01 Correctly close the server side of TLS when forking for delivery.
2151 When a message was received over SMTP with TLS, Exim failed to clear up
2152 the incoming connection properly after forking off the child process to
2153 deliver the message. In some situations the subsequent outgoing
2154 delivery connection happened to have the same fd number as the incoming
2155 connection previously had. Exim would try to use TLS and fail, logging
2156 a "Bad file descriptor" error.
2158 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2159 utilities have not been installed.
2161 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2162 temporary space as the ACL may create new global variables.
2164 TL/01 LDAP support uses per connection or global context settings, depending
2165 upon the detected version of the libraries at build time.
2167 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2168 to extract and use the src ip:port in logging and expansions as if it
2169 were a direct connection from the outside internet. PPv2 support was
2170 updated based on HAProxy spec change in May 2014.
2172 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2174 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2175 Properly escape header and check for NULL return.
2177 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2180 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2182 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2183 characters in header names, implemented as a verify condition.
2184 Contributed by Michael Fischer v. Mollard.
2186 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2187 results permerror and temperror. Previous values are deprecated but
2188 still accepted. In a future release, err_perm and err_temp will be
2189 completely removed, which will be a backward incompatibility if the
2190 ACL tests for either of these two old results. Patch contributed by
2191 user bes-internal on the mailing list.
2193 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2195 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2196 selectors, in both main and reject logs.
2198 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2201 JH/07 Add malware type "sock" for talking to simple daemon.
2203 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2205 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2206 routers/transports under cutthrough routing.
2208 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2209 numbers. Touch up "bool" conditional to keep the same definition.
2211 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2213 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2215 JH/12 Expand items in router/transport headers_add or headers_remove lists
2216 individually rather than the list as a whole. Bug 1452.
2218 Required for reasonable handling of multiple headers_ options when
2219 they may be empty; requires that headers_remove items with embedded
2220 colons must have them doubled (or the list-separator changed).
2222 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2223 view the policy declared in the DMARC record. Currently, $dmarc_status
2224 is a combined value of both the record presence and the result of the
2227 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2229 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2230 dnslookup router and the smtp transport (applying to the forward
2233 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2234 of ldap servers used for a specific lookup. Patch provided by Heiko
2237 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2238 New variable $lookup_dnssec_authenticated for observability.
2240 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2241 Patch submitted by Lars Timman.
2243 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2245 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2246 Requires trusted mode and valid format message id, aborts otherwise.
2247 Patch contributed by Heiko Schlichting.
2249 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2250 certextract with support for various fields. Bug 1358.
2252 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2253 is requested by default, modifiable by smtp transport option
2256 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2257 operate on certificate variables to give certificate fingerprints
2258 Also new ${sha256:cert_variable}.
2260 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2262 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2263 Christian Aistleitner.
2265 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2267 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2268 file. Patch from Wolfgang Breyha.
2270 JH/25 Expand the coverage of the delivery $host and $host_address to
2271 client authenticators run in verify callout. Bug 1476.
2273 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2274 align with daemon_smtp_ports. Bug 72.
2276 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2277 support and error reporting did not work properly.
2279 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2280 and is readable. Patch from Andrew Colin Kissa.
2282 TL/14 Enhance documentation of ${run expansion and how it parses the
2283 commandline after expansion, particularly in the case when an
2284 unquoted variable expansion results in an empty value.
2286 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2288 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2289 by renaming away from it. Addresses GH issue 15, reported by
2292 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2294 TL/15 SECURITY: prevent double expansion in math comparison functions
2295 (can expand unsanitized data). Not remotely exploitable.
2302 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2304 PP/02 Make -n do something, by making it not do something.
2305 When combined with -bP, the name of an option is not output.
2307 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2310 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2311 $sender_host_name and config options to manage this, and basic check
2314 PP/05 DSCP support for outbound connections and control modifier for inbound.
2316 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2317 (Only plugin which currently uses this is kerberos4, which nobody should
2318 be using, but we should make it available and other future plugins might
2319 conceivably use it, even though it would break NAT; stuff *should* be
2320 using channel bindings instead).
2322 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2323 name; added for Sendmail compatibility; requires admin caller.
2324 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2325 just ignore it); requires trusted caller.
2326 Also parse but ignore: -Ac -Am -X<logfile>
2329 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2331 TL/02 Add +smtp_confirmation as a default logging option.
2333 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2334 Patch by Magnus Holmgren from 2007-02-20.
2336 TL/04 Bugzilla 1281 - Spec typo.
2337 Bugzilla 1283 - Spec typo.
2338 Bugzilla 1290 - Spec grammar fixes.
2340 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2342 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2344 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2345 mailing list by by Dmitry Isaikin.
2347 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2349 JH/02 Support "G" suffix to numbers in ${if comparisons.
2351 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2353 NM/01 Bugzilla 1197 - Spec typo
2354 Bugzilla 1196 - Spec examples corrections
2356 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2358 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2359 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2362 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2363 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2365 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2366 "acl {{name}{arg}...}", and optional args on acl condition
2369 JH/05 Permit multiple router/transport headers_add/remove lines.
2371 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2373 JH/07 Avoid using a waiting database for a single-message-only transport.
2374 Performance patch from Paul Fisher. Bugzilla 1262.
2376 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2379 JH/09 Add $headers_added variable, with content from use of ACL modifier
2380 add_header (but not yet added to the message). Bugzilla 199.
2382 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2383 Pulled from Bugzilla 817 by Wolfgang Breyha.
2385 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2387 (nb: this is the same fix as in Exim 4.80.1)
2389 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2392 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2395 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2397 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2398 advertise SMTP AUTH mechanism to us, instead of a generic
2399 protocol violation error. Also, make Exim more robust to bad
2400 data from the Dovecot auth socket.
2402 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2404 When a queue runner is handling a message, Exim first routes the
2405 recipient addresses, during which it prunes them based on the retry
2406 hints database. After that it attempts to deliver the message to
2407 any remaining recipients. It then updates the hints database using
2410 So if a recipient address works intermittently, it can get repeatedly
2411 deferred at routing time. The retry hints record remains fresh so the
2412 address never reaches the final cutoff time.
2414 This is a fairly common occurrence when a user is bumping up against
2415 their storage quota. Exim had some logic in its local delivery code
2416 to deal with this. However it did not apply to per-recipient defers
2417 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2419 This change adds a proper retry rule check during routing so that the
2420 final cutoff time is checked against the message's age. We only do
2421 this check if there is an address retry record and there is not a
2422 domain retry record; this implies that previous attempts to handle
2423 the address had the retry_use_local_parts option turned on. We use
2424 this as an approximation for the destination being like a local
2425 delivery, as in LMTP.
2427 I suspect this new check makes the old local delivery cutoff check
2428 redundant, but I have not verified this so I left the code in place.
2430 TF/02 Correct gecos expansion when From: is a prefix of the username.
2432 Test 0254 submits a message to Exim with the header
2436 When I ran the test suite under the user fanf2, Exim expanded
2437 the header to contain my full name, whereas it should have added
2438 a Resent-Sender: header. It erroneously treats any prefix of the
2439 username as equal to the username.
2441 This change corrects that bug.
2443 GF/01 DCC debug and logging tidyup
2444 Error conditions log to paniclog rather than rejectlog.
2445 Debug lines prefixed by "DCC: " to remove any ambiguity.
2447 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2449 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2450 Bug spotted by Jeremy Harris; was flawed since initial commit.
2451 Would have resulted in OCSP responses post-SNI triggering an Exim
2452 NULL dereference and crash.
2454 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2456 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2457 Bug detection, analysis and fix by Samuel Thibault.
2458 Bugzilla 1331, Debian bug #698092.
2460 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2462 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2463 Server implementation by Todd Lyons, client by JH.
2464 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2465 config variable "prdr_enable" controls whether the server
2466 advertises the facility. If the client requests PRDR a new
2467 acl_data_smtp_prdr ACL is called once for each recipient, after
2468 the body content is received and before the acl_smtp_data ACL.
2469 The client is controlled by both of: a hosts_try_prdr option
2470 on the smtp transport, and the server advertisement.
2471 Default client logging of deliveries and rejections involving
2472 PRDR are flagged with the string "PRDR".
2474 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2475 fclose(). Diagnosis by Todd Lyons.
2477 PP/17 Update configure.default to handle IPv6 localhost better.
2478 Patch by Alain Williams (plus minor tweaks).
2481 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2482 This is now consistent with GnuTLS, and is now documented: the
2483 previous undocumented portable approach to treating the option as
2484 unset was to force an expansion failure. That still works, and
2485 an empty string is now equivalent.
2487 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2488 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2489 not performing validation itself.
2491 PP/20 Added force_command boolean option to pipe transport.
2492 Patch from Nick Koston, of cPanel Inc.
2494 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2497 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2499 PP/21 Fix eximon continuous updating with timestamped log-files.
2500 Broken in a format-string cleanup in 4.80, missed when I repaired the
2501 other false fix of the same issue.
2502 Report and fix from Heiko Schlichting.
2505 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2506 Report from Prashanth Katuri.
2508 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2509 It's SecureTransport, so affects any MacOS clients which use the
2510 system-integrated TLS libraries, including email clients.
2512 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2513 using a MIME ACL for non-SMTP local injection.
2514 Report and assistance in diagnosis by Warren Baker.
2516 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2518 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2520 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2521 last id that failed so it may be referenced in subsequent ACL's.
2523 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2526 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2527 ldap library initialization, allowing self-signed CA's to be
2528 used. Also properly sets require_cert option later in code by
2529 using NULL (global ldap config) instead of ldap handle (per
2530 session). Bug diagnosis and testing by alxgomz.
2532 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2533 the src/util/ subdirectory.
2535 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2536 renamed to Transport Post Delivery Action by Jeremy Harris, as
2539 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2540 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2541 redis_servers = needs to be configured which will be used by the redis
2542 lookup. Patch from Warren Baker, of The Packet Hub.
2544 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2546 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2547 hostname or reverse DNS when processing a host list. Used suggestions
2548 from multiple comments on this bug.
2550 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2552 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2553 Missed a few lines, added it to make the runtest require no keyboard
2556 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2557 contains upper case chars. Make router use caseful_local_part.
2559 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2560 support when GnuTLS has been built with p11-kit.
2566 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2568 This, or similar/improved, will also be change PP/11 of 4.82.
2574 PP/01 Handle short writes when writing local log-files.
2575 In practice, only affects FreeBSD (8 onwards).
2576 Bugzilla 1053, with thanks to Dmitry Isaikin.
2578 NM/01 Bugzilla 949 - Documentation tweak
2580 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2583 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2585 PP/02 Implemented gsasl authenticator.
2587 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2589 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2590 `pkg-config foo` for cflags/libs.
2592 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2593 with rest of GSASL and with heimdal_gssapi.
2595 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2596 `pkg-config foo` for cflags/libs for the TLS implementation.
2598 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2599 properties get this fed in as external SSF. A number of robustness
2600 and debugging improvements to the cyrus_sasl authenticator.
2602 PP/08 cyrus_sasl server now expands the server_realm option.
2604 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2605 Patch by Jeremy Harris.
2607 PP/10 Added dbmjz lookup type.
2609 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2611 PP/12 MAIL args handles TAB as well as SP, for better interop with
2612 non-compliant senders.
2613 Analysis and variant patch by Todd Lyons.
2615 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2616 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2617 Patch from Dirk Mueller <dmueller@suse.com>
2619 PP/13 tls_peerdn now print-escaped for spool files.
2620 Observed some $tls_peerdn in wild which contained \n, which resulted
2621 in spool file corruption.
2623 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2624 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2625 or write after TLS renegotiation, which otherwise led to messages
2628 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2629 as a tracking header (ie: a signed header comes before the signature).
2630 Patch from Wolfgang Breyha.
2632 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2633 comma-sep list; embedded commas doubled.
2635 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2637 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2639 Report and patch from Dmitry Banschikov.
2641 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2642 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2643 failures. We appear to now support TLS1.1+ with Exim.
2645 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2646 lets Exim select keys and certificates based upon TLS SNI from client.
2647 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2648 before an outbound SMTP session. New log_selector, +tls_sni.
2650 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2651 NULL dereference. Report and patch from Alun Jones.
2653 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2654 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2655 Not seeing resolver debug output on NetBSD, but suspect this is a
2656 resolver implementation change.
2658 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2659 Left warnings. Added "eximon gdb" invocation mode.
2661 PP/21 Defaulting "accept_8bitmime" to true, not false.
2663 PP/22 Added -bw for inetd wait mode support.
2665 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2666 locate the relevant includes and libraries. Made this the default.
2668 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2669 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2671 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2672 This may cause build issues on older platforms.
2674 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2675 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2676 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2677 Added SNI support via GnuTLS too.
2678 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2680 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2682 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2683 Applied second patch from Janne, implementing suggestion to default
2684 multiple-strings-in-record handling to match SPF spec.
2686 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2688 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2689 read-only, out of scope).
2690 Patch from Wolfgang Breyha, report from Stuart Northfield.
2692 PP/29 Fix three issues highlighted by clang analyser static analysis.
2693 Only crash-plausible issue would require the Cambridge-specific
2694 iplookup router and a misconfiguration.
2695 Report from Marcin Mirosław.
2697 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2699 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2700 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2701 As part of this, removing so much warning spew let me fix some minor
2702 real issues in debug logging.
2704 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2705 assignment on my part. Fixed.
2707 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2708 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2709 Janne Snabb (who went above and beyond: thank you).
2711 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2712 string otherwise requires a connection and a bunch more work and it's
2713 relatively easy to get wrong. Should also expose TLS library linkage
2716 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2717 64-bit ${eval} (JH/03).
2719 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2720 GNU libc to support some of the 64-bit stuff, should not lead to
2721 conflicts. Defined before os.h is pulled in, so if a given platform
2722 needs to override this, it can.
2724 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2725 protection layer was required, which is not implemented.
2726 Bugzilla 1254, patch from Wolfgang Breyha.
2728 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2729 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2730 tls_dhparam take prime identifiers. Also unbreak combination of
2731 OpenSSL+DH_params+TLSSNI.
2733 PP/39 Disable SSLv2 by default in OpenSSL support.
2739 PP/01 Solaris build fix for Oracle's LDAP libraries.
2740 Bugzilla 1109, patch from Stephen Usher.
2742 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2744 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2747 TF/02 Fix a couple more cases where we did not log the error message
2748 when unlink() failed. See also change 4.74-TF/03.
2750 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2751 lock up or crash if it happened to be inside a call to libc when it
2752 got a SIGUSR1 from exiwhat.
2754 The SIGUSR1 handler appends the current process status to the process
2755 log which is later printed by exiwhat. It used to use the general
2756 purpose logging code to do this, but several functions it calls are
2757 not safe for signals.
2759 The new output code in the SIGUSR1 handler is specific to the process
2760 log, and simple enough that it's easy to inspect for signal safety.
2761 Removing some special cases also simplifies the general logging code.
2762 Removing the spurious timestamps from the process log simplifies
2765 TF/04 Improved ratelimit ACL condition.
2767 The /noupdate option has been deprecated in favour of /readonly which
2768 has clearer semantics. The /leaky, /strict, and /readonly update modes
2769 are mutually exclusive. The update mode is no longer included in the
2770 database key; it just determines when the database is updated. (This
2771 means that when you upgrade Exim will forget old rate measurements.)
2773 Exim now checks that the per_* options are used with an update mode that
2774 makes sense for the current ACL. For example, when Exim is processing a
2775 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2776 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2777 must specify per_mail/readonly. If you omit the update mode it defaults to
2778 /leaky where that makes sense (as before) or /readonly where required.
2780 The /noupdate option is now undocumented but still supported for
2781 backwards compatibility. It is equivalent to /readonly except that in
2782 ACLs where /readonly is required you may specify /leaky/noupdate or
2783 /strict/noupdate which are treated the same as /readonly.
2785 A useful new feature is the /count= option. This is a generalization
2786 of the per_byte option, so that you can measure the throughput of other
2787 aggregate values. For example, the per_byte option is now equivalent
2788 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2790 The per_rcpt option has been generalized using the /count= mechanism
2791 (though it's more complicated than the per_byte equivalence). When it is
2792 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2793 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2794 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2795 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2796 Note that using per_rcpt with a non-readonly update mode in more than
2797 one ACL will cause the recipients to be double-counted. (The per_mail
2798 and per_byte options don't have this problem.)
2800 The handling of very low rates has changed slightly. If the computed rate
2801 is less than the event's count (usually one) then this event is the first
2802 after a long gap. In this case the rate is set to the same as this event's
2803 count, so that the first message of a spam run is counted properly.
2805 The major new feature is a mechanism for counting the rate of unique
2806 events. The new per_addr option counts the number of different
2807 recipients that someone has sent messages to in the last time period. It
2808 behaves like per_rcpt if all the recipient addresses are different, but
2809 duplicate recipient addresses do not increase the measured rate. Like
2810 the /count= option this is a general mechanism, so the per_addr option
2811 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2812 example, measure the rate that a client uses different sender addresses
2813 with the options per_mail/unique=$sender_address. There are further
2814 details in the main documentation.
2816 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2818 TF/06 Removed a few PCRE remnants.
2820 TF/07 Automatically extract Exim's version number from tags in the git
2821 repository when doing development or release builds.
2823 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2824 Bugzilla 879. Patch from Paul Fisher.
2826 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2827 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2830 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2832 PP/05 Variable $av_failed, true if the AV scanner deferred.
2833 Bugzilla 1078. Patch from John Horne.
2835 PP/06 Stop make process more reliably on build failure.
2836 Bugzilla 1087. Patch from Heiko Schlittermann.
2838 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2839 Bugzilla 1089. Patch from Heiko Schlittermann.
2841 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2842 Bugzilla 1131. Patch from Holger Weiß.
2844 PP/09 Handle IPv6 addresses with SPF.
2845 Bugzilla 860. Patch from Wolfgang Breyha.
2847 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2849 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2852 PP/11 match_* no longer expand right-hand-side by default.
2853 New compile-time build option, EXPAND_LISTMATCH_RHS.
2854 New expansion conditions, "inlist", "inlisti".
2856 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2858 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2860 PP/14 fix log_write() format string regression from TF/03.
2861 Bugzilla 1152. Patch from Dmitry Isaikin.
2867 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2869 PP/02 Harmonised TLS library version reporting; only show if debugging.
2870 Layout now matches that introduced for other libraries in 4.74 PP/03.
2872 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2874 PP/04 New "dns_use_edns0" global option.
2876 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2879 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2880 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2882 TK/01 Updated PolarSSL code to 0.14.2.
2883 Bugzilla 1097. Patch from Andreas Metzler.
2885 PP/07 Catch divide-by-zero in ${eval:...}.
2886 Fixes bugzilla 1102.
2888 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2891 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2892 format-string attack -- SECURITY: remote arbitrary code execution.
2894 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2895 time unintentionally subject to list matching rules, letting the header
2896 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2897 arbitrary string expansion). This allowed for information disclosure.
2899 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2900 INT_MIN/-1 -- value coerced to INT_MAX.
2906 NM/01 Workaround for PCRE version dependency in version reporting
2909 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2910 This fixes portability to compilers other than gcc, notably
2911 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2913 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2914 makefiles for portability to HP-UX and POSIX correctness.
2916 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2917 Also via indented variable definition in the Makefile.
2918 (Debugging by Oliver Heesakkers).
2920 PP/02 Restore caching of spamd results with expanded spamd_address.
2921 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2923 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2924 Improves build reliability. Fix from: Frank Elsner
2926 NM/02 Fix wide character breakage in the rfc2047 coding
2927 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2929 NM/03 Allow underscore in dnslist lookups
2930 Fixes bug 1026. Patch from Graeme Fowler
2932 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2933 Code patches from Adam Ciarcinski of NetBSD.
2935 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2938 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2939 is logged, to avoid truncation. Patch from John Horne.
2941 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2942 Patch from Jakob Hirsch.
2944 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2945 SQL string expansion failure details.
2946 Patch from Andrey Oktyabrski.
2948 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2949 Patch from Simon Arlott.
2951 PP/09 New lookups functionality failed to compile on old gcc which rejects
2952 extern declarations in function scope.
2953 Patch from Oliver Fleischmann
2955 PP/10 Use sig_atomic_t for flags set from signal handlers.
2956 Check getgroups() return and improve debugging.
2957 Fixed developed for diagnosis in bug 927 (which turned out to be
2960 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2961 Patch from Mark Zealey.
2963 PP/12 Bugzilla 1056: Improved spamd server selection.
2964 Patch from Mark Zealey.
2966 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2967 Based on patch from Heiko Schlittermann.
2969 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2970 Patch from Uwe Doering, sign-off by Michael Haardt.
2972 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2973 variable declaration deep within a block. Bug and patch from
2976 PP/15 lookups-Makefile IRIX compatibility coercion.
2978 PP/16 Make DISABLE_DKIM build knob functional.
2980 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2981 Patch by Simon Arlott
2983 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2984 variable argument macros. Our copy now differs from upstream.
2990 TF/01 Failure to get a lock on a hints database can have serious
2991 consequences so log it to the panic log.
2993 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2994 controlled using the smtp_confirmation log selector.
2996 TF/03 Include the error message when we fail to unlink a spool file.
2998 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2999 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3000 for maintaining out-of-tree patches for some time.
3002 PP/01 Bugzilla 139: Documentation and portability issues.
3003 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3004 Handle per-OS dynamic-module compilation flags.
3006 PP/02 Let /dev/null have normal permissions.
3007 The 4.73 fixes were a little too stringent and complained about the
3008 permissions on /dev/null. Exempt it from some checks.
3009 Reported by Andreas M. Kirchwitz.
3011 PP/03 Report version information for many libraries, including
3012 Exim version information for dynamically loaded libraries. Created
3013 version.h, now support a version extension string for distributors
3014 who patch heavily. Dynamic module ABI change.
3016 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3017 privilege escalation vulnerability whereby the Exim run-time user
3018 can cause root to append content of the attacker's choosing to
3021 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3024 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3025 If dropping privileges for untrusted macros, we disabled normal logging
3026 on the basis that it would fail; for the Exim run-time user, this is not
3027 the case, and it resulted in successful deliveries going unlogged.
3028 Fixed. Reported by Andreas Metzler.
3034 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3035 only prepend for the Resent-* case. Fixes regression introduced in
3036 Exim 4.70 by NM/22 for Bugzilla 607.
3038 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3039 increasing numbers of administrators be bitten by this.
3041 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3043 PP/03 Bugzilla 994: added openssl_options main configuration option.
3045 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3047 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3049 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3051 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3052 without a peer certificate, leading to a segfault because of an
3053 assumption that peers always have certificates. Be a little more
3054 paranoid. Problem reported by Martin Tscholak.
3056 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3057 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3058 NB: ClamAV planning to remove STREAM in "middle of 2010".
3059 CL also introduces -bmalware, various -d+acl logging additions and
3060 more caution in buffer sizes.
3062 PP/09 Implemented reverse_ip expansion operator.
3064 PP/10 Bugzilla 937: provide a "debug" ACL control.
3066 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3068 PP/12 Bugzilla 973: Implement --version.
3070 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3072 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3074 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3076 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3077 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3078 ignore trailing whitespace.
3080 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3082 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3085 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3086 Notification from Dr Andrew Aitchison.
3088 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3089 ExtendedDetectionInfo response format.
3090 Notification from John Horne.
3092 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3095 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3096 XSL and documented dependency on system catalogs, with examples of how
3099 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3102 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3103 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3104 configuration file which is writeable by the Exim user or group.
3106 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3107 of configuration files to cover files specified with the -C option if
3108 they are going to be used with root privileges, not just the default
3111 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3112 option (effectively making it always true).
3114 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3115 files to be used while preserving root privileges.
3117 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3118 that rogue child processes cannot use them.
3120 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3121 run-time user, instead of root.
3123 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3124 Exim run-time user without dropping privileges.
3126 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3127 result string, instead of calling string_vformat() twice with the same
3130 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3131 for other users. Others should always drop root privileges if they use
3132 -C on the command line, even for a whitelisted configure file.
3134 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3136 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3142 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3143 $data_path, and $header_path variables; fixed documentation bugs and
3146 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3147 exipick to access non-standard spools, including the "frozen" queue
3150 NM/01 Bugzilla 965: Support mysql stored procedures.
3151 Patch from Alain Williams
3153 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3155 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3156 Patch from Andreas Metzler
3158 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3159 Patch from Kirill Miazine
3161 NM/05 Bugzilla 671: Added umask to procmail example.
3163 JJ/03 installed exipick 20100323.0, fixing doc bug
3165 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3166 directory. Notification and patch from Dan Rosenberg.
3168 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3170 TK/02 Improve log output when DKIM signing operation fails.
3172 MH/01 Treat the transport option dkim_domain as a colon separated
3173 list, not as a single string, and sign the message with each element,
3174 omitting multiple occurences of the same signer.
3176 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3177 Bugzilla 985, 986. Patch by Simon Arlott
3179 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3180 Patch by Simon Arlott
3182 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3183 MBX locking. Notification from Dan Rosenberg.
3189 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3191 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3193 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3195 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3197 NM/04 Bugzilla 915: Flag broken perl installation during build.
3203 TK/01 Added patch by Johannes Berg that expands the main option
3204 "spamd_address" if it starts with a dollar sign.
3206 TK/02 Write list of recipients to X-Envelope-Sender header when building
3207 the mbox-format spool file for content scanning (suggested by Jakob
3210 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3211 (http://www.dcc-servers.net/) support via dccifd. Activated by
3212 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3214 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3215 by Mark Daniel Reidel <mr@df.eu>.
3217 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3218 When building exim an external PCRE library is now needed -
3219 PCRE is a system library on the majority of modern systems.
3220 See entry on PCRE_LIBS in EDITME file.
3222 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3223 conversation. Added nologin parameter to request.
3224 Patch contributed by Kirill Miazine.
3226 TF/01 Do not log submission mode rewrites if they do not change the address.
3228 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3230 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3231 log files in place. Contributed by Roberto Lima.
3233 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3235 TF/03 Bugzilla 615: When checking the local_parts router precondition
3236 after a local_part_suffix or local_part_prefix option, Exim now
3237 does not use the address's named list lookup cache, since this
3238 contains cached lookups for the whole local part.
3240 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3241 Robert Millan. Documentation is in experimental-spec.txt.
3243 TF/04 Bugzilla 668: Fix parallel build (make -j).
3245 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3247 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3248 Patch provided by Jan Srzednicki.
3250 TF/05 Leading white space used to be stripped from $spam_report which
3251 wrecked the formatting. Now it is preserved.
3253 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3254 that they are available at delivery time.
3256 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3258 TF/08 TLS error reporting now respects the incoming_interface and
3259 incoming_port log selectors.
3261 TF/09 Produce a more useful error message if an SMTP transport's hosts
3262 setting expands to an empty string.
3264 NM/06 Bugzilla 744: EXPN did not work under TLS.
3265 Patch provided by Phil Pennock.
3267 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3268 Patch provided by Richard Godbee.
3270 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3271 acl_smtp_notquit, added index entry.
3273 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3274 Patch provided by Eugene Bujak.
3276 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3277 accept(). Patch provided by Maxim Dounin.
3279 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3280 Patch provided by Phil Pennock.
3282 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3284 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3285 Patch provided by Brad "anomie" Jorsch.
3287 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3288 Patch provided by Dean Brooks.
3290 TK/05 Add native DKIM support (does not depend on external libraries).
3292 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3293 Patch provided by Graeme Fowler.
3295 NM/16 Bugzilla 851: Documentation example syntax fix.
3297 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3299 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3302 NM/19 Bugzilla 745: TLS version reporting.
3303 Patch provided by Phil Pennock.
3305 NM/20 Bugzilla 167: bool: condition support.
3306 Patch provided by Phil Pennock.
3308 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3309 clients. Patch provided by Phil Pennock.
3311 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3312 Patch provided by Brad "anomie" Jorsch.
3314 NM/23 Bugzilla 687: Fix misparses in eximstats.
3315 Patch provided by Heiko Schlittermann.
3317 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3318 Patch provided by Heiko Schlittermann.
3320 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3321 plus update to original patch.
3323 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3325 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3326 Patch provided by David Brownlee.
3328 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3330 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3332 NM/30 Bugzilla 888: TLS documentation bugfixes.
3334 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3336 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3337 Unlike the original bugzilla I have changed all shell scripts in src tree.
3339 NM/33 Bugzilla 898: Transport filter timeout fix.
3340 Patch by Todd Rinaldo.
3342 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3343 Patch by Serge Demonchaux.
3345 NM/35 Bugzilla 39: Base64 decode bug fixes.
3346 Patch by Jakob Hirsch.
3348 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3350 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3352 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3354 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3360 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3361 ALT-N's libdkim that I have put here:
3362 http://duncanthrax.net/exim-experimental/
3364 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3365 were called 'true' and it seems that C99 defines that as a reserved
3366 keyword to be used with 'bool' variable types. That means you could
3367 not include C99-style headers which use bools without triggering
3368 build errors in sieve.c.
3370 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3371 as mailq or other aliases. Changed the --help handling significantly
3372 to do whats expected. exim_usage() emits usage/help information.
3374 SC/01 Added the -bylocaldomain option to eximstats.
3376 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3378 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3380 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3386 PH/01 Another patch from the Sieve maintainer.
3388 PH/02 When an IPv6 address is converted to a string for single-key lookup
3389 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3390 dots are used instead of colons so that keys in lsearch files need not
3391 contain colons. This was done some time before quoting was made available
3392 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3393 (notated using the quote facility) so as to distinguish them from IPv4
3394 keys. This meant that lookups for IP addresses in host lists did not work
3395 for iplsearch lookups.
3397 This has been fixed by arranging for IPv6 addresses to be expressed with
3398 colons if the lookup type is iplsearch. This is not incompatible, because
3399 previously such lookups could never work.
3401 The situation is now rather anomalous, since one *can* have colons in
3402 ordinary lsearch keys. However, making the change in all cases is
3403 incompatible and would probably break a number of configurations.
3405 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3408 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3409 conversion specification without a maximum field width, thereby enabling
3410 a rogue spamd server to cause a buffer overflow. While nobody in their
3411 right mind would setup Exim to query an untrusted spamd server, an
3412 attacker that gains access to a server running spamd could potentially
3413 exploit this vulnerability to run arbitrary code as the Exim user.
3415 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3416 $primary_hostname instead of what libspf2 thinks the hosts name is.
3418 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3419 a directory entry by the name of the lookup key. Previously, if a
3420 symlink pointed to a non-existing file or a file in a directory that
3421 Exim lacked permissions to read, a lookup for a key matching that
3422 symlink would fail. Now it is enough that a matching directory entry
3423 exists, symlink or not. (Bugzilla 503.)
3425 PH/03 The body_linecount and body_zerocount variables are now exported in the
3428 PH/04 Added the $dnslist_matched variable.
3430 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3431 This means they are set thereafter only if the connection becomes
3434 PH/06 Added the client_condition to authenticators so that some can be skipped
3435 by clients under certain conditions.
3437 PH/07 The error message for a badly-placed control=no_multiline_responses left
3438 "_responses" off the end of the name.
3440 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3442 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3443 (without spaces) instead of just copying the configuration text.
3445 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3447 PH/11 Added $max_received_linelength.
3449 PH/12 Added +ignore_defer and +include_defer to host lists.
3451 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3452 way in which PCRE > 7.0 is built.
3454 PH/14 Implemented queue_only_load_latch.
3456 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3457 MAIL command. The effect was to mangle the value on 64-bit systems.
3459 PH/16 Another patch from the Sieve maintainer.
3461 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3463 PH/18 If a system quota error occurred while trying to create the file for
3464 a maildir delivery, the message "Mailbox is full" was not appended to the
3465 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3466 only to a quota excession during the actual writing of the file.
3468 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3469 characters?) which causes problems in log lines. The DN values are now
3470 passed through string_printing() before being added to log lines.
3472 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3473 and InterBase are left for another time.)
3475 PH/21 Added message_body_newlines option.
3477 PH/22 Guard against possible overflow in moan_check_errorcopy().
3479 PH/23 POSIX allows open() to be a macro; guard against that.
3481 PH/24 If the recipient of an error message contained an @ in the local part
3482 (suitably quoted, of course), incorrect values were put in $domain and
3483 $local_part during the evaluation of errors_copy.
3489 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3490 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3493 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3494 issue a MAIL command.
3496 PH/02 In an ACL statement such as
3498 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3500 if a client was not listed at all, or was listed with a value other than
3501 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3502 the condition was not true (as it should be), so access was not denied.
3503 The bug was that the ! inversion was incorrectly passed on to the second
3504 item. This has been fixed.
3506 PH/03 Added additional dnslists conditions == and =& which are different from
3507 = and & when the dns lookup returns more than one IP address.
3509 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3510 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3512 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3513 FSYNC, which compiles an option called disable_fsync that allows for
3514 bypassing fsync(). The documentation is heavily laced with warnings.
3516 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3518 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3519 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3520 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3521 including adding "make clean"; (3) Added -fPIC when compiling the test
3522 dynamically loaded module, to get rid of a warning.
3524 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3525 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3526 = 0s. The bug is otherwise harmless.
3528 PH/07 There was a bug in the dovecot authenticator such that the value of
3529 $auth1 could be overwritten, and so not correctly preserved, after a
3530 successful authentication. This usually meant that the value preserved by
3531 the server_setid option was incorrect.
3533 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3535 PH/09 Installed PCRE release 7.0.
3537 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3538 run for batched SMTP input. It is now run at the start of every message
3539 in the batch. While fixing this I discovered that the process information
3540 (output by running exiwhat) was not always getting set for -bs and -bS
3541 input. This is fixed, and it now also says "batched" for BSMTP.
3543 PH/11 Added control=no_pipelining.
3545 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3546 patch, slightly modified), and move the expansion of helo_data till after
3547 the connection is made in the smtp transport (so it can use these
3550 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3552 PH/14 Added log_selector = +pid.
3554 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3556 PH/16 Add ${if forany and ${if forall.
3558 PH/17 Added dsn_from option to vary the From: line in DSNs.
3560 PH/18 Flush SMTP output before performing a callout, unless control =
3561 no_callout_flush is set.
3563 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3564 was true (the default) a successful delivery failed to delete the retry
3565 item, thus causing premature timeout of the address. The bug is now
3568 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3570 PH/21 Long custom messages for fakedefer and fakereject are now split up
3571 into multiline responses in the same way that messages for "deny" and
3572 other ACL rejections are.
3574 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3575 with slight modification.
3577 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3578 draft, changing the syntax and factoring some duplicate code".
3580 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3581 for deliveries of the second and subsequent messages over the same SMTP
3584 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3585 ${reduce, with only minor "tidies".
3587 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3589 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3590 expansion side effects.
3592 PH/27 When a message times out after an over-quota error from an Exim-imposed
3593 quota, the bounce message says "mailbox is full". This message was not
3594 being given when it was a system quota that was exceeded. It now should
3597 MH/03 Made $recipients available in local_scan(). local_scan() already has
3598 better access to the recipient list through recipients_list[], but
3599 $recipients can be useful in postmaster-provided expansion strings.
3601 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3602 in the case of a MAIL command with additional options following the
3603 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3604 were accidentally chopped off.
3606 PH/29 SMTP synchronization checks are implemented when a command is read -
3607 there is a check that no more input is waiting when there shouldn't be
3608 any. However, for some commands, a delay in an ACL can mean that it is
3609 some time before the response is written. In this time, more input might
3610 arrive, invalidly. So now there are extra checks after an ACL has run for
3611 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3612 pipelining has not been advertised.
3614 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3616 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3617 returned in the SMTP response when smtp_return_error_details was false.
3618 This has been fixed.
3620 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3621 instead of the C I/O that was originally supplied, because problems were
3622 reported on Solaris.
3624 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3625 Exim which did not show up earlier: it was assuming that a call to
3626 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3627 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3628 was a macro that became an assignment, so it seemed to work. This has
3629 changed to a proper function call with a void return, hence the compile
3630 error. Exim's code has been fixed.
3632 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3635 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3637 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3639 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3640 because it turns out to be too big or there is a timeout) there is no
3641 "Completed" line in the log. When some messages of this type were
3642 selected by exigrep, they were listed as "not completed". Others were
3643 picked up by some special patterns. I have improved the selection
3644 criteria to be more general.
3646 PH/38 The host_find_failed option in the manualroute router can now be set
3647 to "ignore", to completely ignore a host whose IP address cannot be
3648 found. If all hosts are ignored, the behaviour is controlled by the new
3649 host_all_ignored option.
3651 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3652 homing or because of multiple MX records with /mx) generated more than
3653 one IP address, and the following item turned out to be the local host,
3654 all the secondary addresses of the first item were incorrectly removed
3655 from the list, along with the local host and any following hosts (which
3656 is what is supposed to happen).
3658 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3659 whoever called Exim into the -H file. In the case of the daemon it was
3660 behaving confusingly. When first started, it used values for whoever
3661 started the daemon, but after a SIGHUP it used the Exim user (because it
3662 calls itself on a restart). I have changed the code so that it now always
3665 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3666 message are rejected with the same error (e.g. no authentication or bad
3667 sender address), and a DATA command is nevertheless sent (as can happen
3668 with PIPELINING or a stupid MUA), the error message that was given to the
3669 RCPT commands is included in the rejection of the DATA command. This is
3670 intended to be helpful for MUAs that show only the final error to their
3673 PH/42 Another patch from the Sieve maintainer.
3675 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3676 Eximstats - Fixed some broken HTML links and added missing column headers
3678 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3679 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3681 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3683 PH/43 Yet another patch from the Sieve maintainer.
3685 PH/44 I found a way to check for a TCP/IP connection going away before sending
3686 the response to the final '.' that terminates a message, but only in the
3687 case where the client has not sent further data following the '.'
3688 (unfortunately, this is allowed). However, in many cases there won't be
3689 any further data because there won't be any more messages to send. A call
3690 to select() can be used: if it shows that the input is "ready", there is
3691 either input waiting, or the socket has been closed. An attempt to read
3692 the next input character can distinguish the two cases. Previously, Exim
3693 would have sent an OK response which the client would never have see.
3694 This could lead to message repetition. This fix should cure that, at
3695 least in a lot of common cases.
3697 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3698 advertised in response to EHLO.
3704 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3705 fixed by 4.65/MH/01 (is this a record?) are fixed:
3707 (i) An empty string was always treated as zero by the numeric comparison
3708 operators. This behaviour has been restored.
3710 (ii) It is documented that the numeric comparison operators always treat
3711 their arguments as decimal numbers. This was broken in that numbers
3712 starting with 0 were being interpreted as octal.
3714 While fixing these problems I realized that there was another issue that
3715 hadn't been noticed. Values of message_size_limit (both the global option
3716 and the transport option) were treated as octal if they started with 0.
3717 The documentation was vague. These values are now always treated as
3718 decimal, and I will make that clear in the documentation.
3724 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3725 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3728 MH/01 Don't check that the operands of numeric comparison operators are
3729 integers when their expansion is in "skipping" mode (fixes bug
3730 introduced by 4.64-PH/07).
3732 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3733 child addresses, Exim now panics and dies. Previously, because the count
3734 is held in a short int, deliveries were likely to be lost. As such a
3735 large number of recipients for a single message is ridiculous
3736 (performance will be very, very poor), I have chosen to impose a limit
3737 rather than extend the field.
3743 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3744 leftover -K file (the existence of which was triggered by #402).
3745 While we were at it, introduced process PID as part of the -K
3746 filename. This should rule out race conditions when creating
3749 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3750 processing considerably. Previous code took too long for large mails,
3751 triggering a timeout which in turn triggers #401.
3753 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3754 in the DK code in transports.c. sendfile() is not really portable,
3755 hence the _LINUX specificness.
3757 TF/01 In the add_headers option to the mail command in an Exim filter,
3758 there was a bug that Exim would claim a syntax error in any
3759 header after the first one which had an odd number of characters
3762 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3763 callout verification, Exim cached a "reject" for the entire domain. This
3764 is correct for most verifications, but it is not correct for a recipient
3765 verification with use_sender or use_postmaster set, because in that case
3766 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3767 case of MAIL FROM:<> rejection from other early rejections (e.g.
3768 rejection of HELO). When verifying a recipient using a non-null MAIL
3769 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3770 Whatever the result of the callout, the value of the domain cache is
3771 left unchanged (for any other kind of callout, getting as far as trying
3772 RCPT means that the domain itself is ok).
3774 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3777 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3778 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3779 session, the daemon ignores ECONNECT errors and logs others; it now
3780 ignores EPIPE as well.
3782 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3783 (quoted-printable decoding).
3785 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3786 later the small subsequent patch to fix an introduced bug.
3788 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3790 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3792 PH/08 An error is now given if message_size_limit is specified negative.
3794 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3795 to be given (somewhat) arbitrary names.
3797 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3800 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3801 miscellaneous code fixes
3803 PH/10 Added the log_reject_target ACL modifier to specify where to log
3806 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3807 hostname. This is wrong, because it relates to the incoming message (and
3808 probably the interface on which it is arriving) and not to the outgoing
3809 callout (which could be using a different interface). This has been
3810 changed to use the value of the helo_data option from the smtp transport
3811 instead - this is what is used when a message is actually being sent. If
3812 there is no remote transport (possible with a router that sets up host
3813 addresses), $smtp_active_hostname is used.
3815 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3816 tweaks were necessary in order to get it to work (see also 21 below):
3817 (a) The code assumed that strncpy() returns a negative number on buffer
3818 overflow, which isn't the case. Replaced with Exim's string_format()
3820 (b) There were several signed/unsigned issues. I just did the minimum
3821 hacking in of casts. There is scope for a larger refactoring.
3822 (c) The code used strcasecmp() which is not a standard C function.
3823 Replaced with Exim's strcmpic() function.
3824 (d) The code set only $1; it now sets $auth1 as well.
3825 (e) A simple test gave the error "authentication client didn't specify
3826 service in request". It would seem that Dovecot has changed its
3827 interface. Fortunately there's a specification; I followed it and
3828 changed what the client sends and it appears to be working now.
3830 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3833 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3834 address A is aliased to B and C, where B exists and C does not. Without
3835 -v the output is "A verified" because verification stops after a
3836 successful redirection if more than one address is generated. However,
3837 with -v the child addresses are also verified. Exim was outputting "A
3838 failed to verify" and then showing the successful verification for C,
3839 with its parentage. It now outputs "B failed to verify", showing B's
3840 parentage before showing the successful verification of C.
3842 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3843 look up a TXT record in a specific list after matching in a combined
3846 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3847 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3848 they consult the DNS. I had assumed they would set it the way they
3849 wanted; and indeed my experiments on Linux seem to show that in some
3850 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3851 To be on the safe side, however, I have now made the interface to
3852 host_find_byname() similar to host_find_bydns(), with an argument
3853 containing the DNS resolver options. The host_find_byname() function now
3854 sets these options at its start, just as host_find_bydns() does. The smtp
3855 transport options dns_qualify_single and dns_search_parents are passed to
3856 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3857 of host_find_byname() use the default settings of RES_DEFNAMES
3858 (qualify_single) but not RES_DNSRCH (search_parents).
3860 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3861 spool_read_header() do less string testing, by means of a preliminary
3862 switch on the second character of optional "-foo" lines. (This is
3863 overdue, caused by the large number of possibilities that now exist.
3864 Originally there were few.) While I was there, I also converted the
3865 str(n)cmp tests so they don't re-test the leading "-" and the first
3866 character, in the hope this might squeeze out yet more improvement.
3868 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3869 flag allowing group syntax was set by the header_syntax check but not
3870 turned off, possible causing trouble later; (2) The flag was not being
3871 set at all for the header_verify test, causing "group"-style headers to
3872 be rejected. I have now set it in this case, and also caused header_
3873 verify to ignore an empty address taken from a group. While doing this, I
3874 came across some other cases where the code for allowing group syntax
3875 while scanning a header line wasn't quite right (mostly, not resetting
3876 the flag correctly in the right place). These bugs could have caused
3877 trouble for malformed header lines. I hope it is now all correct.
3879 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3880 with the "reply" argument non-NULL. The code, however (which originally
3881 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3882 but it didn't always do it. This confused somebody who was copying the
3883 code for some other use. I have removed all the tests.
3885 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3886 feature that was used to support insecure browsers during the U.S. crypto
3887 embargo. It requires special client support, and Exim is probably the
3888 only MTA that supported it -- and would never use it because real RSA is
3889 always available. This code has been removed, because it had the bad
3890 effect of slowing Exim down by computing (never used) parameters for the
3891 RSA_EXPORT functionality.
3893 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3894 authenticator to fail if there's a tab character in the incoming data
3895 (there should never be unless someone is messing about, as it's supposed
3896 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3897 if the connection is using TLS or if the remote IP is the same as the
3898 local IP, and the "valid-client-cert option" if a client certificate has
3901 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3902 authenticators. This can be used for authorization after authentication
3903 succeeds. (In the case of plaintext, it servers for both authentication
3906 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3907 if any retry times were supplied.
3909 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3910 connection, where there is no client IP address to check. In this
3911 situation, the verify now always succeeds.
3913 PH/25 Applied John Jetmore's -Mset patch.
3915 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3917 PH/27 In a string expansion for a processed (not raw) header when multiple
3918 headers of the same name were present, leading whitespace was being
3919 removed from all of them, but trailing whitespace was being removed only
3920 from the last one. Now trailing whitespace is removed from each header
3921 before concatenation. Completely empty headers in a concatenation (as
3922 before) are ignored.
3924 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3925 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3927 PH/29 [Removed. This was a change that I later backed out, and forgot to
3928 correct the ChangeLog entry (that I had efficiently created) before
3929 committing the later change.]
3931 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3932 address errors (4xx response to RCPT) over the same connection as other
3933 messages routed to the same hosts. Such deliveries are always "forced",
3934 so retry times are not inspected. This resulted in far too many retries
3935 for the affected addresses. The effect occurred only when there were more
3936 hosts than the hosts_max_try setting in the smtp transport when it had
3937 the 4xx errors. Those hosts that it had tried were not added to the list
3938 of hosts for which the message was waiting, so if all were tried, there
3939 was no problem. Two fixes have been applied:
3941 (i) If there are any address or message errors in an SMTP delivery, none
3942 of the hosts (tried or untried) are now added to the list of hosts
3943 for which the message is waiting, so the message should not be a
3944 candidate for sending over the same connection that was used for a
3945 successful delivery of some other message. This seems entirely
3946 reasonable: after all the message is NOT "waiting for some host".
3947 This is so "obvious" that I'm not sure why it wasn't done
3948 previously. Hope I haven't missed anything, but it can't do any
3949 harm, as the worst effect is to miss an optimization.
3951 (ii) If, despite (i), such a delivery is accidentally attempted, the
3952 routing retry time is respected, so at least it doesn't keep
3953 hammering the server.
3955 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3956 in ${readsocket because some servers need this prod.
3958 PH/32 Added some extra debug output when updating a wait-xxx database.
3960 PH/33 The hint "could be header name not terminated by colon", which has been
3961 given for certain expansion errors for a long time, was not being given
3962 for the ${if def:h_colon_omitted{... case.
3964 PH/34 The spec says: "With one important exception, whenever a domain list is
3965 being scanned, $domain contains the subject domain." There was at least
3966 one case where this was not true.
3968 PH/35 The error "getsockname() failed: connection reset by peer" was being
3969 written to the panic log as well as the main log, but it isn't really
3970 panic-worthy as it just means the connection died rather early on. I have
3971 removed the panic log writing for the ECONNRESET error when getsockname()
3974 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3975 runs only) independently of the message's sender address. This meant
3976 that, if the 4xx error was in fact related to the sender, a different
3977 message to the same recipient with a different sender could confuse
3978 things. In particular, this can happen when sending to a greylisting
3979 server, but other circumstances could also provoke similar problems.
3980 I have changed the default so that the retry time for these errors is now
3981 based a combination of the sender and recipient addresses. This change
3982 can be overridden by setting address_retry_include_sender=false in the
3985 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3986 remote server are returned as part of bounce messages. This was not
3987 happening for LMTP over a pipe (the lmtp transport), but now it is the
3988 same for both kinds of LMTP.
3990 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3991 in header lines that were in fact CNAMEs. This is no longer the case.
3993 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3994 and queue runs started by the daemon processed all messages. This has
3995 been fixed so that -R and -S can now usefully be given with -q<time>.
3997 PH/40 Import PCRE release 6.7 (fixes some bugs).
3999 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4001 PH/42 Give an error if -q is specified more than once.
4003 PH/43 Renamed the variables $interface_address and $interface_port as
4004 $received_ip_address and $received_port, to make it clear that these
4005 values apply to message reception, and not to the outgoing interface when
4006 a message is delivered. (The old names remain recognized, of course.)
4008 PH/44 There was no timeout on the connect() call when using a Unix domain
4009 socket in the ${readsocket expansion. There now is.
4011 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4012 be meaningful with "accept".
4014 SC/01 Eximstats V1.43
4015 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4017 SC/02 Eximstats V1.44
4018 Use a glob alias rather than an array ref in the generated
4019 parser. This improves both readability and performance.
4021 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4022 Collect SpamAssassin and rejection statistics.
4023 Don't display local sender or destination tables unless
4024 there is data to show.
4025 Added average volumes into the top table text output.
4027 SC/04 Eximstats V1.46
4028 Collect data on the number of addresses (recipients)
4029 as well as the number of messages.
4031 SC/05 Eximstats V1.47
4032 Added 'Message too big' to the list of mail rejection
4033 reasons (thanks to Marco Gaiarin).
4035 SC/06 Eximstats V1.48
4036 Mainlog lines which have GMT offsets and are too short to
4037 have a flag are now skipped.
4039 SC/07 Eximstats V1.49 (Alain Williams)
4040 Added the -emptyok flag.
4042 SC/08 Eximstats V1.50
4043 Fixes for obtaining the IP address from reject messages.
4045 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4046 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4047 whitespace changes from 4.64-PH/27
4049 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4052 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4053 are found, allow negative numbers in numeric criteria)
4055 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4057 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4058 to match changes made in 4.64-PH/43
4060 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4062 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4063 "domain" is actually an IP address, and force a failure. This locks out
4064 those revolvers/nameservers that support "A-for-A" lookups, in
4065 contravention of the specifications.
4067 PH/48 When a host name was looked up from an IP address, and the subsequent
4068 forward lookup of the name timed out, the host name was left in
4069 $sender_host_name, contrary to the specification.
4071 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4072 restricted to single-key lookups, Exim was not diagnosing an error if
4073 * or *@ was used with a query-style lookup.
4075 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4077 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4078 long ago, but noone interested enough thought of it. Let's just say that
4079 the "1.1" means that there are some new functions that weren't there at
4080 some point in the past.
4082 PH/51 Error processing for expansion failure of helo_data from an smtp
4083 transport during callout processing was broken.
4085 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4086 tested/used via the -bh/-bhc/-bs options.
4088 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4089 bug, fixed in subsequent PCRE releases).
4091 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4092 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4094 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4100 SC/01 Use a glob alias rather than an array ref in eximstats generated
4101 parser. This improves both readability and performance.
4103 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4104 Don't display local sender or destination tables in eximstats unless
4105 there is data to show.
4106 Added average volumes into the eximstats top table text output.
4108 SC/03 Collect data on the number of addresses (recipients) as well
4109 as the number of messages in eximstats.
4111 TF/01 Correct an error in the documentation for the redirect router. Exim
4112 does (usually) call initgroups() when daemonizing.
4114 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4115 with consistent privilege compared to when running as a daemon.
4117 TF/03 Note in the spec that $authenticated_id is not set for local
4118 submissions from trusted users.
4120 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4121 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4123 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4124 by adding some example configuration directives to the default
4125 configuration file. A little bit of work is required to uncomment the
4126 directives and define how usernames and passwords are checked, but
4127 there is now a framework to start from.
4129 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4130 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4131 without this. I don't know how relevant this is to other LDAP libraries.
4133 PH/02 Add the verb name to the "unknown ACL verb" error.
4135 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4137 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4139 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4140 directory not expanded when it should be if an expanded home directory
4141 was set for the address (which is overridden by the transport).
4143 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4146 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4147 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4148 because it is too late at that time, and has no effect.
4150 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4151 security issue with \' (bugzilla #107). I could not use the
4152 PQescapeStringConn() function, because it needs a PGconn value as one of
4155 PH/08 When testing addresses using -bt, indicate those final addresses that
4156 are duplicates that would not cause an additional delivery. At least one
4157 person was confused, thinking that -bt output corresponded to deliveries.
4158 (Suppressing duplicates isn't a good idea as you lose the information
4159 about possibly different redirections that led to the duplicates.)
4161 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4162 systems where poll() doesn't work, in particular OS X.
4164 PH/10 Added more information to debugging output for retry time not reached.
4166 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4167 operations in malware.c.
4169 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4172 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4173 syslog_duplication set false, log lines that would normally be written
4174 both the the main log and to the reject log were not written to syslog at
4177 PH/14 In the default configuration, change the use of "message" in ACL warn
4178 statements to "add_header".
4180 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4181 not followed by a command (e.g. "seen endif").
4183 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4184 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4187 PH/17 Added extra conditions to the default value of delay_warning_condition
4191 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4192 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4193 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4196 The Auto-Submitted: and various List- headers are standardised, whereas I
4197 don't think Precedence: ever was.
4199 PH/18 Refactored debugging code in route_finduser() to show more information,
4200 in particular, the error code if getpwnam() issues one.
4202 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4203 This is apparently needed in addition to the PH/07 change above to avoid
4204 any possible encoding problems.
4206 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4207 but not after initializing Perl.
4209 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4210 output them only if debugging. By default they are written stderr,
4211 apparently, which is not desirable.
4213 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4216 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4219 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4221 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4222 authenticated or an ident call has been made. Suppress the default
4223 values for $authenticated_id and $authenticated_sender (but permit -oMai
4224 and -oMas) when testing with -bh.
4226 PH/24 Re-jigged the order of the tests in the default configuration so that the
4227 tests for valid domains and recipients precede the DNS black list and CSA
4228 tests, on the grounds that those ones are more expensive.
4230 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4231 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4232 This bug exists in every version of Exim that I still have, right back to
4235 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4236 However, an attempt to turn on case-sensitivity in a regex key by
4237 including (?-i) didn't work because the subject string was already
4238 lowercased, and the effects were non-intuitive. It turns out that a
4239 one-line patch can be used to allow (?-i) to work as expected.
4245 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4246 other effects) broke the use of negated acl sub-conditions.
4248 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4251 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4252 "Deny" causes Exim to reject the incoming connection with a 554 error.
4253 Unfortunately, if there is a major crisis, such as a disk failure,
4254 tcp-wrappers gives "deny", whereas what one would like would be some
4255 kind of temporary error. A kludge has been added to help with this.
4256 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4257 554 error is used if errno is still zero or contains ENOENT (which occurs
4258 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4261 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4263 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4264 errors. Otherwise a message that provokes a temporary error (when other
4265 messages do not) can cause a whole host to time out.
4267 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4268 addresses were routed directly to files or pipes from a redirect router.
4269 File deliveries just didn't batch; pipe deliveries might have suffered
4272 PH/06 A failure to get a lock for a hints database would erroneously always say
4273 "Failed to get write lock", even when it was really a read lock.
4275 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4276 of 0600. This has been changed to use the value of the lockfile_mode
4277 option (which defaults to 0600).
4279 PH/08 Applied small patch from the Sieve maintainer.
4281 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4282 folder from quota calculations, a direct delivery into this folder messed
4283 up the contents of the maildirsize file. This was because the regex was
4284 used only to exclude .Trash (or whatever) when the size of the mailbox
4285 was calculated. There was no check that a delivery was happening into an
4286 excluded directory. This bug has been fixed by ignoring all quota
4287 processing for deliveries into excluded directories.
4289 PH/10 Added the maildirfolder_create_regex option to appendfile.
4295 PH/01 The code for finding all the local interface addresses on a FreeBSD
4296 system running IPv6 was broken. This may well have applied to all BSD
4297 systems, as well as to others that have similar system calls. The broken
4298 code found IPv4 interfaces correctly, but gave incorrect values for the
4299 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4300 that it would not match correctly against @[] and not recognize the IPv6
4303 PH/02 The ipliteral router was not recognizing addresses of the form user@
4304 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4306 PH/03 Added disable_ipv6.
4308 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4309 decoded form, because it is most often used to construct To: headers
4310 lines in autoreplies, and the decoded form may well be syntactically
4311 invalid. However, $reply_address has leading white space removed, and all
4312 newlines turned into spaces so that the autoreply transport does not
4315 PH/05 If group was specified without a user on a router, and no group or user
4316 was specified on a transport, the group from the router was ignored.
4318 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4319 for visible compile-time settings that can be used to change these
4320 numbers, for those that want even more. Backwards compatibility with old
4321 spool files has been maintained. However, going back to a previous Exim
4322 release will lost any variables that are in spool files.
4324 PH/07 Two small changes when running in the test harness: increase delay when
4325 passing a TCP/IP connection to a new process, in case the original
4326 process has to generate a bounce, and remove special handling of
4327 127.0.0.2 (sic), which is no longer necessary.
4329 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4330 be the same on different OS.
4332 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4335 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4336 whether --show-vars was specified or not
4338 JJ/02 exipick: Added support for new ACL variable spool format introduced
4341 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4342 syntactically invalid From: or Reply-to: line, and a filter used this to
4343 generate an autoreply, and therefore failed to obtain an address for the
4344 autoreply, Exim could try to deliver to a non-existent relative file
4345 name, causing unrelated and misleading errors. What now happens is that
4346 it logs this as a hard delivery error, but does not attempt to create a
4349 PH/11 The exinext utility has a -C option for testing purposes, but although
4350 the given file was scanned by exinext itself; it wasn't being passed on
4351 when Exim was called.
4353 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4354 an end-of-file indication when reading a command response.
4356 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4357 compiled. In many other places in Exim, IPv6 addresses are always
4358 recognized, so I have changed this. It also means that IPv4 domain
4359 literals of the form [IPV4:n.n.n.n] are now always recognized.
4361 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4362 used if the router is not running as root, for example, when verifying at
4363 ACL time, or when using -bh. The debugging output from this situation was
4364 non-existent - all you got was a failure to exec. I have made two
4367 (a) Failures to set uid/gid, the current directory, or a process leader
4368 in a subprocess such as that created by queryprogram now generate
4369 suitable debugging output when -d is set.
4371 (b) The queryprogram router detects when it is not running as root,
4372 outputs suitable debugging information if -d is set, and then runs
4373 the subprocess without attempting to change uid/gid.
4375 PH/15 Minor change to Makefile for building test_host (undocumented testing
4378 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4379 additional section of a DNS packet that returns MX or SRV records.
4380 Instead, it always explicitly searches for A/AAAA records. This avoids
4381 major problems that occur when a DNS server includes only records of one
4382 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4383 fixed another bug: if SRV records were looked up and the corresponding
4384 address records were *not* found in the additional section, the port
4385 values from the SRV records were lost.
4387 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4388 using the correct key (the original address) when searching the retry
4389 rules in order to find which one to use for generating the retry hint.
4391 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4392 adding the default one. Similarly, if it contains a Reply-To: header, the
4393 errors_reply_to option, if set, is not used.
4395 PH/19 When calculating a retry time, Exim used to measure the "time since
4396 failure" by looking at the "first failed" field in the retry record. Now
4397 it does not use this if it is later than than the arrival time of the
4398 message. Instead it uses the arrival time. This makes for better
4399 behaviour in cases where some deliveries succeed, thus re-setting the
4400 "first failed" field. An example is a quota failure for a huge message
4401 when small messages continue to be delivered. Without this change, the
4402 "time since failure" will always be short, possible causing more frequent
4403 delivery attempts for the huge message than are intended.
4404 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4406 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4407 $1, $2, $3) because the numerical variables can be reset during some
4408 expansion items (e.g. "match"), thereby losing the authentication data.
4410 PH/21 Make -bV show the size of off_t variables so that the test suite can
4411 decide whether to run tests for quotas > 2G.
4413 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4414 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4415 filecount value is greater than 2G or if a quota value is greater than 2G
4416 on a system where the size of off_t is not greater than 4, a panic error
4419 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4420 never match. The debug and -bh output now contains an explicit error
4421 message indicating a malformed IPv4 address or mask.
4423 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4424 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4425 PH/23 above applies.
4427 PH/25 Do not write to syslog when running in the test harness. The only
4428 occasion when this arises is a failure to open the main or panic logs
4429 (for which there is an explicit test).
4431 PH/26 Added the /no_tell option to "control=freeze".
4433 PH/27 If a host name lookup failed very early in a connection, for example, if
4434 the IP address matched host_lookup and the reverse lookup yielded a name
4435 that did not have a forward lookup, an error message of the form "no IP
4436 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4437 could be logged. Now it outputs the IP address instead of "NULL".
4439 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4440 allows the sender and the authenticated sender to be set when
4441 submitting a message from within Exim. Since child_open_exim() is
4442 documented for local_scan(), the new function should be too.
4444 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4445 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4446 results in an empty string is now treated as unset.
4448 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4450 PH/31 Added sender_verify_fail logging option.
4452 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4453 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4454 tidied the source and removed it altogether.
4456 PH/33 When a queue run was abandoned because the load average was too high, a
4457 log line was always written; now it is written only if the queue_run log
4458 selector is set. In addition, the log line for abandonment now contains
4459 information about the queue run such as the pid. This is always present
4460 in "start" and "stop" lines but was omitted from the "abandon" line.
4462 PH/34 Omit spaces between a header name and the colon in the error message that
4463 is given when verify = headers_syntax fails (if there are lots of them,
4464 the message gets confusing).
4466 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4467 names, as there are now some PTR records that contain slashes. This check
4468 is only to protect against broken name servers that fall over on strange
4469 characters, so the fact that it applies to all lookups doesn't matter.
4471 PH/36 Now that the new test suite is complete, we can remove some of the
4472 special code in Exim that was needed for the old test suite. For example,
4473 sorting DNS records because real resolvers return them in an arbitrary
4474 order. The new test suite's fake resolver always returns records in the
4477 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4478 bounces) except when queue_only is set, to avoid logging races between
4479 the different processes.
4481 PH/38 Panic-die if .include specifies a non-absolute path.
4483 PH/39 A tweak to the "H" retry rule from its user.
4485 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4486 a label. They prevented compilation on older perls.
4488 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4489 a warning to be raised on newish perls.
4491 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4492 on queue. Changes to match documented behaviour of showing count of
4493 messages matching specified criteria.
4495 PH/40 Changed the default ident timeout from 30s to 5s.
4497 PH/41 Added support for the use of login_cap features, on those BSD systems
4498 that have them, for controlling the resources used by pipe deliveries.
4500 PH/42 The content-scanning code uses fopen() to create files in which to put
4501 message data. Previously it was not paying any attention to the mode of
4502 the files. Exim runs with umask(0) because the rest of the code creates
4503 files with open(), and sets the required mode explicitly. Thus, these
4504 files were ending up world-writeable. This was not a big issue, because,
4505 being within the spool directory, they were not world-accessible. I have
4506 created a function called modefopen, which takes an additional mode
4507 argument. It sets umask(777), creates the file, chmods it to the required
4508 mode, then resets the umask. All the relevant calls to fopen() in the
4509 content scanning code have been changed to use this function.
4511 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4512 to 24 hours. This avoids potential overflow problems when processing G
4513 and H retry rules. I suspect nobody ever tinkers with this value.
4515 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4517 PH/45 When the plaintext authenticator is running as a client, the server's
4518 challenges are checked to ensure they are valid base64 strings. By
4519 default, the authentication attempt is cancelled if an invalid string is
4520 received. Setting client_ignore_invalid_base64 true ignores these errors.
4521 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4522 they are received. Thus, the responses can be made to depend on the
4523 challenges. If an invalid string is ignored, an empty string is placed in
4526 PH/46 Messages that are created by the autoreply transport now contains a
4527 References: header, in accordance with RFCs 2822 and 3834.
4529 PH/47 Added authenticated_sender_force to the smtp transport.
4531 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4533 PH/49 Installed latest patch from the Sieve maintainer.
4535 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4536 was also set, the appendfile transport was unnecessarily scanning a
4537 directory of message files (e.g. for maildir delivery) to find the count
4538 of files (along with the size), even though it did not need this
4539 information. It now does the scan only if it needs to find either the
4540 size of the count of files.
4542 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4544 PH/52 Two bugs concerned with error handling when the smtp transport is
4547 (i) Exim was not creating retry information for temporary errors given
4548 for individual recipients after the DATA command when the smtp transport
4549 was used in LMTP mode. This meant that they could be retried too
4550 frequently, and not timed out correctly.
4552 (ii) Exim was setting the flag that allows error details to be returned
4553 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4554 recipients that were returned after the DATA command.
4556 PH/53 This is related to PH/52, but is more general: for any failing address,
4557 when detailed error information was permitted to be returned to the
4558 sender, but the error was temporary, then after the final timeout, only
4559 "retry timeout exceeded" was returned. Now it returns the full error as
4560 well as "retry timeout exceeded".
4562 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4563 do this, and (what is worse) MTAs that accept it.
4565 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4566 will now be deprecated.
4568 PH/56 New os.c-cygwin from the Cygwin maintainer.
4570 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4571 formats (previously only available in exim formats via -bpr, -bpru,
4572 and -bpra. Now also available in native and exiqgrep formats)
4574 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4575 with very large, slow to parse queues
4577 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4579 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4581 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4582 responses to authentication challenges, though it was showing the
4583 challenges; (ii) I've removed the CR characters from the debug output for
4586 PH/58 Allow for the insertion of a newline as well as a space when a string
4587 is turned into more than one encoded-word during RFC 2047 encoding. The
4588 Sieve code now uses this.
4590 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4591 data_4xx, lost_connection, tls_required.
4593 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4594 message was being sent as an SMTP response.
4596 PH/61 Add -l and -k options to exicyclog.
4598 PH/62 When verifying, if an address was redirected to one new address, so that
4599 verification continued, and the new address failed or deferred after
4600 having set something in $address_data, the value of $address_data was not
4601 passed back to the ACL. This was different to the case when no
4602 redirection occurred. The value is now passed back in both cases.
4604 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4605 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4606 use login_cap.h, so on its own it isn't the distinguishing feature. The
4607 new name refers directly to the setclassresources() function.
4609 PH/65 Added configuration files for NetBSD3.
4611 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4613 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4614 is preferred over IPv4.
4616 PH/68 The bounce_return_message and bounce_return_body options were not being
4617 honoured for bounces generated during the reception of non-SMTP messages.
4618 In particular, this applied to messages rejected by the ACL. This bug has
4619 been fixed. However, if bounce_return_message is true and bounce_return_
4620 body is false, the headers that are returned for a non-SMTP message
4621 include only those that have been read before the error was detected.
4622 (In the case of an ACL rejection, they have all been read.)
4624 PH/69 The HTML version of the specification is now built in a directory called
4625 spec_html instead of spec.html, because the latter looks like a path with
4626 a MIME-type, and this confuses some software.
4628 PH/70 Catch two compiler warnings in sieve.c.
4630 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4631 function verify_get_ident() calls ip_connect() to connect a socket, but
4632 if the "connect()" function timed out, ip_connect() used to close the
4633 socket. However, verify_get_ident() also closes the socket later, and in
4634 between Exim writes to the log, which may get opened at this point. When
4635 the socket was closed in ip_connect(), the log could get the same file
4636 descriptor number as the socket. This naturally causes chaos. The fix is
4637 not to close the socket in ip_connect(); the socket should be closed by
4638 the function that creates it. There was only one place in the code where
4639 this was missing, in the iplookup router, which I don't think anybody now
4640 uses, but I've fixed it anyway.
4642 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4643 well as to direct DNS lookups. Otherwise the handling of names in host
4644 lists is inconsistent and therefore confusing.
4650 PH/01 Two changes to the default runtime configuration:
4652 (1) Move the checks for relay_from_hosts and authenticated clients from
4653 after to before the (commented out) DNS black list checks.
4655 (2) Add control=submission to the relay_from_hosts and authenticated
4656 clients checks, on the grounds that messages accepted by these
4657 statements are most likely to be submissions.
4659 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4661 (1) Generate an error if the third argument for the ${prvs expansion is
4664 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4667 (3) Reset the variables that are obtained from the first argument of
4668 ${prvscheck and used in the second argument before leaving the code,
4669 because their memory is reclaimed, so using them afterwards may do
4672 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4673 one (it's much easier than Tom thought :-).
4675 (5) Because of (4), we can now allow for the use of $prvscheck_result
4676 inside the third argument.
4678 PH/03 For some reason, the default setting of PATH when running a command from
4679 a pipe transport was just "/usr/bin". I have changed it to
4682 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4683 anything to be listed in the output from -bV.
4685 PH/05 When a filter generated an autoreply, the entire To: header line was
4686 quoted in the delivery log line, like this:
4688 => >A.N.Other <ano@some.domain> <original@ddress> ...
4690 This has been changed so that it extracts the operative address. There
4691 may be more than one such address. If so, they are comma-separated, like
4694 => >ano@some.domain,ona@other.domain <original@ddress> ...
4696 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4697 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4698 not being looked up in the rDNS to get a host name, Exim was showing the
4699 IP address twice in Received: lines, even though the IP addresses were
4700 identical. For example:
4702 Received: from [1.2.3.4] (helo=[1.2.3.4])
4704 However, if the real host name was known, it was omitting the HELO data
4705 if it matched the actual IP address. This has been tidied up so that it
4706 doesn't show the same IP address twice.
4708 PH/07 When both +timestamp and +memory debugging was on, the value given by
4709 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4710 called by the debug printing, thereby overwriting the timestamp buffer.
4711 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4713 PH/08 When the original message was included in an autoreply transport, it
4714 always said "this is a copy of the message, including all the headers",
4715 even if body_only or headers_only was set. It now gives an appropriate
4718 PH/09 Applied a patch from the Sieve maintainer which:
4720 o fixes some comments
4721 o adds the (disabled) notify extension core
4722 o adds some debug output for the result of if/elsif tests
4723 o points to the current vacation draft in the documentation
4724 and documents the missing references header update
4728 o fixes a bug in processing the envelope test (when testing
4729 multiple envelope elements, the last element determined the
4732 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4733 Electronic Mail") by including:
4735 Auto-submitted: auto-generated
4737 in the messages that it generates (bounce messages and others, such as
4738 warnings). In the case of bounce messages for non-SMTP messages, there was
4739 also a typo: it was using "Auto_submitted" (underscore instead of
4740 hyphen). Since every message generated by Exim is necessarily in response
4741 to another message, thes have all been changed to:
4743 Auto-Submitted: auto-replied
4745 in accordance with these statements in the RFC:
4747 The auto-replied keyword:
4749 - SHOULD be used on messages sent in direct response to another
4750 message by an automatic process,
4752 - MUST NOT be used on manually-generated messages,
4754 - MAY be used on Delivery Status Notifications (DSNs) and Message
4755 Disposition Notifications (MDNs),
4757 - MUST NOT be used on messages generated by automatic or periodic
4758 processes, except for messages which are automatic responses to
4761 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4762 to the default Received: header definition.
4764 PH/12 Added log selector acl_warn_skipped (default on).
4766 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4767 variables because (a) they are in the wrong storage pool and (b) even if
4768 they were copied, it wouldn't work properly because of the caching.
4770 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4771 checking when decoding. Apparently there are clients that generate
4772 overlong encoded strings. Why am I not surprised?
4774 PH/15 If the first argument of "${if match_address" was not empty, but did not
4775 contain an "@" character, Exim crashed. Now it writes a panic log message
4776 and treats the condition as false.
4778 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4780 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4781 extension "envelope-auth". The code is finished and in agreement with
4782 other implementations, but there is no documentation so far and in fact,
4783 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4784 not changing the active code.
4786 Print executed "if" and "elsif" statements when debugging is used. This
4787 helps a great deal to understand what a filter does.
4789 Document more things not specified clearly in RFC3028. I had all this
4790 sorted out, when out of a sudden new issues came to my mind. Oops."
4792 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4795 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4796 canonical form (as documented). However, after a host name lookup from
4797 the IP address, check_host() was doing a simple string comparison with
4798 addresses acquired from the DNS when checking that the found name did
4799 have the original IP as one of its addresses. Since any found IPv6
4800 addresses are likely to be in abbreviated form, the comparison could
4801 fail. Luckily, there already exists a function for doing the comparison
4802 by converting both addresses to binary, so now that is used instead of
4803 the text comparison.
4805 PH/20 There was another similar case to PH/19, when a complete host name was
4806 given in a host list; looking up its IP address could give an abbreviated
4807 form, whereas the current host's name might or might not be abbreviated.
4808 The same fix has been applied.
4814 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4815 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4818 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4819 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4821 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4823 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4824 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4825 Also, for "Auto-Submitted", treat anything other than "no" as
4826 non-personal, in accordance with RFC 3834. (Previously it treated
4827 anything starting "auto-" as non-personal.)
4829 TF/01 The control=submission/name=... option had a problem with syntax
4830 errors if the name included a slash character. The /name= option
4831 now slurps the rest of the string, so it can include any characters
4832 but it must come last in the list of options (after /sender_retain
4835 PH/05 Some modifications to the interface to the fake nameserver for the new
4843 TK/01 Added the "success_on_redirect" address verification option. See
4844 NewStuff for rationale and an example.
4846 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4848 PH/02 Patch to exigrep to allow it to work on syslog lines.
4850 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4851 fread() to skip over the body file's header line, because in Cygwin the
4852 header line is locked and is inaccessible.
4854 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4855 co-exist for some time) to make it clear that it is the Exim ID that is
4856 referenced, not the Message-ID: header line.
4858 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4859 string_format() function, because snprintf() does not exist on all
4862 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4863 ${stat: expansion item.
4865 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4866 protocol synchronization error", to keep the pedants happy.
4868 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4869 well as for IRIX systems, when gcc is being used. See the host.c source
4872 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4874 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4877 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4878 transport and to the smtp transport in LMTP mode.
4880 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4882 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4883 run a filter in a subprocess. This could lead to confusion in subsequent
4884 lookups in the parent process. There should also be a search_tidyup() at
4885 the end of the subprocess.
4887 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4888 only if the host matched helo_try_verify_hosts, which caused the
4889 verification to occur when the EHLO/HELO command was issued. The ACL just
4890 tested the remembered result. Now, if a previous verification attempt has
4891 not happened, "verify = helo" does it there and then.
4893 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4895 TK/03 Fix log output including CR from clamd.
4897 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4898 exist provoked a memory error which could cause a segfault.
4900 PH/15 Installed PCRE 6.2
4902 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4904 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4905 of the problem. Specifically, suggested +O2 rather than +O1 for the
4908 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4910 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4911 record, the port was not correctly logged when the outgoing_port log
4912 selector was set (it logged the transort's default port).
4914 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4915 fallback_hosts, and "hosts" in the smtp transport.
4917 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4918 host errors such as "Connection refused".
4920 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4921 authentication with radiusclient 0.4.9:
4923 - Error returned from rc_read_config was caught wrongly
4924 - Username/password not passed on to radius server due to wrong length.
4926 The presumption is that some radiusclient API changes for 4.51/PH/17
4927 were not taken care of correctly. The code is still untested by me (my
4928 Linux distribution still has 0.3.2 of radiusclient), but it was
4929 contributed by a Radius user.
4931 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4932 expanding the "port" option of the smtp transport.
4934 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4935 while reading a MIME header. Thanks to Tom Hughes for a patch.
4937 PH/24 Include config.h inside local_scan.h so that configuration settings are
4940 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4941 that in an ACL for RCPT (for example), you can examine exactly what was
4944 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4945 commands, but it was not correctly comparing the address with the actual
4946 client host address. Thus, it would show the EHLO address in Received:
4947 header lines when this was not necessary.
4949 PH/27 Added the % operator to ${eval:}.
4951 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4952 it should be ignoring failures (because with -C, for example, it has lost
4953 privilege). It wasn't ignoring creation failures other than "already
4956 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4959 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4960 because an input line was too long, either on its own, or by virtue of
4961 too many continuations, the temporary file was not being removed, and the
4962 return code was incorrect.
4964 PH/31 Missing "BOOL" in function definition in filtertest.c.
4966 PH/32 Applied Sieve patches from the maintainer.
4968 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4970 PH/33 Added "verify = not_blind".
4972 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4973 Local/Makefile (with some defaults set). These are used in built scripts
4974 such as exicyclog, but they have never been used in the exim_install
4975 script (though there are many overriding facilities there). I have
4976 arranged that the exim_install script now takes note of these two
4979 PH/35 Installed configuration files for Dragonfly.
4981 PH/36 When a locally submitted message by a trusted user did not contain a
4982 From: header, and the sender address was obtained from -f or from an SMTP
4983 MAIL command, and the trusted user did not use -F to supply a sender
4984 name, $originator_name was incorrectly used when constructing a From:
4985 header. Furthermore, $originator_name was used for submission mode
4986 messages from external hosts without From: headers in a similar way,
4987 which is clearly wrong.
4989 PH/37 Added control=suppress_local_fixups.
4991 PH/38 When log_selector = +received_sender was set, and the addition of the
4992 sender made the log line's construction buffer exactly full, or one byte
4993 less than full, an overflow happened when the terminating "\n" was
4996 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4997 when the result of a list match is failure because a DNS lookup failed.
4999 PH/40 RM_COMMAND is now used in the building process.
5001 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5002 the "build-* directories that it finds.
5004 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5005 address in a domain literal was a prefix of an interface address.
5007 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5008 when verifying a sender address, unless rewrite_headers is false.
5010 PH/44 Wrote a long comment about why errors_to addresses are verified as
5011 recipients, not senders.
5013 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5014 the ratelimit ACL was added.
5016 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5018 PH/46 Added extra information about PostgreSQL errors to the error string.
5020 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5021 avoiding the need to install special zones in a real server. This is
5022 backwards compatible; if it can't find the fake resolver, it drops back.
5023 Thus, both old and new test suites can be run.
5025 TF/02 Added util/ratelimit.pl
5027 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5028 clock is set back in time.
5030 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5031 Candler <B.Candler@pobox.com>.
5033 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5034 OK for addresses that are the result of redirections.
5036 PH/48 A number of further additions for the benefit of the new test suite,
5037 including a fake gethostbyname() that interfaces to the fake DNS resolver
5040 TF/06 The fix for widen_domains has also been applied to qualify_single and
5041 search_parents which are the other dnslookup options that can cause
5044 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5047 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5049 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5050 service" if it encountered an MX record with an empty target hostname.
5051 The message is now "an MX or SRV record indicated no SMTP service".
5053 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5054 if the DNS of the sending site is misconfigured. This is quite a
5055 common situation. This change restores the behaviour of treating a
5056 helo verification defer as a failure.
5058 PH/51 If self=fail was set on a router, the bounce message did not include the
5059 actual error message.
5065 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5067 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5068 command itself ended in error, the underlying message about the transport
5069 filter timeout was being overwritten with the pipe command error. Now the
5070 underlying error message should be appended to the second error message.
5072 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5074 PH/02 Reduce the amount of output that "make" produces by default. Full output
5075 can still be requested.
5077 PH/03 The warning log line about a condition test deferring for a "warn" verb
5078 was being output only once per connection, rather than after each
5079 occurrence (because it was using the same function as for successful
5080 "warn" verbs). This seems wrong, so I have changed it.
5082 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5083 it should not have, which might have caused a crash in the right
5084 circumstances, but probably never did.
5086 PH/04 Installed a modified version of Tony Finch's patch to make submission
5087 mode fix the return path as well as the Sender: header line, and to
5088 add a /name= option so that you can make the user's friendly name appear
5091 TF/03 Added the control = fakedefer ACL modifier.
5093 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5094 Mark Lowes for thorough testing.
5096 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5098 TK/03 Merged latest SRS patch from Miles Wilton.
5100 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5101 with the definition in sysexits.h (which is #included earlier).
5102 Fortunately, Exim does not actually use EX_OK. The code used to try to
5103 preserve the sysexits.h value, by assuming that macro definitions were
5104 scanned for macro replacements. I have been disabused of this notion,
5105 so now the code just undefines EX_OK before #including unistd.h.
5107 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5108 in the smtp transport. When a block could not be written in a single
5109 write() function, the timeout was being re-applied to each part-write.
5110 This seems wrong - if the receiver was accepting one byte at a time it
5111 would take for ever. The timeout is now adjusted when this happens. It
5112 doesn't have to be particularly precise.
5114 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5115 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5117 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5118 without a domain if the check to <postmaster@domain> fails.
5120 SC/01 Eximstats: added -xls and the ability to specify output files
5121 (patch written by Frank Heydlauf).
5123 SC/02 Eximstats: use FileHandles for outputting results.
5125 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5127 SC/04 Eximstats: fixed display of large numbers with -nvr option
5129 SC/05 Eximstats: fixed merging of reports with empty tables.
5131 SC/06 Eximstats: added the -include_original_destination flag
5133 SC/07 Eximstats: removed tabs and trailing whitespace.
5135 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5137 TK/06 MBOX spool code: Add real "From " MBOX separator line
5138 so the .eml file is really in mbox format (even though
5139 most programs do not really care). Patch from Alex Miller.
5141 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5142 The latter is generated from $received_to and is only set if the
5143 message has one envelope recipient. SA can use these headers,
5144 obviously out-of-the-box. Patch from Alex Miller.
5146 PH/08 The ${def test on a variable was returning false if the variable's
5147 value was "0", contrary to what the specification has always said!
5148 The result should be true unless the variable is empty.
5150 PH/09 The syntax error of a character other than { following "${if
5151 def:variable_name" (after optional whitespace) was not being diagnosed.
5152 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5153 accidental colon was present, for example, could give incorrect results.
5155 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5156 result is used (not including appendfile, where other changes are about
5159 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5160 This involved changing a lot of size variables from int to off_t. It
5161 should work with maildirs and everything.
5163 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5164 spamd dying while we are connected to it.
5166 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5169 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5170 function for BDB 4.3.
5172 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5174 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5175 the sources for PCRE have been re-arranged and more files are now
5178 PH/15 The code I had for printing potentially long long variables in PH/11
5179 above was not the best (it lost precision). The length of off_t variables
5180 is now inspected at build time, and an appropriate printing format (%ld
5181 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5182 to be "long long int" or "long int". This is needed for the internal
5183 formatting function string_vformat().
5185 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5186 the configuration file to be ":syslog", then the script "guesses" where
5187 the logs files are, rather than using the compiled in default. In our
5188 case the guess is not the same as the compiled default, so the script
5189 suddenly stopped working when I started to use syslog. The patch checks
5190 to see if log_file_path is "". If so, it attempts to read it from exim
5191 with no configuration file to get the compiled in version, before it
5192 falls back to the previous guessing code."
5194 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5195 implementing BATV in an Exim configuration. See NewStuff for the gory
5198 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5199 Makefile that are specific to HP-UX.
5201 PH/18 If the "use_postmaster" option was set for a recipient callout together
5202 with the "random" option, the postmaster address was used as the MAIL
5203 FROM address for the random test, but not for the subsequent recipient
5204 test. It is now used for both.
5206 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5207 patch removes a few documentation additions to RFC 3028, because the
5208 latest draft now contains them. It adds the new en;ascii-case comparator
5209 and a new error check for 8bit text in MIME parts. Comparator and
5210 require names are now matched exactly. I enabled the subaddress
5211 extension, but it is not well tested yet (read: it works for me)."
5213 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5214 rework some of the code of TK/09 above to avoid the hardwired use of
5215 "%lld" and "long long". Replaced the call to snprintf() with a call to
5218 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5219 records point to non-existent hosts", "retry timeout exceeded", and
5220 "retry time not reached for any host after a long failure period".
5222 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5223 experimental DomainKeys support:
5225 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5226 (2) On an error such as an illegally used "control", the wrong name for
5227 the control was given.
5229 These problems did NOT occur unless DomainKeys support was compiled.
5231 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5233 PH/24 Added ${if match_ip condition.
5235 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5236 databases so that it will be absolutely obvious if a crash occurs in the
5237 DB library. This is a regular occurrence (often caused by mis-matched
5240 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5241 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5242 picked up on a user's system that detects such things. There doesn't seem
5243 to be a gcc warning option for this - only an attribute that has to be
5244 put on the function's prototype. It seems that in Fedora Core 4 they have
5245 set this on a number of new functions. No doubt there will be more in due
5248 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5249 specify a transport. However, if an address that was verified by such a
5250 router was the subject of a callout, Exim crashed because it tried to
5251 read the rcpt_include_affixes from the non-existent transport. Now it
5252 just assumes that the setting of that option is false. This bug was
5253 introduced by 4.51/PH/31.
5255 PH/28 Changed -d+all to exclude +memory, because that information is very
5256 rarely of interest, but it makes the output a lot bigger. People tend to
5257 do -d+all out of habit.
5259 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5260 code in os-type was giving problems when libc.so lives in lib64, like on
5263 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5264 aren't the modern standard, and it seems that some systems' include files
5265 don't always have them. Exim was already checking for some of the newer
5266 ones like T_AAAA, and defining it itself. I've added checks for all the
5267 record types that Exim uses.
5269 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5270 not automatically generating a new one, as it is supposed to. This
5271 prevented TLS from working. If the file did exist, but contained invalid
5272 data, a new version was generated, as expected. It was only the case of a
5273 non-existent file that was broken.
5275 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5276 with a change in libdomainkeys > 0.64.
5278 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5279 from DNS. If the selector record carries the flag, it now has
5280 precedence over the domain-wide flag.
5282 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5284 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5285 the use of a port name that isn't defined in /etc/services) occurred, the
5286 message was deferred as in a normal delivery, and thus remained on the
5287 spool, instead of being failed because of the mua_wrapper setting. This
5288 is now fixed, and I tidied up some of the mua_wrapper messages at the
5291 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5292 the messages in a hash of arrays rather than using individual hashes.
5293 This is a bit cleaner and results in dramatic memory savings, albeit
5294 at a slight CPU cost.
5296 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5297 as requested by Marc Sherman.
5299 SC/10 Eximstats: added histograms for user specified patterns as requested
5302 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5304 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5305 fopen() in the content-scanning modules that did not already have it.
5311 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5312 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5314 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5316 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5318 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5319 to test Sieve filters that use "vacation".
5321 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5322 that changes the way the GnuTLS parameters are stored in the cache file.
5323 The new format can be generated externally. For backward compatibility,
5324 if the data in the cache doesn't make sense, Exim assumes it has read an
5325 old-format file, and it generates new data and writes a new file. This
5326 means that you can't go back to an older release without removing the
5329 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5330 work if there are any delivery delays because "one_time" forces the
5331 parent to be marked "delivered", so its unseen clone is never tried
5332 again. For this reason, Exim now forbids the simultaneous setting of
5335 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5336 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5337 entry if you want to know the details. The fix, however, neglected to
5338 consider the case when local delivery batching is involved. The test for
5339 "previously delivered" was not happening when checking to see if an
5340 address could be batched with a previous (undelivered) one; under
5341 certain circumstances this could lead to multiple deliveries to the same
5344 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5345 in its include files, and this causes problems building Exim.
5347 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5348 header_syntax) but Exim was just ignoring anything given after a slash.
5349 In particular, this caused confusion with an attempt to use "verify =
5350 reverse_host_lookup/defer_ok". An error is now given when options are
5351 supplied for verify items that do not have them. (Maybe reverse_host_
5352 lookup should have a defer_ok option, but that's a different point.)
5354 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5355 defined by RFC 821) to 2048, because there were problems with some AUTH
5356 commands, and RFC 1869 says the size should be increased for extended
5357 SMTP commands that take arguments.
5359 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5362 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5363 "unknown" error; now it says that the functionality isn't in the binary.
5365 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5366 an address' error message when a string expansion fails (syntax or
5367 whatever). Otherwise the password may appear in the log. Following change
5368 PH/42 below, there is no longer a chance of it appearing in a bounce
5371 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5373 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5374 of its addresses was ever tried. (Bugzilla bug #2.)
5376 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5377 the result incorrectly in the debug output. (It correctly added a newline
5378 to what was transported.)
5380 TF/01 Added $received_time.
5382 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5383 commented out examples of how to interface to a virus scanner and to
5384 SpamAssassin. Also added commented examples of av_scanner and
5385 spamd_address settings.
5387 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5388 and controls are allowed in which ACLs. There were a couple of minor
5389 errors. Some of the entries in the conditions table (which is a table of
5390 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5391 negation of where the condition IS allowed.
5393 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5395 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5396 header file does not have a version number, so I've had to invent a new
5397 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5398 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5399 radiusclient), but it was contributed by a Radius user.
5401 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5402 files or directories, for OpenSSL.
5404 PH/19 When an Exim process that is running as root has to create an Exim log
5405 file, it does so in a subprocess that runs as exim:exim so as to get the
5406 ownership right at creation (otherwise, other Exim processes might see
5407 the file with the wrong ownership). There was no test for failure of this
5408 fork() call, which would lead to the process getting stuck as it waited
5409 for a non-existent subprocess. Forks do occasionally fail when resources
5410 run out. I reviewed all the other calls to fork(); they all seem to check
5413 PH/20 When checking for unexpected SMTP input at connect time (before writing
5414 the banner), Exim was not dealing correctly with a non-positive return
5415 from the read() function. If the client had disconnected by this time,
5416 the result was a log entry for a synchronization error with an empty
5417 string after "input=" when read() returned zero. If read() returned -1
5418 (an event I could not check), uninitialized data bytes were printed.
5419 There were reports of junk text (parts of files, etc) appearing after
5422 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5424 PH/22 Added support for macro redefinition, and (re)definition in between
5425 driver and ACL definitions.
5427 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5428 forgetting to use the resulting value; it was using the unexpanded value.
5430 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5431 hadn't been configured. The fix is from Juergen Kreileder, who
5432 understands it better than I do:
5434 "Here's what I see happening with three configured cyrus_sasl
5435 authenticators configured (plain, login, cram-md5):
5437 On startup auth_cyrus_sasl_init() gets called for each of these.
5438 This means three calls to sasl_listmech() without a specified mech_list.
5439 => SASL tests which mechs of all available mechs actually work
5440 => three warnings about OTP not working
5441 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5443 With the patch, sasl_listmech() also gets called three times. But now
5444 SASL's mech_list option is set to the server_mech specified in the the
5445 authenticator. Or in other words, the answer from sasl_listmech()
5446 gets limited to just the mech you're testing for (which is different
5448 => the return list contains just 'plain' or 'login', 'cram-md5' or
5449 nothing depending on the value of ob->server_mech.
5451 I've just tested the patch: Authentication still works fine,
5452 unavailable mechs specified in the exim configuration are still
5453 caught, and the auth.log warnings about OTP are gone."
5455 PH/25 When debugging is enabled, the contents of the command line are added
5456 to the debugging output, even when log_selector=+arguments is not
5459 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5460 answer is "GNU", and only if the return is "GNU/something" is the answer
5463 PH/27 $acl_verify_message is now set immediately after the failure of a
5464 verification in an ACL, and so is available in subsequent modifiers. In
5465 particular, the message can be preserved by coding like this:
5467 warn !verify = sender
5468 set acl_m0 = $acl_verify_message
5470 Previously, $acl_verify_message was set only while expanding "message"
5471 and "log_message" when a very denied access.
5473 PH/28 Modified OS/os.c-Linux with
5475 -#ifndef OS_LOAD_AVERAGE
5476 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5478 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5479 nomenclature these days.)
5481 PH/29 Installed patch from the Sieve maintainer that adds the options
5482 sieve_useraddress and sieve_subaddress to the redirect router.
5484 PH/30 In these circumstances:
5485 . Two addresses routed to the same list of hosts;
5486 . First host does not offer TLS;
5487 . First host accepts first address;
5488 . First host gives temporary error to second address;
5489 . Second host offers TLS and a TLS session is established;
5490 . Second host accepts second address.
5491 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5492 and peerdn, if requested) that were in fact used only for the second
5495 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5496 attention to any local part prefix or suffix that was matched by the
5497 router that accepted the address. It now behaves in the same way as it
5498 does for delivery: the affixes are removed from the local part unless
5499 rcpt_include_affixes is set on the transport.
5501 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5502 timeout during the DATA phase of an incoming message.
5504 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5505 applied a patch sanctioned by the Sieve maintainer.
5507 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5508 the uid or gid is negative. A case of a negative gid caused this to be
5509 noticed. The fix allows for either to be negative.
5511 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5512 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5515 PH/36 The change PH/12 above was broken. Fixed it.
5517 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5518 the grounds that routing the same address twice would always produce the
5519 same answer. This might have been true once, but it is certainly no
5520 longer true now. Routing a child address may depend on the previous
5521 routing that produced that child. Some complicated redirection strategies
5522 went wrong when messages had multiple recipients, and made Exim's
5523 behaviour dependent on the order in which the addresses were given.
5525 I have moved the duplicate checking until after the routing is complete.
5526 Exim scans the addresses that are assigned to local and remote
5527 transports, and removes any duplicates. This means that more work will be
5528 done, as duplicates will always all be routed, but duplicates are
5529 presumably rare, so I don't expect this is of any significance.
5531 For deliveries to pipes, files, and autoreplies, the duplicate checking
5532 still happens during the routing process, since they are not going to be
5535 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5536 It corrects a timeout issue with spamd. This is Ian's comment: "The
5537 background is that sometimes spamd either never reads data from a
5538 connection it has accepted, or it never writes response data. The exiscan
5539 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5540 blindly assumes that writes won't block so it may never time out."
5542 PH/39 Allow G after quota size as well as K and M.
5544 PH/40 The value set for $authenticated_id in an authenticator may not contain
5545 binary zeroes or newlines because the value is written to log lines and
5546 to spool files. There was no check on this. Now the value is run through
5547 the string_printing() function so that such characters are converted to
5548 printable escape sequences.
5550 PH/41 $message_linecount is a new variable that contains the total number of
5551 lines in the message. Compare $body_linecount, which is the count for the
5554 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5555 bounce and delay warning messages, except in certain special cases, which
5558 (a) An SMTP error message from a remote host;
5559 (b) A message specified in a :fail: redirection;
5560 (c) A message specified in a "fail" command in a system filter;
5561 (d) A message specified in a FAIL return from the queryprogram router;
5562 (e) A message specified by the cannot_route_message router option.
5564 In these cases only, Exim does include the error details in bounce and
5565 warning messages. There are also a few cases where bland messages such
5566 as "unrouteable address" or "local delivery error" are given.
5568 PH/43 $value is now also set for the "else" part of a ${run expansion.
5570 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5571 being worked on, but at least Exim now implements the latest version to
5574 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5575 process to complete was treated as a delivery failure, a timeout while
5576 writing the message to the pipe was logged, but erroneously treated as a
5577 successful delivery. Such timeouts include transport filter timeouts. For
5578 consistency with the overall process timeout, these timeouts are now
5579 treated as errors, giving rise to delivery failures by default. However,
5580 there is now a new Boolean option for the pipe transport called
5581 timeout_defer, which, if set TRUE, converts the failures into defers for
5582 both kinds of timeout. A transport filter timeout is now identified in
5585 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5586 systems where "make" and "gmake" are different, calling "gmake" at top
5587 level broke things. I've arranged for the value of $(MAKE) to be passed
5588 from the Makefile to this script so that it can call the same version of
5592 A note about Exim versions 4.44 and 4.50
5593 ----------------------------------------
5595 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5596 changes of various kinds. As a consequence, a big documentation update was
5597 needed. This delayed the release for rather longer than seemed good, especially
5598 in the light of a couple of (minor) security issues. Therefore, the changes
5599 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5600 release. So 4.44 and 4.50 are in effect two different branches that both start
5603 I have left the 4.50 change log unchanged; it contains all the changes since
5604 4.43. The change log for 4.44 is below; many of its items are identical to
5605 those for 4.50. This seems to be the most sensible way to preserve the
5606 historical information.
5612 1. Minor wording change to the doc/README.SIEVE file.
5614 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5615 computation of the current number of files was incorrect.
5617 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5618 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5621 4. Give more explanation in the error message when the command for a transport
5622 filter fails to execute.
5624 5. There are several places where Exim runs a non-Exim command in a
5625 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5626 was being done only for the command run by the queryprogram router. It is
5627 now done for all such subprocesses. The other cases are: ${run, transport
5628 filters, and the commands run by the lmtp and pipe transports.
5630 6. Added CONFIGURE_GROUP build-time option.
5632 7. Some older OS have a limit of 256 on the maximum number of file
5633 descriptors. Exim was using setrlimit() to set 1000 as a large value
5634 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5635 systems. I've change it so that if it can't get 1000, it tries for 256.
5637 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5638 was an oversight, and furthermore, ever since the addition of extra
5639 controls (e.g. 4.43/32), the checks on when to allow different forms of
5640 "control" were broken. There should now be diagnostics for all cases when a
5641 control that does not make sense is encountered.
5643 9. Added the /retain_sender option to "control=submission".
5645 10. $recipients is now available in the predata ACL (oversight).
5647 11. Tidy the search cache before the fork to do a delivery from a message
5648 received from the command line. Otherwise the child will trigger a lookup
5649 failure and thereby defer the delivery if it tries to use (for example) a
5650 cached ldap connection that the parent has called unbind on.
5652 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5653 of $address_data from the recipient verification was clobbered by the
5654 sender verification.
5656 13. The value of address_data from a sender verification is now available in
5657 $sender_address_data in subsequent conditions in the ACL statement.
5659 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5661 15. Added a new option "connect=<time>" to callout options, to set a different
5664 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5665 was its contents. (It was OK if the option was not defined at all.)
5667 17. A "Completed" log line is now written for messages that are removed from
5668 the spool by the -Mrm option.
5670 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5671 information about exactly what failed.
5673 19. Added -dd to debug only the daemon process.
5675 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5676 handles timeouts, both on the server side and network timeouts. Renamed the
5677 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5679 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5680 It is now set to "smtps".
5682 22. $host_address is now set to the target address during the checking of
5683 ignore_target_hosts.
5685 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5686 being passed; this would have caused $sender_host_name to have been used if
5687 matching the list had actually called for a host name (not very likely,
5688 since this list is usually IP addresses). A host name is now passed as
5691 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5692 code that specifies a non-restarting handler (typically sigaction() in
5693 modern systems) in an attempt to fix a rare and obscure crash bug.
5695 25. Narrowed the window for a race in the daemon that could cause it to ignore
5696 SIGCHLD signals. This is not a major problem, because they are used only to
5697 wake it up if nothing else does.
5699 26. A malformed maildirsize file could cause Exim to calculate negative values
5700 for the mailbox size or file count. Odd effects could occur as a result.
5701 The maildirsize information is now recalculated if the size or filecount
5704 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5705 support for a long time. Removed HAVE_SYS_VFS_H.
5707 28. Installed the latest version of exipick from John Jetmore.
5709 29. In an address list, if the pattern was not a regular expression, an empty
5710 subject address (from a bounce message) matched only if the pattern was an
5711 empty string. Non-empty patterns were not even tested. This was the wrong
5712 because it is perfectly reasonable to use an empty address as part of a
5713 database query. An empty address is now tested by patterns that are
5714 lookups. However, all the other forms of pattern expect the subject to
5715 contain a local part and a domain, and therefore, for them, an empty
5716 address still always fails if the pattern is not itself empty.
5718 30. Exim went into a mad DNS loop when attempting to do a callout where the
5719 host was specified on an smtp transport, and looking it up yielded more
5720 than one IP address.
5722 31. Re-factored the code for checking spool and log partition space into a
5723 function that finds that data and another that does the check. The former
5724 is then used to implement four new variables: $spool_space, $log_space,
5725 $spool_inodes, and $log_inodes.
5727 32. The RFC2047 encoding function was originally intended for short strings
5728 such as real names; it was not keeping to the 75-character limit for
5729 encoded words that the RFC imposes. It now respects the limit, and
5730 generates multiple encoded words if necessary. To be on the safe side, I
5731 have increased the buffer size for the ${rfc2047: expansion operator from
5734 33. It is now permitted to omit both strings after an "if" condition; if the
5735 condition is true, the result is "true". As before, when the second string
5736 is omitted, a false condition yields an empty string. This makes it less
5737 cumbersome to write custom ACL and router conditions.
5739 34. Failure to deliver a bounce message always caused it to be frozen, even if
5740 there was an errors_to setting on the router. The errors_to setting is now
5743 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5744 canonical form (fully expanded) before being placed in
5745 $sender_host_address.
5747 36. The table in the code that translates DNS record types into text (T_A to
5748 "A" for instance) was missing entries for NS and CNAME. It is just possible
5749 that this could have caused confusion if both these types were looked up
5750 for the same domain, because the text type is used as part of Exim's
5751 per-process caching. But the chance of anyone hitting this buglet seems
5754 37. The dnsdb lookup has been extended in a number of ways.
5756 (1) There is a new type, "zns", which walks up the domain tree until it
5757 finds some nameserver records. It should be used with care.
5759 (2) There is a new type, "mxh", which is like "mx" except that it returns
5760 just the host names, not the priorities.
5762 (3) It is now possible to give a list of domains (or IP addresses) to be
5763 looked up. The behaviour when one of the lookups defers can be
5764 controlled by a keyword.
5766 (4) It is now possible to specify the separator character for use when
5767 multiple records are returned.
5769 38. The dnslists ACL condition has been extended: it is now possible to supply
5770 a list of IP addresses and/or domains to be looked up in a particular DNS
5773 39. Added log_selector=+queue_time_overall.
5775 40. When running the queue in the test harness, wait just a tad after forking a
5776 delivery process, to get repeatability of debugging output.
5778 41. Include certificate and key file names in error message when GnuTLS fails
5779 to set them up, because the GnuTLS error message doesn't include the name
5780 of the failing file when there is a problem reading it.
5782 42. Allow both -bf and -bF in the same test run.
5784 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5786 44. The "Exiscan patch" is now merged into the mainline Exim source.
5788 45. Sometimes the final signoff response after QUIT could fail to get
5789 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5790 < 0 before doing a fflush(). This bug looks as though it goes back to the
5791 introduction of TLS in release 3.20, but "sometimes" must have been rare
5792 because the tests only now provoked it.
5794 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5795 (this can affect the format of dates).
5797 47. exim_tidydb, when checking for the continued existence of a message for
5798 which it has found a message-specific retry record, was not finding
5799 messages that were in split spool directories. Consequently, it was
5800 deleting retry records that should have stayed in existence.
5802 48. Steve fixed some bugs in eximstats.
5804 49. The SPA authentication driver was not abandoning authentication and moving
5805 on to the next authenticator when an expansion was forced to fail,
5806 contradicting the general specification for all authenticators. Instead it
5807 was generating a temporary error. It now behaves as specified.
5809 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5810 (the order specifies the preference for clients). The order is now AES256,
5811 AES128, 3DES, ARCFOUR128.
5813 51. Small patch to Sieve code - explicitly set From: when generating an
5816 52. Exim crashed if a remote delivery caused a very long error message to be
5817 recorded - for instance if somebody sent an entire SpamAssassin report back
5818 as a large number of 550 error lines. This bug was coincidentally fixed by
5819 increasing the size of one of Exim's internal buffers (big_buffer) that
5820 happened as part of the Exiscan merge. However, to be on the safe side, I
5821 have made the code more robust (and fixed the comments that describe what
5824 53. Now that there can be additional text after "Completed" in log lines (if
5825 the queue_time_overall log selector is set), a one-byte patch to exigrep
5826 was needed to allow it to recognize "Completed" as not the last thing in
5829 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5830 patch that reportedly fixes this has been added. I am not expert enough to
5831 create a test for it. This is what the patch creator wrote:
5833 "I found a little strange behaviour of ldap code when working with
5834 Windows 2003 AD Domain, where users was placed in more than one
5835 Organization Units. When I tried to give exim partial DN, the exit code
5836 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5837 But simultaneously result of request was absolutely normal ldap result,
5838 so I produce this patch..."
5840 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5841 REFERENCE, so I have modified the code to exclude the patch when that macro
5844 55. Some experimental protocols are using DNS PTR records for new purposes. The
5845 keys for these records are domain names, not reversed IP addresses. The
5846 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5847 leaves it alone. Component reversal etc. now happens only for IP addresses.
5850 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5852 57. Double the size of the debug message buffer (to 2048) so that more of very
5853 long debug lines gets shown.
5855 58. The exicyclog utility now does better if the number of log files to keep
5856 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5858 59. Two changes related to the smtp_active_hostname option:
5860 (1) $smtp_active_hostname is now available as a variable.
5861 (2) The default for smtp_banner uses $smtp_active_hostname instead
5862 of $primary_hostname.
5864 60. The host_aton() function is supposed to be passed a string that is known
5865 to be a valid IP address. However, in the case of IPv6 addresses, it was
5866 not checking this. This is a hostage to fortune. Exim now panics and dies
5867 if the condition is not met. A case was found where this could be provoked
5868 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5869 components; fortuitously, this particular loophole had already been fixed
5870 by change 4.50/55 above.
5872 If there are any other similar loopholes, the new check in host_aton()
5873 itself should stop them being exploited. The report I received stated that
5874 data on the command line could provoke the exploit when Exim was running as
5875 exim, but did not say which command line option was involved. All I could
5876 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5877 running as the user.
5880 61. There was a buffer overflow vulnerability in the SPA authentication code
5881 (which came originally from the Samba project). I have added a test to the
5882 spa_base64_to_bits() function which I hope fixes it.
5885 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5886 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5888 63. The daemon start-up calls getloadavg() while still root for those OS that
5889 need the first call to be done as root, but it missed one case: when
5890 deliver_queue_load_max is set with deliver_drop_privilege. This is
5891 necessary for the benefit of the queue runner, because there is no re-exec
5892 when deliver_drop_privilege is set.
5894 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5895 This has been fixed.
5897 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5898 was in use, was not putting the data itself into the right store pool;
5899 consequently, it could be overwritten for a subsequent message in the same
5900 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5903 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5905 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5906 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5907 the function were treating the return as a boolean value, which happened to
5908 work because 0=false and not-0=true, but is not correct code.
5910 68. The host_aton() function was not handling scoped IPv6 addresses (those
5911 with, for example, "%eth0" on the end) correctly.
5913 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5914 negated items (that is, ~something) in unsigned ints. Some compilers
5915 apparently mutter when there is no cast.
5917 70. If an address verification called from an ACL failed, and did not produce a
5918 user-specific message (i.e. there was only a "system" message), nothing was
5919 put in $acl_verify_message. In this situation, it now puts the system
5922 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5923 session; change 4.31/43 added the unwanted input to the log line - except
5924 that it did not do this in the start of session case. It now does.
5926 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5927 This is wrong and can cause the other end to generate a synchronization
5928 error if it is another Exim or anything else that does the synchronization
5929 check. A QUIT command is no longer sent after a timeout.
5931 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5932 during host lookups.
5934 74. The defer_ok option of callout verification was not working if it was used
5935 when verifying addresses in header lines, that is, for this case:
5937 verify = header_sender/callout=defer_ok
5939 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5940 those file descriptors could be used for SMTP connections. If anything
5941 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5942 could be sent to the SMTP client, causing chaos. The daemon now opens
5943 stdin, stdout, and stderr to /dev/null when it puts itself into the
5946 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5947 log by default. The user can override this with suitable Perl magic.
5949 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5950 the log message when discard triggers, was not working for the DATA ACL or
5951 for the non-SMTP ACL.
5953 78. Error message wording change in sieve.c.
5955 79. If smtp_accept_max_per_host was set, the number of connections could be
5956 restricted to fewer than expected, because the daemon was trying to set up
5957 a new connection before checking whether the processes handling previous
5958 connections had finished. The check for completed processes is now done
5959 earlier. On busy systems, this bug wouldn't be noticed because something
5960 else would have woken the daemon, and it would have reaped the completed
5963 80. If a message was submitted locally by a user whose login name contained one
5964 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5965 It caused a spool format error. I have fixed the spool reading code. A
5966 related problem was that the "from" clause in the Received: line became
5967 illegal because of the space(s). It is now covered by ${quote_local_part.
5969 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5972 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5974 83. Patch from Sieve maintainer to fix unterminated string problem in
5975 "vacation" handling.
5977 84. Some minor changes to the Linux configuration files to help with other
5978 OS variants using glibc.
5980 85. One more patch for Sieve to update vacation handling to latest spec.
5983 ----------------------------------------------------
5984 See the note above about the 4.44 and 4.50 releases.
5985 ----------------------------------------------------
5991 1. Change 4.43/35 introduced a bug that caused file counts to be
5992 incorrectly computed when quota_filecount was set in an appendfile
5995 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5996 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5999 3. Give more explanation in the error message when the command for a transport
6000 filter fails to execute.
6002 4. There are several places where Exim runs a non-Exim command in a
6003 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6004 was being done only for the command run by the queryprogram router. It is
6005 now done for all such subprocesses. The other cases are: ${run, transport
6006 filters, and the commands run by the lmtp and pipe transports.
6008 5. Some older OS have a limit of 256 on the maximum number of file
6009 descriptors. Exim was using setrlimit() to set 1000 as a large value
6010 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6011 systems. I've change it so that if it can't get 1000, it tries for 256.
6013 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6014 was an oversight, and furthermore, ever since the addition of extra
6015 controls (e.g. 4.43/32), the checks on when to allow different forms of
6016 "control" were broken. There should now be diagnostics for all cases when a
6017 control that does not make sense is encountered.
6019 7. $recipients is now available in the predata ACL (oversight).
6021 8. Tidy the search cache before the fork to do a delivery from a message
6022 received from the command line. Otherwise the child will trigger a lookup
6023 failure and thereby defer the delivery if it tries to use (for example) a
6024 cached ldap connection that the parent has called unbind on.
6026 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6027 of $address_data from the recipient verification was clobbered by the
6028 sender verification.
6030 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6031 was its contents. (It was OK if the option was not defined at all.)
6033 11. A "Completed" log line is now written for messages that are removed from
6034 the spool by the -Mrm option.
6036 12. $host_address is now set to the target address during the checking of
6037 ignore_target_hosts.
6039 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6040 being passed; this would have caused $sender_host_name to have been used if
6041 matching the list had actually called for a host name (not very likely,
6042 since this list is usually IP addresses). A host name is now passed as
6045 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6046 code that specifies a non-restarting handler (typically sigaction() in
6047 modern systems) in an attempt to fix a rare and obscure crash bug.
6049 15. Narrowed the window for a race in the daemon that could cause it to ignore
6050 SIGCHLD signals. This is not a major problem, because they are used only to
6051 wake it up if nothing else does.
6053 16. A malformed maildirsize file could cause Exim to calculate negative values
6054 for the mailbox size or file count. Odd effects could occur as a result.
6055 The maildirsize information is now recalculated if the size or filecount
6058 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6059 support for a long time. Removed HAVE_SYS_VFS_H.
6061 18. Updated exipick to current release from John Jetmore.
6063 19. Allow an empty sender to be matched against a lookup in an address list.
6064 Previously the only cases considered were a regular expression, or an
6067 20. Exim went into a mad DNS lookup loop when doing a callout where the
6068 host was specified on the transport, if the DNS lookup yielded more than
6071 21. The RFC2047 encoding function was originally intended for short strings
6072 such as real names; it was not keeping to the 75-character limit for
6073 encoded words that the RFC imposes. It now respects the limit, and
6074 generates multiple encoded words if necessary. To be on the safe side, I
6075 have increased the buffer size for the ${rfc2047: expansion operator from
6078 22. Failure to deliver a bounce message always caused it to be frozen, even if
6079 there was an errors_to setting on the router. The errors_to setting is now
6082 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6083 canonical form (fully expanded) before being placed in
6084 $sender_host_address.
6086 24. Updated eximstats to version 1.33
6088 25. Include certificate and key file names in error message when GnuTLS fails
6089 to set them up, because the GnuTLS error message doesn't include the name
6090 of the failing file when there is a problem reading it.
6092 26. Expand error message when OpenSSL has problems setting up cert/key files.
6095 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6096 (this can affect the format of dates).
6098 28. exim_tidydb, when checking for the continued existence of a message for
6099 which it has found a message-specific retry record, was not finding
6100 messages that were in split spool directories. Consequently, it was
6101 deleting retry records that should have stayed in existence.
6103 29. eximstats updated to version 1.35
6104 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6105 1.35 - bugfix such that pie charts by volume are generated correctly
6107 30. The SPA authentication driver was not abandoning authentication and moving
6108 on to the next authenticator when an expansion was forced to fail,
6109 contradicting the general specification for all authenticators. Instead it
6110 was generating a temporary error. It now behaves as specified.
6112 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6113 (the order specifies the preference for clients). The order is now AES256,
6114 AES128, 3DES, ARCFOUR128.
6116 31. Small patch to Sieve code - explicitly set From: when generating an
6119 32. Exim crashed if a remote delivery caused a very long error message to be
6120 recorded - for instance if somebody sent an entire SpamAssassin report back
6121 as a large number of 550 error lines. This bug was coincidentally fixed by
6122 increasing the size of one of Exim's internal buffers (big_buffer) that
6123 happened as part of the Exiscan merge. However, to be on the safe side, I
6124 have made the code more robust (and fixed the comments that describe what
6127 33. Some experimental protocols are using DNS PTR records for new purposes. The
6128 keys for these records are domain names, not reversed IP addresses. The
6129 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6130 leaves it alone. Component reversal etc. now happens only for IP addresses.
6133 34. The host_aton() function is supposed to be passed a string that is known
6134 to be a valid IP address. However, in the case of IPv6 addresses, it was
6135 not checking this. This is a hostage to fortune. Exim now panics and dies
6136 if the condition is not met. A case was found where this could be provoked
6137 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6138 components; fortuitously, this particular loophole had already been fixed
6139 by change 4.50/55 or 4.44/33 above.
6141 If there are any other similar loopholes, the new check in host_aton()
6142 itself should stop them being exploited. The report I received stated that
6143 data on the command line could provoke the exploit when Exim was running as
6144 exim, but did not say which command line option was involved. All I could
6145 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6146 running as the user.
6149 35. There was a buffer overflow vulnerability in the SPA authentication code
6150 (which came originally from the Samba project). I have added a test to the
6151 spa_base64_to_bits() function which I hope fixes it.
6154 36. The daemon start-up calls getloadavg() while still root for those OS that
6155 need the first call to be done as root, but it missed one case: when
6156 deliver_queue_load_max is set with deliver_drop_privilege. This is
6157 necessary for the benefit of the queue runner, because there is no re-exec
6158 when deliver_drop_privilege is set.
6160 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6161 was in use, was not putting the data itself into the right store pool;
6162 consequently, it could be overwritten for a subsequent message in the same
6163 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6166 38. Sometimes the final signoff response after QUIT could fail to get
6167 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6168 < 0 before doing a fflush(). This bug looks as though it goes back to the
6169 introduction of TLS in release 3.20, but "sometimes" must have been rare
6170 because the tests only now provoked it.
6176 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6177 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6178 Now it returns an integer. A number of calls were still expecting a T/F
6179 return. Fortuitously, in all cases, the tests worked in OK situations,
6180 which is the norm. However, things would have gone wrong on any write
6181 failures on the smtp file descriptor. This function is used when sending
6182 messages over SMTP and also when doing verify callouts.
6184 2. When Exim is called to do synchronous delivery of a locally submitted
6185 message (the -odf or -odi options), it no longer closes stderr before doing
6188 3. Implemented the mua_wrapper option.
6190 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6192 5. Implemented the functions header_remove(), header_testname(),
6193 header_add_at_position(), and receive_remove_recipient(), and exported them
6196 6. If an ACL "warn" statement specified the addition of headers, Exim already
6197 inserted X-ACL-Warn: at the start if there was no header name. However, it
6198 was not making this test for the second and subsequent header lines if
6199 there were newlines in the string. This meant that an invalid header could
6200 be inserted if Exim was badly configured.
6202 7. Allow an ACL "warn" statement to add header lines at the start or after all
6203 the Received: headers, as well as at the end.
6205 8. Added the rcpt_4xx retry error code.
6207 9. Added postmaster_mailfrom=xxx to callout verification option.
6209 10. Added mailfrom=xxxx to the callout verification option, for verify=
6212 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6213 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6215 12. Inserted some casts to stop certain compilers warning when using pointer
6216 differences as field lengths or precisions in printf-type calls (mostly
6217 affecting debugging statements).
6219 13. Added optional readline() support for -be (dynamically loaded).
6221 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6222 same clock tick as a message's arrival, so that its received time was the
6223 same as the "first fail" time on the retry record, and that message
6224 remained on the queue past the ultimate address timeout, every queue runner
6225 would try a delivery (because it was past the ultimate address timeout) but
6226 after another failure, the ultimate address timeout, which should have then
6227 bounced the address, did not kick in. This was a "< instead of <=" error;
6228 in most cases the first failure would have been in the next clock tick
6229 after the received time, and all would be well.
6231 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6232 being recognized when the domain list was tested by the match_domain
6233 condition in an expansion string.
6235 16. Added the ${str2b64: operator.
6237 17. Exim was always calling setrlimit() to set a large limit for the number of
6238 processes, without checking whether the existing limit was already
6239 adequate. (It did check for the limit on file descriptors.) Furthermore,
6240 errors from getrlimit() and setrlimit() were being ignored. Now they are
6241 logged to the main and panic logs, but Exim does carry on, to try to do its
6242 job under whatever limits there are.
6244 18. Imported PCRE 5.0.
6246 19. Trivial typo in log message " temporarily refused connection" (the leading
6249 20. If the log selector return_path_on_delivery was set and an address was
6250 redirected to /dev/null, the delivery process crashed because it assumed
6251 that a return path would always be set for a "successful" delivery. In this
6252 case, the whole delivery is bypassed as an optimization, and therefore no
6255 21. Internal re-arrangement: the function for sending a challenge and reading
6256 a response while authentication was assuming a zero-terminated challenge
6257 string. It's now changed to take a pointer and a length, to allow for
6258 binary data in such strings.
6260 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6262 23. Exim was not respecting finduser_retries when seeking the login of the
6263 uid under which it was called; it was always trying 10 times. (The default
6264 setting of finduser_retries is zero.) Also, it was sleeping after the final
6265 failure, which is pointless.
6267 24. Implemented tls_on_connect_ports.
6269 25. Implemented acl_smtp_predata.
6271 26. If the domain in control=submission is set empty, Exim assumes that the
6272 authenticated id is a complete email address when it generates From: or
6273 Sender: header lines.
6275 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6276 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6277 chown and chgrp in /bin and hostname in /usr/bin.
6279 28. Exim was keeping the "process log" file open after each use, just as it
6280 does for the main log. This opens the possibility of it remaining open for
6281 long periods when the USR1 signal hits a daemon. Occasional processlog
6282 errors were reported, that could have been caused by this. Anyway, it seems
6283 much more sensible not to leave this file open at all, so that is what now
6286 29. The long-running daemon process does not normally write to the log once it
6287 has entered its main loop, and it closes the log before doing so. This is
6288 so that log files can straightforwardly be renamed and moved. However,
6289 there are a couple of unusual error situations where the daemon does write
6290 log entries, and I had neglected to close the log afterwards.
6292 30. The text of an SMTP error response that was received during a remote
6293 delivery was being truncated at 512 bytes. This is too short for some of
6294 the long messages that one sometimes sees. I've increased the limit to
6297 31. It is now possible to make retry rules that apply only when a message has a
6298 specific sender, in particular, an empty sender.
6300 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6301 it possible to be selective about when SMTP synchronization is enforced.
6303 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6305 32. Implemented hosts_connection_nolog.
6307 33. Added an ACL for QUIT.
6309 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6312 35. Added mailbox_size and mailbox_filecount to appendfile.
6314 36. Added control = no_multiline_responses to ACLs.
6316 37. There was a bug in the logic of the code that waits for the clock to tick
6317 in the case where the clock went backwards by a substantial amount such
6318 that the microsecond fraction of "now" was more than the microsecond
6319 fraction of "then" (but the whole seconds number was less).
6321 38. Added support for the libradius Radius client library this is found on
6322 FreeBSD (previously only the radiusclient library was supported).
6328 1. When certain lookups returned multiple values in the form name=value, the
6329 quoting of the values was not always being done properly. Specifically:
6330 (a) If the value started with a double quote, but contained no whitespace,
6332 (b) If the value contained whitespace other than a space character (i.e.
6333 tabs or newlines or carriage returns) it was not quoted.
6334 This fix has been applied to the mysql and pgsql lookups by writing a
6335 separate quoting function and calling it from the lookup code. The fix
6336 should probably also be applied to nisplus, ibase and oracle lookups, but
6337 since I cannot test any of those, I have not disturbed their existing code.
6339 2. A hit in the callout cache for a specific address caused a log line with no
6340 reason for rejecting RCPT. Now it says "Previous (cached) callout
6341 verification failure".
6343 3. There was an off-by-one bug in the queryprogram router. An over-long
6344 return line was truncated at 256 instead of 255 characters, thereby
6345 overflowing its buffer with the terminating zero. As well as fixing this, I
6346 have increased the buffer size to 1024 (and made a note to document this).
6348 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6349 when Exim is waiting for an SMTP response from a remote server, Exim
6350 restarts its select() call on the socket, thereby resetting its timeout.
6351 This is not a problem when such interrupts are rare. Somebody set up a cron
6352 job to run exiwhat every 2 minutes, which is less than the normal select()
6353 timeout (5 or 10 minutes). This meant that the select() timeout never
6354 kicked in because it was always reset. I have fixed this by comparing the
6355 time when an interrupt arrives with the time at the start of the first call
6356 to select(). If more time than the timeout has elapsed, the interrupt is
6357 treated as a timeout.
6359 5. Some internal re-factoring in preparation for the addition of Sieve
6360 extensions (by MH). In particular, the "personal" test is moved to a
6361 separate function, and given an option for scanning Cc: and Bcc: (which is
6362 not set for Exim filters).
6364 6. When Exim created an email address using the login of the caller as the
6365 local part (e.g. when creating a From: or Sender: header line), it was not
6366 quoting the local part when it contained special characters such as @.
6368 7. Installed new OpenBSD configuration files.
6370 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6371 try to make them clearer.
6373 9. Callout options, other than the timeout value, were being ignored when
6374 verifying sender addresses in header lines. For example, when using
6376 verify = header_sender/callout=no_cache
6378 the cache was (incorrectly) being used.
6380 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6381 only the Cygwin environment.
6383 11. When return_path_on_delivery was set as a log selector, if different remote
6384 addresses in the same message used different return paths and parallel
6385 remote delivery occurred, the wrong values would sometimes be logged.
6386 (Whenever a remote delivery process finished, the return path value from
6387 the most recently started remote delivery process was logged.)
6389 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6390 lines when AUTH and/or TLS are in use. This is the "received protocol"
6391 field. Exim used to use "asmtp" for authenticated SMTP, without any
6392 indication (in the protocol name) for TLS use. Now it follows the RFC and
6393 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6394 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6395 names appear in log lines as well as in Received: header lines.
6397 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6398 extensions, and comparison tests, and to fix some bugs.
6400 14. Changes to the "personal" filter test:
6402 (1) The test was buggy in that it was just doing the equivalent of
6403 "contains" tests on header lines. For example, if a user's address was
6404 anne@some.where, the "personal" test would incorrectly be true for
6406 To: susanne@some.where
6408 This test is now done by extracting each address from the header in turn,
6409 and checking the entire address. Other tests that are part of "personal"
6410 are now done using regular expressions (for example, to check local parts
6411 of addresses in From: header lines).
6413 (2) The list of non-personal local parts in From: addresses has been
6414 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6415 taken from the Sieve specification recommendations.
6417 (3) If the message contains any header line starting with "List-" it is
6418 treated as non-personal.
6420 (4) The test for "circular" in the Subject: header line has been removed
6421 because it now seems ill-conceived.
6423 15. Minor typos in src/EDITME comments corrected.
6425 16. Installed latest exipick from John Jetmore.
6427 17. If headers_add on a router specified a text string that was too long for
6428 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6429 of string_sprintf() is now avoided.
6431 18. $message_body_size was not set (it was always zero) when running the DATA
6432 ACL and the local_scan() function.
6434 19. For the "mail" command in an Exim filter, no default was being set for
6435 the once_repeat time, causing a random time value to be used if "once" was
6436 specified. (If the value happened to be <= 0, no repeat happened.) The
6437 default is now 0s, meaning "never repeat". The "vacation" command was OK
6438 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6439 (I found it when inspecting the code).
6441 20. There is now an overall timeout for performing a callout verification. It
6442 defaults to 4 times the callout timeout, which applies to individual SMTP
6443 commands during the callout. The overall timeout applies when there is more
6444 than one host that can be tried. The timeout is checked before trying the
6445 next host. This prevents very long delays if there are a large number of
6446 hosts and all are timing out (e.g. when the network connections are timing
6447 out). The value of the overall timeout can be changed by specifying an
6448 additional sub-option for "callout", called "maxwait". For example:
6450 verify = sender/callout=5s,maxwait=20s
6452 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6453 to the end before writing, but this should make it even safer).
6455 22. Exim was forgetting that it had advertised PIPELINING for the second and
6456 subsequent messages on an SMTP connection. It was also not resetting its
6457 memory on STARTTLS and an internal HELO.
6459 23. When Exim logs an SMTP synchronization error within a session, it now
6460 records whether PIPELINING has been advertised or not.
6462 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6463 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6464 rather than long int.
6466 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6468 26. Added the never_mail option to autoreply.
6474 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6475 crash if the getsockname() call failed; this can happen if a connection is
6476 closed very soon after it is established. The problem was simply in the
6477 order in which certain operations were done, causing Exim to try to write
6478 to the SMTP stream before it had set up the file descriptor. The bug has
6479 been fixed by making things happen in the correct order.
6485 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6486 before the connection was closed, thus losing the rejection response.
6488 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6489 some early Solaris releases, but causes trouble in current releases where
6490 socklen_t is defined.
6492 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6495 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6498 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6499 that releasing the top of it at the end releases what was used for sub-
6500 expansions (unless the block got too big). However, discard this block if
6501 the first thing is a variable or header, so that we can use its block when
6502 it is dynamic (useful for very large $message_headers, for example).
6504 6. Lookups now cache *every* query, not just the most recent. A new, separate
6505 store pool is used for this. It can be recovered when all lookup caches are
6506 flushed. Lookups now release memory at the end of their result strings.
6507 This has involved some general refactoring of the lookup sources.
6509 7. Some code has been added to the store_xxx() functions to reduce the amount
6510 of flapping under certain conditions.
6512 8. log_incoming_interface used to affect only the <= reception log lines. Now
6513 it causes the local interface and port to be added to several more SMTP log
6514 lines, for example "SMTP connection from", and rejection lines.
6516 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6518 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6520 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6521 could be overwritten at the end of the current message (or the start of a
6522 new message if it was set in a HELO ACL). The value is now preserved for
6523 the duration of the SMTP connection.
6525 12. If a transport had a headers_rewrite setting, and a matching header line
6526 contained an unqualified address, that address was qualified, even if it
6527 did not match any rewriting rules. The underlying bug was that the values
6528 of the flags that permit the existence of unqualified sender and recipient
6529 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6530 non-local messages, and by -bnq for local messages) were not being
6531 preserved with the message after it was received.
6533 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6534 "next input=" as part of the text comprising the host identity instead of
6535 the correct text. The code was using the same buffer for two different
6536 strings. However, depending on which order the printing function evaluated
6537 its arguments, the bug did not always show up. Under Linux, for example, my
6538 test suite worked just fine.
6540 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6541 doesn't work with some older versions of Perl. It has been changed to "my",
6542 which in any case is probably the better facility to use.
6544 15. A really picky compiler found some instances of statements for creating
6545 error messages that either had too many or two few arguments for the format
6548 16. The size of the buffer for calls to the DNS resolver has been increased
6549 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6550 for addresses that have a lot of PTR records. This alleviates a problem; it
6551 does not fully solve it.
6553 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6554 buffer now truncates the list and logs the incident, which is the same
6555 action as happens when Exim is looking up a host name and its aliases.
6556 Previously in this situation something unpredictable would happen;
6557 sometimes it was "internal error: store_reset failed".
6559 18. If a server dropped the connection unexpectedly when an Exim client was
6560 using GnuTLS and trying to read a response, the client delivery process
6561 crashed while trying to generate an error log message.
6563 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6564 string, for example:
6566 warn message = H1: something\nH2: something
6568 the text was added as a single header line from Exim's point of view
6569 though it ended up OK in the delivered message. However, searching for the
6570 second and subsequent header lines using $h_h2: did not work. This has been
6571 fixed. Similarly, if a system filter added multiple headers in this way,
6572 the routers could not see them.
6574 20. Expanded the error message when iplsearch is called with an invalid key to
6575 suggest using net-iplsearch in a host list.
6577 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6578 ACLs are no longer actually imposed (and a message to that effect is
6581 22. If a "gecos" field in a passwd entry contained escaped characters, in
6582 particular, if it contained a \" sequence, Exim got it wrong when building
6583 a From: or a Sender: header from that name. A second bug also caused
6584 incorrect handling when an unquoted " was present following a character
6585 that needed quoting.
6587 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6588 was not being matched caselessly.
6590 24. Arranged for all hyphens in the exim.8 source to be escaped with
6593 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6594 properly. Recipient callout cache records were still being keyed to include
6595 the sender, even when use_sender was set false. This led to far more
6596 callouts that were necessary. The sender is no longer included in the key
6597 when use_sender is false.
6599 26. Added "control = submission" modifier to ACLs.
6601 27. Added the ${base62d: operator to decode base 62 numbers.
6603 28. dnsdb lookups can now access SRV records.
6605 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6606 the configuration file.
6608 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6609 (-v) mode. This makes the output for a verbose queue run more intelligible.
6611 31. Added a use_postmaster feature to recipient callouts.
6613 32. Added the $body_zerocount variable, containing the number of binary zero
6614 bytes in the message body.
6616 33. The time of last modification of the "new" subdirectory is now used as the
6617 "mailbox time last read" when there is a quota error for a maildir
6620 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6622 35. Added +ignore_unknown as a special item in host lists.
6624 36. Code for decoding IPv6 addresses in host lists is now included, even if
6625 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6626 address was recognized as an IP address, but was then not correctly decoded
6627 into binary, causing unexpected and incorrect effects when compared with
6634 1. Very minor rewording of debugging text in manualroute to say "list of
6635 hosts" instead of "hostlist".
6637 2. If verify=header_syntax was set, and a header line with an unqualified
6638 address (no domain) and a large number of spaces between the end of the
6639 name and the colon was received, the reception process suffered a buffer
6640 overflow, and (when I tested it) crashed. This was caused by some obsolete
6641 code that should have been removed. The fix is to remove it!
6643 3. When running in the test harness, delay a bit after writing a bounce
6644 message to get a bit more predictability in the log output.
6646 4. Added a call to search_tidyup() just before forking a reception process. In
6647 theory, someone could use a lookup in the expansion of smtp_accept_max_
6648 per_host which, without the tidyup, could leave open a database connection.
6650 5. Added the variables $recipient_data and $sender_data which get set from a
6651 lookup success in an ACL "recipients" or "senders" condition, or a router
6652 "senders" option, similar to $domain_data and $local_part_data.
6654 6. Moved the writing of debug_print from before to after the "senders" test
6657 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6658 problems for message scanning, either using a data ACL, or using
6659 local_scan() because the Received: header was not generated till after they
6660 were called (in order to set the time as the time of reception completion).
6661 I have revised the way this works. The header is now generated after the
6662 body is received, but before the ACL or local_scan() are called. After they
6663 are run, the timestamp in the header is updated.
6669 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6670 before starting a queue runner without re-exec. This happened only when
6671 deliver_drop_privilege was set or when the Exim user was set to root. The
6672 effect of the bug was that timeouts during subsequent deliveries caused
6673 crashes instead of being properly handled. The handler is now left at its
6674 default (and expected) setting.
6676 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6677 message, again when deliver_drop_privilege is set or Exim is run as root.
6678 The bug described in (1) was not present in this case, but the tidying up
6679 of the other signals was missing. I have made the two cases consistent.
6681 3. The ignore_target_hosts setting on a manualroute router was being ignored
6682 for hosts that were looked up using the /MX notation.
6684 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6687 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6688 operated on the sender address. After changing the $sender_address to <>
6689 for the sender address verify, Exim was re-instated it as the original
6690 (before rewriting) address, but remembering that it had rewritten it, so it
6691 wasn't rewriting it again. This bug also had the effect of breaking the
6692 sender address verification caching when the sender address was rewritten.
6694 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6695 This has been changed so that if the ip literal address matches
6696 ignore_target_hosts, the router declines.
6698 7. Added expansion conditions match_domain, match_address, and match_local_
6699 part (NOT match_host).
6701 8. The placeholder for the Received: header didn't have a length field set.
6703 9. Added code to Exim itself and to exim_lock to test for a specific race
6704 condition that could lead to file corruption when using MBX delivery. The
6705 issue is with the lockfile that is created in /tmp. If this file is removed
6706 after a process has opened it but before that process has acquired a lock,
6707 there is the potential for a second process to recreate the file and also
6708 acquire a lock. This could lead to two Exim processes writing to the file
6709 at the same time. The added code performs the same test as UW imapd; it
6710 checks after acquiring the lock that its file descriptor still refers to
6711 the same named file.
6713 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6714 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6717 11. Added the smtp_active_hostname option. If used, this will typically be made
6718 to depend on the incoming interface address. Because $interface_address is
6719 not set up until the daemon has forked a reception process, error responses
6720 that can happen earlier (such as "too many connections") no longer contain
6723 12. If an expansion in a condition on a "warn" statement fails because a lookup
6724 defers, the "warn" statement is abandoned, and the next ACL statement is
6725 processed. Previously this caused the whole ACL to be aborted.
6727 13. Added the iplsearch lookup type.
6729 14. Added ident_timeout as a log selector.
6731 15. Added tls_certificate_verified as a log selector.
6733 16. Added a global option tls_require_ciphers (compare the smtp transport
6734 option of the same name). This controls incoming TLS connections.
6736 17. I finally figured out how to make tls_require_ciphers do a similar thing
6737 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6738 before starting the TLS session.
6740 18. Tabs are now shown as \t in -bP output.
6742 19. If the log selector return_path_on_delivery was set, Exim crashed when
6743 bouncing a message because it had too many Received: header lines.
6745 20. If two routers both had headers_remove settings, and the first one included
6746 a superfluous trailing colon, the final name in the first list and the
6747 first name in the second list were incorrectly joined into one item (with a
6748 colon in the middle).
6754 1. Added -C and -D options to the exinext utility, mainly to make it easier
6755 to include in the automated testing, but these could be helpful when
6756 multiple configurations are in use.
6758 2. The exinext utility was not formatting the output nicely when there was
6759 an alternate port involved in the retry record key, nor when there was a
6760 message id as well (for retries that were specific to a specific message
6761 and a specific host). It was also confused by IPv6 addresses, because of
6762 the additional colons they contain. I have fixed the IPv4 problem, and
6763 patched it up to do a reasonable job for IPv6.
6765 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6766 delivery, the log line now contains "pipelined" if PIPELINING was used.
6768 4. An SMTP transport process used to panic and die if the bind() call to set
6769 an explicit outgoing interface failed. This has been changed; it is now
6770 treated in the same way as a connect() failure.
6772 5. A reference to $sender_host_name in the part of a conditional expansion
6773 that was being skipped was still causing a DNS lookup. This no longer
6776 6. The def: expansion condition was not recognizing references to header lines
6777 that used bh_ and bheader_.
6779 7. Added the _cache feature to named lists.
6781 8. The code for checking quota_filecount in the appendfile transport was
6782 allowing one more file than it should have been.
6784 9. For compatibility with Sendmail, the command line option
6792 and sets the incoming protocol and host name (for trusted callers). The
6793 host name and its colon can be omitted when only the protocol is to be set.
6794 Note the Exim already has two private options, -pd and -ps, that refer to
6795 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6796 "s", but I don't think that's a major issue.
6798 10. A number of refactoring changes to the code, none of which should affect
6801 (a) The number of logging options was getting close to filling up the
6802 32-bit word that was used as a bit map. I have split them into two classes:
6803 those that are passed in the argument to log_write(), and those that are
6804 only ever tested independently outside of that function. These are now in
6805 separate 32-bit words, so there is plenty of room for expansion again.
6806 There is no change in the user interface or the logging behaviour.
6808 (b) When building, for example, log lines, the code previously used a
6809 macro that called string_cat() twice, in order to add two strings. This is
6810 not really sufficiently general. Furthermore, there was one instance where
6811 it was actually wrong because one of the argument was used twice, and in
6812 one call a function was used. (As it happened, calling the function twice
6813 did not affect the overall behaviour.) The macro has been replaced by a
6814 function that can join an arbitrary number of extra strings onto a growing
6817 (c) The code for expansion conditions now uses a table and a binary chop
6818 instead of a serial search (which was left over from when there were very
6819 few conditions). Also, it now recognizes conditions like "pam" even when
6820 the relevant support is not compiled in: a suitably worded error message is
6821 given if an attempt is made to use such a condition.
6823 11. Added ${time_interval:xxxxx}.
6825 12. A bug was causing one of the ddress fields not to be passed back correctly
6826 from remote delivery subprocesses. The field in question was not being
6827 subsequently used, so this caused to problems in practice.
6829 13. Added new log selectors queue_time and deliver_time.
6831 14. Might have fixed a bug in maildirsizefile handling that threw up
6832 "unexpected character" debug warnings, and recalculated the data
6833 unnecessarily. In any case, I expanded the warning message to give more
6836 15. Added the message "Restricted characters in address" to the statements in
6837 the default ACL that block characters like @ and % in local parts.
6839 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6840 Three changes have been made:
6842 (a) There was a serious bug; a negative response to MAIL caused the whole
6843 recipient domain to be cached as invalid, thereby blocking all messages
6844 to all local parts at the same domain, from all senders. This bug has
6845 been fixed. The domain is no longer cached after a negative response to
6846 MAIL if the sender used is not empty.
6848 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6851 (c) A new callout option, "use_sender" has been added for people who want
6852 the modified behaviour.
6858 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6861 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6862 indeed breaks things for older releases.
6864 3. Added additional logging to the case where there is a problem reading data
6865 from a filter that is running in a subprocess using a pipe, in order to
6866 try to track down a specific problem.
6868 4. Testing facility fudge: when running in the test harness and attempting
6869 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6870 getting "No route to host". Convert this to a timeout.
6872 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6875 6. Some OS don't have socklen_t but use size_t instead. This affects the
6876 fifth argument of getsockopt() amongst other things. This is now
6877 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6878 can be set for individual OS. I have set it for SunOS5, OSF1, and
6879 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6880 some earlier ones do not.
6882 7. Change 4.30/15 was not doing the test caselessly.
6884 8. The standard form for an IPv6 address literal was being rejected by address
6885 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6886 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6887 this, as well as the form without the "IPv6" on the front (but only when
6888 address literals are enabled, of course).
6890 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6892 10. Exim crashed if a message with an empty sender address specified by -f
6893 encountered a router with an errors_to setting. This could be provoked only
6894 by a command such as
6898 where an empty string was supplied; "<>" did not hit this bug.
6900 11. Installed PCRE release 4.5.
6902 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6903 remained set. It is now erased.
6905 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6906 times from message ids (which are base 36 rather than the normal 62).
6908 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6909 were being counted as actual protocol errors, and logged if the log
6910 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6911 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6912 and DATA following a set of rejected RCPTs do not count as protocol errors.
6913 In other words, Exim assumes they were pipelined, though this may not
6914 actually be the case. Of course, in all cases the client gets an
6915 appropriate error code.
6917 15. If a lookup fails in an ACL condition, a message about the failure may
6918 be available; it is used if testing the ACL cannot continue, because most
6919 such messages specify what the cause of the deferral is. However, some
6920 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6921 that caused an old message to be retained and used if a later statement
6922 caused a defer, replacing the real cause of the deferral.
6924 16. If an IP address had so many PTR records that the DNS lookup buffer
6925 was not large enough to hold them, Exim could crash while trying to process
6926 the truncated data. It now detects and logs this case.
6928 17. Further to 4.21/58, another change has been made: if (and only if) the
6929 first line of a message (the first header line) ends with CRLF, a bare LF
6930 in a subsequent header line has a space inserted after it, so as not to
6931 terminate the header.
6933 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6934 unnecessarily, used atoi() instead of strtol(), and didn't check the
6935 termination when getting file sizes from file names by regex.
6937 19. Completely re-implemented the support for maildirsize files, in the light
6938 of a number of problems with the previous contributed implementation
6939 (4.30/29). In particular:
6941 . If the quota is zero, the maildirsize file is maintained, but no quota is
6944 . If the maildir directory does not exist, it is created before any attempt
6945 to write a maildirsize file.
6947 . The quota value in the file is just a cache; if the quota is changed in
6948 the transport, the new value overrides.
6950 . A regular expression is available for excluding directories from the
6953 20. The autoreply transport checks the characters in options that define the
6954 message's headers; it allows continued headers, but it was checking with
6955 isspace() after an embedded newline instead of explicitly looking for a
6958 21. If all the "regular" hosts to which an address was routed had passed their
6959 expiry times, and had not reached their retry times, the address was
6960 bounced, even if fallback hosts were defined. Now Exim should go on to try
6963 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6964 equivalent code in the SMTP transport. Some hosts send humungous responses
6965 to HELO/EHLO, more than 1024 it seems.
6967 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6968 gives compiler warnings in some environments. I've now done it "properly",
6971 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6972 (because of problems with the built-in one) was declared to return uschar *
6973 instead of char *, causing compiler failure.
6975 25. Fixed a file descriptor leak when processing alias/forward files.
6977 26. Fixed a minor format string issue in dbfn.c.
6979 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6981 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6982 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6985 29. When a sender address is verified, it is cached, to save repeating the test
6986 when there is more than one recipient in a message. However, when the
6987 verification involves a callout, it is possible for different callout
6988 options to be set for different recipients. It is too complicated to keep
6989 track of this in the cache, so now Exim always runs a verification when a
6990 callout is required, relying on the callout cache for the optimization.
6991 The overhead is duplication of the address routing, but this should not be
6994 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6995 to be verified with callout=postmaster, and the main callout worked but the
6996 postmaster check failed, the verification correctly failed. However, if a
6997 subsequent RCPT command asked for sender verification *without* the
6998 postmaster check, incorrect caching caused this verification also to fail,
7001 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7002 it was not caching the DNS options (qualify_single, search_parents) that
7003 were used when the lookup failed. A subsequent lookup with different
7004 options therefore always gave the same answer, though there were cases
7005 where it should not have. (Example: a "domains = !$mx_any" option on a
7006 dnslookup router: the "domains" option is always processed without any
7007 widening, but the router might have qualify_single set.) Now Exim uses the
7008 cached value only when the same options are set.
7010 32. Added John Jetmore's "exipick" utility to the distribution.
7012 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7013 than a timeout (e.g. a certificate is required, and is not provided), an
7014 Exim server now closes the connection immediately. Previously it waited for
7015 the client to close - but if the client is SSL, it seems that they each
7016 wait for each other, leading to a delay before one of them times out.
7018 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7019 maintained 0.8.x compatibility because I don't think many are using it, and
7020 it is clearly obsolete.
7022 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7025 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7026 client certificate was expired. A simple patch fixes this, though I don't
7027 understand the full logic of why the verify callback is called multiple
7030 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7031 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7032 which causes problems with some clients (such as the Certicom SSL Plus
7033 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7034 disables the coutermeasure allowing Eudora to connect."
7036 38. Exim was not checking that a write() to a log file succeeded. This could
7037 lead to Bad Things if a log got too big, in particular if it hit a file
7038 size limit. Exim now panics and dies if it cannot write to a log file, just
7039 as it does if it cannot open a log file.
7041 39. Modified OS/Makefile-Linux so that it now contains
7043 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7045 The two -D definitions ensure that Exim is compiled with large file
7046 support, which makes it possible to handle log files that are bigger than
7049 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7050 instance) a domain was checked against a named list that involved a lookup,
7051 causing $domain_data to be set, then another domain was checked against the
7052 same list, then the first domain was re-checked, the value of $domain_data
7053 after the final check could be wrong. In particular, if the second check
7054 failed, it could be set empty. This bug probably also applied to
7057 41. The strip_trailing_dot option was not being applied to the address given
7058 with the -f command-line option.
7060 42. The code for reading a message's header from the spool was incrementing
7061 $received_count, but never initializing it. This meant that the value was
7062 incorrect (doubled) while delivering a message in the same process in which
7063 it was received. In the most common configuration of Exim, this never
7064 happens - a fresh exec is done - but it can happen when
7065 deliver_drop_privilege is set.
7067 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7068 it now includes up to 150 characters of the unexpected data in the log
7071 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7072 and building data strings. The size of both of these buffers was 10 000
7073 bytes - far larger than anybody would *ever* want, thought I. Needless to
7074 say, somebody hit the limit. I have increased the maximum line length to
7075 20 000 and the maximum data length of concatenated lines to 100 000. I have
7076 also fixed two bugs, because there was no checking on these buffers. Tsk,
7077 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7078 buffer is too small.
7080 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7081 lsearch lookups. Now it does.
7083 46. When parsing a route_list item in a manualroute router, a fixed-length
7084 buffer was used for the list of hosts. I made this 1024 bytes long,
7085 thinking that nobody would ever have a list of hosts that long. Wrong.
7086 Somebody had a whole pile of complicated expansion conditions, and the
7087 string was silently truncated, leading to an expansion error. It turns out
7088 that it is easier to change to an unlimited length (owing to other changes
7089 that have happened since this code was originally written) than to build
7090 structure for giving a limitation error. The length of the item that
7091 expands into the list of hosts is now unlimited.
7093 47. The lsearch lookup could not handle data where the length of text line was
7094 more than 4095 characters. Such lines were truncated, leading to shortened
7095 data being returned. It should now handle lines of any length.
7097 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7098 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7101 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7102 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7103 no longer happens in comment lines. A list of replacements is now placed
7104 at the head of all of the source files, except those whose only change is
7105 to replace PERL_COMMAND in the very first #! line.
7107 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7108 messages on the queue, with a bottom-up merge sort, using code contributed
7109 by Michael Haardt. This should make operations like -bp somewhat faster on
7110 large queues. It won't affect queue runners, except when queue_run_in_order
7113 51. Installed eximstats 1.31 in the distribution.
7115 52. Added support for SRV lookups to the dnslookup router.
7117 53. If an ACL referred to $message_body or $message_body_end, the value was not
7118 reset for any messages that followed in the same SMTP session.
7120 54. The store-handling optimization for building very long strings was not
7121 differentiating between the different store pools. I don't think this
7122 actually made any difference in practice, but I've tidied it.
7124 55. While running the routers to verify a sender address, $sender_address
7125 was still set to the sender address. This is wrong, because when routing to
7126 send a bounce to the sender, it would be empty. Therefore, I have changed
7127 it so that, while verifying a sender address, $sender_address is set to <>.
7128 (There is no change to what happens when verifying a recipient address.)
7130 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7131 target A or AAAA records (if not already returned) without resetting the
7132 qualify_single or search_parents options of the DNS resolver. These are
7133 inappropriate in this case because the targets of MX and SRV records must
7134 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7135 target that, when qualified, matched something in the local domain. These
7136 two options are now turned off when doing these lookups.
7138 57. It seems that at least some releases of Reiserfs (which does not have the
7139 concept of a fixed number of inodes) returns zero and not -1 for the
7140 number of available inodes. This interacted badly with check_spool_inodes,
7141 which assumed that -1 was the "no such thing" setting. What I have done is
7142 to check that the total number of inodes is greater than zero before doing
7143 the test of how many are available.
7145 58. When a "warn" ACL statement has a log_message modifier, the message is
7146 remembered, and not repeated. This is to avoid a lot of repetition when a
7147 message has many recipients that cause the same warning to be written.
7148 However, Exim was preserving the list of already written lines for an
7149 entire SMTP session, which doesn't seem right. The memory is now reset if a
7150 new message is started.
7152 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7153 debugging output unless log_rewrite was also set.
7155 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7156 of (int)(handle) when we know that handle contains (void *)(-1).
7158 61. The Exim daemon panic-logs an error return when it closes the incoming
7159 connection. However "connection reset by peer" seems to be common, and
7160 isn't really an error worthy of noting specially, so that particular error
7163 62. When Exim is trying to find all the local interfaces, it used to panic and
7164 die if the ioctl to get the interface flags failed. However, it seems that
7165 on at least one OS (Solaris 9) it is possible to have an interface that is
7166 included in the list of interfaces, but for which you get a failure error
7167 for this call. This happens when the interface is not "plumbed" into a
7168 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7169 failure of the "get flags" call assumes that the interface is down.
7171 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7172 makes life easier for people who are doing arithmetic on fields extracted
7173 from dates, where you often get leading zeros that should not be
7174 interpreted as octal.
7176 64. Added qualify_domain to the redirect router, to override the global
7179 65. If a pathologically long header line contained very many addresses (the
7180 report of this problem mentioned 10 000) and each of them was rewritten,
7181 Exim could use up a very large amount of memory. (It kept on making new
7182 copies of the header line as it rewrote, and never released the old ones.)
7183 At the expense of a bit more processing, the header rewriting function has
7184 been changed so that it no longer eats memory in this way.
7186 66. The generation of the Received: header has been moved from the time that a
7187 message starts to be received, to the time that it finishes. The timestamp
7188 in the Received: header should now be very close to that of the <= log
7189 line. There are two side-effects of this change:
7191 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7192 logged header lines no longer include the local Received: line, because
7193 it has not yet been created. The same applies to a copy of the message
7194 that is returned to a non-SMTP sender when a message is rejected.
7196 (b) When a filter file is tested using -bf, no additional Received: header
7197 is added to the test message. After some thought, I decided that this
7200 This change does not affect the value of $received_for. It is still set
7201 after address rewriting, but before local_scan() is called.
7203 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7205 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7206 gave an unhelpful panic error message, and a defer error. I have managed to
7207 change this behaviour so that it now rejects any supplied certificate,
7208 which seems right, as the list of acceptable certificates is empty.
7210 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7211 gave an unhelpful defer error. I have not managed to make this reject any
7212 supplied certificates, but the error message it gives is "no certificate
7213 supplied", which is not helpful.
7215 70. exigrep's output now also includes lines that are not associated with any
7216 message, but which match the given pattern. Implemented by a patch from
7217 Martin Sluka, which also tidied up the Perl a bit.
7219 71. Recipient callout verification, like sender verification, was using <> in
7220 the MAIL FROM command. This isn't really the right thing, since the actual
7221 sender may affect whether the remote host accepts the recipient or not. I
7222 have changed it to use the actual sender in the callout; this means that
7223 the cache record is now keyed on a recipient/sender pair, not just the
7224 recipient address. There doesn't seem to be a real danger of callout loops,
7225 since a callout by the remote host to check the sender would use <>.
7226 [SEE ABOVE: changed after hitting problems.]
7228 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7229 temporary errors. However, in the case of such a code being given after
7230 the end of a data transmission (i.e. after ".") Exim was failing to write
7231 a retry record for the message. (Yes, there was some broken host that was
7232 actually sending 8xx at this point.)
7234 73. An unknown lookup type in a host list could cause Exim to panic-die when
7235 the list was checked. (An example that provoked this was putting <; in the
7236 middle of a list instead of at the start.) If this happened during a DATA
7237 ACL check, a -D file could be left lying around. This kind of configuration
7238 error no longer causes Exim to die; instead it causes a defer error. The
7239 incident is still logged to the main and panic logs.
7241 74. Buglet left over from Exim 3 conversion. The message "too many messages
7242 in one connection" was written to the rejectlog but not the mainlog, except
7243 when address rewriting (yes!) was being logged.
7245 75. Added write_rejectlog option.
7247 76. When a system filter was run not as root (that is, when system_filter_user
7248 was set), the values of the $n variables were not being returned to the
7249 main process; thus, they were not subsequently available in the $sn
7252 77. Added +return_path_on_delivery log selector.
7254 78. A connection timeout was being treated differently from recipients deferred
7255 when testing hosts_max_try with a message that was older than the host's
7256 retry timeout. (The host should not be counted, thus allowing all hosts to
7257 be tried at least once before bouncing.) This may have been the cause of an
7258 occasionally reported bug whereby a message would remain on the queue
7259 longer than the retry timeout, but would be bounced if a delivery was
7260 forced. I say "may" because I never totally pinned down the problem;
7261 setting up timeout/retry tests is difficult. See also the next item.
7263 79. The ultimate address timeout was not being applied to errors that involved
7264 a combination of host plus message (for example, a timeout on a MAIL
7265 command). When an address resolved to a number of possible hosts, and they
7266 were not all tried for each delivery (e.g. because of hosts_max_try), a
7267 message could remain on the queue longer than the retry timeout.
7269 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7272 81. Fixed an obscure SMTP outgoing bug which required at least the following
7273 conditions: (a) there was another message waiting for the same server;
7274 (b) the server returned 5xx to all RCPT commands in the first message so
7275 that the message was not completed; (c) the server dropped the connection
7276 or gave a negative response to the RSET that Exim sends to abort the
7277 transaction. The observed case was a dropped connection after DATA that had
7278 been sent in pipelining mode. That is, the server had advertised PIPELINING
7279 but was not implementing it correctly. The effect of the bug was incorrect
7280 behaviour, such as trying another host, and this could lead to a crash.
7286 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7287 and daemon.c were passed as pointers to ints; they should have been
7288 pointers to socklen_t variables (which are typically unsigned ints).
7290 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7293 3. Fixed a really odd bug that affected only the testing scheme; patching a
7294 certain fixed string in the binary changed the value of another string that
7295 happened to be identical to the end of the original first string.
7297 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7298 name", it returns that address as the IP address. On some operating
7299 systems (e.g. Solaris), it also passes back the IP address string as the
7300 "host name". However, on others (e.g. Linux), it passes back an empty
7301 string. Exim wasn't checking for this, and was changing the host name to an
7302 empty string, assuming it had been canonicalized.
7304 5. Although rare, it is permitted to have more than one PTR record for a given
7305 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7306 all the names associated with an address, because they do in Solaris.
7307 However, it seems that they do not in Linux for data that comes from the
7308 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7309 I found this out when I moved to a new Linux workstation and tried to run
7310 the Exim test suite.
7312 To get round this problem I have changed the code so that it now does its
7313 own call to the DNS to look up PTR records when searching for a host name.
7314 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7315 addresses that are only in /etc/hosts are still found.
7317 This behaviour is, however, controlled by an option called host_lookup_
7318 order, which defaults to "bydns:byaddr". If people want to use the other
7319 order, or indeed, just use one or the other means of lookup, they can
7320 specify it in this variable.
7322 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7323 some operating systems, this comes back from gethostbyaddr() as an empty
7324 string, and this is what Exim used to test for. However, it seems that in
7325 other systems, "." is yielded. Exim now tests for this case too.
7327 7. The values of check_spool_space and check_log_space are now held internally
7328 as a number of kilobytes instead of an absolute number of bytes. If a
7329 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7330 kilobyte. This means that much larger values can be stored.
7332 8. Exim monitor: an attempt to get the action menu when not actually pointing
7333 at a message produces an empty menu entitled "No message selected". This
7334 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7335 no entries in it ("Shell widget menu has zero width and/or height"). So I
7336 have added a single, blank menu entry in this case.
7338 9. Added ${quote_local_part.
7340 10. MIME decoding is now applied to the contents of Subject: header lines when
7343 11. Now that a reference to $sender_host_address automatically causes a reverse
7344 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7345 host lookup before query-style lookups in lists that might use this
7346 variable. This has therefore been abolished, and the "net-" prefix is no
7347 longer necessary for query-style lookups.
7349 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7350 have been a typo for LFLAGS, so it has been changed.
7352 13. The install script calls Exim with "-C /dev/null" in order to find the
7353 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7354 to be output. However, since Exim outputs its version number before the
7355 error, it didn't break the script. It just looked ugly. I fixed this by
7356 always allowing "-C /dev/null" if the caller is root.
7358 14. Ignore overlarge ACL variable number when reading spool file - insurance
7359 against a later release with more variables having written the file.
7361 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7362 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7363 this, as well as the form without the "IPv6" on the front.
7365 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7366 OS/Makefile-Darwin file.
7368 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7369 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7371 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7372 variable. However, some OS use ints for inodes. I've added cast to long int
7373 to get rid of the compiler warning.
7375 19. I had forgotten to lock out "/../" in configuration file names when
7376 ALT_CONFIG_PREFIX was set.
7378 20. Routers used for verification do not need to specify transports. However,
7379 if such a router generated a host list, and callout was configured, Exim
7380 crashed, because it could not find a port number from the (non-existent)
7381 transport. It now assumes port 25 in this circumstance.
7383 21. Added the -t option to exigrep.
7385 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7386 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7387 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7388 them is set, LOOKUP_LSEARCH is forced.
7390 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7392 24. Added sender and host information to the "rejected by local_scan()" log
7393 line; previously there was no indication of these.
7395 25. Added .include_if_exists.
7397 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7398 Linux. It turns out that not all file systems support this. Apparently some
7399 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7400 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7401 not supported on the file descriptor, is now ignored when Exim is trying to
7402 sync a directory. This applies only to Linux.
7404 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7406 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7407 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7410 29. Added support for maildirsize files from supplied patch (modified a bit).
7412 30. The use of :fail: followed by an empty string could lead Exim to respond to
7413 sender verification failures with (e.g.):
7415 550 Verification failed for <xxx>
7416 550 Sender verify failed
7418 where the first response line was missing the '-' that indicates it is not
7419 the final line of the response.
7421 31. The loop for finding the name of the user that called Exim had a hardwired
7422 limit of 10; it now uses the value of finduser_retries, which is used for
7423 all other user lookups.
7425 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7428 33. Exim was neglecting to zero errno before one call of strtol() when
7429 expanding a string and expecting an integer value. On some systems this
7430 resulted in spurious "integer overflow" errors. Also, it was casting the
7431 result into an int without checking.
7433 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7434 did not work. The code looks as if it has *never* worked, though it appears
7435 to have been documented since at least release 1.62. I have made it work.
7437 35. The "timeout_DNS" error in retry rules, also documented since at least
7438 1.62, also never worked. As it isn't clear exactly what this means, and
7439 clearly it isn't a major issue, I have abolished the feature by treating it
7440 as "timeout", and writing a warning to the main and panic logs.
7442 36. The display of retry rules for -brt wasn't always showing the error code
7445 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7446 timeout_connect_A, timeout_connect_MX.
7448 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7449 to the empty sender.
7451 39. The daemon was not analysing the content of -oX till after it had closed
7452 stderr and disconnected from the controlling terminal. This meant that any
7453 syntax errors were only noted on the panic log, and the return code from
7454 the command was 0. By re-arranging the code a little, I've made the
7455 decoding happen first, so such errors now appear on stderr, and the return
7456 code is 1. However, the actual setting up of the sockets still happens in
7457 the disconnected process, so errors there are still only recorded on the
7460 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7461 connections (as happens on some IP stacks) was logged at start up time as
7462 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7463 it from "IPv6 and IPv4", which means that two separate sockets are being
7466 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7467 says whether AF_INET or AF_INET6 was passed as an argument.
7469 42. Exiwhat output was messed up when time zones were included in log
7472 43. Exiwhat now gives more information about the daemon's listening ports,
7473 and whether -tls-on-connect was used.
7475 44. The "port" option of the smtp transport is now expanded.
7477 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7478 silently ignored. Now an error message is written to the main and panic
7481 46. There's a new ACL modifier called "logwrite" which writes to a log file
7482 as soon as it is encountered.
7484 47. Added $local_user_uid and $local_user_gid at routing time.
7486 48. Exim crashed when trying to verify a sender address that was being
7489 49. Exim was recognizing only a space character after ".include". It now also
7490 recognizes a tab character.
7492 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7493 extracting the relevant information from the specification. The man page no
7494 longer contains scrambled data for the -d option, and I've added a section
7495 at the front about calling Exim under different names.
7497 51. Added "extra_headers" argument to the "mail" command in filter files.
7499 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7502 53. Installed eximstats 1.29.
7504 54. Added transport_filter_timeout as a generic transport option.
7506 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7507 Cc: header lines. This was required by RFC 822, but it not required by RFC
7510 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7511 incoming messages that did not have them. Now it does so only if the
7512 message originates locally, that is, if there is no associated remote host
7513 address. When Resent- header lines are present, this applies to the Resent-
7514 lines rather than the non-Resent- lines.
7516 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7517 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7519 58. Messages for configuration errors now include the name of the main
7520 configuration file - useful now that there may be more than one file in a
7521 list (.included file names were always shown).
7523 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7524 for those rare installations that do not start the daemon as root or run it
7525 setuid root. I've cut out the call to initgroups() if the daemon is not
7528 60. The Exim user and group can now be bound into the binary as text strings
7529 that are looked up at the start of Exim's processing.
7531 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7533 62. Added $mailstore_basename variable.
7535 63. Installed patch to sieve.c from Michael Haardt.
7537 64. When Exim failed to open the panic log after failing to open the main log,
7538 the original message it was trying to log was written to stderr and debug
7539 output, but if they were not available (the usual case in production), it
7540 was lost. Now it is written to syslog before the two lines that record the
7541 failures to open the logs.
7543 65. Users' Exim filters run in subprocesses under the user's uid. It is
7544 possible for a "deliver" command or an alias in a "personal" command to
7545 provoke an address rewrite. If logging of address rewriting is configured,
7546 this fails because the process is not running as root or exim. There may be
7547 a better way of dealing with this, but for the moment (because 4.30 needs
7548 to be released), I have disabled address rewrite logging when running a
7549 filter in a non-root, non-exim process.
7555 1. The buildconfig auxiliary program wasn't quoting the value set for
7556 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7557 not defined. This bug was masked in 4.22 by the effect that was fixed in
7560 2. Some messages that were rejected after a message id was allocated were
7561 shown as "incomplete" by exigrep. It no longer does this for messages that
7562 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7564 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7565 have allow_domain_literals set, the ID did not get logged in the <= line.
7566 Domain literals are now always recognized in Message-ID: header lines.
7568 4. The first argument for a ${extract expansion item is the key name or field
7569 number. Leading and trailing spaces in this item were not being ignored,
7570 causing some misleading effects.
7572 5. When deliver_drop_privilege was set, single queue runner processes started
7573 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7574 same command in the process it spins off) were not dropping privilege.
7576 6. When the daemon running as "exim" started a queue runner, it always
7577 re-executed Exim in the spun-off process. This is a waste of effort when
7578 deliver_drop_privilege is set. The new process now just calls the
7579 queue-runner function directly.
7585 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7588 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7589 ignored. Though the use of -oP was forcing the writing of a pid file, it
7590 was always written to the default place.
7592 3. If the message "no IP address found for host xxxx" is generated during
7593 incoming verification, it is now followed by identification of the incoming
7594 connection (so you can more easily find what provoked it).
7596 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7598 5. Added some features to "harden" Exim a bit more against certain attacks:
7600 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7601 be put in Local/Makefile. This is like the never_users runtime option,
7602 but it cannot be overridden. The default setting is "root".
7604 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7605 prefix string with which any file named in a -C command line option
7608 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7609 is retained for -C and -D only if the caller of Exim is root. Without
7610 it, the exim user may also use -C and -D and retain privilege.
7612 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7613 command line option is disabled.
7615 6. Macro names set by the -D option must start with an upper case letter, just
7616 like macro names defined in the configuration file.
7618 7. Added "dereference=" facility to LDAP.
7620 8. Two instances of the typo "uknown" in the source files are fixed.
7622 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7623 the Configure-Makefile script screwed up while processing it.
7625 10. Incorporated PCRE 4.4.
7627 11. The SMTP synchronization check was not operating right at the start of an
7628 SMTP session. For example, it could not catch a HELO sent before the client
7629 waited for the greeting. There is now a check for outstanding input at the
7630 point when the greeting is written. Because of the duplex, asynchronous
7631 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7632 way, but not yet received, when the check is performed.
7634 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7635 on TCP/IP sockets, because this apparently causes some broken clients to
7638 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7639 unchanged) from the Cygwin maintainer.
7641 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7642 was supported instead of testing for mbx. Effectively a typo.
7644 15. The spa authenticator server code was not checking that the input it
7645 received was valid base64.
7647 16. The debug output line for the "set" modifier in ACLs was not showing the
7648 name of the variable that was being set.
7650 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7652 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7653 lookup on its own. Something else was needed to trigger the lookup. For
7654 example, a match in host_lookup or the need for a host name in a host list.
7655 Now, if $sender_host_name is referenced and the host name has not yet been
7656 looked up, a lookup is performed. If the lookup fails, the variable remains
7657 empty, and $host_lookup_failed is set to "1".
7659 19. Added "eqi" as a case-independent comparison operator.
7661 20. The saslauthd authentication condition could segfault if neither service
7662 nor realm was specified.
7664 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7665 error message that was logged was misleading, and incoming SMTP
7666 connections were dropped. The message is now more accurate, and temporary
7667 errors are given to SMTP connections.
7669 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7670 (except RSET) with a 421 error, until QUIT is received. However, it was
7671 failing to send a response to QUIT.
7673 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7674 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7675 ACL is now run after the helo_try_verify_hosts code.
7677 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7678 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7679 case-independent, so other case variants are also recognized). Apparently
7680 some systems use these upper case variants.
7682 25. If more than two messages were waiting for the same host, and a transport
7683 filter was specified for the transport, Exim sent two messages over the
7684 same TCP/IP connection, and then failed with "socket operation on non-
7685 socket" when it tried to send the third.
7687 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7689 27. The extern definition of crypt16() in expand.c was not being excluded when
7690 the OS had its own crypt16() function.
7692 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7693 as a preferred synonym for return_size_limit, both as an option and as an
7696 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7698 30. Changed the default configuration ACL to relax the local part checking rule
7699 for addresses that are not in any local domains. For these addresses,
7700 slashes and pipe symbols are allowed within local parts, but the sequence
7701 /../ is explicitly forbidden.
7703 31. SPA server authentication was not clearing the challenge buffer before
7706 32. log_message in a "warn" ACL statement was writing to the reject log as
7707 well as to the main log, which contradicts the documentation and doesn't
7708 seem right (because no rejection is happening). So I have stopped it.
7710 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7711 However, I am unable to do any testing of this.
7713 34. Fixed an infelicity in the appendfile transport. When checking directories
7714 for a mailbox, to see if any needed to be created, it was accidentally
7715 using path names with one or more superfluous leading slashes; tracing
7716 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7718 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7719 contents are added to the log line that is written for every discarded
7720 recipient. (Previously a log_message setting was ignored.)
7722 36. The ${quote: operator now quotes the string if it is empty.
7724 37. The install script runs exim in order to find its version number. If for
7725 some reason other than non-existence or emptiness, which it checks, it
7726 could not run './exim', it was installing it with an empty version number,
7727 i.e. as "exim-". This error state is now caught, and the installation is
7730 38. An argument was missing from the function that creates an error message
7731 when Exim fails to connect to the socket for saslauthd authentication.
7732 This could cause Exim to crash, or give a corrupted message.
7734 39. Added isip, isip4, and isip6 to ${if conditions.
7736 40. The ACL variables $acl_xx are now saved with the message, and can be
7737 accessed later in routers, transports, and filters.
7739 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7740 strings in the file are not string-expanded.
7742 42. If a MAIL command specified a SIZE value that was too large to fit into an
7743 int variable, the check against message_size_limit failed. Such values are
7744 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7745 day this will have to be increased, but I don't think I want to be around
7746 when emails are that large.
7753 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7754 iconv() is not standard in FreeBSD.
7756 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7757 IPv6 enabled. The observed symptom was a segmentation fault on return from
7758 the function os_common_find_running_interfaces() in src/os.c.
7760 3. In the check_special_case() function in daemon.c I had used "errno" as an
7761 argument name, which causes warnings on some systems. This was basically a
7762 typo, since it was named "eno" in the comments!
7764 4. The code that waits for the clock to tick (at a resolution of some fraction
7765 of a second) so as to ensure message-id uniqueness was always waiting for
7766 at least one whole tick, when it could have waited for less. [This is
7767 almost certainly not relevant at current processor speeds, where it is
7768 unlikely to ever wait at all. But we try to future-proof.]
7770 5. The function that sleeps for a time interval that includes fractions of a
7771 second contained a race. It did not block SIGALRM between setting the
7772 timer, and suspending (a couple of lines later). If the interval was short
7773 and the sigsuspend() was delayed until after it had expired, the suspension
7774 never ended. On busy systems this could lead to processes getting stuck for
7777 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7778 process, before it forks any delivery processes. The open lookup caching
7779 mechanism meant that the open file or database connection was passed into
7780 the delivery process. The problem was that delivery processes always tidy
7781 up cached lookup data. This could cause a problem for the next delivery
7782 process started by the queue runner, because the external queue runner
7783 process does not know about the closure. So the next delivery process
7784 still has data in the lookup cache. In the case of a file lookup, there was
7785 no problem because closing a file descriptor in a subprocess doesn't affect
7786 the parent. However, if the lookup was caching a connection to a database,
7787 the connection was closed, and the second delivery process was likely to
7788 see errors such as "PGSQL: query failed: server closed the connection
7789 unexpectedly". The problem has been fixed by closing all cached lookups
7790 in a queue runner before running a delivery process.
7792 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7793 seem to have the "const" qualifier which it has on other OS. I've
7796 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7797 *of the same type* (client or server) with the same public name that an
7798 error should be diagnosed.
7800 9. When Exim looked up a host name for an IP address, but failed to find the
7801 original IP address when looking up the host name (a safety check), it
7802 output the message "<ip address> does not match any IP for NULL", which was
7803 confusing, to say the least. The bug was that the host name should have
7804 appeared instead of "NULL".
7806 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7807 user that is built into the binary, and the -C or -D options is used, root
7808 privilege is dropped before the configuration file is read. In addition,
7809 logging is switched to stderr instead of the normal log files. If the
7810 configuration then re-defines the Exim user, the unprivileged environment
7811 is probably not what is expected, so Exim logs a panic warning message (but
7814 However, if deliver_drop_privilege is set, the unprivileged state may well
7815 be exactly what is intended, so the warning has been cut out in that case,
7816 and Exim is allowed to try to write to its normal log files.
7822 1. smtp_return_error_details was not giving details for temporary sender
7823 or receiver verification errors.
7825 2. Diagnose a configuration error if two authenticators have the same public
7828 3. Exim used not to create the message log file for a message until the first
7829 delivery attempt. This could be confusing when incoming messages were held
7830 for policy or load reasons. The message log file is now created at the time
7831 the message is received, and an initial "Received" line is written to it.
7833 4. The automatically generated man page for command line options had a minor
7834 bug that caused no ill effects; however, a more serious problem was that
7835 the procedure for building the man page automatically didn't always
7836 operate. Consequently, release 4.20 contains an out-of-date version. This
7837 shouldn't happen again.
7839 5. When building Exim with embedded Perl support, the script that builds the
7840 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7841 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7843 6. The freeze_tell option was not being used for messages that were frozen on
7844 arrival, either by an ACL or by local_scan().
7846 7. Added the smtp_incomplete_transaction log selector.
7848 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7849 was accepting AUTH without a new EHLO.
7851 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7852 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7855 10. Logging of TCP/IP connections (when configured) now happens in the main
7856 daemon process instead of the child process, so that the TCP/IP connection
7857 count is more accurate (but it can never be perfect).
7859 11. The use of "drop" in a nested ACL was not being handled correctly in the
7860 outer ACL. Now, if condition failure induced by the nested "drop" causes
7861 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7862 or "require"), the connection is dropped.
7864 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7865 that yield "discard" can now be used with an "accept" or a "discard" verb,
7866 but an error is generated for any others (because I can't see a useful way
7867 to define what should happen).
7869 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7870 are now processed in the same way as lines in the Exim configuration file.
7871 In particular, continuation lines are supported.
7873 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7875 15. Added -ti meaning -t -i.
7877 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7878 domains, and warn by logging if others are found.
7880 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7881 returned from GIFCONF when Exim is trying to find the list of interfaces on
7882 a host. The code in os.c has been modified to copy each ifreq to an aligned
7883 structure in all cases.
7885 Also, in some cases, the returned ifreq's were being copied to a 'struct
7886 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7887 means the last couple of bytes of an IPv6 address could be chopped if the
7888 ifreq contained only a normal sockaddr (14 bytes storage).
7890 18. Named domain lists were not supported in the hosts_treat_as_local option.
7891 An entry such as +xxxx was not recognized, and was treated as a literal
7894 19. Ensure that header lines added by a DATA ACL are included in the reject log
7895 if the ACL subsequently rejects the message.
7897 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7898 MD5 (which is deprecated).
7900 21. When testing a filter file using -bf, Exim was writing a message when it
7901 took the sender from a "From " line in the message, but it was not doing so
7902 when it took $return_path from a Return-Path: header line. It now does.
7904 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7905 with a valid header line field name (a series of printing characters
7906 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7908 23. Changed "disc" in the source to "disk" to conform to the documentation and
7909 the book and for uniformity.
7911 24. Ignore Sendmail's -Ooption=value command line item.
7913 25. When execve() failed while trying to run a command in a pipe transport,
7914 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7915 could be confused with a return value of 69 from the command itself. This
7916 has been changed to 127, the value the shell returns if it is asked to run
7917 a non-existent command. The wording for the related log line suggests a
7918 non-existent command as the problem.
7920 26. If received_header_text expands to an empty string, do not add a Received:
7921 header line to the message. (Well, it adds a token one on the spool, but
7922 marks it "old" so that it doesn't get used or transmitted.)
7924 27. Installed eximstats 1.28 (addition of -nt option).
7926 28. There was no check for failure on the call to getsockname() in the daemon
7927 code. This can fail if there is a shortage of resources on the system, with
7928 ENOMEM, for example. A temporary error is now given on failure.
7930 29. Contrary to the C standard, it seems that in some environments, the
7931 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7932 program. Exim now does this explicitly; it affects the formatting of
7933 timestamps using strftime().
7935 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7936 complaints. I've now initialized all the variables, to avoid this.
7938 32. Header lines added by a system filter were not being "seen" during
7939 transport-time rewrites.
7941 33. The info_callback() function passed to OpenSSL is set up with type void
7942 (*)(SSL *, int, int), as described somewhere. However, when calling the
7943 function (actually a macro) that sets it up, the type void(*)() is
7944 expected. I've put in a cast to prevent warnings from picky compilers.
7946 34. If a DNS black list lookup found a CNAME record, but there were no A
7947 records associated with the domain it pointed at, Exim crashed.
7949 35. If a DNS black list lookup returned more than one A record, Exim ignored
7950 all but the first. It now scans all returned addresses if a particular IP
7951 value is being sought. In this situation, the contents of the
7952 $dnslist_value variable are a list of all the addresses, separated by a
7955 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7956 to accept a host name and all its aliases if the forward lookup for any of
7957 them yielded the IP address of the incoming connection. Now it accepts only
7958 those names whose forward lookup yields the correct IP address. Any other
7959 names are discarded. This closes a loophole whereby a rogue DNS
7960 administrator could create reverse DNS records to break through a
7961 wildcarded host restriction in an ACL.
7963 37. If a user filter or a system filter that ran in a subprocess used any of
7964 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7965 the wrong values were passed to the pipe command ($thisaddress had the
7966 value of $0, $0 had the value of $1, etc). This bug was introduced by
7967 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7969 38. Improved the line breaking for long SMTP error messages from ACLs.
7970 Previously, if there was no break point between 40 and 75 characters, Exim
7971 left the rest of the message alone. Two changes have been made: (a) I've
7972 reduced the minimum length to 35 characters; (b) if it can't find a break
7973 point between 35 and 75 characters, it looks ahead and uses the first one
7974 that it finds. This may give the occasional overlong line, but at least the
7975 remaining text gets split now.
7977 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7978 file descriptors might be low, and that setting 1000 would always raise it.
7979 It turns out that in some environments, the limit is already over 1000 and
7980 that lowering it causes trouble. So now Exim takes care not to decrease it.
7982 40. When delivering a message, the value of $return_path is set to $sender_
7983 address at the start of routing (routers may change the value). By an
7984 oversight, this default was not being set up when an address was tested by
7985 -bt or -bv, which affected the outcome if any router or filter referred to
7988 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7989 log only when "message" or "log_message" are set. However, if one of the
7990 conditions was an address verification, or a call to a nested ACL, the
7991 messages generated by the underlying test were being passed through. This
7992 no longer happens. The underlying message is available in $acl_verify_
7993 message for both "message" and "log_message" expansions, so it can be
7994 passed through if needed.
7996 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7997 new expansion $bh_ to give the encoded byte string without charset
7998 translation. Translation happens only if iconv() is available; HAVE_ICONV
7999 indicates this at build time. HEADERS_CHARSET gives the charset to
8000 translate to; headers_charset can change it in the configuration, and
8001 "headers charset" can change it in an individual filter file.
8003 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8004 that creates RFC 2047 encoded "words" labels them as that charset instead
8005 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8006 expansion operator; (ii) when Exim creates a From: line for a local
8007 message; (iii) when a header line is rewritten to include a "phrase" part.
8009 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8010 buggy, causing it to skip the first lines of messages whose message ID
8011 ended in 'D'. This would not have bitten before Exim release 4.14, because
8012 message IDs were unlikely to end in 'D' before then. The effect was to have
8013 incorrect size information for certain domains.
8015 45. #include "config.h" was missing at the start of the crypt16.c module. This
8016 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8019 46. If there was a timeout during a "random" callout check, Exim treated it as
8020 a failure of the random address, and carried on sending RSET and the real
8021 address. If the delay was just some slowness somewhere, the response to the
8022 original RCPT would be taken as a response to RSET and so on, causing
8023 mayhem of various kinds.
8025 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8026 when I implemented it. It didn't allow for the fact that some option values
8027 may legitimately be negative (e.g. size_addition), and it didn't even do
8028 the right test for positive values.
8030 48. Domain names in DNS records are case-independent. Exim always looks them up
8031 in lower case. Some resolvers return domain names in exactly the case they
8032 appear in the zone file, that is, they may contain uppercase letters. Not
8033 all resolvers do this - some return always lower case. Exim was treating a
8034 change of case by a resolver as a change of domain, similar to a widening
8035 of a domain abbreviation. This triggered its re-routing code and so it was
8036 trying to route what was effectively the same domain again. This normally
8037 caused routing to fail (because the router wouldn't handle the domain
8038 twice). Now Exim checks for this case specially, and just changes the
8039 casing of the domain that it ultimately uses when it transmits the message
8042 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8045 50. If a filter generated a file delivery with a non-absolute name (possible if
8046 no home directory exists for the router), the forbid_file option was not
8049 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8050 the existing equality matching.
8052 52. Exim was using ints instead of ino_t variables in some places where it was
8053 dealing with inode numbers.
8055 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8056 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8057 called TMPDIR, and if it finds it is different, it changes its value.
8059 54. The smtp_printf() function is now made available to local_scan() so
8060 additional output lines can be written before returning. There is also an
8061 smtp_fflush() function to enable the detection of a dropped connection.
8062 The variables smtp_input and smtp_batched_input are exported to
8065 55. Changed the default runtime configuration: the message "Unknown user"
8066 has been removed from the ACL, and instead placed on the localuser router,
8067 using the cannot_route_message feature. This means that any verification
8068 failures that generate their own messages won't get overridden. Similarly,
8069 the "Unrouteable address" message that was in the ACL for unverifiable
8070 relay addresses has also been removed.
8072 56. Added hosts_avoid_esmtp to the smtp transport.
8074 57. The exicyclog script was not checking for the esoteric option
8075 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8076 will work only if exicyclog is run under the appropriate euid.
8078 58. Following a discussion on the list, the rules by which Exim recognises line
8079 endings on incoming messages have been changed. The -dropcr and drop_cr
8080 options are now no-ops, retained only for backwards compatibility. The
8081 following line terminators are recognized: LF CRLF CR. However, special
8082 processing applies to CR:
8084 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8085 nor a local message in the state where . is a terminator.
8087 (ii) If a bare CR is encountered in a header line, an extra space is added
8088 after the line terminator so as not to end the header. The reasoning
8089 behind this is that bare CRs in header lines are most likely either
8090 to be mistakes, or people trying to play silly games.
8092 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8093 was being incorrectly given as 18 bytes larger than it should have been.
8094 This is a VOB (very old bug).
8096 60. This may never have affected anything current, but just in case it has:
8097 When the local host is found other than at the start of a list of hosts,
8098 the local host, those with the same MX, and any that follow, are discarded.
8099 When the list in question was part of a longer list of hosts, the following
8100 hosts (not currently being processed) were also being discarded. This no
8101 longer happens. I'm not sure if this situation could ever has previously
8104 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8107 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8108 header. This is something that is recommended in a new Internet Draft, and
8109 is something that is documented as being done by Sendmail. There are two
8110 possible values. For messages generated by the autoreply transport, Exim
8113 Auto-Submitted: auto-replied
8115 whereas for all other generated messages (e.g. bounces) it adds
8117 Auto-Submitted: auto-generated
8119 63. The "personal" condition in filters now includes a test for the
8120 Auto-Submitted: header. If it contains the string "auto-" the message it
8121 not considered personal.
8123 64. Added rcpt_include_affixes as a generic transport option.
8125 65. Added queue_only_override (default true).
8127 66. Added the syslog_duplication option.
8129 67. If what should have been the first header line of a message consisted of
8130 a space followed by a colon, Exim was mis-interpreting it as a header line.
8131 It isn't of course - it is syntactically invalid and should therefore be
8132 treated as the start of the message body. The misbehaviour could have
8133 caused a number of strange effects, including loss of data in subsequent
8134 header lines, and spool format errors.
8136 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8137 client host had authenticated. This control can now be exercised by an ACL
8138 for more flexibility.
8140 69. By default, callouts do not happen when testing with -bh. There is now a
8141 variant, -bhc, which does actually run the callout code, including
8142 consulting and updating the callout cache.
8144 70. Added support for saslauthd authentication, courtesy of Alexander
8147 71. If statvfs() failed on the spool or log directories while checking their
8148 size for availability, Exim confusingly gave the error "space shortage".
8149 Furthermore, in debugging mode it crashed with a floating point exception.
8150 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8151 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8152 serious problem, Exim now writes to the main and panic logs when this
8153 happens, with details of the failure. It then refuses to accept the
8154 incoming message, giving the message "spool directory problem" or "log
8155 directory problem" with a 421 code for SMTP messages.
8157 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8158 0, 1, and 2 exist, because some OS complain for execs without them (see
8159 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8160 descriptors. However, the code omitted to check that the open succeeded,
8161 causing mysterious errors if for some reason the permissions on /dev/null
8162 got screwed. Now Exim writes a message to the main and panic logs, and
8163 bombs out if it can't open /dev/null.
8165 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8166 interact so that it is all more flexible. It is supposed to remain
8167 backwards compatible. Also added extra_local_interfaces.
8169 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8170 to bomb out with an assertion failure - to the client this appears as a
8171 connection drop. This problem occurs in the part of the code that was taken
8172 from the Samba project. Fortunately, the assertion is in a very simple
8173 function, so I have fixed this by reproducing the function inline in the
8174 one place where it is called, and arranging for authentication to fail
8175 instead of killing the process with assert().
8177 75. The SPA client code was not working when the server requested OEM rather
8178 than Unicode encoding.
8180 76. Added code to make require_files with a specific uid setting more usable in
8181 the case where statting the file as root fails - usually a non-root-mounted
8182 NFS file system. When this happens and the failure is EACCES, Exim now
8183 forks a subprocess and does the per-uid checking as the relevant uid.
8185 77. Added process_log_path.
8187 78. If log_file_path was not explicitly set, a setting of check_log_space or
8188 check_log_inodes was ignored.
8190 79. If a space check for the spool or log partitions fails, the incident is now
8191 logged. Of course, in the latter case the data may get lost...
8193 80. Added the %p formatting code to string_format() so that it can be used to
8194 print addresses in debug_print(). Adjusted all the address printing in the
8195 debugging in store.c to use %p rather than %d.
8197 81. There was a concern that a line of code in smtp_in.c could overflow a
8198 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8199 initially expressed, the concern was not well-founded, because trailing
8200 spaces are removed early. However, if the trailing spaces were followed by
8201 a NULL, they did not get removed, so the overflow was possible. Two fixes
8204 (a) I re-wrote the offending code in a cleaner fashion.
8205 (b) If an incoming SMTP command contains a NULL character, it is rejected
8208 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8209 runs initgroups(), so that if the Exim user is in any additional groups,
8210 they will be used during message reception.
8216 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob