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.
95 JH/22 Fix static address-list lookups to properly return the matched item.
96 Previously only the domain part was returned.
102 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
103 related applications. Previously an "H" was used where available info
104 says that "M" should be, so change to match.
106 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
107 as arguments, so an implementation trying to copy these into a local
108 buffer was taking a taint-enforcement trap. Fix by using dynamically
109 created buffers. Similar fix for radius expansion condition.
111 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
112 reasonable, eg. to count headers. Fix by using dynamically created
113 buffers rather than a local. Do similar fixes for ACL actions "dcc",
114 "log_reject_target", "malware" and "spam"; the arguments are expanded
115 so could be handling tainted values.
117 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
118 broken the (no-op) support for this sendmail command. Restore it
119 to doing nothing, silently, and returning good status.
121 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
122 record path was given (or the default used) without a leading directory
123 path, an error occurred on trying to open it. Use the transport's working
126 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
127 transport. Previously it was the name on the DNS A-record; use instead
128 the head of the CNAME chain leading there (if there is one). This seems
129 to align better with RFC 6125.
131 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
132 smtp_accept_max_per_host allocated resources which were not released
133 when the limit was exceeded. This eventually crashed the daemon. Fix
134 by adding a release action in that path.
136 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
137 expanded; previously using tainted values was rejected. Fix by using
138 dynamically-created buffers.
140 JH/09 Relax restrictions on ACL verify condition needing access to message
141 headers. Previously they were only permitted in data and non-smtp ACLs;
142 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
143 not_blind, header_sender and header_names_ascii verification.
145 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
146 Previously a macro used one argument twice; when called with the
147 argument as an expression having side-effects, incorrect operation
148 resulted. Use an inlineable function.
150 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
151 held open for a verify callout. Previously this wan not accounted for
152 and a corrupt onward SMTP conversation resulted.
154 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
155 Previously they were embedded naively in the constructed address; when
156 needed, strip the quoting and quote the entire local_part.
157 Also make the inbound_srs expansion condition handle quoting.
159 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
160 excluded, not matching the documentation.
162 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
163 was given for the sqlite_dbfile a trap resulted.
165 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
166 "name" argument resulted in a trap. There is no reason to disallow such;
167 this was a coding error.
169 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
170 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
171 spent suspended, ignoring the POSIX definition. Previously we assumed
172 it did and a constant offset from real time could be used as a correction.
173 Change to using the same clock source for the start-of-message and the
174 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
175 exists, just to get a clock slightly more aligned to reality.
177 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
178 RFC says it is optional some validators care. The missing char was not
179 intended but triggered by a line-wrap alignment. Discovery and fix by
180 Guillaume Outters, hacked on by JH.
182 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
183 name being quoted was tainted a trap would be taken. Fix by using
184 dynamically created buffers. The routine could have been called by a
185 rewrite with the "h" flag, by using the "-F" command-line option, or
186 by using a "name=" option on a control=submission ACL modifier.
188 JH/19 SPF: change the Authentication-Results expansion component to give
189 smtp.helo when the sender domain is empty. Previously it gave
192 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
193 not in 127.0.0.0/8 to help in spotting list domains taken over by a
194 domain-parking registrar.
196 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
197 Previously when a whitespace character was specified it was not inserted
198 after removing the newline.
200 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
201 the domain part of the recipient address. This overrides any tls_sni
202 option set, which was previously used.
204 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
207 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
208 is_tainted() had an off-by-one error in the overenthusiastic direction.
209 Find and fix by Gavan. Although NetBSD is not a supported platform for
210 4.94 this bug could affect other platforms.
212 PP/01 Fix default prime selection to be consistent.
213 One path used ike23 still, instead of exim.dev.20160529.3; now both
214 execution flows will use the same DH primes (currently
215 exim.dev.20160529.3).
217 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
218 option in smtp transport, to match the documentation. Previously
219 verification was not being done in some cases where it should have been.
221 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
222 than one server was defined and depending on the platform memory layout
223 details, an internal consistency trap could be hit while walking the list
226 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
227 files. The value is used by the authresults expansion item. Previously
228 if this was used in a router or transport, a crash could result.
230 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
231 have a sufficient incidence of EINTR returns from select that an
232 interaction with other operations done by the main daemon loop exposed
233 a bug in the error-handling. This was benign apart from the log
236 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
237 for consistency with delivered "=>" and failed "**" lines. While we're
238 there, handle PRX and TFO.
240 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
241 applied. This resulted, if any header-line rewrite rules were configured,
242 in a panic-log triggerable by sending a message with a long address in
243 a header. Fix by increasing the arbitrary limit to larger than a single
244 (dewrapped) 5322 header line maximum size.
246 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
247 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
248 RFC 6648 which deprecates X- options in protocols as a general practice.
249 Changeover between the implementations is handled by the mechanisms
252 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
253 a remote which uses callout/hold. Previously the local was queued.
255 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
258 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
259 log line. The asterisk is now consistently appended to the remote IP
260 (and port, if given), and will also be provided on defer and fail log
261 lines. Previously it could be placed on the local IP if that was being
262 logged, and was only provided on delivery lines.
264 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
266 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
267 authenticator client_send option. Previously the next char, after a pair
268 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
269 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
270 documentation. There is still no way to get a leading ^ immediately
271 after a NUL (ie. for the password of a PLAIN method authenticator.
273 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
274 files. For bad sizes read, delete the record and whine to paniclog.
276 JH/38 When logging an AUTH failure, as server, do not include sensitive
277 information. Previously, the credentials would be included if given
278 as part of the AUTH command line and an ACL denied authentication.
280 JH/39 Bug 2691: fix $local_part_data. When the matching list element
281 referred to a file, bad data was returned. This likely also affected
284 JH/40 The gsasl authenticator now supports caching of the salted password
285 generated by the client-side implementation. This required the addition
286 of a new variable: $auth4.
288 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
289 left undeleted; the attempt to re-create it then failed - resulting in
290 the usual "SIGHUP tp have daemon reload configuration" to not work.
291 This affected any platform not supporting "abstract" Unix-domain
292 sockets (i.e. not Linux).
294 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
295 recipients" response to RCPT in a later response, with a 250. The
296 previous coding assumed this would not happen, and under PIPELINING
297 would result in both lost and duplicate recipients for a message.
299 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
300 Previously the weighting was incorrectly applied. Similar fix for socks
301 proxies. Found and fixed by Heiko Schlichting.
303 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
304 not handle sub-lists included using the +namedlist syntax. While
305 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
306 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
309 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
310 Previously the permanent pool was used, so the sore could not be freed.
311 This meant a connection with many messages would use continually-growing
314 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
315 per-pool so as not to waste too much space. Previously a constant size
316 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
317 DOS attacks harder. The cost is wasted memory use in the larger blocks.
319 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
320 same space repeatedly, and a smaller process footprint.
322 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
323 for GnuTLS. Previously the name started "%s" which, while not a bug,
324 looked as if if might be one.
326 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
327 "continued connection") make the $proxy_* variables available. Previously
328 the information was not passed across the exec() call for subsequent
329 transport executions. This also mean that the log lines for the
330 messages can show the proxy information.
332 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
333 receive time. With modern systems the difference is significant.
334 The historical behaviour can be restored by disabling (a new) log_selector
335 "queue_time_exclusive".
337 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
338 content) were specifically tested for. Now, also cover expansions
339 resulting in ACL names and inline ACL content.
341 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
342 making it unusable in complex expressions.
344 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
345 fail with a taint-check on the spoolfile name, and leave the message
348 HS/01 Enforce absolute PID file path name.
350 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
352 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
354 PP/02 Bug 2643: Correct TLS DH constants.
355 A missing NUL termination in our code-generation tool had led to some
356 incorrect Diffie-Hellman constants in the Exim source.
357 Reported by kylon94, code-gen tool fix by Simon Arlott.
359 PP/03 Impose security length checks on various command-line options.
360 Fixes CVE-2020-SPRSS reported by Qualys.
362 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
363 better. Reported by Qualys.
365 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
366 providing a particularly obnoxious sender full name.
369 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
371 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
374 PP/08 Change default for recipients_max from unlimited to 50,000.
376 PP/09 Fix security issue with too many recipients on a message (to remove a
377 known security problem if someone does set recipients_max to unlimited,
378 or if local additions add to the recipient list).
379 Fixes CVE-2020-RCPTL reported by Qualys.
381 PP/10 Fix security issue in SMTP verb option parsing
382 Fixes CVE-2020-EXOPT reported by Qualys.
384 PP/11 Fix security issue in BDAT state confusion.
385 Ensure we reset known-good where we know we need to not be reading BDAT
386 data, as a general case fix, and move the places where we switch to BDAT
387 mode until after various protocol state checks.
388 Fixes CVE-2020-BDATA reported by Qualys.
390 HS/03 Die on "/../" in msglog file names
392 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
393 the Exim runtime user are allowed to create files.
395 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
398 QS/03 When reading the output from interpreted forward files we do not
399 pass the pipe between the parent and the interpreting process to
400 executed child processes (if any).
402 QS/04 Always die if requested from internal logging, even is logging is
405 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
406 the API; compilation noo longer completes with DMARC support included.
407 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
408 been reported on other platforms.
410 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
412 JH/56 Make the majority of info read from config files readonly, for defence-in-
413 depth against exploits. Suggestion by Qualys.
414 Not supported on Solaris 10.
416 JH/57 Fix control=fakreject for a custom message containing tainted data.
417 Previously this resulted in a log complaint, due to a re-expansion present
418 since fakereject was originally introduced.
420 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
421 modifier was given, a loop resulted.
423 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
424 pipelined SMTP command followed the BDAT LAST then it would be
425 incorrectly treated as part of the message body, causing a verification
428 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
429 require looser validation rules than those for 821-level addresses,
430 which only permit IP addresses.
436 JH/01 Avoid costly startup code when not strictly needed. This reduces time
437 for some exim process initialisations. It does mean that the logging
438 of TLS configuration problems is only done for the daemon startup.
440 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
442 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
443 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
446 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
447 DKIM signing is being done. Previously a transport_filter would always
448 disable CHUNKING, falling back to traditional DATA.
450 JH/05 Regard command-line recipients as tainted.
452 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
454 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
455 PAM library frees one of the arguments given to it, despite the
456 documentation. Therefore a plain malloc must be used.
458 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
459 on-stack buffers were used, resulting in a taint trap when DSN information
460 copied from a received message was written into the buffer.
462 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
463 the ordering of its ARC headers. This caused a crash.
465 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
466 a new record was being constructed with information from the peer, a trap
469 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
470 installation would get error messages from DMARC verify, when it hit the
471 nonexistent file indicated by the default. Distros wanting DMARC enabled
472 should both provide the file and set the option.
473 Also enforce no DMARC verification for command-line sourced messages.
475 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
476 could, depending on the platform, hang at the STARTTLS response.
478 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
479 message on a connection. Previously if one message had ARC headers and
480 the following one did not, a crash could result when adding an
481 Authentication-Results: header.
483 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
484 Exim main code and Exim-related utities. The introduction of taint
485 tracking also did many adjustments to string handling. Since then, eximon
486 frequently terminated with an assert failure.
488 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
489 check for 452 responses. This slightly helps the inefficieny of doing
490 a large alias-expansion into a recipient-limited target. The max_rcpt
491 transport option still applies (and at the current default, will override
492 the new feature). The check is done for either cause of synch, and forces
493 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
494 connection. The new facility is not tunable at this time.
496 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
497 library live data was being used, so the results became garbage. Make
498 copies while it is still usable.
500 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
501 on delivery deferred (==) and failed (**) lines (if a delivery was
502 attemtped). Previously it was only on completion (=>) lines.
504 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
505 for the expansion of the server_scram_iter and server_scram_salt options.
507 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
508 are now specifically given a NO_DATA response without hitting the system
509 resolver. The library goes on to do the now-standard TXT lookup.
510 Use of dnsdb lookups is not affected.
512 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
513 only retrieve the errormessage once. Previously two calls to dlerror()
514 were used, and the second one (for mainlog/paniclog) retrieved null
517 JH/20 Taint checking: disallow use of tainted data for
518 - the appendfile transport file and directory options
519 - the pipe transport command
520 - the autoreply transport file, log and once options
521 - file names used by the redirect router (including filter files)
523 - paths used by single-key lookups
524 Previously this was permitted.
526 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
527 adjusted the size of a major service buffer; this failed because the
528 buffer was in use at the time. Change to a compile-time increase in the
529 buffer size, when this authenticator is compiled into exim.
531 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
532 previous fast-mode was untenable in the face of glibs using mmap to
533 support larger malloc requests.
535 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
536 New values supported, if defined on system where compiled:
537 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
538 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
540 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
541 running a limited number of proceses in parallel, a benefit is gained. The
542 amount varies with the platform hardware and load. The use of the option
543 queue_run_in_order means we cannot do this, as ordering becomes
546 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
547 had introduced a string-copy (for ensuring NUL-termination) which was not
548 appropriate for that case, which can include embedded NUL bytes in the
549 block of data. Investigation showed the copy to actually be needless, the
550 data being length-specified.
552 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
553 done during a receiving connection, and both used TLS, global info was
554 used rather than per-connection info for tracking the state of data
555 queued for transmission. This could result in a connection hang.
557 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
558 Previously, when delivering serveral messages down a single connection
559 only the first would provide a SIZE. This was due to the size information
560 not being properly tracked.
562 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
563 TAI (at 37 seconds currently), pretend to be in UTC for time-related
564 expansion and logging. Previously, spurious values such as a future
565 minute could be seen.
567 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
568 it could crash from a null-deref. This could also affect the
569 ${addresses: } operator and ${readsock } item.
571 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
572 message following a mime one, the variable was not reset.
574 JH/30 When an pipelined-connect fails at the first response, assume incorrect
575 cached capability (perhaps the peer reneged?) and immediately retry in
578 JH/31 Fix spurious detection of timeout while writing to transport filter.
580 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
581 an attempt to copy the string was made before checking it.
583 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
584 taint of the lookup key was maintained; we now regard the presence in the
585 filesystem as sufficient validation.
587 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
588 argument is supplied.
590 JH/35 The dsearch lookup now requires that the directory is an absolute path.
591 Previously this was not checked, and nonempty relative paths made an
592 access under Exim's current working directory.
594 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
595 Previously no event was raised.
597 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
598 parameter supplied by the sender MAIL FROM command. Previously it was
599 ignored, and only the check_spool_space option value for the required
602 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
603 the size of the signing public-key. Previously it was instead giving
604 the size of the signature hash.
606 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
607 the default. See the (new) dkim_verify_min_keysizes option.
609 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
610 and an ACL use a lookup for checking either the local_part or domain,
611 stale data could be accessed. Ensure that variable references are
612 dropped between messages.
614 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
615 by the client was not checked as pointing within response data before
616 being used. A malicious client could thus cause an out-of-bounds read and
617 possibly gain authentication. Fix by adding the check.
619 JH/42 Internationalisation: change the default for downconversion in the smtp
620 transport to be "if needed". Previously it was "as previously set" for
621 the message, which usually meant "if needed" for message-submission but
622 "no" for everything else. However, MTAs have been seen using SMTPUTF8
623 even when the envelope addresses did not need it, resulting in forwarding
624 failures to non-supporting MTAs. A downconvert in such cases will be
625 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
626 transport. The change does mean that addresses needing conversion will
627 be converted when previously a delivery failure would occur.
629 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
630 response was received it would be used unchecked in a fail-DSN, violating
631 standards on line-length limits. Truncate if needed.
633 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
634 doesn't work. (Additionally add a new main config option to configure the
641 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
642 side, to decode a TLS 1.3 packet capture.
644 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
645 Previously the default library behaviour applied, sending two, each in
648 JH/03 Debug output for ACL now gives the config file name and line number for
651 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
653 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
655 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
656 buffer overrun for (non-chunking) other transports.
658 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
659 TLS1.3, means that a server rejecting a client certificate is not visible
660 to the client until the first read of encrypted data (typically the
661 response to EHLO). Add detection for that case and treat it as a failed
662 TLS connection attempt, so that the normal retry-in-clear can work (if
663 suitably configured).
665 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
666 and/or domain. Found and fixed by Jason Betts.
668 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
669 configuration). If a CNAME target was not a wellformed name pattern, a
672 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
673 the OS reports them interleaved with other addresses.
675 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
676 used both for input and for a verify callout, both encrypted, SMTP
677 responses being sent by the server could be lost. This resulted in
678 dropped connections and sometimes bounces generated by a peer sending
681 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
682 string. Previously it was possible to cause undefined behaviour in a
683 library routine (usually a crash). Found by "zerons".
685 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
688 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
689 API was removed, so update to use the newer ones.
691 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
692 any timeout set, is taking a long time. Previously we would hang on to a
693 rotated logfile "forever" if the input was arriving with long gaps
694 (a previous attempt to fix addressed lack, for a long time, of initial
697 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
698 shared (NFS) environment. The length of the tempfile name is now
699 4 + 16 ("hdr.$message_exim_id") which might break on file
700 systems which restrict the file name length to lower values.
703 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
704 shared (NFS) environment.
706 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
707 did for all versions <4.90). Notably -M, -m, --invert, -I may be
710 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
711 on some platforms for bit 31.
713 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
714 to changes apparently associated with TLS1.3 handling some of the APIs
715 previously used were either nonfunctional or inappropriate. Strings
716 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
717 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
718 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
719 This affects log line X= elements, the $tls_{in,out}_cipher variables,
720 and the use of specific cipher names in the encrypted= ACL condition.
722 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
724 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
725 verification result was not updated unless hosts_require_ocsp applied.
727 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
728 queue_list_requires_admin set to false, non-admin users were denied the
731 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
732 directory-of-certs mode. Previously they were advertised despite the
735 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
736 A single TCP connection by a client will now hold a TLS connection open
737 for multiple message deliveries, by default. Previously the default was to
740 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
741 default. If built with the facility, DANE will be used. The facility
742 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
744 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
745 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
746 must be defined and you must still, unless you define DISABLE_TLS, manage
747 the the include-dir and library-file requirements that go with that
748 choice. Non-TLS builds are still supported.
750 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
753 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
754 default. If the platform supports and has the facility enabled, it will
755 be requested on all coneections.
757 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
758 controlled by the build-time option SUPPORT_PIPE_CONNECT.
760 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
762 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
763 success-DSN messages. Previously the From: header was always the default
764 one for these; the option was ignored.
766 JH/28 Fix the timeout on smtp response to apply to the whole response.
767 Previously it was reset for every read, so a teergrubing peer sending
768 single bytes within the time limit could extend the connection for a
769 long time. Credit to Qualsys Security Advisory Team for the discovery.
771 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
772 delivery address, which leaked information of the results of local
773 forwarding. Change to the original envelope recipient address, per
776 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
777 requested. Previously not bounce was generated and a log entry of
778 error ignored was made.
780 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
782 JH/32 Introduce a general tainting mechanism for values read from the input
783 channel, and values derived from them. Refuse to expand any tainted
784 values, to catch one form of exploit.
786 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
787 was unused and the unexpanded text used for the test. Found and
788 fixed by Ruben Jenster.
790 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
791 an attempt to use a TLS library read routine dereffed a nul pointer,
794 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
795 them in our smtp response.
797 JH/36 Have the general router option retry_use_local_part default to true when
798 any of the restrictive preconditions are set (to anything). Previously it
799 was only for check_local user. The change removes one item of manual
800 configuration which is required for proper retries when a remote router
801 handles a subset of addresses for a domain.
803 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
804 link count into consideration.
806 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
807 caused the extension of big_buffer, the following lines were ignored.
809 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
810 accordance with RFC 2308. Previously there was no expiry, so a longlived
811 receive process (eg. due to ACL delays) versus a short SOA value could
814 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
816 JH/39 Promote DMARC support to mainline.
818 JH/40 Bug 2452: Add a References: header to DSNs.
820 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
821 parameters. The relevant library call is documented as "Deprecated: This
822 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
823 3.6.0, DH parameters are negotiated following RFC7919."
825 HS/06 Change the default of dnssec_request_domains to "*"
827 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
828 carried on and emitted a BDAT command, even when PIPELINING was not
831 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
832 buffer was used for the filename, resulting in a trap when tainted
833 arguments (eg. $domain) were used.
835 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
836 recommended to avoid a possible server-load attack. The feature can be
837 re-enabled via the openssl_options main cofiguration option.
839 JH/45 local_scan API: documented the current smtp_printf() call. This changed
840 for version 4.90 - adding a "more data" boolean to the arguments.
841 Bumped the ABI version number also, this having been missed previously;
842 release versions 4.90 to 4.92.3 inclusive were effectively broken in
843 respect of usage of smtp_printf() by either local_scan code or libraries
844 accessed via the ${dlfunc } expansion item. Both will need coding
845 adjustment for any calls to smtp_printf() to match the new function
846 signature; a FALSE value for the new argument is always safe.
848 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
849 the file-offset (which the Linux syscall does, and exim expects); this
850 resulted in an indefinite loop.
852 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
853 to do ARC verification. The Authentication-Results: header line added
854 by the configuration then had no ARC item.
856 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
857 an incorrect mode was used for reading the result, resulting in it being
864 JH/01 Remove code calling the customisable local_scan function, unless a new
865 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
867 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
868 non-signal-safe functions being used.
870 JH/03 Bug 2269: When presented with a received message having a stupidly large
871 number of DKIM-Signature headers, disable DKIM verification to avoid
872 a resource-consumption attack. The limit is set at twenty.
874 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
875 report of oldest_pass in ${authres } in consequence, and separate out
876 some descriptions of reasons for verification fail.
878 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
879 files in the spool were present and unlocked. A queue-runner could spot
880 them, resulting in a duplicate delivery. Fix that by doing the unlock
881 after the unlink. Investigation by Tim Stewart. Take the opportunity to
882 add more error-checking on spoolfile handling while that code is being
885 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
886 No known attacks, no CVE, this is defensive hardening.
888 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
889 a queue-runner could start a delivery while other operations were ongoing.
890 Cutthrough delivery was a common victim, resulting in duplicate delivery.
891 Found and investigated by Tim Stewart. Fix by using the open message data
892 file handle rather than opening another, and not locally closing it (which
893 releases a lock) for that case, while creating the temporary .eml format
894 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
896 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
897 $sender_verify_failure/$recipient_verify_failure to "random".
899 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
902 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
903 Previously this would segfault.
905 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
908 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
909 like zero, since the resolver should be doing this for us, But we need one
910 as a CNAME but no MX presence gets the CNAME returned; we need to check
911 that doesn't point to an MX to declare it "no MX returned" rather than
912 "error, loop". A new main option is added so the older capability of
913 following some limited number of chain links is maintained.
915 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
917 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
918 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
919 now-deprecated earlier definitions used only specified the range up to TLS
920 1.2 (in the older-version library docs).
922 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
924 JH/15 Rework TLS client-side context management. Stop using a global, and
925 explicitly pass a context around. This enables future use of TLS for
926 connections to service-daemons (eg. malware scanning) while a client smtp
927 connection is using TLS; with cutthrough connections this is quite likely.
929 JH/16 Fix ARC verification to do AS checks in reverse order.
931 JH/17 Support a "tls" option on the ${readsocket } expansion item.
933 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
934 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
935 Previously the "utf8" would be re-prepended for every additional message.
937 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
938 Previously thery were accepted, resulting in issues when attempting to
939 forward messages to a non-supporting MTA.
941 PP/02 Let -n work with printing macros too, not just options.
943 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
944 one parent address was copied, and bogus data was used at delivery-logging
945 time. Either a crash (after delivery) or bogus log data could result.
946 Discovery and analysis by Tim Stewart.
948 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
949 Previously if the string ended mid-character, we did not insert the
950 promised '?' replacement.
952 PP/04 Documentation: current string operators work on bytes, not codepoints.
954 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
955 should pack well giving a smaller memory footprint so better caching and
956 therefore performance. Group the declarations where this can't be done so
957 that the byte-sized flag variables are not interspersed among pointer
958 variables, giving a better chance of good packing by the compiler.
960 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
961 non-null, to avoid issues with sites running BATV. Previously reports were
962 sent with an empty envelope sender so looked like bounces.
964 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
965 The ignore_error flag wasn't being returned from the filter subprocess so
966 was not set for later routers. Investigation and fix by Matthias Kurz.
968 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
969 and a msg:complete for the whole, when a message is manually removed using
970 -Mrm. Developement by Matthias Kurz, hacked on by JH.
972 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
973 a "Gnu special" function, asprintf() in the DB utility binary builds; I
974 hope that is portable enough.
976 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
977 requiring a known-CA anchor certificate; make it now rely entirely on the
978 TLSA as an anchor. Checking the name on the leaf cert against the name
979 on the A-record for the host is still done for TA (but not for EE mode).
981 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
982 included in delivery lines for non-proxied connections, when compiled with
983 SUPPORT_SOCKS and running with proxy logging enabled.
985 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
986 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
987 move the existing event to fire before the normal logging of message
988 failure so that custom logging is bracketed by normal logging.
990 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
991 msg:fail:internal event. Developement by Matthias Kurz.
993 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
994 far too small for todays use of crypto signatures stored there. Go all
995 the way to the max DNS message size of 64kB, even though this might be
996 overmuch for IOT constrained device use.
998 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
999 copy a string over itself. The library routine is documented as not
1000 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1002 JH/32 For main options check_spool_space and check_inode_space, where the
1003 platform supports 64b integers, support more than the previous 2^31 kB
1004 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1005 the previous G, M, k.
1007 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1008 $authenticated_fail_id variable on authentication failure. Previously
1011 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1012 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1013 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1014 GNUTLS_SEC_PARAM_MEDIUM.
1016 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1017 side. Previously we would continue as if no SNI had been received.
1019 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1020 "<" character, which should be a list-separator specification, we walked
1021 off past the nul-terimation.
1023 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1024 causes) even when the retry time is not yet met. Previously they were
1025 not, meaning that when (say) an account was over-quota and temp-rejecting,
1026 and multiple senders' messages were queued, only one sender would get
1027 notified on each configured delay_warning cycle.
1029 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1031 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1032 cipher-suites, an error can be left on the stack even for a succeeding
1033 accept; this results in impossible error messages when a later operation
1036 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1037 return error codes indicating retry. Under TLS1.3 this becomes required.
1039 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1040 it only wrote the new authenticators, resulting in a lack of tracking of
1041 peer changes of ESMTP extensions until the next cache flush.
1043 JH/41 Fix the loop reading a message header line to check for integer overflow,
1044 and more-often against header_maxsize. Previously a crafted message could
1045 induce a crash of the recive process; now the message is cleanly rejected.
1047 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1048 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1054 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1055 When redis_servers is set to a list of > 1 element, and the Redis servers
1056 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1057 case of MOVED into a DEFER case instead, thus moving the query onto the
1058 next server in the list. For a cluster of N elements, all N servers must
1059 be defined in redis_servers.
1061 GF/02 Catch and remove uninitialized value warning in exiqsumm
1062 Check for existence of @ARGV before looking at $ARGV[0]
1064 JH/01 Replace the store_release() internal interface with store_newblock(),
1065 which internalises the check required to safely use the old one, plus
1066 the allocate and data copy operations duplicated in both (!) of the
1067 extant use locations.
1069 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1070 modifier. This matches the restriction on the commandline.
1072 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1073 Previously only the last row was returned.
1075 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1076 we assumed that tags in the header were well-formed, and parsed the
1077 element content after inspecting only the first char of the tag.
1078 Assumptions at that stage could crash the receive process on malformed
1081 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1082 While running the DKIM ACL we operate on the Permanent memory pool so that
1083 variables created with "set" persist to the DATA ACL. Also (at any time)
1084 DNS lookups that fail create cache records using the Permanent pool. But
1085 expansions release any allocations made on the current pool - so a dnsdb
1086 lookup expansion done in the DKIM ACL releases the memory used for the
1087 DNS negative-cache, and bad things result. Solution is to switch to the
1088 Main pool for expansions.
1089 While we're in that code, add checks on the DNS cache during store_reset,
1090 active in the testsuite.
1091 Problem spotted, and debugging aided, by Wolfgang Breyha.
1093 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1094 When none of the hosts presented to a transport match an already-open
1095 connection, close it and proceed with the list. Previously we would
1096 queue the message. Spotted by Lena with Yahoo, probably involving
1099 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1100 Previously a spurious "250 OK id=" response was appended to the proper
1103 JH/08 The "support for" informational output now, which built with Content
1104 Scanning support, has a line for the malware scanner interfaces compiled
1105 in. Interface can be individually included or not at build time.
1107 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1108 by the template makefile "src/EDITME". The "STREAM" support for an older
1109 ClamAV interface method is removed.
1111 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1112 rows affected is given instead).
1114 JH/11 The runtime Berkeley DB library version is now additionally output by
1115 "exim -d -bV". Previously only the compile-time version was shown.
1117 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1118 SMTP connection. Previously, when one had more recipients than the
1119 first, an abortive onward connection was made. Move to full support for
1120 multiple onward connections in sequence, handling cutthrough connection
1121 for all multi-message initiating connections.
1123 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1124 routers. Previously, a multi-recipient message would fail to match the
1125 onward-connection opened for the first recipient, and cause its closure.
1127 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1128 a timeout on read on a GnuTLS initiating connection, resulting in the
1129 initiating connection being dropped. This mattered most when the callout
1130 was marked defer_ok. Fix to keep the two timeout-detection methods
1133 JH/15 Relax results from ACL control request to enable cutthrough, in
1134 unsupported situations, from error to silently (except under debug)
1135 ignoring. This covers use with PRDR, frozen messages, queue-only and
1138 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1140 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1141 metadata, resulting in a crash in free().
1143 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1144 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1145 Broken also in d185889f4, with init system revamp.
1147 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1148 Previously we abruptly closed the connection after reading a malware-
1149 found indication; now we go on to read the "scan ok" response line,
1152 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1153 ACL. Previously, a crash would result.
1155 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1156 macro text after a replacement (previously it was only once per line) and
1157 by skipping builtin macros when searching for an uppercase lead character.
1159 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1160 for the build is renamed.
1162 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1163 was allocated for every new TLS startup, meaning one per message. Fix
1164 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1166 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1167 reported the original. Fix to report (as far as possible) the ACL
1168 result replacing the original.
1170 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1171 OpenSSL. Certificate information is loaded for every new TLS startup,
1172 and the resources needed to be freed.
1174 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1176 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1179 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1180 DATA response info to the (existing) per-recipient response info for
1181 the "C=" log element. It can have useful tracking info from the
1182 destination system. Patch from Simon Arlott.
1184 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1185 length value. Previously this would segfault.
1187 HS/02 Support Avast multiline protoocol, this allows passing flags to
1188 newer versions of the scanner.
1190 JH/28 Ensure that variables possibly set during message acceptance are marked
1191 dead before release of memory in the daemon loop. This stops complaints
1192 about them when the debug_store option is enabled. Discovered specifically
1193 for sender_rate_period, but applies to a whole set of variables.
1194 Do the same for the queue-runner and queue-list loops, for variables set
1195 from spool message files. Do the same for the SMTP per-message loop, for
1196 certain variables indirectly set in ACL operations.
1198 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1199 as a multi-recipient message from a mailinglist manager). The coding had
1200 an arbitrary cutoff number of characters while checking for more input;
1201 enforced by writing a NUL into the buffer. This corrupted long / fast
1202 input. The problem was exposed more widely when more pipelineing of SMTP
1203 responses was introduced, and one Exim system was feeding another.
1204 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1205 receiving system, and refused recipients seen by the sending system
1206 (propating to people being dropped from mailing lists).
1207 Discovered and pinpointed by David Carter.
1209 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1210 replaced by the ${authresults } expansion.
1212 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1214 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1215 allows proper process termination in container environments.
1217 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1218 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1220 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1221 and "err_perm", deprecated since 4.83 when the RFC-defined words
1222 "temperror" and "permerror" were introduced.
1224 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1225 transport-filters or DKIM-signing. The restriction was lost in the
1226 consolidation of verify-callout and delivery SMTP handling.
1227 Extend the restriction to also cover ARC-signing.
1229 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1230 in defer=pass mode supply a 450 to the initiator. Previously the message
1233 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1234 tls_require_ciphers is used as before.
1236 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1237 "pass_unscanned". Only tmpfails from the scanner are written to
1238 the paniclog, as they may require admin intervention (permission
1239 denied, license issues). Other scanner errors (like decompression
1240 bombs) do not cause a paniclog entry.
1242 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1243 Previously it was possible to log spurious information in receive log
1246 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1247 triggered odd behaviour from Outlook Express clients.
1249 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1252 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1253 since the IETF WG has not yet settled on that versus the original
1254 "bare" representation.
1256 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1257 Previously the millisecond value corrupted the output.
1258 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1259 corrupted the output.
1265 JH/01 Rework error string handling in TLS interface so that the caller in
1266 more cases is responsible for logging. This permits library-sourced
1267 string to be attached to addresses during delivery, and collapses
1268 pairs of long lines into single ones.
1270 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1271 during configuration. Wildcards are allowed and expanded.
1273 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1274 This permits better logging.
1276 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1277 do not close it down and have the child transport start it up again on
1278 the passed-on TCP connection. Instead, proxy the child (and any
1279 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1280 affected: the continued delivery log lines do not have any DNSSEC, TLS
1281 Certificate or OCSP information. TLS cipher information is still logged.
1283 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1284 identical IP addresses on different listening ports. Will also affect
1287 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1288 add noisy ifdef guards to special-case this sillyness.
1289 Patch from Bernd Kuhls.
1291 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1292 than 255 are no longer allowed.
1294 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1295 Disable the session-cache too, which might reduce our load. Since we
1296 currrectly use a new context for every connection, both as server and
1297 client, there is no benefit for these.
1298 GnuTLS appears to not support tickets server-side by default (we don't
1299 call gnutls_session_ticket_enable_server()) but client side is enabled
1300 by default on recent versions (3.1.3 +) unless the PFS priority string
1303 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1304 <https://reproducible-builds.org/specs/source-date-epoch/>.
1306 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1307 the check for any unsuccessful recipients did not notice the limit, and
1308 erroneously found still-pending ones.
1310 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1311 MSG_MORE. Only in-clear (not on TLS connections).
1313 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1314 transport-filter is involved we can buffer the headers in memory for
1315 creating the signature, and read the spool data file once for the
1316 signature and again for transmission.
1318 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1319 4.77 as the kernel support then wasn't solid, having issues in 64bit
1320 mode. Now, it's been long enough. Add support for FreeBSD also.
1322 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1323 case where the routing stage had gathered several addresses to send to
1324 a host before calling the transport for the first, we previously failed
1325 to close down TLS in the old transport process before passing the TCP
1326 connection to the new process. The new one sent a STARTTLS command
1327 which naturally failed, giving a failed delivery and bloating the retry
1328 database. Investigation and fix prototype from Wolfgang Breyha.
1330 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1331 false-negatives in the check that the sender had not preempted a response
1332 or prompt from Exim (running as a server), due to that code's lack of
1333 awareness of the SMTP input buffering.
1335 PP/04 Add commandline_checks_require_admin option.
1336 Exim drops privileges sanely, various checks such as -be aren't a
1337 security problem, as long as you trust local users with access to their
1338 own account. When invoked by services which pass untrusted data to
1339 Exim, this might be an issue. Set this option in main configuration
1340 AND make fixes to the calling application, such as using `--` to stop
1343 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1344 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1345 responses to those, into a single TLS record each way (this usually means
1346 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1349 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1350 affects you only if you're dancing at the edge of the param size limits.
1351 If you are, and this message makes sense to you, then: raise the
1352 configured limit or use OpenSSL 1.1. Nothing we can do for older
1355 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1356 cmdline element to get the documented default one. Previously it was
1359 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1360 are unsuable for non-message-specific reasons - eg. connection timeout,
1361 banner-time rejection.
1363 JH/16 Fix logging of delivery remote port, when specified by router, under
1366 PP/06 Repair manualroute's ability to take options in any order, even if one
1367 is the name of a transport.
1370 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1372 JH/17 Change the list-building routines interface to use the expanding-string
1373 triplet model, for better allocation and copying behaviour.
1375 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1376 Previously it was constructed the first time a possibly-matching string
1377 was met in the configuration file input during startup; now it is done
1380 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1381 compatible one, to avoid the (poorly documented) possibility of a config
1382 file in the working directory redirecting the DB files, possibly correpting
1383 some existing file. CVE-2017-10140 assigned for BDB.
1385 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1386 cache-hot. Previously, although the result was properly cached, the
1387 initial verify call returned a defer.
1389 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1390 the main verify for receipient in uncached-mode.
1392 JH/22 Retire historical build files to an "unsupported" subdir. These are
1393 defined as "ones for which we have no current evidence of testing".
1395 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1396 if present. Previously it was ignored.
1398 JH/24 Start using specified-initialisers in C structure init coding. This is
1399 a C99 feature (it's 2017, so now considered safe).
1401 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1402 if was a fixed-sized field and bitmask ops via macros; it is now more
1405 PP/07 GitHub PR 56: Apply MariaDB build fix.
1406 Patch provided by Jaroslav Škarvada.
1408 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1409 during Coverity cleanups [4.87 JH/47]
1410 Diagnosis and fix provided by Michael Fischer v. Mollard.
1412 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1413 the right size to place the terminating semicolon on its own folded
1414 line, the header hash was calculated to an incorrect value thanks to
1415 the (relaxed) space the fold became.
1417 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1418 and confused the parent.
1420 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1421 which could crash as a result. This could lead to undeliverable messages.
1423 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1426 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1427 matters on fast-turnover and PID-randomising systems, which were getting
1428 out-of-order delivery.
1430 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1431 a possibly-overlapping copy. The symptom was that "Remote host closed
1432 connection in response to HELO" was logged instead of the actual 4xx
1435 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1436 Previously only that bufferd was discarded, resulting in SYMTP command
1439 JH/32 DKIM: when a message has multiple signatures matching an identity given
1440 in dkim_verify_signers, run the dkim acl once for each. Previously only
1441 one run was done. Bug 2189.
1443 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1444 "panic the current process" to "deliberately defer". The panic log is
1445 still written with the problem list name; the mail and reject logs now
1446 get a temp-reject line for the message that was being handled, saying
1447 something like "domains check lookup or other defer". The SMTP 451
1448 message is still "Temporary local problem".
1450 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1451 A crafted sequence of BDAT commands could result in in-use memory beeing
1452 freed. CVE-2017-16943.
1454 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1455 from SMTP input. Previously it was always done; now only done for DATA
1456 and not BDAT commands. CVE-2017-16944.
1458 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1459 to the message (such as an overlong header line). Previously this was
1460 not done and we did not exit BDAT mode. Followon from the previous item
1461 though a different problem.
1467 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1468 than -2003 did; needs libidn2 in addition to libidn.
1470 JH/02 The path option on a pipe transport is now expanded before use.
1472 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1473 Patch provided by "Björn", documentation fix added too.
1475 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1476 missing a wire-to-host endian conversion.
1478 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1479 close after a BDAT command line could be taken as a following command,
1480 giving a synch failure. Fix by only checking for synch immediately
1481 before acknowledging the chunk.
1483 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1484 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1485 macro. Patches provided by Josh Soref.
1487 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1488 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1489 by IANA as a service extension. However, John Klensin suggests that we
1492 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1493 the dkim code may be unix-mode line endings rather than smtp wire-format
1494 CRLF, so prepend a CR to any bare LF.
1496 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1497 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1499 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1500 the first were themselves being wrongly included in the feed into dkim
1501 processing; with most chunk sizes in use this resulted in an incorrect
1502 body hash calculated value.
1504 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1505 DKIM signature block, for verification. Although advised against by
1506 standards it is specifically not ruled illegal.
1508 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1510 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1511 missing a body hash (the bh= tag).
1513 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1514 It seems that HAProxy sends the Proxy Protocol information in clear and
1515 only then does a TLS startup, so do the same.
1517 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1518 TCP connections (such as for Spamd) unless the daemon successfully set
1519 Fast Open mode on its listening sockets. This fixes breakage seen on
1520 too-old kernels or those not configured for Fast Open, at the cost of
1521 requiring both directions being enabled for TFO, and TFO never being used
1522 by non-daemon-related Exim processes.
1524 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1525 endings, at least on the first header line. Try to canonify any that get
1526 past that check, despite the cost.
1528 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1529 now limited to an arbitrary five deep, while parsing addresses with the
1530 strip_excess_angle_brackets option enabled.
1532 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1533 instead leave the unprompted TLS handshake in socket buffer for the
1534 TLS library to consume.
1536 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1538 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1540 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1541 region in one operation, for speed, and this leaves assigned pointers
1542 dangling. Add checks run only under the testsuite which checks all
1543 variables at a store-reset and panics on a dangling pointer; add code
1544 explicitly nulling out all the variables discovered. Fixes one known
1545 bug: a transport crash, where a dangling pointer for $sending_ip_address
1546 originally assigned in a verify callout, is re-used.
1548 PP/06 Drop '.' from @INC in various Perl scripts.
1550 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1552 PP/08 Reduce a number of compilation warnings under clang; building with
1553 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1554 should be warning-free.
1556 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1558 HS/01 Fix portability problems introduced by PP/08 for platforms where
1559 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1561 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1562 chunk. This allows us to accept broken chunked messages. We need a more
1563 general solution here.
1565 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1566 already-broken messages in the queue.
1568 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1570 JH/19 Fix reference counting bug in routing-generated-address tracking.
1576 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1577 supports it and a size is available (ie. the sending peer gave us one).
1579 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1580 years of being deprecated). The replacements are the ACLs
1581 acl_smtp_mime and acl_not_smtp_mime.
1583 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1584 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1585 or even an in-clear connection were permitted. Now, if the host lookup
1586 was dnssec and dane was requested then the host is only used if the
1587 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1588 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1589 if one fails this test.
1590 This means that a poorly-configured remote DNS will make it incommunicado;
1591 but it protects against a DNS-interception attack on it.
1593 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1594 non-noisy when a race steals the message being considered.
1596 JH/05 If main configuration option tls_certificate is unset, generate a
1597 self-signed certificate for inbound TLS connections.
1599 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1600 in rewrites and routers.
1602 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1603 and logged a warning sing 4.83; now they are a configuration file error.
1605 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1606 (lacking @domain). Apply the same qualification processing as RCPT.
1608 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1610 JH/10 Support ${sha256:} applied to a string (as well as the previous
1613 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1614 a cutthrough deliver is pending, as we always want to make a connection.
1615 This also avoids re-routing the message when later placing the cutthrough
1616 connection after a verify cache hit.
1617 Do not update it with the verify result either.
1619 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1620 when routing results in more than one destination address.
1622 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1623 signing (which inhibits the cutthrough capability). Previously only
1624 the presence of an option was tested; now an expansion evaluating as
1625 empty is permissible (obviously it should depend only on data available
1626 when the cutthrough connection is made).
1628 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1629 the relevant preceding SMTP command did not note the pipelining mode.
1631 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1632 Previously they were not counted.
1634 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1635 as one having no matching records. Previously we deferred the message
1636 that needed the lookup.
1638 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1639 distinguished as "(=".
1641 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1642 for missing MX records. Previously it only worked for missing A records.
1644 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1646 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1647 after the data-go-ahead and data-ack. Patch from Jason Betts.
1649 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1650 even for a "none" policy. Patch from Tony Meyer.
1652 JH/22 Fix continued use of a connection for further deliveries. If a port was
1653 specified by a router, it must also match for the delivery to be
1656 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1657 When one of the recipients of a message was unsuitable for the connection
1658 (has no matching addresses), we lost track of needing to mark it
1659 deferred. As a result mail would be lost.
1661 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1663 JH/25 Decoding ACL controls is now done using a binary search; the source code
1664 takes up less space and should be simpler to maintain. Merge the ACL
1665 condition decode tables also, with similar effect.
1667 JH/26 Fix problem with one_time used on a redirect router which returned the
1668 parent address unchanged. A retry would see the parent address marked as
1669 delivered, so not attempt the (identical) child. As a result mail would
1672 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1673 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1674 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1677 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1678 Main config options check_{log,spool}_{inodes,space} are now
1679 100 inodes, 10MB unless set otherwise in the configuration.
1681 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1682 Previously it only applied to the main-section connection policy
1685 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1687 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1688 by me. Added RFC7919 DH primes as an alternative.
1690 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1691 are not in the system include path.
1693 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1694 GnuTLS, when a session startup failed (eg because the client disconnected)
1695 Exim did stdio operations after fclose. This was exposed by a recent
1696 change which nulled out the file handle after the fclose.
1698 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1699 signed directly by the cert-signing cert, rather than an intermediate
1700 OCSP-signing cert. This is the model used by LetsEncrypt.
1702 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1704 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1705 an incoming connection.
1707 HS/02 Bug 1802: Do not half-close the connection after sending a request
1710 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1711 fallback to "prime256v1".
1713 JH/34 SECURITY: Use proper copy of DATA command in error message.
1714 Could leak key material. Remotely exploitable. CVE-2016-9963.
1720 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1721 and 3.4.4 - once the server is enabled to respond to an OCSP request
1722 it does even when not requested, resulting in a stapling non-aware
1723 client dropping the TLS connection.
1725 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1726 support variable-length bit vectors. No functional change.
1728 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1729 The I= interface field on outgoing lines is now after the H= remote
1730 host field, same as incoming lines. There is a separate
1731 outgoing_interface log selector which allows you to disable the
1734 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1735 If not running log_selector +smtp_connection the mainlog would be held
1736 open indefinitely after a "too many connections" event, including to a
1737 deleted file after a log rotate. Leave the per net connection logging
1738 leaving it open for efficiency as that will be quickly detected by the
1739 check on the next write.
1741 HS/01 Bug 1671: Fix post transport crash.
1742 Processing the wait-<transport> messages could crash the delivery
1743 process if the message IDs didn't exist for some reason. When
1744 using 'split_spool_directory=yes' the construction of the spool
1745 file name failed already, exposing the same netto behaviour.
1747 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1748 mime_regex ACL conditions.
1750 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1751 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1752 to HELO, local diagnostic string.
1754 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1755 log line to debug. Even when configured with a tls authenticator many
1756 client connections are expected to not authenticate in this way, so
1757 an authenticate fail is not an error.
1759 HS/02 Add the Exim version string to the process info. This way exiwhat
1760 gives some more detail about the running daemon.
1762 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1763 matter for fast-change records such as DNSBLs.
1765 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1766 retry record, even if constant. There may be multiple transports with
1767 different interface settings and the retry behaviour needs to be kept
1770 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1772 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1774 JH/10 Bug 840: fix log_defer_output option of pipe transport
1776 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1777 hosts_randomize. This matters a lot when combined with mua_wrapper.
1779 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1780 ${quote_pgsql:<string>} operator.
1782 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1784 JH/14 Change header file name for freeradius-client. Relevant if compiling
1785 with Radius support; from the Gentoo tree and checked under Fedora.
1787 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1789 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1790 option transport_filter, meaning no filtering.
1792 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1794 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1795 defaults to "*" (all hosts). The variable is now available when not built
1796 with TLS, default unset, mainly to enable keeping the testsuite sane.
1797 If a server certificate is not supplied (via tls_certificate) an error is
1798 logged, and clients will find TLS connections fail on startup. Presumably
1799 they will retry in-clear.
1800 Packagers of Exim are strongly encouraged to create a server certificate
1801 at installation time.
1803 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1804 with the $config_file variable.
1806 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1807 in transport context, after the attempt, and per-recipient. The latter type
1808 is per host attempted. The event data is the error message, and the errno
1809 information encodes the lookup type (A vs. MX) used for the (first) host,
1810 and the trailing two digits of the smtp 4xx response.
1812 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1813 to write to mainlog (or rejectlog, paniclog) in the window between file
1814 creation and permissions/ownership being changed. Particularly affects
1815 installations where exicyclog is run as root, rather than exim user;
1816 result is that the running daemon panics and dies.
1818 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1820 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1821 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1822 "pri" and "weight". Note that the previous implicit priority given by the
1823 list order is no longer honoured.
1825 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1826 for DKIM processing.
1828 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1829 by defining SUPPORT_SOCKS.
1831 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1832 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1833 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1834 variables are renamed to proxy_{local,external}_{address,port}.
1836 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1837 for a build by defining SUPPORT_I18N
1839 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1840 of the query string, and make ${quote_redis:} do that quoting.
1842 JH/29 Move Events support from Experimental to mainline, enabled by default
1843 and removable for a build by defining DISABLE_EVENT.
1845 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1847 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1848 cached by the daemon.
1850 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1851 by defining LOOKUP_REDIS. The libhiredis library is required.
1853 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1854 keys are given for lookup.
1856 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1857 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1858 only supported when built with TLS support. The PolarSSL SHA routines
1859 are still used when the TLS library is too old for convenient support.
1861 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1862 openssl_options), for security. OpenSSL forces this from version 1.1.0
1863 server-side so match that on older versions.
1865 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1866 allocation for $value could be released as the expansion processing
1867 concluded, but leaving the global pointer active for it.
1869 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1870 and to use the domains and local_parts ACL conditions.
1872 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1873 incorrectly not doubled on cutthrough transmission, hence seen as a
1874 body-termination at the receiving system - resulting in truncated mails.
1875 Commonly the sender saw a TCP-level error, and retransmitted the message
1876 via the normal store-and-forward channel. This could result in duplicates
1877 received - but deduplicating mailstores were liable to retain only the
1878 initial truncated version.
1880 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1882 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1884 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1885 we're in there, support oversigning also; bug 1309.
1887 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1889 HS/04 Add support for keep_environment and add_environment options.
1891 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1892 either intentional arithmetic overflow during PRNG, or testing config-
1895 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1896 delivery resulted in actual delivery. Cancel cutthrough before DATA
1899 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1900 rejects a recipient: pass the reject to the originator.
1902 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1903 Many were false-positives and ignorable, but it's worth fixing the
1906 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1907 for the new environment-manipulation done at startup. Move the routines
1908 from being local to tls.c to being global via the os.c file.
1910 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1911 an extract embedded as result-arg for a map, the first arg for extract
1912 is unavailable so we cannot tell if this is a numbered or keyed
1913 extraction. Accept either.
1919 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1922 JH/02 The smtp transport option "multi_domain" is now expanded.
1924 JH/03 The smtp transport now requests PRDR by default, if the server offers
1927 JH/04 Certificate name checking on server certificates, when exim is a client,
1928 is now done by default. The transport option tls_verify_cert_hostnames
1929 can be used to disable this per-host. The build option
1930 EXPERIMENTAL_CERTNAMES is withdrawn.
1932 JH/05 The value of the tls_verify_certificates smtp transport and main options
1933 default to the word "system" to access the system default CA bundle.
1934 For GnuTLS, only version 3.0.20 or later.
1936 JH/06 Verification of the server certificate for a TLS connection is now tried
1937 (but not required) by default. The verification status is now logged by
1938 default, for both outbound TLS and client-certificate supplying inbound
1941 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1944 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1945 Status Notification (bounce) messages are now MIME format per RFC 3464.
1946 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1947 under the control of the dsn_advertise_hosts option, and routers may
1948 have a dsn_lasthop option.
1950 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1951 default, modifiable by a malware= option. The list separator for
1952 the options can now be changed in the usual way. Bug 68.
1954 JH/10 The smtp_receive_timeout main option is now expanded before use.
1956 JH/11 The incoming_interface log option now also enables logging of the
1957 local interface on delivery outgoing connections.
1959 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1960 if the interface and destination host and port all match.
1962 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1965 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1966 Patch from Andrew Lewis.
1968 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1969 now supports optional time-restrictions, weighting, and priority
1970 modifiers per server. Patch originally by <rommer@active.by>.
1972 JH/16 The spamd_address main option now supports a mixed list of local
1973 and remote servers. Remote servers can be IPv6 addresses, and
1974 specify a port-range.
1976 JH/17 Bug 68: The spamd_address main option now supports an optional
1977 timeout value per server.
1979 JH/18 Bug 1581: Router and transport options headers_add/remove can
1980 now have the list separator specified.
1982 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1985 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1988 JH/21 Support for the A6 type of dns record is withdrawn.
1990 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1991 rather than the verbs used.
1993 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1994 from 255 to 1024 chars.
1996 JH/24 Verification callouts now attempt to use TLS by default.
1998 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1999 are generic router options now. The defaults didn't change.
2001 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2002 Original patch from Alexander Shikoff, worked over by JH.
2004 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2005 files if ZCAT_COMMAND is not executable.
2007 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2009 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2011 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2012 Normally benign, it bites when the pair was led to by a CNAME;
2013 modern usage is to not canonicalize the domain to a CNAME target
2014 (and we were inconsistent anyway for A-only vs AAAA+A).
2016 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2018 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2019 when evaluating $sender_host_dnssec.
2021 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2022 $sender_helo_dnssec variable.
2024 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2026 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2028 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2030 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2031 documented as working, but never had. Support all but $spam_report.
2033 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2034 added for tls authenticator.
2036 HS/03 Add perl_taintmode main config option
2042 TL/01 When running the test suite, the README says that variables such as
2043 no_msglog_check are global and can be placed anywhere in a specific
2044 test's script, however it was observed that placement needed to be near
2045 the beginning for it to behave that way. Changed the runtest perl
2046 script to read through the entire script once to detect and set these
2047 variables, reset to the beginning of the script, and then run through
2048 the script parsing/test process like normal.
2050 TL/02 The BSD's have an arc4random API. One of the functions to induce
2051 adding randomness was arc4random_stir(), but it has been removed in
2052 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2053 function when detected.
2055 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2056 cause callback expansion.
2058 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2059 syntax errors in an expansion can be treated as a string instead of
2060 logging or causing an error, due to the internal use of bool_lax
2061 instead of bool when processing it.
2063 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2064 server certificates when making smtp deliveries.
2066 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2068 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2070 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2072 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2073 Merged patch from Sebastian Wiedenroth.
2075 JH/05 Fix results-pipe from transport process. Several recipients, combined
2076 with certificate use, exposed issues where response data items split
2077 over buffer boundaries were not parsed properly. This eventually
2078 resulted in duplicates being sent. This issue only became common enough
2079 to notice due to the introduction of connection certificate information,
2080 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2082 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2083 size buffer was used, resulting in syntax errors when an expansion
2086 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2087 version 3.3.6 or later.
2089 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2090 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2091 both become "event_action", the variables become $event_name, $event_data
2092 and $event_defer_errno. There is a new variable $verify_mode, usable in
2093 routers, transports and related events. The tls:cert event is now also
2094 raised for inbound connections, if the main configuration event_action
2097 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2098 early OCSP support, but no stapling (appears to be less than 1.0.0).
2100 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2101 server certificate names available under the smtp transport option
2102 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2105 JH/10 Time-related extraction expansions from certificates now use the main
2106 option "timezone" setting for output formatting, and are consistent
2107 between OpenSSL and GnuTLS compilations. Bug 1541.
2109 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2110 encoded parameter in the incoming message. Bug 1558.
2112 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2113 include certificate info, eximon was claiming there were spoolfile
2116 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2118 JH/14 Log delivery-related information more consistently, using the sequence
2119 "H=<name> [<ip>]" wherever possible.
2121 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2122 are problematic for Debian distribution, omit them from the release
2125 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2127 JH/16 Fix string representation of time values on 64bit time_t architectures.
2130 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2131 output list separator was used.
2136 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2137 checkers that were complaining about end of non-void function with no
2140 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2141 This was a regression introduced in 4.83 by another bugfix.
2143 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2145 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2146 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2152 TF/01 Correctly close the server side of TLS when forking for delivery.
2154 When a message was received over SMTP with TLS, Exim failed to clear up
2155 the incoming connection properly after forking off the child process to
2156 deliver the message. In some situations the subsequent outgoing
2157 delivery connection happened to have the same fd number as the incoming
2158 connection previously had. Exim would try to use TLS and fail, logging
2159 a "Bad file descriptor" error.
2161 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2162 utilities have not been installed.
2164 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2165 temporary space as the ACL may create new global variables.
2167 TL/01 LDAP support uses per connection or global context settings, depending
2168 upon the detected version of the libraries at build time.
2170 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2171 to extract and use the src ip:port in logging and expansions as if it
2172 were a direct connection from the outside internet. PPv2 support was
2173 updated based on HAProxy spec change in May 2014.
2175 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2177 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2178 Properly escape header and check for NULL return.
2180 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2183 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2185 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2186 characters in header names, implemented as a verify condition.
2187 Contributed by Michael Fischer v. Mollard.
2189 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2190 results permerror and temperror. Previous values are deprecated but
2191 still accepted. In a future release, err_perm and err_temp will be
2192 completely removed, which will be a backward incompatibility if the
2193 ACL tests for either of these two old results. Patch contributed by
2194 user bes-internal on the mailing list.
2196 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2198 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2199 selectors, in both main and reject logs.
2201 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2204 JH/07 Add malware type "sock" for talking to simple daemon.
2206 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2208 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2209 routers/transports under cutthrough routing.
2211 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2212 numbers. Touch up "bool" conditional to keep the same definition.
2214 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2216 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2218 JH/12 Expand items in router/transport headers_add or headers_remove lists
2219 individually rather than the list as a whole. Bug 1452.
2221 Required for reasonable handling of multiple headers_ options when
2222 they may be empty; requires that headers_remove items with embedded
2223 colons must have them doubled (or the list-separator changed).
2225 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2226 view the policy declared in the DMARC record. Currently, $dmarc_status
2227 is a combined value of both the record presence and the result of the
2230 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2232 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2233 dnslookup router and the smtp transport (applying to the forward
2236 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2237 of ldap servers used for a specific lookup. Patch provided by Heiko
2240 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2241 New variable $lookup_dnssec_authenticated for observability.
2243 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2244 Patch submitted by Lars Timman.
2246 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2248 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2249 Requires trusted mode and valid format message id, aborts otherwise.
2250 Patch contributed by Heiko Schlichting.
2252 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2253 certextract with support for various fields. Bug 1358.
2255 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2256 is requested by default, modifiable by smtp transport option
2259 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2260 operate on certificate variables to give certificate fingerprints
2261 Also new ${sha256:cert_variable}.
2263 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2265 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2266 Christian Aistleitner.
2268 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2270 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2271 file. Patch from Wolfgang Breyha.
2273 JH/25 Expand the coverage of the delivery $host and $host_address to
2274 client authenticators run in verify callout. Bug 1476.
2276 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2277 align with daemon_smtp_ports. Bug 72.
2279 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2280 support and error reporting did not work properly.
2282 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2283 and is readable. Patch from Andrew Colin Kissa.
2285 TL/14 Enhance documentation of ${run expansion and how it parses the
2286 commandline after expansion, particularly in the case when an
2287 unquoted variable expansion results in an empty value.
2289 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2291 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2292 by renaming away from it. Addresses GH issue 15, reported by
2295 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2297 TL/15 SECURITY: prevent double expansion in math comparison functions
2298 (can expand unsanitized data). Not remotely exploitable.
2305 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2307 PP/02 Make -n do something, by making it not do something.
2308 When combined with -bP, the name of an option is not output.
2310 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2313 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2314 $sender_host_name and config options to manage this, and basic check
2317 PP/05 DSCP support for outbound connections and control modifier for inbound.
2319 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2320 (Only plugin which currently uses this is kerberos4, which nobody should
2321 be using, but we should make it available and other future plugins might
2322 conceivably use it, even though it would break NAT; stuff *should* be
2323 using channel bindings instead).
2325 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2326 name; added for Sendmail compatibility; requires admin caller.
2327 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2328 just ignore it); requires trusted caller.
2329 Also parse but ignore: -Ac -Am -X<logfile>
2332 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2334 TL/02 Add +smtp_confirmation as a default logging option.
2336 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2337 Patch by Magnus Holmgren from 2007-02-20.
2339 TL/04 Bugzilla 1281 - Spec typo.
2340 Bugzilla 1283 - Spec typo.
2341 Bugzilla 1290 - Spec grammar fixes.
2343 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2345 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2347 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2348 mailing list by by Dmitry Isaikin.
2350 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2352 JH/02 Support "G" suffix to numbers in ${if comparisons.
2354 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2356 NM/01 Bugzilla 1197 - Spec typo
2357 Bugzilla 1196 - Spec examples corrections
2359 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2361 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2362 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2365 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2366 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2368 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2369 "acl {{name}{arg}...}", and optional args on acl condition
2372 JH/05 Permit multiple router/transport headers_add/remove lines.
2374 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2376 JH/07 Avoid using a waiting database for a single-message-only transport.
2377 Performance patch from Paul Fisher. Bugzilla 1262.
2379 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2382 JH/09 Add $headers_added variable, with content from use of ACL modifier
2383 add_header (but not yet added to the message). Bugzilla 199.
2385 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2386 Pulled from Bugzilla 817 by Wolfgang Breyha.
2388 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2390 (nb: this is the same fix as in Exim 4.80.1)
2392 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2395 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2398 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2400 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2401 advertise SMTP AUTH mechanism to us, instead of a generic
2402 protocol violation error. Also, make Exim more robust to bad
2403 data from the Dovecot auth socket.
2405 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2407 When a queue runner is handling a message, Exim first routes the
2408 recipient addresses, during which it prunes them based on the retry
2409 hints database. After that it attempts to deliver the message to
2410 any remaining recipients. It then updates the hints database using
2413 So if a recipient address works intermittently, it can get repeatedly
2414 deferred at routing time. The retry hints record remains fresh so the
2415 address never reaches the final cutoff time.
2417 This is a fairly common occurrence when a user is bumping up against
2418 their storage quota. Exim had some logic in its local delivery code
2419 to deal with this. However it did not apply to per-recipient defers
2420 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2422 This change adds a proper retry rule check during routing so that the
2423 final cutoff time is checked against the message's age. We only do
2424 this check if there is an address retry record and there is not a
2425 domain retry record; this implies that previous attempts to handle
2426 the address had the retry_use_local_parts option turned on. We use
2427 this as an approximation for the destination being like a local
2428 delivery, as in LMTP.
2430 I suspect this new check makes the old local delivery cutoff check
2431 redundant, but I have not verified this so I left the code in place.
2433 TF/02 Correct gecos expansion when From: is a prefix of the username.
2435 Test 0254 submits a message to Exim with the header
2439 When I ran the test suite under the user fanf2, Exim expanded
2440 the header to contain my full name, whereas it should have added
2441 a Resent-Sender: header. It erroneously treats any prefix of the
2442 username as equal to the username.
2444 This change corrects that bug.
2446 GF/01 DCC debug and logging tidyup
2447 Error conditions log to paniclog rather than rejectlog.
2448 Debug lines prefixed by "DCC: " to remove any ambiguity.
2450 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2452 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2453 Bug spotted by Jeremy Harris; was flawed since initial commit.
2454 Would have resulted in OCSP responses post-SNI triggering an Exim
2455 NULL dereference and crash.
2457 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2459 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2460 Bug detection, analysis and fix by Samuel Thibault.
2461 Bugzilla 1331, Debian bug #698092.
2463 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2465 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2466 Server implementation by Todd Lyons, client by JH.
2467 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2468 config variable "prdr_enable" controls whether the server
2469 advertises the facility. If the client requests PRDR a new
2470 acl_data_smtp_prdr ACL is called once for each recipient, after
2471 the body content is received and before the acl_smtp_data ACL.
2472 The client is controlled by both of: a hosts_try_prdr option
2473 on the smtp transport, and the server advertisement.
2474 Default client logging of deliveries and rejections involving
2475 PRDR are flagged with the string "PRDR".
2477 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2478 fclose(). Diagnosis by Todd Lyons.
2480 PP/17 Update configure.default to handle IPv6 localhost better.
2481 Patch by Alain Williams (plus minor tweaks).
2484 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2485 This is now consistent with GnuTLS, and is now documented: the
2486 previous undocumented portable approach to treating the option as
2487 unset was to force an expansion failure. That still works, and
2488 an empty string is now equivalent.
2490 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2491 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2492 not performing validation itself.
2494 PP/20 Added force_command boolean option to pipe transport.
2495 Patch from Nick Koston, of cPanel Inc.
2497 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2500 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2502 PP/21 Fix eximon continuous updating with timestamped log-files.
2503 Broken in a format-string cleanup in 4.80, missed when I repaired the
2504 other false fix of the same issue.
2505 Report and fix from Heiko Schlichting.
2508 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2509 Report from Prashanth Katuri.
2511 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2512 It's SecureTransport, so affects any MacOS clients which use the
2513 system-integrated TLS libraries, including email clients.
2515 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2516 using a MIME ACL for non-SMTP local injection.
2517 Report and assistance in diagnosis by Warren Baker.
2519 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2521 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2523 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2524 last id that failed so it may be referenced in subsequent ACL's.
2526 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2529 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2530 ldap library initialization, allowing self-signed CA's to be
2531 used. Also properly sets require_cert option later in code by
2532 using NULL (global ldap config) instead of ldap handle (per
2533 session). Bug diagnosis and testing by alxgomz.
2535 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2536 the src/util/ subdirectory.
2538 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2539 renamed to Transport Post Delivery Action by Jeremy Harris, as
2542 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2543 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2544 redis_servers = needs to be configured which will be used by the redis
2545 lookup. Patch from Warren Baker, of The Packet Hub.
2547 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2549 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2550 hostname or reverse DNS when processing a host list. Used suggestions
2551 from multiple comments on this bug.
2553 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2555 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2556 Missed a few lines, added it to make the runtest require no keyboard
2559 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2560 contains upper case chars. Make router use caseful_local_part.
2562 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2563 support when GnuTLS has been built with p11-kit.
2569 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2571 This, or similar/improved, will also be change PP/11 of 4.82.
2577 PP/01 Handle short writes when writing local log-files.
2578 In practice, only affects FreeBSD (8 onwards).
2579 Bugzilla 1053, with thanks to Dmitry Isaikin.
2581 NM/01 Bugzilla 949 - Documentation tweak
2583 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2586 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2588 PP/02 Implemented gsasl authenticator.
2590 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2592 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2593 `pkg-config foo` for cflags/libs.
2595 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2596 with rest of GSASL and with heimdal_gssapi.
2598 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2599 `pkg-config foo` for cflags/libs for the TLS implementation.
2601 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2602 properties get this fed in as external SSF. A number of robustness
2603 and debugging improvements to the cyrus_sasl authenticator.
2605 PP/08 cyrus_sasl server now expands the server_realm option.
2607 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2608 Patch by Jeremy Harris.
2610 PP/10 Added dbmjz lookup type.
2612 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2614 PP/12 MAIL args handles TAB as well as SP, for better interop with
2615 non-compliant senders.
2616 Analysis and variant patch by Todd Lyons.
2618 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2619 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2620 Patch from Dirk Mueller <dmueller@suse.com>
2622 PP/13 tls_peerdn now print-escaped for spool files.
2623 Observed some $tls_peerdn in wild which contained \n, which resulted
2624 in spool file corruption.
2626 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2627 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2628 or write after TLS renegotiation, which otherwise led to messages
2631 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2632 as a tracking header (ie: a signed header comes before the signature).
2633 Patch from Wolfgang Breyha.
2635 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2636 comma-sep list; embedded commas doubled.
2638 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2640 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2642 Report and patch from Dmitry Banschikov.
2644 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2645 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2646 failures. We appear to now support TLS1.1+ with Exim.
2648 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2649 lets Exim select keys and certificates based upon TLS SNI from client.
2650 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2651 before an outbound SMTP session. New log_selector, +tls_sni.
2653 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2654 NULL dereference. Report and patch from Alun Jones.
2656 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2657 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2658 Not seeing resolver debug output on NetBSD, but suspect this is a
2659 resolver implementation change.
2661 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2662 Left warnings. Added "eximon gdb" invocation mode.
2664 PP/21 Defaulting "accept_8bitmime" to true, not false.
2666 PP/22 Added -bw for inetd wait mode support.
2668 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2669 locate the relevant includes and libraries. Made this the default.
2671 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2672 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2674 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2675 This may cause build issues on older platforms.
2677 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2678 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2679 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2680 Added SNI support via GnuTLS too.
2681 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2683 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2685 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2686 Applied second patch from Janne, implementing suggestion to default
2687 multiple-strings-in-record handling to match SPF spec.
2689 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2691 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2692 read-only, out of scope).
2693 Patch from Wolfgang Breyha, report from Stuart Northfield.
2695 PP/29 Fix three issues highlighted by clang analyser static analysis.
2696 Only crash-plausible issue would require the Cambridge-specific
2697 iplookup router and a misconfiguration.
2698 Report from Marcin Mirosław.
2700 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2702 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2703 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2704 As part of this, removing so much warning spew let me fix some minor
2705 real issues in debug logging.
2707 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2708 assignment on my part. Fixed.
2710 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2711 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2712 Janne Snabb (who went above and beyond: thank you).
2714 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2715 string otherwise requires a connection and a bunch more work and it's
2716 relatively easy to get wrong. Should also expose TLS library linkage
2719 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2720 64-bit ${eval} (JH/03).
2722 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2723 GNU libc to support some of the 64-bit stuff, should not lead to
2724 conflicts. Defined before os.h is pulled in, so if a given platform
2725 needs to override this, it can.
2727 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2728 protection layer was required, which is not implemented.
2729 Bugzilla 1254, patch from Wolfgang Breyha.
2731 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2732 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2733 tls_dhparam take prime identifiers. Also unbreak combination of
2734 OpenSSL+DH_params+TLSSNI.
2736 PP/39 Disable SSLv2 by default in OpenSSL support.
2742 PP/01 Solaris build fix for Oracle's LDAP libraries.
2743 Bugzilla 1109, patch from Stephen Usher.
2745 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2747 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2750 TF/02 Fix a couple more cases where we did not log the error message
2751 when unlink() failed. See also change 4.74-TF/03.
2753 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2754 lock up or crash if it happened to be inside a call to libc when it
2755 got a SIGUSR1 from exiwhat.
2757 The SIGUSR1 handler appends the current process status to the process
2758 log which is later printed by exiwhat. It used to use the general
2759 purpose logging code to do this, but several functions it calls are
2760 not safe for signals.
2762 The new output code in the SIGUSR1 handler is specific to the process
2763 log, and simple enough that it's easy to inspect for signal safety.
2764 Removing some special cases also simplifies the general logging code.
2765 Removing the spurious timestamps from the process log simplifies
2768 TF/04 Improved ratelimit ACL condition.
2770 The /noupdate option has been deprecated in favour of /readonly which
2771 has clearer semantics. The /leaky, /strict, and /readonly update modes
2772 are mutually exclusive. The update mode is no longer included in the
2773 database key; it just determines when the database is updated. (This
2774 means that when you upgrade Exim will forget old rate measurements.)
2776 Exim now checks that the per_* options are used with an update mode that
2777 makes sense for the current ACL. For example, when Exim is processing a
2778 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2779 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2780 must specify per_mail/readonly. If you omit the update mode it defaults to
2781 /leaky where that makes sense (as before) or /readonly where required.
2783 The /noupdate option is now undocumented but still supported for
2784 backwards compatibility. It is equivalent to /readonly except that in
2785 ACLs where /readonly is required you may specify /leaky/noupdate or
2786 /strict/noupdate which are treated the same as /readonly.
2788 A useful new feature is the /count= option. This is a generalization
2789 of the per_byte option, so that you can measure the throughput of other
2790 aggregate values. For example, the per_byte option is now equivalent
2791 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2793 The per_rcpt option has been generalized using the /count= mechanism
2794 (though it's more complicated than the per_byte equivalence). When it is
2795 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2796 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2797 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2798 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2799 Note that using per_rcpt with a non-readonly update mode in more than
2800 one ACL will cause the recipients to be double-counted. (The per_mail
2801 and per_byte options don't have this problem.)
2803 The handling of very low rates has changed slightly. If the computed rate
2804 is less than the event's count (usually one) then this event is the first
2805 after a long gap. In this case the rate is set to the same as this event's
2806 count, so that the first message of a spam run is counted properly.
2808 The major new feature is a mechanism for counting the rate of unique
2809 events. The new per_addr option counts the number of different
2810 recipients that someone has sent messages to in the last time period. It
2811 behaves like per_rcpt if all the recipient addresses are different, but
2812 duplicate recipient addresses do not increase the measured rate. Like
2813 the /count= option this is a general mechanism, so the per_addr option
2814 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2815 example, measure the rate that a client uses different sender addresses
2816 with the options per_mail/unique=$sender_address. There are further
2817 details in the main documentation.
2819 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2821 TF/06 Removed a few PCRE remnants.
2823 TF/07 Automatically extract Exim's version number from tags in the git
2824 repository when doing development or release builds.
2826 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2827 Bugzilla 879. Patch from Paul Fisher.
2829 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2830 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2833 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2835 PP/05 Variable $av_failed, true if the AV scanner deferred.
2836 Bugzilla 1078. Patch from John Horne.
2838 PP/06 Stop make process more reliably on build failure.
2839 Bugzilla 1087. Patch from Heiko Schlittermann.
2841 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2842 Bugzilla 1089. Patch from Heiko Schlittermann.
2844 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2845 Bugzilla 1131. Patch from Holger Weiß.
2847 PP/09 Handle IPv6 addresses with SPF.
2848 Bugzilla 860. Patch from Wolfgang Breyha.
2850 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2852 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2855 PP/11 match_* no longer expand right-hand-side by default.
2856 New compile-time build option, EXPAND_LISTMATCH_RHS.
2857 New expansion conditions, "inlist", "inlisti".
2859 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2861 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2863 PP/14 fix log_write() format string regression from TF/03.
2864 Bugzilla 1152. Patch from Dmitry Isaikin.
2870 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2872 PP/02 Harmonised TLS library version reporting; only show if debugging.
2873 Layout now matches that introduced for other libraries in 4.74 PP/03.
2875 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2877 PP/04 New "dns_use_edns0" global option.
2879 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2882 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2883 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2885 TK/01 Updated PolarSSL code to 0.14.2.
2886 Bugzilla 1097. Patch from Andreas Metzler.
2888 PP/07 Catch divide-by-zero in ${eval:...}.
2889 Fixes bugzilla 1102.
2891 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2894 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2895 format-string attack -- SECURITY: remote arbitrary code execution.
2897 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2898 time unintentionally subject to list matching rules, letting the header
2899 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2900 arbitrary string expansion). This allowed for information disclosure.
2902 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2903 INT_MIN/-1 -- value coerced to INT_MAX.
2909 NM/01 Workaround for PCRE version dependency in version reporting
2912 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2913 This fixes portability to compilers other than gcc, notably
2914 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2916 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2917 makefiles for portability to HP-UX and POSIX correctness.
2919 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2920 Also via indented variable definition in the Makefile.
2921 (Debugging by Oliver Heesakkers).
2923 PP/02 Restore caching of spamd results with expanded spamd_address.
2924 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2926 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2927 Improves build reliability. Fix from: Frank Elsner
2929 NM/02 Fix wide character breakage in the rfc2047 coding
2930 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2932 NM/03 Allow underscore in dnslist lookups
2933 Fixes bug 1026. Patch from Graeme Fowler
2935 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2936 Code patches from Adam Ciarcinski of NetBSD.
2938 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2941 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2942 is logged, to avoid truncation. Patch from John Horne.
2944 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2945 Patch from Jakob Hirsch.
2947 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2948 SQL string expansion failure details.
2949 Patch from Andrey Oktyabrski.
2951 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2952 Patch from Simon Arlott.
2954 PP/09 New lookups functionality failed to compile on old gcc which rejects
2955 extern declarations in function scope.
2956 Patch from Oliver Fleischmann
2958 PP/10 Use sig_atomic_t for flags set from signal handlers.
2959 Check getgroups() return and improve debugging.
2960 Fixed developed for diagnosis in bug 927 (which turned out to be
2963 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2964 Patch from Mark Zealey.
2966 PP/12 Bugzilla 1056: Improved spamd server selection.
2967 Patch from Mark Zealey.
2969 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2970 Based on patch from Heiko Schlittermann.
2972 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2973 Patch from Uwe Doering, sign-off by Michael Haardt.
2975 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2976 variable declaration deep within a block. Bug and patch from
2979 PP/15 lookups-Makefile IRIX compatibility coercion.
2981 PP/16 Make DISABLE_DKIM build knob functional.
2983 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2984 Patch by Simon Arlott
2986 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2987 variable argument macros. Our copy now differs from upstream.
2993 TF/01 Failure to get a lock on a hints database can have serious
2994 consequences so log it to the panic log.
2996 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2997 controlled using the smtp_confirmation log selector.
2999 TF/03 Include the error message when we fail to unlink a spool file.
3001 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3002 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3003 for maintaining out-of-tree patches for some time.
3005 PP/01 Bugzilla 139: Documentation and portability issues.
3006 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3007 Handle per-OS dynamic-module compilation flags.
3009 PP/02 Let /dev/null have normal permissions.
3010 The 4.73 fixes were a little too stringent and complained about the
3011 permissions on /dev/null. Exempt it from some checks.
3012 Reported by Andreas M. Kirchwitz.
3014 PP/03 Report version information for many libraries, including
3015 Exim version information for dynamically loaded libraries. Created
3016 version.h, now support a version extension string for distributors
3017 who patch heavily. Dynamic module ABI change.
3019 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3020 privilege escalation vulnerability whereby the Exim run-time user
3021 can cause root to append content of the attacker's choosing to
3024 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3027 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3028 If dropping privileges for untrusted macros, we disabled normal logging
3029 on the basis that it would fail; for the Exim run-time user, this is not
3030 the case, and it resulted in successful deliveries going unlogged.
3031 Fixed. Reported by Andreas Metzler.
3037 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3038 only prepend for the Resent-* case. Fixes regression introduced in
3039 Exim 4.70 by NM/22 for Bugzilla 607.
3041 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3042 increasing numbers of administrators be bitten by this.
3044 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3046 PP/03 Bugzilla 994: added openssl_options main configuration option.
3048 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3050 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3052 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3054 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3055 without a peer certificate, leading to a segfault because of an
3056 assumption that peers always have certificates. Be a little more
3057 paranoid. Problem reported by Martin Tscholak.
3059 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3060 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3061 NB: ClamAV planning to remove STREAM in "middle of 2010".
3062 CL also introduces -bmalware, various -d+acl logging additions and
3063 more caution in buffer sizes.
3065 PP/09 Implemented reverse_ip expansion operator.
3067 PP/10 Bugzilla 937: provide a "debug" ACL control.
3069 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3071 PP/12 Bugzilla 973: Implement --version.
3073 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3075 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3077 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3079 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3080 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3081 ignore trailing whitespace.
3083 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3085 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3088 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3089 Notification from Dr Andrew Aitchison.
3091 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3092 ExtendedDetectionInfo response format.
3093 Notification from John Horne.
3095 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3098 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3099 XSL and documented dependency on system catalogs, with examples of how
3102 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3105 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3106 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3107 configuration file which is writeable by the Exim user or group.
3109 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3110 of configuration files to cover files specified with the -C option if
3111 they are going to be used with root privileges, not just the default
3114 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3115 option (effectively making it always true).
3117 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3118 files to be used while preserving root privileges.
3120 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3121 that rogue child processes cannot use them.
3123 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3124 run-time user, instead of root.
3126 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3127 Exim run-time user without dropping privileges.
3129 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3130 result string, instead of calling string_vformat() twice with the same
3133 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3134 for other users. Others should always drop root privileges if they use
3135 -C on the command line, even for a whitelisted configure file.
3137 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3139 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3145 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3146 $data_path, and $header_path variables; fixed documentation bugs and
3149 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3150 exipick to access non-standard spools, including the "frozen" queue
3153 NM/01 Bugzilla 965: Support mysql stored procedures.
3154 Patch from Alain Williams
3156 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3158 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3159 Patch from Andreas Metzler
3161 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3162 Patch from Kirill Miazine
3164 NM/05 Bugzilla 671: Added umask to procmail example.
3166 JJ/03 installed exipick 20100323.0, fixing doc bug
3168 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3169 directory. Notification and patch from Dan Rosenberg.
3171 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3173 TK/02 Improve log output when DKIM signing operation fails.
3175 MH/01 Treat the transport option dkim_domain as a colon separated
3176 list, not as a single string, and sign the message with each element,
3177 omitting multiple occurences of the same signer.
3179 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3180 Bugzilla 985, 986. Patch by Simon Arlott
3182 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3183 Patch by Simon Arlott
3185 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3186 MBX locking. Notification from Dan Rosenberg.
3192 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3194 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3196 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3198 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3200 NM/04 Bugzilla 915: Flag broken perl installation during build.
3206 TK/01 Added patch by Johannes Berg that expands the main option
3207 "spamd_address" if it starts with a dollar sign.
3209 TK/02 Write list of recipients to X-Envelope-Sender header when building
3210 the mbox-format spool file for content scanning (suggested by Jakob
3213 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3214 (http://www.dcc-servers.net/) support via dccifd. Activated by
3215 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3217 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3218 by Mark Daniel Reidel <mr@df.eu>.
3220 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3221 When building exim an external PCRE library is now needed -
3222 PCRE is a system library on the majority of modern systems.
3223 See entry on PCRE_LIBS in EDITME file.
3225 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3226 conversation. Added nologin parameter to request.
3227 Patch contributed by Kirill Miazine.
3229 TF/01 Do not log submission mode rewrites if they do not change the address.
3231 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3233 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3234 log files in place. Contributed by Roberto Lima.
3236 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3238 TF/03 Bugzilla 615: When checking the local_parts router precondition
3239 after a local_part_suffix or local_part_prefix option, Exim now
3240 does not use the address's named list lookup cache, since this
3241 contains cached lookups for the whole local part.
3243 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3244 Robert Millan. Documentation is in experimental-spec.txt.
3246 TF/04 Bugzilla 668: Fix parallel build (make -j).
3248 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3250 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3251 Patch provided by Jan Srzednicki.
3253 TF/05 Leading white space used to be stripped from $spam_report which
3254 wrecked the formatting. Now it is preserved.
3256 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3257 that they are available at delivery time.
3259 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3261 TF/08 TLS error reporting now respects the incoming_interface and
3262 incoming_port log selectors.
3264 TF/09 Produce a more useful error message if an SMTP transport's hosts
3265 setting expands to an empty string.
3267 NM/06 Bugzilla 744: EXPN did not work under TLS.
3268 Patch provided by Phil Pennock.
3270 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3271 Patch provided by Richard Godbee.
3273 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3274 acl_smtp_notquit, added index entry.
3276 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3277 Patch provided by Eugene Bujak.
3279 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3280 accept(). Patch provided by Maxim Dounin.
3282 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3283 Patch provided by Phil Pennock.
3285 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3287 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3288 Patch provided by Brad "anomie" Jorsch.
3290 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3291 Patch provided by Dean Brooks.
3293 TK/05 Add native DKIM support (does not depend on external libraries).
3295 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3296 Patch provided by Graeme Fowler.
3298 NM/16 Bugzilla 851: Documentation example syntax fix.
3300 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3302 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3305 NM/19 Bugzilla 745: TLS version reporting.
3306 Patch provided by Phil Pennock.
3308 NM/20 Bugzilla 167: bool: condition support.
3309 Patch provided by Phil Pennock.
3311 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3312 clients. Patch provided by Phil Pennock.
3314 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3315 Patch provided by Brad "anomie" Jorsch.
3317 NM/23 Bugzilla 687: Fix misparses in eximstats.
3318 Patch provided by Heiko Schlittermann.
3320 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3321 Patch provided by Heiko Schlittermann.
3323 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3324 plus update to original patch.
3326 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3328 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3329 Patch provided by David Brownlee.
3331 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3333 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3335 NM/30 Bugzilla 888: TLS documentation bugfixes.
3337 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3339 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3340 Unlike the original bugzilla I have changed all shell scripts in src tree.
3342 NM/33 Bugzilla 898: Transport filter timeout fix.
3343 Patch by Todd Rinaldo.
3345 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3346 Patch by Serge Demonchaux.
3348 NM/35 Bugzilla 39: Base64 decode bug fixes.
3349 Patch by Jakob Hirsch.
3351 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3353 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3355 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3357 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3363 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3364 ALT-N's libdkim that I have put here:
3365 http://duncanthrax.net/exim-experimental/
3367 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3368 were called 'true' and it seems that C99 defines that as a reserved
3369 keyword to be used with 'bool' variable types. That means you could
3370 not include C99-style headers which use bools without triggering
3371 build errors in sieve.c.
3373 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3374 as mailq or other aliases. Changed the --help handling significantly
3375 to do whats expected. exim_usage() emits usage/help information.
3377 SC/01 Added the -bylocaldomain option to eximstats.
3379 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3381 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3383 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3389 PH/01 Another patch from the Sieve maintainer.
3391 PH/02 When an IPv6 address is converted to a string for single-key lookup
3392 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3393 dots are used instead of colons so that keys in lsearch files need not
3394 contain colons. This was done some time before quoting was made available
3395 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3396 (notated using the quote facility) so as to distinguish them from IPv4
3397 keys. This meant that lookups for IP addresses in host lists did not work
3398 for iplsearch lookups.
3400 This has been fixed by arranging for IPv6 addresses to be expressed with
3401 colons if the lookup type is iplsearch. This is not incompatible, because
3402 previously such lookups could never work.
3404 The situation is now rather anomalous, since one *can* have colons in
3405 ordinary lsearch keys. However, making the change in all cases is
3406 incompatible and would probably break a number of configurations.
3408 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3411 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3412 conversion specification without a maximum field width, thereby enabling
3413 a rogue spamd server to cause a buffer overflow. While nobody in their
3414 right mind would setup Exim to query an untrusted spamd server, an
3415 attacker that gains access to a server running spamd could potentially
3416 exploit this vulnerability to run arbitrary code as the Exim user.
3418 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3419 $primary_hostname instead of what libspf2 thinks the hosts name is.
3421 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3422 a directory entry by the name of the lookup key. Previously, if a
3423 symlink pointed to a non-existing file or a file in a directory that
3424 Exim lacked permissions to read, a lookup for a key matching that
3425 symlink would fail. Now it is enough that a matching directory entry
3426 exists, symlink or not. (Bugzilla 503.)
3428 PH/03 The body_linecount and body_zerocount variables are now exported in the
3431 PH/04 Added the $dnslist_matched variable.
3433 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3434 This means they are set thereafter only if the connection becomes
3437 PH/06 Added the client_condition to authenticators so that some can be skipped
3438 by clients under certain conditions.
3440 PH/07 The error message for a badly-placed control=no_multiline_responses left
3441 "_responses" off the end of the name.
3443 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3445 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3446 (without spaces) instead of just copying the configuration text.
3448 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3450 PH/11 Added $max_received_linelength.
3452 PH/12 Added +ignore_defer and +include_defer to host lists.
3454 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3455 way in which PCRE > 7.0 is built.
3457 PH/14 Implemented queue_only_load_latch.
3459 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3460 MAIL command. The effect was to mangle the value on 64-bit systems.
3462 PH/16 Another patch from the Sieve maintainer.
3464 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3466 PH/18 If a system quota error occurred while trying to create the file for
3467 a maildir delivery, the message "Mailbox is full" was not appended to the
3468 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3469 only to a quota excession during the actual writing of the file.
3471 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3472 characters?) which causes problems in log lines. The DN values are now
3473 passed through string_printing() before being added to log lines.
3475 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3476 and InterBase are left for another time.)
3478 PH/21 Added message_body_newlines option.
3480 PH/22 Guard against possible overflow in moan_check_errorcopy().
3482 PH/23 POSIX allows open() to be a macro; guard against that.
3484 PH/24 If the recipient of an error message contained an @ in the local part
3485 (suitably quoted, of course), incorrect values were put in $domain and
3486 $local_part during the evaluation of errors_copy.
3492 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3493 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3496 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3497 issue a MAIL command.
3499 PH/02 In an ACL statement such as
3501 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3503 if a client was not listed at all, or was listed with a value other than
3504 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3505 the condition was not true (as it should be), so access was not denied.
3506 The bug was that the ! inversion was incorrectly passed on to the second
3507 item. This has been fixed.
3509 PH/03 Added additional dnslists conditions == and =& which are different from
3510 = and & when the dns lookup returns more than one IP address.
3512 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3513 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3515 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3516 FSYNC, which compiles an option called disable_fsync that allows for
3517 bypassing fsync(). The documentation is heavily laced with warnings.
3519 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3521 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3522 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3523 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3524 including adding "make clean"; (3) Added -fPIC when compiling the test
3525 dynamically loaded module, to get rid of a warning.
3527 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3528 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3529 = 0s. The bug is otherwise harmless.
3531 PH/07 There was a bug in the dovecot authenticator such that the value of
3532 $auth1 could be overwritten, and so not correctly preserved, after a
3533 successful authentication. This usually meant that the value preserved by
3534 the server_setid option was incorrect.
3536 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3538 PH/09 Installed PCRE release 7.0.
3540 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3541 run for batched SMTP input. It is now run at the start of every message
3542 in the batch. While fixing this I discovered that the process information
3543 (output by running exiwhat) was not always getting set for -bs and -bS
3544 input. This is fixed, and it now also says "batched" for BSMTP.
3546 PH/11 Added control=no_pipelining.
3548 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3549 patch, slightly modified), and move the expansion of helo_data till after
3550 the connection is made in the smtp transport (so it can use these
3553 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3555 PH/14 Added log_selector = +pid.
3557 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3559 PH/16 Add ${if forany and ${if forall.
3561 PH/17 Added dsn_from option to vary the From: line in DSNs.
3563 PH/18 Flush SMTP output before performing a callout, unless control =
3564 no_callout_flush is set.
3566 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3567 was true (the default) a successful delivery failed to delete the retry
3568 item, thus causing premature timeout of the address. The bug is now
3571 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3573 PH/21 Long custom messages for fakedefer and fakereject are now split up
3574 into multiline responses in the same way that messages for "deny" and
3575 other ACL rejections are.
3577 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3578 with slight modification.
3580 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3581 draft, changing the syntax and factoring some duplicate code".
3583 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3584 for deliveries of the second and subsequent messages over the same SMTP
3587 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3588 ${reduce, with only minor "tidies".
3590 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3592 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3593 expansion side effects.
3595 PH/27 When a message times out after an over-quota error from an Exim-imposed
3596 quota, the bounce message says "mailbox is full". This message was not
3597 being given when it was a system quota that was exceeded. It now should
3600 MH/03 Made $recipients available in local_scan(). local_scan() already has
3601 better access to the recipient list through recipients_list[], but
3602 $recipients can be useful in postmaster-provided expansion strings.
3604 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3605 in the case of a MAIL command with additional options following the
3606 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3607 were accidentally chopped off.
3609 PH/29 SMTP synchronization checks are implemented when a command is read -
3610 there is a check that no more input is waiting when there shouldn't be
3611 any. However, for some commands, a delay in an ACL can mean that it is
3612 some time before the response is written. In this time, more input might
3613 arrive, invalidly. So now there are extra checks after an ACL has run for
3614 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3615 pipelining has not been advertised.
3617 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3619 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3620 returned in the SMTP response when smtp_return_error_details was false.
3621 This has been fixed.
3623 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3624 instead of the C I/O that was originally supplied, because problems were
3625 reported on Solaris.
3627 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3628 Exim which did not show up earlier: it was assuming that a call to
3629 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3630 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3631 was a macro that became an assignment, so it seemed to work. This has
3632 changed to a proper function call with a void return, hence the compile
3633 error. Exim's code has been fixed.
3635 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3638 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3640 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3642 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3643 because it turns out to be too big or there is a timeout) there is no
3644 "Completed" line in the log. When some messages of this type were
3645 selected by exigrep, they were listed as "not completed". Others were
3646 picked up by some special patterns. I have improved the selection
3647 criteria to be more general.
3649 PH/38 The host_find_failed option in the manualroute router can now be set
3650 to "ignore", to completely ignore a host whose IP address cannot be
3651 found. If all hosts are ignored, the behaviour is controlled by the new
3652 host_all_ignored option.
3654 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3655 homing or because of multiple MX records with /mx) generated more than
3656 one IP address, and the following item turned out to be the local host,
3657 all the secondary addresses of the first item were incorrectly removed
3658 from the list, along with the local host and any following hosts (which
3659 is what is supposed to happen).
3661 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3662 whoever called Exim into the -H file. In the case of the daemon it was
3663 behaving confusingly. When first started, it used values for whoever
3664 started the daemon, but after a SIGHUP it used the Exim user (because it
3665 calls itself on a restart). I have changed the code so that it now always
3668 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3669 message are rejected with the same error (e.g. no authentication or bad
3670 sender address), and a DATA command is nevertheless sent (as can happen
3671 with PIPELINING or a stupid MUA), the error message that was given to the
3672 RCPT commands is included in the rejection of the DATA command. This is
3673 intended to be helpful for MUAs that show only the final error to their
3676 PH/42 Another patch from the Sieve maintainer.
3678 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3679 Eximstats - Fixed some broken HTML links and added missing column headers
3681 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3682 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3684 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3686 PH/43 Yet another patch from the Sieve maintainer.
3688 PH/44 I found a way to check for a TCP/IP connection going away before sending
3689 the response to the final '.' that terminates a message, but only in the
3690 case where the client has not sent further data following the '.'
3691 (unfortunately, this is allowed). However, in many cases there won't be
3692 any further data because there won't be any more messages to send. A call
3693 to select() can be used: if it shows that the input is "ready", there is
3694 either input waiting, or the socket has been closed. An attempt to read
3695 the next input character can distinguish the two cases. Previously, Exim
3696 would have sent an OK response which the client would never have see.
3697 This could lead to message repetition. This fix should cure that, at
3698 least in a lot of common cases.
3700 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3701 advertised in response to EHLO.
3707 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3708 fixed by 4.65/MH/01 (is this a record?) are fixed:
3710 (i) An empty string was always treated as zero by the numeric comparison
3711 operators. This behaviour has been restored.
3713 (ii) It is documented that the numeric comparison operators always treat
3714 their arguments as decimal numbers. This was broken in that numbers
3715 starting with 0 were being interpreted as octal.
3717 While fixing these problems I realized that there was another issue that
3718 hadn't been noticed. Values of message_size_limit (both the global option
3719 and the transport option) were treated as octal if they started with 0.
3720 The documentation was vague. These values are now always treated as
3721 decimal, and I will make that clear in the documentation.
3727 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3728 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3731 MH/01 Don't check that the operands of numeric comparison operators are
3732 integers when their expansion is in "skipping" mode (fixes bug
3733 introduced by 4.64-PH/07).
3735 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3736 child addresses, Exim now panics and dies. Previously, because the count
3737 is held in a short int, deliveries were likely to be lost. As such a
3738 large number of recipients for a single message is ridiculous
3739 (performance will be very, very poor), I have chosen to impose a limit
3740 rather than extend the field.
3746 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3747 leftover -K file (the existence of which was triggered by #402).
3748 While we were at it, introduced process PID as part of the -K
3749 filename. This should rule out race conditions when creating
3752 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3753 processing considerably. Previous code took too long for large mails,
3754 triggering a timeout which in turn triggers #401.
3756 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3757 in the DK code in transports.c. sendfile() is not really portable,
3758 hence the _LINUX specificness.
3760 TF/01 In the add_headers option to the mail command in an Exim filter,
3761 there was a bug that Exim would claim a syntax error in any
3762 header after the first one which had an odd number of characters
3765 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3766 callout verification, Exim cached a "reject" for the entire domain. This
3767 is correct for most verifications, but it is not correct for a recipient
3768 verification with use_sender or use_postmaster set, because in that case
3769 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3770 case of MAIL FROM:<> rejection from other early rejections (e.g.
3771 rejection of HELO). When verifying a recipient using a non-null MAIL
3772 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3773 Whatever the result of the callout, the value of the domain cache is
3774 left unchanged (for any other kind of callout, getting as far as trying
3775 RCPT means that the domain itself is ok).
3777 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3780 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3781 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3782 session, the daemon ignores ECONNECT errors and logs others; it now
3783 ignores EPIPE as well.
3785 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3786 (quoted-printable decoding).
3788 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3789 later the small subsequent patch to fix an introduced bug.
3791 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3793 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3795 PH/08 An error is now given if message_size_limit is specified negative.
3797 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3798 to be given (somewhat) arbitrary names.
3800 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3803 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3804 miscellaneous code fixes
3806 PH/10 Added the log_reject_target ACL modifier to specify where to log
3809 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3810 hostname. This is wrong, because it relates to the incoming message (and
3811 probably the interface on which it is arriving) and not to the outgoing
3812 callout (which could be using a different interface). This has been
3813 changed to use the value of the helo_data option from the smtp transport
3814 instead - this is what is used when a message is actually being sent. If
3815 there is no remote transport (possible with a router that sets up host
3816 addresses), $smtp_active_hostname is used.
3818 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3819 tweaks were necessary in order to get it to work (see also 21 below):
3820 (a) The code assumed that strncpy() returns a negative number on buffer
3821 overflow, which isn't the case. Replaced with Exim's string_format()
3823 (b) There were several signed/unsigned issues. I just did the minimum
3824 hacking in of casts. There is scope for a larger refactoring.
3825 (c) The code used strcasecmp() which is not a standard C function.
3826 Replaced with Exim's strcmpic() function.
3827 (d) The code set only $1; it now sets $auth1 as well.
3828 (e) A simple test gave the error "authentication client didn't specify
3829 service in request". It would seem that Dovecot has changed its
3830 interface. Fortunately there's a specification; I followed it and
3831 changed what the client sends and it appears to be working now.
3833 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3836 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3837 address A is aliased to B and C, where B exists and C does not. Without
3838 -v the output is "A verified" because verification stops after a
3839 successful redirection if more than one address is generated. However,
3840 with -v the child addresses are also verified. Exim was outputting "A
3841 failed to verify" and then showing the successful verification for C,
3842 with its parentage. It now outputs "B failed to verify", showing B's
3843 parentage before showing the successful verification of C.
3845 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3846 look up a TXT record in a specific list after matching in a combined
3849 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3850 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3851 they consult the DNS. I had assumed they would set it the way they
3852 wanted; and indeed my experiments on Linux seem to show that in some
3853 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3854 To be on the safe side, however, I have now made the interface to
3855 host_find_byname() similar to host_find_bydns(), with an argument
3856 containing the DNS resolver options. The host_find_byname() function now
3857 sets these options at its start, just as host_find_bydns() does. The smtp
3858 transport options dns_qualify_single and dns_search_parents are passed to
3859 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3860 of host_find_byname() use the default settings of RES_DEFNAMES
3861 (qualify_single) but not RES_DNSRCH (search_parents).
3863 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3864 spool_read_header() do less string testing, by means of a preliminary
3865 switch on the second character of optional "-foo" lines. (This is
3866 overdue, caused by the large number of possibilities that now exist.
3867 Originally there were few.) While I was there, I also converted the
3868 str(n)cmp tests so they don't re-test the leading "-" and the first
3869 character, in the hope this might squeeze out yet more improvement.
3871 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3872 flag allowing group syntax was set by the header_syntax check but not
3873 turned off, possible causing trouble later; (2) The flag was not being
3874 set at all for the header_verify test, causing "group"-style headers to
3875 be rejected. I have now set it in this case, and also caused header_
3876 verify to ignore an empty address taken from a group. While doing this, I
3877 came across some other cases where the code for allowing group syntax
3878 while scanning a header line wasn't quite right (mostly, not resetting
3879 the flag correctly in the right place). These bugs could have caused
3880 trouble for malformed header lines. I hope it is now all correct.
3882 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3883 with the "reply" argument non-NULL. The code, however (which originally
3884 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3885 but it didn't always do it. This confused somebody who was copying the
3886 code for some other use. I have removed all the tests.
3888 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3889 feature that was used to support insecure browsers during the U.S. crypto
3890 embargo. It requires special client support, and Exim is probably the
3891 only MTA that supported it -- and would never use it because real RSA is
3892 always available. This code has been removed, because it had the bad
3893 effect of slowing Exim down by computing (never used) parameters for the
3894 RSA_EXPORT functionality.
3896 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3897 authenticator to fail if there's a tab character in the incoming data
3898 (there should never be unless someone is messing about, as it's supposed
3899 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3900 if the connection is using TLS or if the remote IP is the same as the
3901 local IP, and the "valid-client-cert option" if a client certificate has
3904 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3905 authenticators. This can be used for authorization after authentication
3906 succeeds. (In the case of plaintext, it servers for both authentication
3909 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3910 if any retry times were supplied.
3912 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3913 connection, where there is no client IP address to check. In this
3914 situation, the verify now always succeeds.
3916 PH/25 Applied John Jetmore's -Mset patch.
3918 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3920 PH/27 In a string expansion for a processed (not raw) header when multiple
3921 headers of the same name were present, leading whitespace was being
3922 removed from all of them, but trailing whitespace was being removed only
3923 from the last one. Now trailing whitespace is removed from each header
3924 before concatenation. Completely empty headers in a concatenation (as
3925 before) are ignored.
3927 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3928 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3930 PH/29 [Removed. This was a change that I later backed out, and forgot to
3931 correct the ChangeLog entry (that I had efficiently created) before
3932 committing the later change.]
3934 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3935 address errors (4xx response to RCPT) over the same connection as other
3936 messages routed to the same hosts. Such deliveries are always "forced",
3937 so retry times are not inspected. This resulted in far too many retries
3938 for the affected addresses. The effect occurred only when there were more
3939 hosts than the hosts_max_try setting in the smtp transport when it had
3940 the 4xx errors. Those hosts that it had tried were not added to the list
3941 of hosts for which the message was waiting, so if all were tried, there
3942 was no problem. Two fixes have been applied:
3944 (i) If there are any address or message errors in an SMTP delivery, none
3945 of the hosts (tried or untried) are now added to the list of hosts
3946 for which the message is waiting, so the message should not be a
3947 candidate for sending over the same connection that was used for a
3948 successful delivery of some other message. This seems entirely
3949 reasonable: after all the message is NOT "waiting for some host".
3950 This is so "obvious" that I'm not sure why it wasn't done
3951 previously. Hope I haven't missed anything, but it can't do any
3952 harm, as the worst effect is to miss an optimization.
3954 (ii) If, despite (i), such a delivery is accidentally attempted, the
3955 routing retry time is respected, so at least it doesn't keep
3956 hammering the server.
3958 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3959 in ${readsocket because some servers need this prod.
3961 PH/32 Added some extra debug output when updating a wait-xxx database.
3963 PH/33 The hint "could be header name not terminated by colon", which has been
3964 given for certain expansion errors for a long time, was not being given
3965 for the ${if def:h_colon_omitted{... case.
3967 PH/34 The spec says: "With one important exception, whenever a domain list is
3968 being scanned, $domain contains the subject domain." There was at least
3969 one case where this was not true.
3971 PH/35 The error "getsockname() failed: connection reset by peer" was being
3972 written to the panic log as well as the main log, but it isn't really
3973 panic-worthy as it just means the connection died rather early on. I have
3974 removed the panic log writing for the ECONNRESET error when getsockname()
3977 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3978 runs only) independently of the message's sender address. This meant
3979 that, if the 4xx error was in fact related to the sender, a different
3980 message to the same recipient with a different sender could confuse
3981 things. In particular, this can happen when sending to a greylisting
3982 server, but other circumstances could also provoke similar problems.
3983 I have changed the default so that the retry time for these errors is now
3984 based a combination of the sender and recipient addresses. This change
3985 can be overridden by setting address_retry_include_sender=false in the
3988 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3989 remote server are returned as part of bounce messages. This was not
3990 happening for LMTP over a pipe (the lmtp transport), but now it is the
3991 same for both kinds of LMTP.
3993 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3994 in header lines that were in fact CNAMEs. This is no longer the case.
3996 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3997 and queue runs started by the daemon processed all messages. This has
3998 been fixed so that -R and -S can now usefully be given with -q<time>.
4000 PH/40 Import PCRE release 6.7 (fixes some bugs).
4002 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4004 PH/42 Give an error if -q is specified more than once.
4006 PH/43 Renamed the variables $interface_address and $interface_port as
4007 $received_ip_address and $received_port, to make it clear that these
4008 values apply to message reception, and not to the outgoing interface when
4009 a message is delivered. (The old names remain recognized, of course.)
4011 PH/44 There was no timeout on the connect() call when using a Unix domain
4012 socket in the ${readsocket expansion. There now is.
4014 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4015 be meaningful with "accept".
4017 SC/01 Eximstats V1.43
4018 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4020 SC/02 Eximstats V1.44
4021 Use a glob alias rather than an array ref in the generated
4022 parser. This improves both readability and performance.
4024 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4025 Collect SpamAssassin and rejection statistics.
4026 Don't display local sender or destination tables unless
4027 there is data to show.
4028 Added average volumes into the top table text output.
4030 SC/04 Eximstats V1.46
4031 Collect data on the number of addresses (recipients)
4032 as well as the number of messages.
4034 SC/05 Eximstats V1.47
4035 Added 'Message too big' to the list of mail rejection
4036 reasons (thanks to Marco Gaiarin).
4038 SC/06 Eximstats V1.48
4039 Mainlog lines which have GMT offsets and are too short to
4040 have a flag are now skipped.
4042 SC/07 Eximstats V1.49 (Alain Williams)
4043 Added the -emptyok flag.
4045 SC/08 Eximstats V1.50
4046 Fixes for obtaining the IP address from reject messages.
4048 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4049 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4050 whitespace changes from 4.64-PH/27
4052 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4055 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4056 are found, allow negative numbers in numeric criteria)
4058 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4060 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4061 to match changes made in 4.64-PH/43
4063 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4065 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4066 "domain" is actually an IP address, and force a failure. This locks out
4067 those revolvers/nameservers that support "A-for-A" lookups, in
4068 contravention of the specifications.
4070 PH/48 When a host name was looked up from an IP address, and the subsequent
4071 forward lookup of the name timed out, the host name was left in
4072 $sender_host_name, contrary to the specification.
4074 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4075 restricted to single-key lookups, Exim was not diagnosing an error if
4076 * or *@ was used with a query-style lookup.
4078 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4080 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4081 long ago, but noone interested enough thought of it. Let's just say that
4082 the "1.1" means that there are some new functions that weren't there at
4083 some point in the past.
4085 PH/51 Error processing for expansion failure of helo_data from an smtp
4086 transport during callout processing was broken.
4088 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4089 tested/used via the -bh/-bhc/-bs options.
4091 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4092 bug, fixed in subsequent PCRE releases).
4094 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4095 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4097 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4103 SC/01 Use a glob alias rather than an array ref in eximstats generated
4104 parser. This improves both readability and performance.
4106 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4107 Don't display local sender or destination tables in eximstats unless
4108 there is data to show.
4109 Added average volumes into the eximstats top table text output.
4111 SC/03 Collect data on the number of addresses (recipients) as well
4112 as the number of messages in eximstats.
4114 TF/01 Correct an error in the documentation for the redirect router. Exim
4115 does (usually) call initgroups() when daemonizing.
4117 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4118 with consistent privilege compared to when running as a daemon.
4120 TF/03 Note in the spec that $authenticated_id is not set for local
4121 submissions from trusted users.
4123 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4124 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4126 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4127 by adding some example configuration directives to the default
4128 configuration file. A little bit of work is required to uncomment the
4129 directives and define how usernames and passwords are checked, but
4130 there is now a framework to start from.
4132 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4133 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4134 without this. I don't know how relevant this is to other LDAP libraries.
4136 PH/02 Add the verb name to the "unknown ACL verb" error.
4138 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4140 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4142 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4143 directory not expanded when it should be if an expanded home directory
4144 was set for the address (which is overridden by the transport).
4146 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4149 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4150 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4151 because it is too late at that time, and has no effect.
4153 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4154 security issue with \' (bugzilla #107). I could not use the
4155 PQescapeStringConn() function, because it needs a PGconn value as one of
4158 PH/08 When testing addresses using -bt, indicate those final addresses that
4159 are duplicates that would not cause an additional delivery. At least one
4160 person was confused, thinking that -bt output corresponded to deliveries.
4161 (Suppressing duplicates isn't a good idea as you lose the information
4162 about possibly different redirections that led to the duplicates.)
4164 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4165 systems where poll() doesn't work, in particular OS X.
4167 PH/10 Added more information to debugging output for retry time not reached.
4169 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4170 operations in malware.c.
4172 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4175 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4176 syslog_duplication set false, log lines that would normally be written
4177 both the the main log and to the reject log were not written to syslog at
4180 PH/14 In the default configuration, change the use of "message" in ACL warn
4181 statements to "add_header".
4183 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4184 not followed by a command (e.g. "seen endif").
4186 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4187 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4190 PH/17 Added extra conditions to the default value of delay_warning_condition
4194 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4195 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4196 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4199 The Auto-Submitted: and various List- headers are standardised, whereas I
4200 don't think Precedence: ever was.
4202 PH/18 Refactored debugging code in route_finduser() to show more information,
4203 in particular, the error code if getpwnam() issues one.
4205 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4206 This is apparently needed in addition to the PH/07 change above to avoid
4207 any possible encoding problems.
4209 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4210 but not after initializing Perl.
4212 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4213 output them only if debugging. By default they are written stderr,
4214 apparently, which is not desirable.
4216 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4219 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4222 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4224 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4225 authenticated or an ident call has been made. Suppress the default
4226 values for $authenticated_id and $authenticated_sender (but permit -oMai
4227 and -oMas) when testing with -bh.
4229 PH/24 Re-jigged the order of the tests in the default configuration so that the
4230 tests for valid domains and recipients precede the DNS black list and CSA
4231 tests, on the grounds that those ones are more expensive.
4233 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4234 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4235 This bug exists in every version of Exim that I still have, right back to
4238 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4239 However, an attempt to turn on case-sensitivity in a regex key by
4240 including (?-i) didn't work because the subject string was already
4241 lowercased, and the effects were non-intuitive. It turns out that a
4242 one-line patch can be used to allow (?-i) to work as expected.
4248 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4249 other effects) broke the use of negated acl sub-conditions.
4251 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4254 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4255 "Deny" causes Exim to reject the incoming connection with a 554 error.
4256 Unfortunately, if there is a major crisis, such as a disk failure,
4257 tcp-wrappers gives "deny", whereas what one would like would be some
4258 kind of temporary error. A kludge has been added to help with this.
4259 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4260 554 error is used if errno is still zero or contains ENOENT (which occurs
4261 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4264 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4266 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4267 errors. Otherwise a message that provokes a temporary error (when other
4268 messages do not) can cause a whole host to time out.
4270 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4271 addresses were routed directly to files or pipes from a redirect router.
4272 File deliveries just didn't batch; pipe deliveries might have suffered
4275 PH/06 A failure to get a lock for a hints database would erroneously always say
4276 "Failed to get write lock", even when it was really a read lock.
4278 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4279 of 0600. This has been changed to use the value of the lockfile_mode
4280 option (which defaults to 0600).
4282 PH/08 Applied small patch from the Sieve maintainer.
4284 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4285 folder from quota calculations, a direct delivery into this folder messed
4286 up the contents of the maildirsize file. This was because the regex was
4287 used only to exclude .Trash (or whatever) when the size of the mailbox
4288 was calculated. There was no check that a delivery was happening into an
4289 excluded directory. This bug has been fixed by ignoring all quota
4290 processing for deliveries into excluded directories.
4292 PH/10 Added the maildirfolder_create_regex option to appendfile.
4298 PH/01 The code for finding all the local interface addresses on a FreeBSD
4299 system running IPv6 was broken. This may well have applied to all BSD
4300 systems, as well as to others that have similar system calls. The broken
4301 code found IPv4 interfaces correctly, but gave incorrect values for the
4302 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4303 that it would not match correctly against @[] and not recognize the IPv6
4306 PH/02 The ipliteral router was not recognizing addresses of the form user@
4307 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4309 PH/03 Added disable_ipv6.
4311 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4312 decoded form, because it is most often used to construct To: headers
4313 lines in autoreplies, and the decoded form may well be syntactically
4314 invalid. However, $reply_address has leading white space removed, and all
4315 newlines turned into spaces so that the autoreply transport does not
4318 PH/05 If group was specified without a user on a router, and no group or user
4319 was specified on a transport, the group from the router was ignored.
4321 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4322 for visible compile-time settings that can be used to change these
4323 numbers, for those that want even more. Backwards compatibility with old
4324 spool files has been maintained. However, going back to a previous Exim
4325 release will lost any variables that are in spool files.
4327 PH/07 Two small changes when running in the test harness: increase delay when
4328 passing a TCP/IP connection to a new process, in case the original
4329 process has to generate a bounce, and remove special handling of
4330 127.0.0.2 (sic), which is no longer necessary.
4332 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4333 be the same on different OS.
4335 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4338 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4339 whether --show-vars was specified or not
4341 JJ/02 exipick: Added support for new ACL variable spool format introduced
4344 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4345 syntactically invalid From: or Reply-to: line, and a filter used this to
4346 generate an autoreply, and therefore failed to obtain an address for the
4347 autoreply, Exim could try to deliver to a non-existent relative file
4348 name, causing unrelated and misleading errors. What now happens is that
4349 it logs this as a hard delivery error, but does not attempt to create a
4352 PH/11 The exinext utility has a -C option for testing purposes, but although
4353 the given file was scanned by exinext itself; it wasn't being passed on
4354 when Exim was called.
4356 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4357 an end-of-file indication when reading a command response.
4359 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4360 compiled. In many other places in Exim, IPv6 addresses are always
4361 recognized, so I have changed this. It also means that IPv4 domain
4362 literals of the form [IPV4:n.n.n.n] are now always recognized.
4364 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4365 used if the router is not running as root, for example, when verifying at
4366 ACL time, or when using -bh. The debugging output from this situation was
4367 non-existent - all you got was a failure to exec. I have made two
4370 (a) Failures to set uid/gid, the current directory, or a process leader
4371 in a subprocess such as that created by queryprogram now generate
4372 suitable debugging output when -d is set.
4374 (b) The queryprogram router detects when it is not running as root,
4375 outputs suitable debugging information if -d is set, and then runs
4376 the subprocess without attempting to change uid/gid.
4378 PH/15 Minor change to Makefile for building test_host (undocumented testing
4381 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4382 additional section of a DNS packet that returns MX or SRV records.
4383 Instead, it always explicitly searches for A/AAAA records. This avoids
4384 major problems that occur when a DNS server includes only records of one
4385 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4386 fixed another bug: if SRV records were looked up and the corresponding
4387 address records were *not* found in the additional section, the port
4388 values from the SRV records were lost.
4390 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4391 using the correct key (the original address) when searching the retry
4392 rules in order to find which one to use for generating the retry hint.
4394 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4395 adding the default one. Similarly, if it contains a Reply-To: header, the
4396 errors_reply_to option, if set, is not used.
4398 PH/19 When calculating a retry time, Exim used to measure the "time since
4399 failure" by looking at the "first failed" field in the retry record. Now
4400 it does not use this if it is later than than the arrival time of the
4401 message. Instead it uses the arrival time. This makes for better
4402 behaviour in cases where some deliveries succeed, thus re-setting the
4403 "first failed" field. An example is a quota failure for a huge message
4404 when small messages continue to be delivered. Without this change, the
4405 "time since failure" will always be short, possible causing more frequent
4406 delivery attempts for the huge message than are intended.
4407 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4409 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4410 $1, $2, $3) because the numerical variables can be reset during some
4411 expansion items (e.g. "match"), thereby losing the authentication data.
4413 PH/21 Make -bV show the size of off_t variables so that the test suite can
4414 decide whether to run tests for quotas > 2G.
4416 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4417 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4418 filecount value is greater than 2G or if a quota value is greater than 2G
4419 on a system where the size of off_t is not greater than 4, a panic error
4422 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4423 never match. The debug and -bh output now contains an explicit error
4424 message indicating a malformed IPv4 address or mask.
4426 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4427 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4428 PH/23 above applies.
4430 PH/25 Do not write to syslog when running in the test harness. The only
4431 occasion when this arises is a failure to open the main or panic logs
4432 (for which there is an explicit test).
4434 PH/26 Added the /no_tell option to "control=freeze".
4436 PH/27 If a host name lookup failed very early in a connection, for example, if
4437 the IP address matched host_lookup and the reverse lookup yielded a name
4438 that did not have a forward lookup, an error message of the form "no IP
4439 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4440 could be logged. Now it outputs the IP address instead of "NULL".
4442 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4443 allows the sender and the authenticated sender to be set when
4444 submitting a message from within Exim. Since child_open_exim() is
4445 documented for local_scan(), the new function should be too.
4447 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4448 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4449 results in an empty string is now treated as unset.
4451 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4453 PH/31 Added sender_verify_fail logging option.
4455 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4456 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4457 tidied the source and removed it altogether.
4459 PH/33 When a queue run was abandoned because the load average was too high, a
4460 log line was always written; now it is written only if the queue_run log
4461 selector is set. In addition, the log line for abandonment now contains
4462 information about the queue run such as the pid. This is always present
4463 in "start" and "stop" lines but was omitted from the "abandon" line.
4465 PH/34 Omit spaces between a header name and the colon in the error message that
4466 is given when verify = headers_syntax fails (if there are lots of them,
4467 the message gets confusing).
4469 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4470 names, as there are now some PTR records that contain slashes. This check
4471 is only to protect against broken name servers that fall over on strange
4472 characters, so the fact that it applies to all lookups doesn't matter.
4474 PH/36 Now that the new test suite is complete, we can remove some of the
4475 special code in Exim that was needed for the old test suite. For example,
4476 sorting DNS records because real resolvers return them in an arbitrary
4477 order. The new test suite's fake resolver always returns records in the
4480 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4481 bounces) except when queue_only is set, to avoid logging races between
4482 the different processes.
4484 PH/38 Panic-die if .include specifies a non-absolute path.
4486 PH/39 A tweak to the "H" retry rule from its user.
4488 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4489 a label. They prevented compilation on older perls.
4491 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4492 a warning to be raised on newish perls.
4494 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4495 on queue. Changes to match documented behaviour of showing count of
4496 messages matching specified criteria.
4498 PH/40 Changed the default ident timeout from 30s to 5s.
4500 PH/41 Added support for the use of login_cap features, on those BSD systems
4501 that have them, for controlling the resources used by pipe deliveries.
4503 PH/42 The content-scanning code uses fopen() to create files in which to put
4504 message data. Previously it was not paying any attention to the mode of
4505 the files. Exim runs with umask(0) because the rest of the code creates
4506 files with open(), and sets the required mode explicitly. Thus, these
4507 files were ending up world-writeable. This was not a big issue, because,
4508 being within the spool directory, they were not world-accessible. I have
4509 created a function called modefopen, which takes an additional mode
4510 argument. It sets umask(777), creates the file, chmods it to the required
4511 mode, then resets the umask. All the relevant calls to fopen() in the
4512 content scanning code have been changed to use this function.
4514 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4515 to 24 hours. This avoids potential overflow problems when processing G
4516 and H retry rules. I suspect nobody ever tinkers with this value.
4518 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4520 PH/45 When the plaintext authenticator is running as a client, the server's
4521 challenges are checked to ensure they are valid base64 strings. By
4522 default, the authentication attempt is cancelled if an invalid string is
4523 received. Setting client_ignore_invalid_base64 true ignores these errors.
4524 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4525 they are received. Thus, the responses can be made to depend on the
4526 challenges. If an invalid string is ignored, an empty string is placed in
4529 PH/46 Messages that are created by the autoreply transport now contains a
4530 References: header, in accordance with RFCs 2822 and 3834.
4532 PH/47 Added authenticated_sender_force to the smtp transport.
4534 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4536 PH/49 Installed latest patch from the Sieve maintainer.
4538 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4539 was also set, the appendfile transport was unnecessarily scanning a
4540 directory of message files (e.g. for maildir delivery) to find the count
4541 of files (along with the size), even though it did not need this
4542 information. It now does the scan only if it needs to find either the
4543 size of the count of files.
4545 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4547 PH/52 Two bugs concerned with error handling when the smtp transport is
4550 (i) Exim was not creating retry information for temporary errors given
4551 for individual recipients after the DATA command when the smtp transport
4552 was used in LMTP mode. This meant that they could be retried too
4553 frequently, and not timed out correctly.
4555 (ii) Exim was setting the flag that allows error details to be returned
4556 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4557 recipients that were returned after the DATA command.
4559 PH/53 This is related to PH/52, but is more general: for any failing address,
4560 when detailed error information was permitted to be returned to the
4561 sender, but the error was temporary, then after the final timeout, only
4562 "retry timeout exceeded" was returned. Now it returns the full error as
4563 well as "retry timeout exceeded".
4565 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4566 do this, and (what is worse) MTAs that accept it.
4568 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4569 will now be deprecated.
4571 PH/56 New os.c-cygwin from the Cygwin maintainer.
4573 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4574 formats (previously only available in exim formats via -bpr, -bpru,
4575 and -bpra. Now also available in native and exiqgrep formats)
4577 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4578 with very large, slow to parse queues
4580 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4582 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4584 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4585 responses to authentication challenges, though it was showing the
4586 challenges; (ii) I've removed the CR characters from the debug output for
4589 PH/58 Allow for the insertion of a newline as well as a space when a string
4590 is turned into more than one encoded-word during RFC 2047 encoding. The
4591 Sieve code now uses this.
4593 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4594 data_4xx, lost_connection, tls_required.
4596 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4597 message was being sent as an SMTP response.
4599 PH/61 Add -l and -k options to exicyclog.
4601 PH/62 When verifying, if an address was redirected to one new address, so that
4602 verification continued, and the new address failed or deferred after
4603 having set something in $address_data, the value of $address_data was not
4604 passed back to the ACL. This was different to the case when no
4605 redirection occurred. The value is now passed back in both cases.
4607 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4608 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4609 use login_cap.h, so on its own it isn't the distinguishing feature. The
4610 new name refers directly to the setclassresources() function.
4612 PH/65 Added configuration files for NetBSD3.
4614 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4616 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4617 is preferred over IPv4.
4619 PH/68 The bounce_return_message and bounce_return_body options were not being
4620 honoured for bounces generated during the reception of non-SMTP messages.
4621 In particular, this applied to messages rejected by the ACL. This bug has
4622 been fixed. However, if bounce_return_message is true and bounce_return_
4623 body is false, the headers that are returned for a non-SMTP message
4624 include only those that have been read before the error was detected.
4625 (In the case of an ACL rejection, they have all been read.)
4627 PH/69 The HTML version of the specification is now built in a directory called
4628 spec_html instead of spec.html, because the latter looks like a path with
4629 a MIME-type, and this confuses some software.
4631 PH/70 Catch two compiler warnings in sieve.c.
4633 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4634 function verify_get_ident() calls ip_connect() to connect a socket, but
4635 if the "connect()" function timed out, ip_connect() used to close the
4636 socket. However, verify_get_ident() also closes the socket later, and in
4637 between Exim writes to the log, which may get opened at this point. When
4638 the socket was closed in ip_connect(), the log could get the same file
4639 descriptor number as the socket. This naturally causes chaos. The fix is
4640 not to close the socket in ip_connect(); the socket should be closed by
4641 the function that creates it. There was only one place in the code where
4642 this was missing, in the iplookup router, which I don't think anybody now
4643 uses, but I've fixed it anyway.
4645 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4646 well as to direct DNS lookups. Otherwise the handling of names in host
4647 lists is inconsistent and therefore confusing.
4653 PH/01 Two changes to the default runtime configuration:
4655 (1) Move the checks for relay_from_hosts and authenticated clients from
4656 after to before the (commented out) DNS black list checks.
4658 (2) Add control=submission to the relay_from_hosts and authenticated
4659 clients checks, on the grounds that messages accepted by these
4660 statements are most likely to be submissions.
4662 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4664 (1) Generate an error if the third argument for the ${prvs expansion is
4667 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4670 (3) Reset the variables that are obtained from the first argument of
4671 ${prvscheck and used in the second argument before leaving the code,
4672 because their memory is reclaimed, so using them afterwards may do
4675 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4676 one (it's much easier than Tom thought :-).
4678 (5) Because of (4), we can now allow for the use of $prvscheck_result
4679 inside the third argument.
4681 PH/03 For some reason, the default setting of PATH when running a command from
4682 a pipe transport was just "/usr/bin". I have changed it to
4685 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4686 anything to be listed in the output from -bV.
4688 PH/05 When a filter generated an autoreply, the entire To: header line was
4689 quoted in the delivery log line, like this:
4691 => >A.N.Other <ano@some.domain> <original@ddress> ...
4693 This has been changed so that it extracts the operative address. There
4694 may be more than one such address. If so, they are comma-separated, like
4697 => >ano@some.domain,ona@other.domain <original@ddress> ...
4699 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4700 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4701 not being looked up in the rDNS to get a host name, Exim was showing the
4702 IP address twice in Received: lines, even though the IP addresses were
4703 identical. For example:
4705 Received: from [1.2.3.4] (helo=[1.2.3.4])
4707 However, if the real host name was known, it was omitting the HELO data
4708 if it matched the actual IP address. This has been tidied up so that it
4709 doesn't show the same IP address twice.
4711 PH/07 When both +timestamp and +memory debugging was on, the value given by
4712 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4713 called by the debug printing, thereby overwriting the timestamp buffer.
4714 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4716 PH/08 When the original message was included in an autoreply transport, it
4717 always said "this is a copy of the message, including all the headers",
4718 even if body_only or headers_only was set. It now gives an appropriate
4721 PH/09 Applied a patch from the Sieve maintainer which:
4723 o fixes some comments
4724 o adds the (disabled) notify extension core
4725 o adds some debug output for the result of if/elsif tests
4726 o points to the current vacation draft in the documentation
4727 and documents the missing references header update
4731 o fixes a bug in processing the envelope test (when testing
4732 multiple envelope elements, the last element determined the
4735 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4736 Electronic Mail") by including:
4738 Auto-submitted: auto-generated
4740 in the messages that it generates (bounce messages and others, such as
4741 warnings). In the case of bounce messages for non-SMTP messages, there was
4742 also a typo: it was using "Auto_submitted" (underscore instead of
4743 hyphen). Since every message generated by Exim is necessarily in response
4744 to another message, thes have all been changed to:
4746 Auto-Submitted: auto-replied
4748 in accordance with these statements in the RFC:
4750 The auto-replied keyword:
4752 - SHOULD be used on messages sent in direct response to another
4753 message by an automatic process,
4755 - MUST NOT be used on manually-generated messages,
4757 - MAY be used on Delivery Status Notifications (DSNs) and Message
4758 Disposition Notifications (MDNs),
4760 - MUST NOT be used on messages generated by automatic or periodic
4761 processes, except for messages which are automatic responses to
4764 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4765 to the default Received: header definition.
4767 PH/12 Added log selector acl_warn_skipped (default on).
4769 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4770 variables because (a) they are in the wrong storage pool and (b) even if
4771 they were copied, it wouldn't work properly because of the caching.
4773 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4774 checking when decoding. Apparently there are clients that generate
4775 overlong encoded strings. Why am I not surprised?
4777 PH/15 If the first argument of "${if match_address" was not empty, but did not
4778 contain an "@" character, Exim crashed. Now it writes a panic log message
4779 and treats the condition as false.
4781 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4783 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4784 extension "envelope-auth". The code is finished and in agreement with
4785 other implementations, but there is no documentation so far and in fact,
4786 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4787 not changing the active code.
4789 Print executed "if" and "elsif" statements when debugging is used. This
4790 helps a great deal to understand what a filter does.
4792 Document more things not specified clearly in RFC3028. I had all this
4793 sorted out, when out of a sudden new issues came to my mind. Oops."
4795 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4798 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4799 canonical form (as documented). However, after a host name lookup from
4800 the IP address, check_host() was doing a simple string comparison with
4801 addresses acquired from the DNS when checking that the found name did
4802 have the original IP as one of its addresses. Since any found IPv6
4803 addresses are likely to be in abbreviated form, the comparison could
4804 fail. Luckily, there already exists a function for doing the comparison
4805 by converting both addresses to binary, so now that is used instead of
4806 the text comparison.
4808 PH/20 There was another similar case to PH/19, when a complete host name was
4809 given in a host list; looking up its IP address could give an abbreviated
4810 form, whereas the current host's name might or might not be abbreviated.
4811 The same fix has been applied.
4817 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4818 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4821 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4822 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4824 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4826 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4827 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4828 Also, for "Auto-Submitted", treat anything other than "no" as
4829 non-personal, in accordance with RFC 3834. (Previously it treated
4830 anything starting "auto-" as non-personal.)
4832 TF/01 The control=submission/name=... option had a problem with syntax
4833 errors if the name included a slash character. The /name= option
4834 now slurps the rest of the string, so it can include any characters
4835 but it must come last in the list of options (after /sender_retain
4838 PH/05 Some modifications to the interface to the fake nameserver for the new
4846 TK/01 Added the "success_on_redirect" address verification option. See
4847 NewStuff for rationale and an example.
4849 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4851 PH/02 Patch to exigrep to allow it to work on syslog lines.
4853 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4854 fread() to skip over the body file's header line, because in Cygwin the
4855 header line is locked and is inaccessible.
4857 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4858 co-exist for some time) to make it clear that it is the Exim ID that is
4859 referenced, not the Message-ID: header line.
4861 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4862 string_format() function, because snprintf() does not exist on all
4865 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4866 ${stat: expansion item.
4868 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4869 protocol synchronization error", to keep the pedants happy.
4871 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4872 well as for IRIX systems, when gcc is being used. See the host.c source
4875 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4877 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4880 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4881 transport and to the smtp transport in LMTP mode.
4883 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4885 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4886 run a filter in a subprocess. This could lead to confusion in subsequent
4887 lookups in the parent process. There should also be a search_tidyup() at
4888 the end of the subprocess.
4890 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4891 only if the host matched helo_try_verify_hosts, which caused the
4892 verification to occur when the EHLO/HELO command was issued. The ACL just
4893 tested the remembered result. Now, if a previous verification attempt has
4894 not happened, "verify = helo" does it there and then.
4896 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4898 TK/03 Fix log output including CR from clamd.
4900 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4901 exist provoked a memory error which could cause a segfault.
4903 PH/15 Installed PCRE 6.2
4905 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4907 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4908 of the problem. Specifically, suggested +O2 rather than +O1 for the
4911 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4913 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4914 record, the port was not correctly logged when the outgoing_port log
4915 selector was set (it logged the transort's default port).
4917 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4918 fallback_hosts, and "hosts" in the smtp transport.
4920 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4921 host errors such as "Connection refused".
4923 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4924 authentication with radiusclient 0.4.9:
4926 - Error returned from rc_read_config was caught wrongly
4927 - Username/password not passed on to radius server due to wrong length.
4929 The presumption is that some radiusclient API changes for 4.51/PH/17
4930 were not taken care of correctly. The code is still untested by me (my
4931 Linux distribution still has 0.3.2 of radiusclient), but it was
4932 contributed by a Radius user.
4934 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4935 expanding the "port" option of the smtp transport.
4937 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4938 while reading a MIME header. Thanks to Tom Hughes for a patch.
4940 PH/24 Include config.h inside local_scan.h so that configuration settings are
4943 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4944 that in an ACL for RCPT (for example), you can examine exactly what was
4947 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4948 commands, but it was not correctly comparing the address with the actual
4949 client host address. Thus, it would show the EHLO address in Received:
4950 header lines when this was not necessary.
4952 PH/27 Added the % operator to ${eval:}.
4954 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4955 it should be ignoring failures (because with -C, for example, it has lost
4956 privilege). It wasn't ignoring creation failures other than "already
4959 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4962 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4963 because an input line was too long, either on its own, or by virtue of
4964 too many continuations, the temporary file was not being removed, and the
4965 return code was incorrect.
4967 PH/31 Missing "BOOL" in function definition in filtertest.c.
4969 PH/32 Applied Sieve patches from the maintainer.
4971 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4973 PH/33 Added "verify = not_blind".
4975 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4976 Local/Makefile (with some defaults set). These are used in built scripts
4977 such as exicyclog, but they have never been used in the exim_install
4978 script (though there are many overriding facilities there). I have
4979 arranged that the exim_install script now takes note of these two
4982 PH/35 Installed configuration files for Dragonfly.
4984 PH/36 When a locally submitted message by a trusted user did not contain a
4985 From: header, and the sender address was obtained from -f or from an SMTP
4986 MAIL command, and the trusted user did not use -F to supply a sender
4987 name, $originator_name was incorrectly used when constructing a From:
4988 header. Furthermore, $originator_name was used for submission mode
4989 messages from external hosts without From: headers in a similar way,
4990 which is clearly wrong.
4992 PH/37 Added control=suppress_local_fixups.
4994 PH/38 When log_selector = +received_sender was set, and the addition of the
4995 sender made the log line's construction buffer exactly full, or one byte
4996 less than full, an overflow happened when the terminating "\n" was
4999 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5000 when the result of a list match is failure because a DNS lookup failed.
5002 PH/40 RM_COMMAND is now used in the building process.
5004 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5005 the "build-* directories that it finds.
5007 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5008 address in a domain literal was a prefix of an interface address.
5010 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5011 when verifying a sender address, unless rewrite_headers is false.
5013 PH/44 Wrote a long comment about why errors_to addresses are verified as
5014 recipients, not senders.
5016 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5017 the ratelimit ACL was added.
5019 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5021 PH/46 Added extra information about PostgreSQL errors to the error string.
5023 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5024 avoiding the need to install special zones in a real server. This is
5025 backwards compatible; if it can't find the fake resolver, it drops back.
5026 Thus, both old and new test suites can be run.
5028 TF/02 Added util/ratelimit.pl
5030 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5031 clock is set back in time.
5033 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5034 Candler <B.Candler@pobox.com>.
5036 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5037 OK for addresses that are the result of redirections.
5039 PH/48 A number of further additions for the benefit of the new test suite,
5040 including a fake gethostbyname() that interfaces to the fake DNS resolver
5043 TF/06 The fix for widen_domains has also been applied to qualify_single and
5044 search_parents which are the other dnslookup options that can cause
5047 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5050 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5052 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5053 service" if it encountered an MX record with an empty target hostname.
5054 The message is now "an MX or SRV record indicated no SMTP service".
5056 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5057 if the DNS of the sending site is misconfigured. This is quite a
5058 common situation. This change restores the behaviour of treating a
5059 helo verification defer as a failure.
5061 PH/51 If self=fail was set on a router, the bounce message did not include the
5062 actual error message.
5068 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5070 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5071 command itself ended in error, the underlying message about the transport
5072 filter timeout was being overwritten with the pipe command error. Now the
5073 underlying error message should be appended to the second error message.
5075 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5077 PH/02 Reduce the amount of output that "make" produces by default. Full output
5078 can still be requested.
5080 PH/03 The warning log line about a condition test deferring for a "warn" verb
5081 was being output only once per connection, rather than after each
5082 occurrence (because it was using the same function as for successful
5083 "warn" verbs). This seems wrong, so I have changed it.
5085 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5086 it should not have, which might have caused a crash in the right
5087 circumstances, but probably never did.
5089 PH/04 Installed a modified version of Tony Finch's patch to make submission
5090 mode fix the return path as well as the Sender: header line, and to
5091 add a /name= option so that you can make the user's friendly name appear
5094 TF/03 Added the control = fakedefer ACL modifier.
5096 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5097 Mark Lowes for thorough testing.
5099 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5101 TK/03 Merged latest SRS patch from Miles Wilton.
5103 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5104 with the definition in sysexits.h (which is #included earlier).
5105 Fortunately, Exim does not actually use EX_OK. The code used to try to
5106 preserve the sysexits.h value, by assuming that macro definitions were
5107 scanned for macro replacements. I have been disabused of this notion,
5108 so now the code just undefines EX_OK before #including unistd.h.
5110 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5111 in the smtp transport. When a block could not be written in a single
5112 write() function, the timeout was being re-applied to each part-write.
5113 This seems wrong - if the receiver was accepting one byte at a time it
5114 would take for ever. The timeout is now adjusted when this happens. It
5115 doesn't have to be particularly precise.
5117 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5118 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5120 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5121 without a domain if the check to <postmaster@domain> fails.
5123 SC/01 Eximstats: added -xls and the ability to specify output files
5124 (patch written by Frank Heydlauf).
5126 SC/02 Eximstats: use FileHandles for outputting results.
5128 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5130 SC/04 Eximstats: fixed display of large numbers with -nvr option
5132 SC/05 Eximstats: fixed merging of reports with empty tables.
5134 SC/06 Eximstats: added the -include_original_destination flag
5136 SC/07 Eximstats: removed tabs and trailing whitespace.
5138 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5140 TK/06 MBOX spool code: Add real "From " MBOX separator line
5141 so the .eml file is really in mbox format (even though
5142 most programs do not really care). Patch from Alex Miller.
5144 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5145 The latter is generated from $received_to and is only set if the
5146 message has one envelope recipient. SA can use these headers,
5147 obviously out-of-the-box. Patch from Alex Miller.
5149 PH/08 The ${def test on a variable was returning false if the variable's
5150 value was "0", contrary to what the specification has always said!
5151 The result should be true unless the variable is empty.
5153 PH/09 The syntax error of a character other than { following "${if
5154 def:variable_name" (after optional whitespace) was not being diagnosed.
5155 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5156 accidental colon was present, for example, could give incorrect results.
5158 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5159 result is used (not including appendfile, where other changes are about
5162 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5163 This involved changing a lot of size variables from int to off_t. It
5164 should work with maildirs and everything.
5166 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5167 spamd dying while we are connected to it.
5169 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5172 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5173 function for BDB 4.3.
5175 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5177 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5178 the sources for PCRE have been re-arranged and more files are now
5181 PH/15 The code I had for printing potentially long long variables in PH/11
5182 above was not the best (it lost precision). The length of off_t variables
5183 is now inspected at build time, and an appropriate printing format (%ld
5184 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5185 to be "long long int" or "long int". This is needed for the internal
5186 formatting function string_vformat().
5188 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5189 the configuration file to be ":syslog", then the script "guesses" where
5190 the logs files are, rather than using the compiled in default. In our
5191 case the guess is not the same as the compiled default, so the script
5192 suddenly stopped working when I started to use syslog. The patch checks
5193 to see if log_file_path is "". If so, it attempts to read it from exim
5194 with no configuration file to get the compiled in version, before it
5195 falls back to the previous guessing code."
5197 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5198 implementing BATV in an Exim configuration. See NewStuff for the gory
5201 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5202 Makefile that are specific to HP-UX.
5204 PH/18 If the "use_postmaster" option was set for a recipient callout together
5205 with the "random" option, the postmaster address was used as the MAIL
5206 FROM address for the random test, but not for the subsequent recipient
5207 test. It is now used for both.
5209 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5210 patch removes a few documentation additions to RFC 3028, because the
5211 latest draft now contains them. It adds the new en;ascii-case comparator
5212 and a new error check for 8bit text in MIME parts. Comparator and
5213 require names are now matched exactly. I enabled the subaddress
5214 extension, but it is not well tested yet (read: it works for me)."
5216 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5217 rework some of the code of TK/09 above to avoid the hardwired use of
5218 "%lld" and "long long". Replaced the call to snprintf() with a call to
5221 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5222 records point to non-existent hosts", "retry timeout exceeded", and
5223 "retry time not reached for any host after a long failure period".
5225 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5226 experimental DomainKeys support:
5228 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5229 (2) On an error such as an illegally used "control", the wrong name for
5230 the control was given.
5232 These problems did NOT occur unless DomainKeys support was compiled.
5234 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5236 PH/24 Added ${if match_ip condition.
5238 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5239 databases so that it will be absolutely obvious if a crash occurs in the
5240 DB library. This is a regular occurrence (often caused by mis-matched
5243 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5244 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5245 picked up on a user's system that detects such things. There doesn't seem
5246 to be a gcc warning option for this - only an attribute that has to be
5247 put on the function's prototype. It seems that in Fedora Core 4 they have
5248 set this on a number of new functions. No doubt there will be more in due
5251 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5252 specify a transport. However, if an address that was verified by such a
5253 router was the subject of a callout, Exim crashed because it tried to
5254 read the rcpt_include_affixes from the non-existent transport. Now it
5255 just assumes that the setting of that option is false. This bug was
5256 introduced by 4.51/PH/31.
5258 PH/28 Changed -d+all to exclude +memory, because that information is very
5259 rarely of interest, but it makes the output a lot bigger. People tend to
5260 do -d+all out of habit.
5262 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5263 code in os-type was giving problems when libc.so lives in lib64, like on
5266 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5267 aren't the modern standard, and it seems that some systems' include files
5268 don't always have them. Exim was already checking for some of the newer
5269 ones like T_AAAA, and defining it itself. I've added checks for all the
5270 record types that Exim uses.
5272 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5273 not automatically generating a new one, as it is supposed to. This
5274 prevented TLS from working. If the file did exist, but contained invalid
5275 data, a new version was generated, as expected. It was only the case of a
5276 non-existent file that was broken.
5278 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5279 with a change in libdomainkeys > 0.64.
5281 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5282 from DNS. If the selector record carries the flag, it now has
5283 precedence over the domain-wide flag.
5285 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5287 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5288 the use of a port name that isn't defined in /etc/services) occurred, the
5289 message was deferred as in a normal delivery, and thus remained on the
5290 spool, instead of being failed because of the mua_wrapper setting. This
5291 is now fixed, and I tidied up some of the mua_wrapper messages at the
5294 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5295 the messages in a hash of arrays rather than using individual hashes.
5296 This is a bit cleaner and results in dramatic memory savings, albeit
5297 at a slight CPU cost.
5299 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5300 as requested by Marc Sherman.
5302 SC/10 Eximstats: added histograms for user specified patterns as requested
5305 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5307 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5308 fopen() in the content-scanning modules that did not already have it.
5314 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5315 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5317 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5319 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5321 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5322 to test Sieve filters that use "vacation".
5324 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5325 that changes the way the GnuTLS parameters are stored in the cache file.
5326 The new format can be generated externally. For backward compatibility,
5327 if the data in the cache doesn't make sense, Exim assumes it has read an
5328 old-format file, and it generates new data and writes a new file. This
5329 means that you can't go back to an older release without removing the
5332 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5333 work if there are any delivery delays because "one_time" forces the
5334 parent to be marked "delivered", so its unseen clone is never tried
5335 again. For this reason, Exim now forbids the simultaneous setting of
5338 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5339 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5340 entry if you want to know the details. The fix, however, neglected to
5341 consider the case when local delivery batching is involved. The test for
5342 "previously delivered" was not happening when checking to see if an
5343 address could be batched with a previous (undelivered) one; under
5344 certain circumstances this could lead to multiple deliveries to the same
5347 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5348 in its include files, and this causes problems building Exim.
5350 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5351 header_syntax) but Exim was just ignoring anything given after a slash.
5352 In particular, this caused confusion with an attempt to use "verify =
5353 reverse_host_lookup/defer_ok". An error is now given when options are
5354 supplied for verify items that do not have them. (Maybe reverse_host_
5355 lookup should have a defer_ok option, but that's a different point.)
5357 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5358 defined by RFC 821) to 2048, because there were problems with some AUTH
5359 commands, and RFC 1869 says the size should be increased for extended
5360 SMTP commands that take arguments.
5362 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5365 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5366 "unknown" error; now it says that the functionality isn't in the binary.
5368 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5369 an address' error message when a string expansion fails (syntax or
5370 whatever). Otherwise the password may appear in the log. Following change
5371 PH/42 below, there is no longer a chance of it appearing in a bounce
5374 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5376 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5377 of its addresses was ever tried. (Bugzilla bug #2.)
5379 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5380 the result incorrectly in the debug output. (It correctly added a newline
5381 to what was transported.)
5383 TF/01 Added $received_time.
5385 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5386 commented out examples of how to interface to a virus scanner and to
5387 SpamAssassin. Also added commented examples of av_scanner and
5388 spamd_address settings.
5390 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5391 and controls are allowed in which ACLs. There were a couple of minor
5392 errors. Some of the entries in the conditions table (which is a table of
5393 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5394 negation of where the condition IS allowed.
5396 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5398 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5399 header file does not have a version number, so I've had to invent a new
5400 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5401 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5402 radiusclient), but it was contributed by a Radius user.
5404 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5405 files or directories, for OpenSSL.
5407 PH/19 When an Exim process that is running as root has to create an Exim log
5408 file, it does so in a subprocess that runs as exim:exim so as to get the
5409 ownership right at creation (otherwise, other Exim processes might see
5410 the file with the wrong ownership). There was no test for failure of this
5411 fork() call, which would lead to the process getting stuck as it waited
5412 for a non-existent subprocess. Forks do occasionally fail when resources
5413 run out. I reviewed all the other calls to fork(); they all seem to check
5416 PH/20 When checking for unexpected SMTP input at connect time (before writing
5417 the banner), Exim was not dealing correctly with a non-positive return
5418 from the read() function. If the client had disconnected by this time,
5419 the result was a log entry for a synchronization error with an empty
5420 string after "input=" when read() returned zero. If read() returned -1
5421 (an event I could not check), uninitialized data bytes were printed.
5422 There were reports of junk text (parts of files, etc) appearing after
5425 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5427 PH/22 Added support for macro redefinition, and (re)definition in between
5428 driver and ACL definitions.
5430 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5431 forgetting to use the resulting value; it was using the unexpanded value.
5433 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5434 hadn't been configured. The fix is from Juergen Kreileder, who
5435 understands it better than I do:
5437 "Here's what I see happening with three configured cyrus_sasl
5438 authenticators configured (plain, login, cram-md5):
5440 On startup auth_cyrus_sasl_init() gets called for each of these.
5441 This means three calls to sasl_listmech() without a specified mech_list.
5442 => SASL tests which mechs of all available mechs actually work
5443 => three warnings about OTP not working
5444 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5446 With the patch, sasl_listmech() also gets called three times. But now
5447 SASL's mech_list option is set to the server_mech specified in the the
5448 authenticator. Or in other words, the answer from sasl_listmech()
5449 gets limited to just the mech you're testing for (which is different
5451 => the return list contains just 'plain' or 'login', 'cram-md5' or
5452 nothing depending on the value of ob->server_mech.
5454 I've just tested the patch: Authentication still works fine,
5455 unavailable mechs specified in the exim configuration are still
5456 caught, and the auth.log warnings about OTP are gone."
5458 PH/25 When debugging is enabled, the contents of the command line are added
5459 to the debugging output, even when log_selector=+arguments is not
5462 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5463 answer is "GNU", and only if the return is "GNU/something" is the answer
5466 PH/27 $acl_verify_message is now set immediately after the failure of a
5467 verification in an ACL, and so is available in subsequent modifiers. In
5468 particular, the message can be preserved by coding like this:
5470 warn !verify = sender
5471 set acl_m0 = $acl_verify_message
5473 Previously, $acl_verify_message was set only while expanding "message"
5474 and "log_message" when a very denied access.
5476 PH/28 Modified OS/os.c-Linux with
5478 -#ifndef OS_LOAD_AVERAGE
5479 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5481 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5482 nomenclature these days.)
5484 PH/29 Installed patch from the Sieve maintainer that adds the options
5485 sieve_useraddress and sieve_subaddress to the redirect router.
5487 PH/30 In these circumstances:
5488 . Two addresses routed to the same list of hosts;
5489 . First host does not offer TLS;
5490 . First host accepts first address;
5491 . First host gives temporary error to second address;
5492 . Second host offers TLS and a TLS session is established;
5493 . Second host accepts second address.
5494 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5495 and peerdn, if requested) that were in fact used only for the second
5498 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5499 attention to any local part prefix or suffix that was matched by the
5500 router that accepted the address. It now behaves in the same way as it
5501 does for delivery: the affixes are removed from the local part unless
5502 rcpt_include_affixes is set on the transport.
5504 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5505 timeout during the DATA phase of an incoming message.
5507 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5508 applied a patch sanctioned by the Sieve maintainer.
5510 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5511 the uid or gid is negative. A case of a negative gid caused this to be
5512 noticed. The fix allows for either to be negative.
5514 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5515 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5518 PH/36 The change PH/12 above was broken. Fixed it.
5520 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5521 the grounds that routing the same address twice would always produce the
5522 same answer. This might have been true once, but it is certainly no
5523 longer true now. Routing a child address may depend on the previous
5524 routing that produced that child. Some complicated redirection strategies
5525 went wrong when messages had multiple recipients, and made Exim's
5526 behaviour dependent on the order in which the addresses were given.
5528 I have moved the duplicate checking until after the routing is complete.
5529 Exim scans the addresses that are assigned to local and remote
5530 transports, and removes any duplicates. This means that more work will be
5531 done, as duplicates will always all be routed, but duplicates are
5532 presumably rare, so I don't expect this is of any significance.
5534 For deliveries to pipes, files, and autoreplies, the duplicate checking
5535 still happens during the routing process, since they are not going to be
5538 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5539 It corrects a timeout issue with spamd. This is Ian's comment: "The
5540 background is that sometimes spamd either never reads data from a
5541 connection it has accepted, or it never writes response data. The exiscan
5542 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5543 blindly assumes that writes won't block so it may never time out."
5545 PH/39 Allow G after quota size as well as K and M.
5547 PH/40 The value set for $authenticated_id in an authenticator may not contain
5548 binary zeroes or newlines because the value is written to log lines and
5549 to spool files. There was no check on this. Now the value is run through
5550 the string_printing() function so that such characters are converted to
5551 printable escape sequences.
5553 PH/41 $message_linecount is a new variable that contains the total number of
5554 lines in the message. Compare $body_linecount, which is the count for the
5557 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5558 bounce and delay warning messages, except in certain special cases, which
5561 (a) An SMTP error message from a remote host;
5562 (b) A message specified in a :fail: redirection;
5563 (c) A message specified in a "fail" command in a system filter;
5564 (d) A message specified in a FAIL return from the queryprogram router;
5565 (e) A message specified by the cannot_route_message router option.
5567 In these cases only, Exim does include the error details in bounce and
5568 warning messages. There are also a few cases where bland messages such
5569 as "unrouteable address" or "local delivery error" are given.
5571 PH/43 $value is now also set for the "else" part of a ${run expansion.
5573 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5574 being worked on, but at least Exim now implements the latest version to
5577 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5578 process to complete was treated as a delivery failure, a timeout while
5579 writing the message to the pipe was logged, but erroneously treated as a
5580 successful delivery. Such timeouts include transport filter timeouts. For
5581 consistency with the overall process timeout, these timeouts are now
5582 treated as errors, giving rise to delivery failures by default. However,
5583 there is now a new Boolean option for the pipe transport called
5584 timeout_defer, which, if set TRUE, converts the failures into defers for
5585 both kinds of timeout. A transport filter timeout is now identified in
5588 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5589 systems where "make" and "gmake" are different, calling "gmake" at top
5590 level broke things. I've arranged for the value of $(MAKE) to be passed
5591 from the Makefile to this script so that it can call the same version of
5595 A note about Exim versions 4.44 and 4.50
5596 ----------------------------------------
5598 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5599 changes of various kinds. As a consequence, a big documentation update was
5600 needed. This delayed the release for rather longer than seemed good, especially
5601 in the light of a couple of (minor) security issues. Therefore, the changes
5602 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5603 release. So 4.44 and 4.50 are in effect two different branches that both start
5606 I have left the 4.50 change log unchanged; it contains all the changes since
5607 4.43. The change log for 4.44 is below; many of its items are identical to
5608 those for 4.50. This seems to be the most sensible way to preserve the
5609 historical information.
5615 1. Minor wording change to the doc/README.SIEVE file.
5617 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5618 computation of the current number of files was incorrect.
5620 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5621 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5624 4. Give more explanation in the error message when the command for a transport
5625 filter fails to execute.
5627 5. There are several places where Exim runs a non-Exim command in a
5628 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5629 was being done only for the command run by the queryprogram router. It is
5630 now done for all such subprocesses. The other cases are: ${run, transport
5631 filters, and the commands run by the lmtp and pipe transports.
5633 6. Added CONFIGURE_GROUP build-time option.
5635 7. Some older OS have a limit of 256 on the maximum number of file
5636 descriptors. Exim was using setrlimit() to set 1000 as a large value
5637 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5638 systems. I've change it so that if it can't get 1000, it tries for 256.
5640 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5641 was an oversight, and furthermore, ever since the addition of extra
5642 controls (e.g. 4.43/32), the checks on when to allow different forms of
5643 "control" were broken. There should now be diagnostics for all cases when a
5644 control that does not make sense is encountered.
5646 9. Added the /retain_sender option to "control=submission".
5648 10. $recipients is now available in the predata ACL (oversight).
5650 11. Tidy the search cache before the fork to do a delivery from a message
5651 received from the command line. Otherwise the child will trigger a lookup
5652 failure and thereby defer the delivery if it tries to use (for example) a
5653 cached ldap connection that the parent has called unbind on.
5655 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5656 of $address_data from the recipient verification was clobbered by the
5657 sender verification.
5659 13. The value of address_data from a sender verification is now available in
5660 $sender_address_data in subsequent conditions in the ACL statement.
5662 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5664 15. Added a new option "connect=<time>" to callout options, to set a different
5667 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5668 was its contents. (It was OK if the option was not defined at all.)
5670 17. A "Completed" log line is now written for messages that are removed from
5671 the spool by the -Mrm option.
5673 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5674 information about exactly what failed.
5676 19. Added -dd to debug only the daemon process.
5678 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5679 handles timeouts, both on the server side and network timeouts. Renamed the
5680 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5682 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5683 It is now set to "smtps".
5685 22. $host_address is now set to the target address during the checking of
5686 ignore_target_hosts.
5688 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5689 being passed; this would have caused $sender_host_name to have been used if
5690 matching the list had actually called for a host name (not very likely,
5691 since this list is usually IP addresses). A host name is now passed as
5694 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5695 code that specifies a non-restarting handler (typically sigaction() in
5696 modern systems) in an attempt to fix a rare and obscure crash bug.
5698 25. Narrowed the window for a race in the daemon that could cause it to ignore
5699 SIGCHLD signals. This is not a major problem, because they are used only to
5700 wake it up if nothing else does.
5702 26. A malformed maildirsize file could cause Exim to calculate negative values
5703 for the mailbox size or file count. Odd effects could occur as a result.
5704 The maildirsize information is now recalculated if the size or filecount
5707 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5708 support for a long time. Removed HAVE_SYS_VFS_H.
5710 28. Installed the latest version of exipick from John Jetmore.
5712 29. In an address list, if the pattern was not a regular expression, an empty
5713 subject address (from a bounce message) matched only if the pattern was an
5714 empty string. Non-empty patterns were not even tested. This was the wrong
5715 because it is perfectly reasonable to use an empty address as part of a
5716 database query. An empty address is now tested by patterns that are
5717 lookups. However, all the other forms of pattern expect the subject to
5718 contain a local part and a domain, and therefore, for them, an empty
5719 address still always fails if the pattern is not itself empty.
5721 30. Exim went into a mad DNS loop when attempting to do a callout where the
5722 host was specified on an smtp transport, and looking it up yielded more
5723 than one IP address.
5725 31. Re-factored the code for checking spool and log partition space into a
5726 function that finds that data and another that does the check. The former
5727 is then used to implement four new variables: $spool_space, $log_space,
5728 $spool_inodes, and $log_inodes.
5730 32. The RFC2047 encoding function was originally intended for short strings
5731 such as real names; it was not keeping to the 75-character limit for
5732 encoded words that the RFC imposes. It now respects the limit, and
5733 generates multiple encoded words if necessary. To be on the safe side, I
5734 have increased the buffer size for the ${rfc2047: expansion operator from
5737 33. It is now permitted to omit both strings after an "if" condition; if the
5738 condition is true, the result is "true". As before, when the second string
5739 is omitted, a false condition yields an empty string. This makes it less
5740 cumbersome to write custom ACL and router conditions.
5742 34. Failure to deliver a bounce message always caused it to be frozen, even if
5743 there was an errors_to setting on the router. The errors_to setting is now
5746 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5747 canonical form (fully expanded) before being placed in
5748 $sender_host_address.
5750 36. The table in the code that translates DNS record types into text (T_A to
5751 "A" for instance) was missing entries for NS and CNAME. It is just possible
5752 that this could have caused confusion if both these types were looked up
5753 for the same domain, because the text type is used as part of Exim's
5754 per-process caching. But the chance of anyone hitting this buglet seems
5757 37. The dnsdb lookup has been extended in a number of ways.
5759 (1) There is a new type, "zns", which walks up the domain tree until it
5760 finds some nameserver records. It should be used with care.
5762 (2) There is a new type, "mxh", which is like "mx" except that it returns
5763 just the host names, not the priorities.
5765 (3) It is now possible to give a list of domains (or IP addresses) to be
5766 looked up. The behaviour when one of the lookups defers can be
5767 controlled by a keyword.
5769 (4) It is now possible to specify the separator character for use when
5770 multiple records are returned.
5772 38. The dnslists ACL condition has been extended: it is now possible to supply
5773 a list of IP addresses and/or domains to be looked up in a particular DNS
5776 39. Added log_selector=+queue_time_overall.
5778 40. When running the queue in the test harness, wait just a tad after forking a
5779 delivery process, to get repeatability of debugging output.
5781 41. Include certificate and key file names in error message when GnuTLS fails
5782 to set them up, because the GnuTLS error message doesn't include the name
5783 of the failing file when there is a problem reading it.
5785 42. Allow both -bf and -bF in the same test run.
5787 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5789 44. The "Exiscan patch" is now merged into the mainline Exim source.
5791 45. Sometimes the final signoff response after QUIT could fail to get
5792 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5793 < 0 before doing a fflush(). This bug looks as though it goes back to the
5794 introduction of TLS in release 3.20, but "sometimes" must have been rare
5795 because the tests only now provoked it.
5797 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5798 (this can affect the format of dates).
5800 47. exim_tidydb, when checking for the continued existence of a message for
5801 which it has found a message-specific retry record, was not finding
5802 messages that were in split spool directories. Consequently, it was
5803 deleting retry records that should have stayed in existence.
5805 48. Steve fixed some bugs in eximstats.
5807 49. The SPA authentication driver was not abandoning authentication and moving
5808 on to the next authenticator when an expansion was forced to fail,
5809 contradicting the general specification for all authenticators. Instead it
5810 was generating a temporary error. It now behaves as specified.
5812 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5813 (the order specifies the preference for clients). The order is now AES256,
5814 AES128, 3DES, ARCFOUR128.
5816 51. Small patch to Sieve code - explicitly set From: when generating an
5819 52. Exim crashed if a remote delivery caused a very long error message to be
5820 recorded - for instance if somebody sent an entire SpamAssassin report back
5821 as a large number of 550 error lines. This bug was coincidentally fixed by
5822 increasing the size of one of Exim's internal buffers (big_buffer) that
5823 happened as part of the Exiscan merge. However, to be on the safe side, I
5824 have made the code more robust (and fixed the comments that describe what
5827 53. Now that there can be additional text after "Completed" in log lines (if
5828 the queue_time_overall log selector is set), a one-byte patch to exigrep
5829 was needed to allow it to recognize "Completed" as not the last thing in
5832 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5833 patch that reportedly fixes this has been added. I am not expert enough to
5834 create a test for it. This is what the patch creator wrote:
5836 "I found a little strange behaviour of ldap code when working with
5837 Windows 2003 AD Domain, where users was placed in more than one
5838 Organization Units. When I tried to give exim partial DN, the exit code
5839 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5840 But simultaneously result of request was absolutely normal ldap result,
5841 so I produce this patch..."
5843 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5844 REFERENCE, so I have modified the code to exclude the patch when that macro
5847 55. Some experimental protocols are using DNS PTR records for new purposes. The
5848 keys for these records are domain names, not reversed IP addresses. The
5849 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5850 leaves it alone. Component reversal etc. now happens only for IP addresses.
5853 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5855 57. Double the size of the debug message buffer (to 2048) so that more of very
5856 long debug lines gets shown.
5858 58. The exicyclog utility now does better if the number of log files to keep
5859 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5861 59. Two changes related to the smtp_active_hostname option:
5863 (1) $smtp_active_hostname is now available as a variable.
5864 (2) The default for smtp_banner uses $smtp_active_hostname instead
5865 of $primary_hostname.
5867 60. The host_aton() function is supposed to be passed a string that is known
5868 to be a valid IP address. However, in the case of IPv6 addresses, it was
5869 not checking this. This is a hostage to fortune. Exim now panics and dies
5870 if the condition is not met. A case was found where this could be provoked
5871 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5872 components; fortuitously, this particular loophole had already been fixed
5873 by change 4.50/55 above.
5875 If there are any other similar loopholes, the new check in host_aton()
5876 itself should stop them being exploited. The report I received stated that
5877 data on the command line could provoke the exploit when Exim was running as
5878 exim, but did not say which command line option was involved. All I could
5879 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5880 running as the user.
5883 61. There was a buffer overflow vulnerability in the SPA authentication code
5884 (which came originally from the Samba project). I have added a test to the
5885 spa_base64_to_bits() function which I hope fixes it.
5888 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5889 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5891 63. The daemon start-up calls getloadavg() while still root for those OS that
5892 need the first call to be done as root, but it missed one case: when
5893 deliver_queue_load_max is set with deliver_drop_privilege. This is
5894 necessary for the benefit of the queue runner, because there is no re-exec
5895 when deliver_drop_privilege is set.
5897 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5898 This has been fixed.
5900 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5901 was in use, was not putting the data itself into the right store pool;
5902 consequently, it could be overwritten for a subsequent message in the same
5903 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5906 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5908 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5909 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5910 the function were treating the return as a boolean value, which happened to
5911 work because 0=false and not-0=true, but is not correct code.
5913 68. The host_aton() function was not handling scoped IPv6 addresses (those
5914 with, for example, "%eth0" on the end) correctly.
5916 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5917 negated items (that is, ~something) in unsigned ints. Some compilers
5918 apparently mutter when there is no cast.
5920 70. If an address verification called from an ACL failed, and did not produce a
5921 user-specific message (i.e. there was only a "system" message), nothing was
5922 put in $acl_verify_message. In this situation, it now puts the system
5925 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5926 session; change 4.31/43 added the unwanted input to the log line - except
5927 that it did not do this in the start of session case. It now does.
5929 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5930 This is wrong and can cause the other end to generate a synchronization
5931 error if it is another Exim or anything else that does the synchronization
5932 check. A QUIT command is no longer sent after a timeout.
5934 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5935 during host lookups.
5937 74. The defer_ok option of callout verification was not working if it was used
5938 when verifying addresses in header lines, that is, for this case:
5940 verify = header_sender/callout=defer_ok
5942 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5943 those file descriptors could be used for SMTP connections. If anything
5944 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5945 could be sent to the SMTP client, causing chaos. The daemon now opens
5946 stdin, stdout, and stderr to /dev/null when it puts itself into the
5949 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5950 log by default. The user can override this with suitable Perl magic.
5952 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5953 the log message when discard triggers, was not working for the DATA ACL or
5954 for the non-SMTP ACL.
5956 78. Error message wording change in sieve.c.
5958 79. If smtp_accept_max_per_host was set, the number of connections could be
5959 restricted to fewer than expected, because the daemon was trying to set up
5960 a new connection before checking whether the processes handling previous
5961 connections had finished. The check for completed processes is now done
5962 earlier. On busy systems, this bug wouldn't be noticed because something
5963 else would have woken the daemon, and it would have reaped the completed
5966 80. If a message was submitted locally by a user whose login name contained one
5967 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5968 It caused a spool format error. I have fixed the spool reading code. A
5969 related problem was that the "from" clause in the Received: line became
5970 illegal because of the space(s). It is now covered by ${quote_local_part.
5972 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5975 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5977 83. Patch from Sieve maintainer to fix unterminated string problem in
5978 "vacation" handling.
5980 84. Some minor changes to the Linux configuration files to help with other
5981 OS variants using glibc.
5983 85. One more patch for Sieve to update vacation handling to latest spec.
5986 ----------------------------------------------------
5987 See the note above about the 4.44 and 4.50 releases.
5988 ----------------------------------------------------
5994 1. Change 4.43/35 introduced a bug that caused file counts to be
5995 incorrectly computed when quota_filecount was set in an appendfile
5998 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5999 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6002 3. Give more explanation in the error message when the command for a transport
6003 filter fails to execute.
6005 4. There are several places where Exim runs a non-Exim command in a
6006 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6007 was being done only for the command run by the queryprogram router. It is
6008 now done for all such subprocesses. The other cases are: ${run, transport
6009 filters, and the commands run by the lmtp and pipe transports.
6011 5. Some older OS have a limit of 256 on the maximum number of file
6012 descriptors. Exim was using setrlimit() to set 1000 as a large value
6013 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6014 systems. I've change it so that if it can't get 1000, it tries for 256.
6016 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6017 was an oversight, and furthermore, ever since the addition of extra
6018 controls (e.g. 4.43/32), the checks on when to allow different forms of
6019 "control" were broken. There should now be diagnostics for all cases when a
6020 control that does not make sense is encountered.
6022 7. $recipients is now available in the predata ACL (oversight).
6024 8. Tidy the search cache before the fork to do a delivery from a message
6025 received from the command line. Otherwise the child will trigger a lookup
6026 failure and thereby defer the delivery if it tries to use (for example) a
6027 cached ldap connection that the parent has called unbind on.
6029 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6030 of $address_data from the recipient verification was clobbered by the
6031 sender verification.
6033 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6034 was its contents. (It was OK if the option was not defined at all.)
6036 11. A "Completed" log line is now written for messages that are removed from
6037 the spool by the -Mrm option.
6039 12. $host_address is now set to the target address during the checking of
6040 ignore_target_hosts.
6042 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6043 being passed; this would have caused $sender_host_name to have been used if
6044 matching the list had actually called for a host name (not very likely,
6045 since this list is usually IP addresses). A host name is now passed as
6048 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6049 code that specifies a non-restarting handler (typically sigaction() in
6050 modern systems) in an attempt to fix a rare and obscure crash bug.
6052 15. Narrowed the window for a race in the daemon that could cause it to ignore
6053 SIGCHLD signals. This is not a major problem, because they are used only to
6054 wake it up if nothing else does.
6056 16. A malformed maildirsize file could cause Exim to calculate negative values
6057 for the mailbox size or file count. Odd effects could occur as a result.
6058 The maildirsize information is now recalculated if the size or filecount
6061 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6062 support for a long time. Removed HAVE_SYS_VFS_H.
6064 18. Updated exipick to current release from John Jetmore.
6066 19. Allow an empty sender to be matched against a lookup in an address list.
6067 Previously the only cases considered were a regular expression, or an
6070 20. Exim went into a mad DNS lookup loop when doing a callout where the
6071 host was specified on the transport, if the DNS lookup yielded more than
6074 21. The RFC2047 encoding function was originally intended for short strings
6075 such as real names; it was not keeping to the 75-character limit for
6076 encoded words that the RFC imposes. It now respects the limit, and
6077 generates multiple encoded words if necessary. To be on the safe side, I
6078 have increased the buffer size for the ${rfc2047: expansion operator from
6081 22. Failure to deliver a bounce message always caused it to be frozen, even if
6082 there was an errors_to setting on the router. The errors_to setting is now
6085 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6086 canonical form (fully expanded) before being placed in
6087 $sender_host_address.
6089 24. Updated eximstats to version 1.33
6091 25. Include certificate and key file names in error message when GnuTLS fails
6092 to set them up, because the GnuTLS error message doesn't include the name
6093 of the failing file when there is a problem reading it.
6095 26. Expand error message when OpenSSL has problems setting up cert/key files.
6098 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6099 (this can affect the format of dates).
6101 28. exim_tidydb, when checking for the continued existence of a message for
6102 which it has found a message-specific retry record, was not finding
6103 messages that were in split spool directories. Consequently, it was
6104 deleting retry records that should have stayed in existence.
6106 29. eximstats updated to version 1.35
6107 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6108 1.35 - bugfix such that pie charts by volume are generated correctly
6110 30. The SPA authentication driver was not abandoning authentication and moving
6111 on to the next authenticator when an expansion was forced to fail,
6112 contradicting the general specification for all authenticators. Instead it
6113 was generating a temporary error. It now behaves as specified.
6115 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6116 (the order specifies the preference for clients). The order is now AES256,
6117 AES128, 3DES, ARCFOUR128.
6119 31. Small patch to Sieve code - explicitly set From: when generating an
6122 32. Exim crashed if a remote delivery caused a very long error message to be
6123 recorded - for instance if somebody sent an entire SpamAssassin report back
6124 as a large number of 550 error lines. This bug was coincidentally fixed by
6125 increasing the size of one of Exim's internal buffers (big_buffer) that
6126 happened as part of the Exiscan merge. However, to be on the safe side, I
6127 have made the code more robust (and fixed the comments that describe what
6130 33. Some experimental protocols are using DNS PTR records for new purposes. The
6131 keys for these records are domain names, not reversed IP addresses. The
6132 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6133 leaves it alone. Component reversal etc. now happens only for IP addresses.
6136 34. The host_aton() function is supposed to be passed a string that is known
6137 to be a valid IP address. However, in the case of IPv6 addresses, it was
6138 not checking this. This is a hostage to fortune. Exim now panics and dies
6139 if the condition is not met. A case was found where this could be provoked
6140 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6141 components; fortuitously, this particular loophole had already been fixed
6142 by change 4.50/55 or 4.44/33 above.
6144 If there are any other similar loopholes, the new check in host_aton()
6145 itself should stop them being exploited. The report I received stated that
6146 data on the command line could provoke the exploit when Exim was running as
6147 exim, but did not say which command line option was involved. All I could
6148 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6149 running as the user.
6152 35. There was a buffer overflow vulnerability in the SPA authentication code
6153 (which came originally from the Samba project). I have added a test to the
6154 spa_base64_to_bits() function which I hope fixes it.
6157 36. The daemon start-up calls getloadavg() while still root for those OS that
6158 need the first call to be done as root, but it missed one case: when
6159 deliver_queue_load_max is set with deliver_drop_privilege. This is
6160 necessary for the benefit of the queue runner, because there is no re-exec
6161 when deliver_drop_privilege is set.
6163 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6164 was in use, was not putting the data itself into the right store pool;
6165 consequently, it could be overwritten for a subsequent message in the same
6166 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6169 38. Sometimes the final signoff response after QUIT could fail to get
6170 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6171 < 0 before doing a fflush(). This bug looks as though it goes back to the
6172 introduction of TLS in release 3.20, but "sometimes" must have been rare
6173 because the tests only now provoked it.
6179 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6180 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6181 Now it returns an integer. A number of calls were still expecting a T/F
6182 return. Fortuitously, in all cases, the tests worked in OK situations,
6183 which is the norm. However, things would have gone wrong on any write
6184 failures on the smtp file descriptor. This function is used when sending
6185 messages over SMTP and also when doing verify callouts.
6187 2. When Exim is called to do synchronous delivery of a locally submitted
6188 message (the -odf or -odi options), it no longer closes stderr before doing
6191 3. Implemented the mua_wrapper option.
6193 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6195 5. Implemented the functions header_remove(), header_testname(),
6196 header_add_at_position(), and receive_remove_recipient(), and exported them
6199 6. If an ACL "warn" statement specified the addition of headers, Exim already
6200 inserted X-ACL-Warn: at the start if there was no header name. However, it
6201 was not making this test for the second and subsequent header lines if
6202 there were newlines in the string. This meant that an invalid header could
6203 be inserted if Exim was badly configured.
6205 7. Allow an ACL "warn" statement to add header lines at the start or after all
6206 the Received: headers, as well as at the end.
6208 8. Added the rcpt_4xx retry error code.
6210 9. Added postmaster_mailfrom=xxx to callout verification option.
6212 10. Added mailfrom=xxxx to the callout verification option, for verify=
6215 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6216 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6218 12. Inserted some casts to stop certain compilers warning when using pointer
6219 differences as field lengths or precisions in printf-type calls (mostly
6220 affecting debugging statements).
6222 13. Added optional readline() support for -be (dynamically loaded).
6224 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6225 same clock tick as a message's arrival, so that its received time was the
6226 same as the "first fail" time on the retry record, and that message
6227 remained on the queue past the ultimate address timeout, every queue runner
6228 would try a delivery (because it was past the ultimate address timeout) but
6229 after another failure, the ultimate address timeout, which should have then
6230 bounced the address, did not kick in. This was a "< instead of <=" error;
6231 in most cases the first failure would have been in the next clock tick
6232 after the received time, and all would be well.
6234 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6235 being recognized when the domain list was tested by the match_domain
6236 condition in an expansion string.
6238 16. Added the ${str2b64: operator.
6240 17. Exim was always calling setrlimit() to set a large limit for the number of
6241 processes, without checking whether the existing limit was already
6242 adequate. (It did check for the limit on file descriptors.) Furthermore,
6243 errors from getrlimit() and setrlimit() were being ignored. Now they are
6244 logged to the main and panic logs, but Exim does carry on, to try to do its
6245 job under whatever limits there are.
6247 18. Imported PCRE 5.0.
6249 19. Trivial typo in log message " temporarily refused connection" (the leading
6252 20. If the log selector return_path_on_delivery was set and an address was
6253 redirected to /dev/null, the delivery process crashed because it assumed
6254 that a return path would always be set for a "successful" delivery. In this
6255 case, the whole delivery is bypassed as an optimization, and therefore no
6258 21. Internal re-arrangement: the function for sending a challenge and reading
6259 a response while authentication was assuming a zero-terminated challenge
6260 string. It's now changed to take a pointer and a length, to allow for
6261 binary data in such strings.
6263 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6265 23. Exim was not respecting finduser_retries when seeking the login of the
6266 uid under which it was called; it was always trying 10 times. (The default
6267 setting of finduser_retries is zero.) Also, it was sleeping after the final
6268 failure, which is pointless.
6270 24. Implemented tls_on_connect_ports.
6272 25. Implemented acl_smtp_predata.
6274 26. If the domain in control=submission is set empty, Exim assumes that the
6275 authenticated id is a complete email address when it generates From: or
6276 Sender: header lines.
6278 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6279 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6280 chown and chgrp in /bin and hostname in /usr/bin.
6282 28. Exim was keeping the "process log" file open after each use, just as it
6283 does for the main log. This opens the possibility of it remaining open for
6284 long periods when the USR1 signal hits a daemon. Occasional processlog
6285 errors were reported, that could have been caused by this. Anyway, it seems
6286 much more sensible not to leave this file open at all, so that is what now
6289 29. The long-running daemon process does not normally write to the log once it
6290 has entered its main loop, and it closes the log before doing so. This is
6291 so that log files can straightforwardly be renamed and moved. However,
6292 there are a couple of unusual error situations where the daemon does write
6293 log entries, and I had neglected to close the log afterwards.
6295 30. The text of an SMTP error response that was received during a remote
6296 delivery was being truncated at 512 bytes. This is too short for some of
6297 the long messages that one sometimes sees. I've increased the limit to
6300 31. It is now possible to make retry rules that apply only when a message has a
6301 specific sender, in particular, an empty sender.
6303 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6304 it possible to be selective about when SMTP synchronization is enforced.
6306 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6308 32. Implemented hosts_connection_nolog.
6310 33. Added an ACL for QUIT.
6312 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6315 35. Added mailbox_size and mailbox_filecount to appendfile.
6317 36. Added control = no_multiline_responses to ACLs.
6319 37. There was a bug in the logic of the code that waits for the clock to tick
6320 in the case where the clock went backwards by a substantial amount such
6321 that the microsecond fraction of "now" was more than the microsecond
6322 fraction of "then" (but the whole seconds number was less).
6324 38. Added support for the libradius Radius client library this is found on
6325 FreeBSD (previously only the radiusclient library was supported).
6331 1. When certain lookups returned multiple values in the form name=value, the
6332 quoting of the values was not always being done properly. Specifically:
6333 (a) If the value started with a double quote, but contained no whitespace,
6335 (b) If the value contained whitespace other than a space character (i.e.
6336 tabs or newlines or carriage returns) it was not quoted.
6337 This fix has been applied to the mysql and pgsql lookups by writing a
6338 separate quoting function and calling it from the lookup code. The fix
6339 should probably also be applied to nisplus, ibase and oracle lookups, but
6340 since I cannot test any of those, I have not disturbed their existing code.
6342 2. A hit in the callout cache for a specific address caused a log line with no
6343 reason for rejecting RCPT. Now it says "Previous (cached) callout
6344 verification failure".
6346 3. There was an off-by-one bug in the queryprogram router. An over-long
6347 return line was truncated at 256 instead of 255 characters, thereby
6348 overflowing its buffer with the terminating zero. As well as fixing this, I
6349 have increased the buffer size to 1024 (and made a note to document this).
6351 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6352 when Exim is waiting for an SMTP response from a remote server, Exim
6353 restarts its select() call on the socket, thereby resetting its timeout.
6354 This is not a problem when such interrupts are rare. Somebody set up a cron
6355 job to run exiwhat every 2 minutes, which is less than the normal select()
6356 timeout (5 or 10 minutes). This meant that the select() timeout never
6357 kicked in because it was always reset. I have fixed this by comparing the
6358 time when an interrupt arrives with the time at the start of the first call
6359 to select(). If more time than the timeout has elapsed, the interrupt is
6360 treated as a timeout.
6362 5. Some internal re-factoring in preparation for the addition of Sieve
6363 extensions (by MH). In particular, the "personal" test is moved to a
6364 separate function, and given an option for scanning Cc: and Bcc: (which is
6365 not set for Exim filters).
6367 6. When Exim created an email address using the login of the caller as the
6368 local part (e.g. when creating a From: or Sender: header line), it was not
6369 quoting the local part when it contained special characters such as @.
6371 7. Installed new OpenBSD configuration files.
6373 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6374 try to make them clearer.
6376 9. Callout options, other than the timeout value, were being ignored when
6377 verifying sender addresses in header lines. For example, when using
6379 verify = header_sender/callout=no_cache
6381 the cache was (incorrectly) being used.
6383 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6384 only the Cygwin environment.
6386 11. When return_path_on_delivery was set as a log selector, if different remote
6387 addresses in the same message used different return paths and parallel
6388 remote delivery occurred, the wrong values would sometimes be logged.
6389 (Whenever a remote delivery process finished, the return path value from
6390 the most recently started remote delivery process was logged.)
6392 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6393 lines when AUTH and/or TLS are in use. This is the "received protocol"
6394 field. Exim used to use "asmtp" for authenticated SMTP, without any
6395 indication (in the protocol name) for TLS use. Now it follows the RFC and
6396 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6397 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6398 names appear in log lines as well as in Received: header lines.
6400 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6401 extensions, and comparison tests, and to fix some bugs.
6403 14. Changes to the "personal" filter test:
6405 (1) The test was buggy in that it was just doing the equivalent of
6406 "contains" tests on header lines. For example, if a user's address was
6407 anne@some.where, the "personal" test would incorrectly be true for
6409 To: susanne@some.where
6411 This test is now done by extracting each address from the header in turn,
6412 and checking the entire address. Other tests that are part of "personal"
6413 are now done using regular expressions (for example, to check local parts
6414 of addresses in From: header lines).
6416 (2) The list of non-personal local parts in From: addresses has been
6417 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6418 taken from the Sieve specification recommendations.
6420 (3) If the message contains any header line starting with "List-" it is
6421 treated as non-personal.
6423 (4) The test for "circular" in the Subject: header line has been removed
6424 because it now seems ill-conceived.
6426 15. Minor typos in src/EDITME comments corrected.
6428 16. Installed latest exipick from John Jetmore.
6430 17. If headers_add on a router specified a text string that was too long for
6431 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6432 of string_sprintf() is now avoided.
6434 18. $message_body_size was not set (it was always zero) when running the DATA
6435 ACL and the local_scan() function.
6437 19. For the "mail" command in an Exim filter, no default was being set for
6438 the once_repeat time, causing a random time value to be used if "once" was
6439 specified. (If the value happened to be <= 0, no repeat happened.) The
6440 default is now 0s, meaning "never repeat". The "vacation" command was OK
6441 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6442 (I found it when inspecting the code).
6444 20. There is now an overall timeout for performing a callout verification. It
6445 defaults to 4 times the callout timeout, which applies to individual SMTP
6446 commands during the callout. The overall timeout applies when there is more
6447 than one host that can be tried. The timeout is checked before trying the
6448 next host. This prevents very long delays if there are a large number of
6449 hosts and all are timing out (e.g. when the network connections are timing
6450 out). The value of the overall timeout can be changed by specifying an
6451 additional sub-option for "callout", called "maxwait". For example:
6453 verify = sender/callout=5s,maxwait=20s
6455 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6456 to the end before writing, but this should make it even safer).
6458 22. Exim was forgetting that it had advertised PIPELINING for the second and
6459 subsequent messages on an SMTP connection. It was also not resetting its
6460 memory on STARTTLS and an internal HELO.
6462 23. When Exim logs an SMTP synchronization error within a session, it now
6463 records whether PIPELINING has been advertised or not.
6465 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6466 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6467 rather than long int.
6469 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6471 26. Added the never_mail option to autoreply.
6477 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6478 crash if the getsockname() call failed; this can happen if a connection is
6479 closed very soon after it is established. The problem was simply in the
6480 order in which certain operations were done, causing Exim to try to write
6481 to the SMTP stream before it had set up the file descriptor. The bug has
6482 been fixed by making things happen in the correct order.
6488 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6489 before the connection was closed, thus losing the rejection response.
6491 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6492 some early Solaris releases, but causes trouble in current releases where
6493 socklen_t is defined.
6495 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6498 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6501 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6502 that releasing the top of it at the end releases what was used for sub-
6503 expansions (unless the block got too big). However, discard this block if
6504 the first thing is a variable or header, so that we can use its block when
6505 it is dynamic (useful for very large $message_headers, for example).
6507 6. Lookups now cache *every* query, not just the most recent. A new, separate
6508 store pool is used for this. It can be recovered when all lookup caches are
6509 flushed. Lookups now release memory at the end of their result strings.
6510 This has involved some general refactoring of the lookup sources.
6512 7. Some code has been added to the store_xxx() functions to reduce the amount
6513 of flapping under certain conditions.
6515 8. log_incoming_interface used to affect only the <= reception log lines. Now
6516 it causes the local interface and port to be added to several more SMTP log
6517 lines, for example "SMTP connection from", and rejection lines.
6519 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6521 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6523 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6524 could be overwritten at the end of the current message (or the start of a
6525 new message if it was set in a HELO ACL). The value is now preserved for
6526 the duration of the SMTP connection.
6528 12. If a transport had a headers_rewrite setting, and a matching header line
6529 contained an unqualified address, that address was qualified, even if it
6530 did not match any rewriting rules. The underlying bug was that the values
6531 of the flags that permit the existence of unqualified sender and recipient
6532 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6533 non-local messages, and by -bnq for local messages) were not being
6534 preserved with the message after it was received.
6536 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6537 "next input=" as part of the text comprising the host identity instead of
6538 the correct text. The code was using the same buffer for two different
6539 strings. However, depending on which order the printing function evaluated
6540 its arguments, the bug did not always show up. Under Linux, for example, my
6541 test suite worked just fine.
6543 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6544 doesn't work with some older versions of Perl. It has been changed to "my",
6545 which in any case is probably the better facility to use.
6547 15. A really picky compiler found some instances of statements for creating
6548 error messages that either had too many or two few arguments for the format
6551 16. The size of the buffer for calls to the DNS resolver has been increased
6552 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6553 for addresses that have a lot of PTR records. This alleviates a problem; it
6554 does not fully solve it.
6556 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6557 buffer now truncates the list and logs the incident, which is the same
6558 action as happens when Exim is looking up a host name and its aliases.
6559 Previously in this situation something unpredictable would happen;
6560 sometimes it was "internal error: store_reset failed".
6562 18. If a server dropped the connection unexpectedly when an Exim client was
6563 using GnuTLS and trying to read a response, the client delivery process
6564 crashed while trying to generate an error log message.
6566 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6567 string, for example:
6569 warn message = H1: something\nH2: something
6571 the text was added as a single header line from Exim's point of view
6572 though it ended up OK in the delivered message. However, searching for the
6573 second and subsequent header lines using $h_h2: did not work. This has been
6574 fixed. Similarly, if a system filter added multiple headers in this way,
6575 the routers could not see them.
6577 20. Expanded the error message when iplsearch is called with an invalid key to
6578 suggest using net-iplsearch in a host list.
6580 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6581 ACLs are no longer actually imposed (and a message to that effect is
6584 22. If a "gecos" field in a passwd entry contained escaped characters, in
6585 particular, if it contained a \" sequence, Exim got it wrong when building
6586 a From: or a Sender: header from that name. A second bug also caused
6587 incorrect handling when an unquoted " was present following a character
6588 that needed quoting.
6590 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6591 was not being matched caselessly.
6593 24. Arranged for all hyphens in the exim.8 source to be escaped with
6596 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6597 properly. Recipient callout cache records were still being keyed to include
6598 the sender, even when use_sender was set false. This led to far more
6599 callouts that were necessary. The sender is no longer included in the key
6600 when use_sender is false.
6602 26. Added "control = submission" modifier to ACLs.
6604 27. Added the ${base62d: operator to decode base 62 numbers.
6606 28. dnsdb lookups can now access SRV records.
6608 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6609 the configuration file.
6611 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6612 (-v) mode. This makes the output for a verbose queue run more intelligible.
6614 31. Added a use_postmaster feature to recipient callouts.
6616 32. Added the $body_zerocount variable, containing the number of binary zero
6617 bytes in the message body.
6619 33. The time of last modification of the "new" subdirectory is now used as the
6620 "mailbox time last read" when there is a quota error for a maildir
6623 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6625 35. Added +ignore_unknown as a special item in host lists.
6627 36. Code for decoding IPv6 addresses in host lists is now included, even if
6628 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6629 address was recognized as an IP address, but was then not correctly decoded
6630 into binary, causing unexpected and incorrect effects when compared with
6637 1. Very minor rewording of debugging text in manualroute to say "list of
6638 hosts" instead of "hostlist".
6640 2. If verify=header_syntax was set, and a header line with an unqualified
6641 address (no domain) and a large number of spaces between the end of the
6642 name and the colon was received, the reception process suffered a buffer
6643 overflow, and (when I tested it) crashed. This was caused by some obsolete
6644 code that should have been removed. The fix is to remove it!
6646 3. When running in the test harness, delay a bit after writing a bounce
6647 message to get a bit more predictability in the log output.
6649 4. Added a call to search_tidyup() just before forking a reception process. In
6650 theory, someone could use a lookup in the expansion of smtp_accept_max_
6651 per_host which, without the tidyup, could leave open a database connection.
6653 5. Added the variables $recipient_data and $sender_data which get set from a
6654 lookup success in an ACL "recipients" or "senders" condition, or a router
6655 "senders" option, similar to $domain_data and $local_part_data.
6657 6. Moved the writing of debug_print from before to after the "senders" test
6660 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6661 problems for message scanning, either using a data ACL, or using
6662 local_scan() because the Received: header was not generated till after they
6663 were called (in order to set the time as the time of reception completion).
6664 I have revised the way this works. The header is now generated after the
6665 body is received, but before the ACL or local_scan() are called. After they
6666 are run, the timestamp in the header is updated.
6672 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6673 before starting a queue runner without re-exec. This happened only when
6674 deliver_drop_privilege was set or when the Exim user was set to root. The
6675 effect of the bug was that timeouts during subsequent deliveries caused
6676 crashes instead of being properly handled. The handler is now left at its
6677 default (and expected) setting.
6679 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6680 message, again when deliver_drop_privilege is set or Exim is run as root.
6681 The bug described in (1) was not present in this case, but the tidying up
6682 of the other signals was missing. I have made the two cases consistent.
6684 3. The ignore_target_hosts setting on a manualroute router was being ignored
6685 for hosts that were looked up using the /MX notation.
6687 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6690 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6691 operated on the sender address. After changing the $sender_address to <>
6692 for the sender address verify, Exim was re-instated it as the original
6693 (before rewriting) address, but remembering that it had rewritten it, so it
6694 wasn't rewriting it again. This bug also had the effect of breaking the
6695 sender address verification caching when the sender address was rewritten.
6697 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6698 This has been changed so that if the ip literal address matches
6699 ignore_target_hosts, the router declines.
6701 7. Added expansion conditions match_domain, match_address, and match_local_
6702 part (NOT match_host).
6704 8. The placeholder for the Received: header didn't have a length field set.
6706 9. Added code to Exim itself and to exim_lock to test for a specific race
6707 condition that could lead to file corruption when using MBX delivery. The
6708 issue is with the lockfile that is created in /tmp. If this file is removed
6709 after a process has opened it but before that process has acquired a lock,
6710 there is the potential for a second process to recreate the file and also
6711 acquire a lock. This could lead to two Exim processes writing to the file
6712 at the same time. The added code performs the same test as UW imapd; it
6713 checks after acquiring the lock that its file descriptor still refers to
6714 the same named file.
6716 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6717 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6720 11. Added the smtp_active_hostname option. If used, this will typically be made
6721 to depend on the incoming interface address. Because $interface_address is
6722 not set up until the daemon has forked a reception process, error responses
6723 that can happen earlier (such as "too many connections") no longer contain
6726 12. If an expansion in a condition on a "warn" statement fails because a lookup
6727 defers, the "warn" statement is abandoned, and the next ACL statement is
6728 processed. Previously this caused the whole ACL to be aborted.
6730 13. Added the iplsearch lookup type.
6732 14. Added ident_timeout as a log selector.
6734 15. Added tls_certificate_verified as a log selector.
6736 16. Added a global option tls_require_ciphers (compare the smtp transport
6737 option of the same name). This controls incoming TLS connections.
6739 17. I finally figured out how to make tls_require_ciphers do a similar thing
6740 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6741 before starting the TLS session.
6743 18. Tabs are now shown as \t in -bP output.
6745 19. If the log selector return_path_on_delivery was set, Exim crashed when
6746 bouncing a message because it had too many Received: header lines.
6748 20. If two routers both had headers_remove settings, and the first one included
6749 a superfluous trailing colon, the final name in the first list and the
6750 first name in the second list were incorrectly joined into one item (with a
6751 colon in the middle).
6757 1. Added -C and -D options to the exinext utility, mainly to make it easier
6758 to include in the automated testing, but these could be helpful when
6759 multiple configurations are in use.
6761 2. The exinext utility was not formatting the output nicely when there was
6762 an alternate port involved in the retry record key, nor when there was a
6763 message id as well (for retries that were specific to a specific message
6764 and a specific host). It was also confused by IPv6 addresses, because of
6765 the additional colons they contain. I have fixed the IPv4 problem, and
6766 patched it up to do a reasonable job for IPv6.
6768 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6769 delivery, the log line now contains "pipelined" if PIPELINING was used.
6771 4. An SMTP transport process used to panic and die if the bind() call to set
6772 an explicit outgoing interface failed. This has been changed; it is now
6773 treated in the same way as a connect() failure.
6775 5. A reference to $sender_host_name in the part of a conditional expansion
6776 that was being skipped was still causing a DNS lookup. This no longer
6779 6. The def: expansion condition was not recognizing references to header lines
6780 that used bh_ and bheader_.
6782 7. Added the _cache feature to named lists.
6784 8. The code for checking quota_filecount in the appendfile transport was
6785 allowing one more file than it should have been.
6787 9. For compatibility with Sendmail, the command line option
6795 and sets the incoming protocol and host name (for trusted callers). The
6796 host name and its colon can be omitted when only the protocol is to be set.
6797 Note the Exim already has two private options, -pd and -ps, that refer to
6798 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6799 "s", but I don't think that's a major issue.
6801 10. A number of refactoring changes to the code, none of which should affect
6804 (a) The number of logging options was getting close to filling up the
6805 32-bit word that was used as a bit map. I have split them into two classes:
6806 those that are passed in the argument to log_write(), and those that are
6807 only ever tested independently outside of that function. These are now in
6808 separate 32-bit words, so there is plenty of room for expansion again.
6809 There is no change in the user interface or the logging behaviour.
6811 (b) When building, for example, log lines, the code previously used a
6812 macro that called string_cat() twice, in order to add two strings. This is
6813 not really sufficiently general. Furthermore, there was one instance where
6814 it was actually wrong because one of the argument was used twice, and in
6815 one call a function was used. (As it happened, calling the function twice
6816 did not affect the overall behaviour.) The macro has been replaced by a
6817 function that can join an arbitrary number of extra strings onto a growing
6820 (c) The code for expansion conditions now uses a table and a binary chop
6821 instead of a serial search (which was left over from when there were very
6822 few conditions). Also, it now recognizes conditions like "pam" even when
6823 the relevant support is not compiled in: a suitably worded error message is
6824 given if an attempt is made to use such a condition.
6826 11. Added ${time_interval:xxxxx}.
6828 12. A bug was causing one of the ddress fields not to be passed back correctly
6829 from remote delivery subprocesses. The field in question was not being
6830 subsequently used, so this caused to problems in practice.
6832 13. Added new log selectors queue_time and deliver_time.
6834 14. Might have fixed a bug in maildirsizefile handling that threw up
6835 "unexpected character" debug warnings, and recalculated the data
6836 unnecessarily. In any case, I expanded the warning message to give more
6839 15. Added the message "Restricted characters in address" to the statements in
6840 the default ACL that block characters like @ and % in local parts.
6842 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6843 Three changes have been made:
6845 (a) There was a serious bug; a negative response to MAIL caused the whole
6846 recipient domain to be cached as invalid, thereby blocking all messages
6847 to all local parts at the same domain, from all senders. This bug has
6848 been fixed. The domain is no longer cached after a negative response to
6849 MAIL if the sender used is not empty.
6851 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6854 (c) A new callout option, "use_sender" has been added for people who want
6855 the modified behaviour.
6861 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6864 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6865 indeed breaks things for older releases.
6867 3. Added additional logging to the case where there is a problem reading data
6868 from a filter that is running in a subprocess using a pipe, in order to
6869 try to track down a specific problem.
6871 4. Testing facility fudge: when running in the test harness and attempting
6872 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6873 getting "No route to host". Convert this to a timeout.
6875 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6878 6. Some OS don't have socklen_t but use size_t instead. This affects the
6879 fifth argument of getsockopt() amongst other things. This is now
6880 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6881 can be set for individual OS. I have set it for SunOS5, OSF1, and
6882 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6883 some earlier ones do not.
6885 7. Change 4.30/15 was not doing the test caselessly.
6887 8. The standard form for an IPv6 address literal was being rejected by address
6888 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6889 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6890 this, as well as the form without the "IPv6" on the front (but only when
6891 address literals are enabled, of course).
6893 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6895 10. Exim crashed if a message with an empty sender address specified by -f
6896 encountered a router with an errors_to setting. This could be provoked only
6897 by a command such as
6901 where an empty string was supplied; "<>" did not hit this bug.
6903 11. Installed PCRE release 4.5.
6905 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6906 remained set. It is now erased.
6908 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6909 times from message ids (which are base 36 rather than the normal 62).
6911 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6912 were being counted as actual protocol errors, and logged if the log
6913 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6914 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6915 and DATA following a set of rejected RCPTs do not count as protocol errors.
6916 In other words, Exim assumes they were pipelined, though this may not
6917 actually be the case. Of course, in all cases the client gets an
6918 appropriate error code.
6920 15. If a lookup fails in an ACL condition, a message about the failure may
6921 be available; it is used if testing the ACL cannot continue, because most
6922 such messages specify what the cause of the deferral is. However, some
6923 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6924 that caused an old message to be retained and used if a later statement
6925 caused a defer, replacing the real cause of the deferral.
6927 16. If an IP address had so many PTR records that the DNS lookup buffer
6928 was not large enough to hold them, Exim could crash while trying to process
6929 the truncated data. It now detects and logs this case.
6931 17. Further to 4.21/58, another change has been made: if (and only if) the
6932 first line of a message (the first header line) ends with CRLF, a bare LF
6933 in a subsequent header line has a space inserted after it, so as not to
6934 terminate the header.
6936 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6937 unnecessarily, used atoi() instead of strtol(), and didn't check the
6938 termination when getting file sizes from file names by regex.
6940 19. Completely re-implemented the support for maildirsize files, in the light
6941 of a number of problems with the previous contributed implementation
6942 (4.30/29). In particular:
6944 . If the quota is zero, the maildirsize file is maintained, but no quota is
6947 . If the maildir directory does not exist, it is created before any attempt
6948 to write a maildirsize file.
6950 . The quota value in the file is just a cache; if the quota is changed in
6951 the transport, the new value overrides.
6953 . A regular expression is available for excluding directories from the
6956 20. The autoreply transport checks the characters in options that define the
6957 message's headers; it allows continued headers, but it was checking with
6958 isspace() after an embedded newline instead of explicitly looking for a
6961 21. If all the "regular" hosts to which an address was routed had passed their
6962 expiry times, and had not reached their retry times, the address was
6963 bounced, even if fallback hosts were defined. Now Exim should go on to try
6966 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6967 equivalent code in the SMTP transport. Some hosts send humungous responses
6968 to HELO/EHLO, more than 1024 it seems.
6970 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6971 gives compiler warnings in some environments. I've now done it "properly",
6974 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6975 (because of problems with the built-in one) was declared to return uschar *
6976 instead of char *, causing compiler failure.
6978 25. Fixed a file descriptor leak when processing alias/forward files.
6980 26. Fixed a minor format string issue in dbfn.c.
6982 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6984 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6985 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6988 29. When a sender address is verified, it is cached, to save repeating the test
6989 when there is more than one recipient in a message. However, when the
6990 verification involves a callout, it is possible for different callout
6991 options to be set for different recipients. It is too complicated to keep
6992 track of this in the cache, so now Exim always runs a verification when a
6993 callout is required, relying on the callout cache for the optimization.
6994 The overhead is duplication of the address routing, but this should not be
6997 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6998 to be verified with callout=postmaster, and the main callout worked but the
6999 postmaster check failed, the verification correctly failed. However, if a
7000 subsequent RCPT command asked for sender verification *without* the
7001 postmaster check, incorrect caching caused this verification also to fail,
7004 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7005 it was not caching the DNS options (qualify_single, search_parents) that
7006 were used when the lookup failed. A subsequent lookup with different
7007 options therefore always gave the same answer, though there were cases
7008 where it should not have. (Example: a "domains = !$mx_any" option on a
7009 dnslookup router: the "domains" option is always processed without any
7010 widening, but the router might have qualify_single set.) Now Exim uses the
7011 cached value only when the same options are set.
7013 32. Added John Jetmore's "exipick" utility to the distribution.
7015 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7016 than a timeout (e.g. a certificate is required, and is not provided), an
7017 Exim server now closes the connection immediately. Previously it waited for
7018 the client to close - but if the client is SSL, it seems that they each
7019 wait for each other, leading to a delay before one of them times out.
7021 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7022 maintained 0.8.x compatibility because I don't think many are using it, and
7023 it is clearly obsolete.
7025 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7028 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7029 client certificate was expired. A simple patch fixes this, though I don't
7030 understand the full logic of why the verify callback is called multiple
7033 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7034 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7035 which causes problems with some clients (such as the Certicom SSL Plus
7036 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7037 disables the coutermeasure allowing Eudora to connect."
7039 38. Exim was not checking that a write() to a log file succeeded. This could
7040 lead to Bad Things if a log got too big, in particular if it hit a file
7041 size limit. Exim now panics and dies if it cannot write to a log file, just
7042 as it does if it cannot open a log file.
7044 39. Modified OS/Makefile-Linux so that it now contains
7046 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7048 The two -D definitions ensure that Exim is compiled with large file
7049 support, which makes it possible to handle log files that are bigger than
7052 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7053 instance) a domain was checked against a named list that involved a lookup,
7054 causing $domain_data to be set, then another domain was checked against the
7055 same list, then the first domain was re-checked, the value of $domain_data
7056 after the final check could be wrong. In particular, if the second check
7057 failed, it could be set empty. This bug probably also applied to
7060 41. The strip_trailing_dot option was not being applied to the address given
7061 with the -f command-line option.
7063 42. The code for reading a message's header from the spool was incrementing
7064 $received_count, but never initializing it. This meant that the value was
7065 incorrect (doubled) while delivering a message in the same process in which
7066 it was received. In the most common configuration of Exim, this never
7067 happens - a fresh exec is done - but it can happen when
7068 deliver_drop_privilege is set.
7070 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7071 it now includes up to 150 characters of the unexpected data in the log
7074 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7075 and building data strings. The size of both of these buffers was 10 000
7076 bytes - far larger than anybody would *ever* want, thought I. Needless to
7077 say, somebody hit the limit. I have increased the maximum line length to
7078 20 000 and the maximum data length of concatenated lines to 100 000. I have
7079 also fixed two bugs, because there was no checking on these buffers. Tsk,
7080 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7081 buffer is too small.
7083 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7084 lsearch lookups. Now it does.
7086 46. When parsing a route_list item in a manualroute router, a fixed-length
7087 buffer was used for the list of hosts. I made this 1024 bytes long,
7088 thinking that nobody would ever have a list of hosts that long. Wrong.
7089 Somebody had a whole pile of complicated expansion conditions, and the
7090 string was silently truncated, leading to an expansion error. It turns out
7091 that it is easier to change to an unlimited length (owing to other changes
7092 that have happened since this code was originally written) than to build
7093 structure for giving a limitation error. The length of the item that
7094 expands into the list of hosts is now unlimited.
7096 47. The lsearch lookup could not handle data where the length of text line was
7097 more than 4095 characters. Such lines were truncated, leading to shortened
7098 data being returned. It should now handle lines of any length.
7100 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7101 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7104 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7105 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7106 no longer happens in comment lines. A list of replacements is now placed
7107 at the head of all of the source files, except those whose only change is
7108 to replace PERL_COMMAND in the very first #! line.
7110 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7111 messages on the queue, with a bottom-up merge sort, using code contributed
7112 by Michael Haardt. This should make operations like -bp somewhat faster on
7113 large queues. It won't affect queue runners, except when queue_run_in_order
7116 51. Installed eximstats 1.31 in the distribution.
7118 52. Added support for SRV lookups to the dnslookup router.
7120 53. If an ACL referred to $message_body or $message_body_end, the value was not
7121 reset for any messages that followed in the same SMTP session.
7123 54. The store-handling optimization for building very long strings was not
7124 differentiating between the different store pools. I don't think this
7125 actually made any difference in practice, but I've tidied it.
7127 55. While running the routers to verify a sender address, $sender_address
7128 was still set to the sender address. This is wrong, because when routing to
7129 send a bounce to the sender, it would be empty. Therefore, I have changed
7130 it so that, while verifying a sender address, $sender_address is set to <>.
7131 (There is no change to what happens when verifying a recipient address.)
7133 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7134 target A or AAAA records (if not already returned) without resetting the
7135 qualify_single or search_parents options of the DNS resolver. These are
7136 inappropriate in this case because the targets of MX and SRV records must
7137 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7138 target that, when qualified, matched something in the local domain. These
7139 two options are now turned off when doing these lookups.
7141 57. It seems that at least some releases of Reiserfs (which does not have the
7142 concept of a fixed number of inodes) returns zero and not -1 for the
7143 number of available inodes. This interacted badly with check_spool_inodes,
7144 which assumed that -1 was the "no such thing" setting. What I have done is
7145 to check that the total number of inodes is greater than zero before doing
7146 the test of how many are available.
7148 58. When a "warn" ACL statement has a log_message modifier, the message is
7149 remembered, and not repeated. This is to avoid a lot of repetition when a
7150 message has many recipients that cause the same warning to be written.
7151 However, Exim was preserving the list of already written lines for an
7152 entire SMTP session, which doesn't seem right. The memory is now reset if a
7153 new message is started.
7155 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7156 debugging output unless log_rewrite was also set.
7158 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7159 of (int)(handle) when we know that handle contains (void *)(-1).
7161 61. The Exim daemon panic-logs an error return when it closes the incoming
7162 connection. However "connection reset by peer" seems to be common, and
7163 isn't really an error worthy of noting specially, so that particular error
7166 62. When Exim is trying to find all the local interfaces, it used to panic and
7167 die if the ioctl to get the interface flags failed. However, it seems that
7168 on at least one OS (Solaris 9) it is possible to have an interface that is
7169 included in the list of interfaces, but for which you get a failure error
7170 for this call. This happens when the interface is not "plumbed" into a
7171 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7172 failure of the "get flags" call assumes that the interface is down.
7174 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7175 makes life easier for people who are doing arithmetic on fields extracted
7176 from dates, where you often get leading zeros that should not be
7177 interpreted as octal.
7179 64. Added qualify_domain to the redirect router, to override the global
7182 65. If a pathologically long header line contained very many addresses (the
7183 report of this problem mentioned 10 000) and each of them was rewritten,
7184 Exim could use up a very large amount of memory. (It kept on making new
7185 copies of the header line as it rewrote, and never released the old ones.)
7186 At the expense of a bit more processing, the header rewriting function has
7187 been changed so that it no longer eats memory in this way.
7189 66. The generation of the Received: header has been moved from the time that a
7190 message starts to be received, to the time that it finishes. The timestamp
7191 in the Received: header should now be very close to that of the <= log
7192 line. There are two side-effects of this change:
7194 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7195 logged header lines no longer include the local Received: line, because
7196 it has not yet been created. The same applies to a copy of the message
7197 that is returned to a non-SMTP sender when a message is rejected.
7199 (b) When a filter file is tested using -bf, no additional Received: header
7200 is added to the test message. After some thought, I decided that this
7203 This change does not affect the value of $received_for. It is still set
7204 after address rewriting, but before local_scan() is called.
7206 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7208 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7209 gave an unhelpful panic error message, and a defer error. I have managed to
7210 change this behaviour so that it now rejects any supplied certificate,
7211 which seems right, as the list of acceptable certificates is empty.
7213 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7214 gave an unhelpful defer error. I have not managed to make this reject any
7215 supplied certificates, but the error message it gives is "no certificate
7216 supplied", which is not helpful.
7218 70. exigrep's output now also includes lines that are not associated with any
7219 message, but which match the given pattern. Implemented by a patch from
7220 Martin Sluka, which also tidied up the Perl a bit.
7222 71. Recipient callout verification, like sender verification, was using <> in
7223 the MAIL FROM command. This isn't really the right thing, since the actual
7224 sender may affect whether the remote host accepts the recipient or not. I
7225 have changed it to use the actual sender in the callout; this means that
7226 the cache record is now keyed on a recipient/sender pair, not just the
7227 recipient address. There doesn't seem to be a real danger of callout loops,
7228 since a callout by the remote host to check the sender would use <>.
7229 [SEE ABOVE: changed after hitting problems.]
7231 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7232 temporary errors. However, in the case of such a code being given after
7233 the end of a data transmission (i.e. after ".") Exim was failing to write
7234 a retry record for the message. (Yes, there was some broken host that was
7235 actually sending 8xx at this point.)
7237 73. An unknown lookup type in a host list could cause Exim to panic-die when
7238 the list was checked. (An example that provoked this was putting <; in the
7239 middle of a list instead of at the start.) If this happened during a DATA
7240 ACL check, a -D file could be left lying around. This kind of configuration
7241 error no longer causes Exim to die; instead it causes a defer error. The
7242 incident is still logged to the main and panic logs.
7244 74. Buglet left over from Exim 3 conversion. The message "too many messages
7245 in one connection" was written to the rejectlog but not the mainlog, except
7246 when address rewriting (yes!) was being logged.
7248 75. Added write_rejectlog option.
7250 76. When a system filter was run not as root (that is, when system_filter_user
7251 was set), the values of the $n variables were not being returned to the
7252 main process; thus, they were not subsequently available in the $sn
7255 77. Added +return_path_on_delivery log selector.
7257 78. A connection timeout was being treated differently from recipients deferred
7258 when testing hosts_max_try with a message that was older than the host's
7259 retry timeout. (The host should not be counted, thus allowing all hosts to
7260 be tried at least once before bouncing.) This may have been the cause of an
7261 occasionally reported bug whereby a message would remain on the queue
7262 longer than the retry timeout, but would be bounced if a delivery was
7263 forced. I say "may" because I never totally pinned down the problem;
7264 setting up timeout/retry tests is difficult. See also the next item.
7266 79. The ultimate address timeout was not being applied to errors that involved
7267 a combination of host plus message (for example, a timeout on a MAIL
7268 command). When an address resolved to a number of possible hosts, and they
7269 were not all tried for each delivery (e.g. because of hosts_max_try), a
7270 message could remain on the queue longer than the retry timeout.
7272 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7275 81. Fixed an obscure SMTP outgoing bug which required at least the following
7276 conditions: (a) there was another message waiting for the same server;
7277 (b) the server returned 5xx to all RCPT commands in the first message so
7278 that the message was not completed; (c) the server dropped the connection
7279 or gave a negative response to the RSET that Exim sends to abort the
7280 transaction. The observed case was a dropped connection after DATA that had
7281 been sent in pipelining mode. That is, the server had advertised PIPELINING
7282 but was not implementing it correctly. The effect of the bug was incorrect
7283 behaviour, such as trying another host, and this could lead to a crash.
7289 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7290 and daemon.c were passed as pointers to ints; they should have been
7291 pointers to socklen_t variables (which are typically unsigned ints).
7293 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7296 3. Fixed a really odd bug that affected only the testing scheme; patching a
7297 certain fixed string in the binary changed the value of another string that
7298 happened to be identical to the end of the original first string.
7300 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7301 name", it returns that address as the IP address. On some operating
7302 systems (e.g. Solaris), it also passes back the IP address string as the
7303 "host name". However, on others (e.g. Linux), it passes back an empty
7304 string. Exim wasn't checking for this, and was changing the host name to an
7305 empty string, assuming it had been canonicalized.
7307 5. Although rare, it is permitted to have more than one PTR record for a given
7308 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7309 all the names associated with an address, because they do in Solaris.
7310 However, it seems that they do not in Linux for data that comes from the
7311 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7312 I found this out when I moved to a new Linux workstation and tried to run
7313 the Exim test suite.
7315 To get round this problem I have changed the code so that it now does its
7316 own call to the DNS to look up PTR records when searching for a host name.
7317 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7318 addresses that are only in /etc/hosts are still found.
7320 This behaviour is, however, controlled by an option called host_lookup_
7321 order, which defaults to "bydns:byaddr". If people want to use the other
7322 order, or indeed, just use one or the other means of lookup, they can
7323 specify it in this variable.
7325 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7326 some operating systems, this comes back from gethostbyaddr() as an empty
7327 string, and this is what Exim used to test for. However, it seems that in
7328 other systems, "." is yielded. Exim now tests for this case too.
7330 7. The values of check_spool_space and check_log_space are now held internally
7331 as a number of kilobytes instead of an absolute number of bytes. If a
7332 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7333 kilobyte. This means that much larger values can be stored.
7335 8. Exim monitor: an attempt to get the action menu when not actually pointing
7336 at a message produces an empty menu entitled "No message selected". This
7337 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7338 no entries in it ("Shell widget menu has zero width and/or height"). So I
7339 have added a single, blank menu entry in this case.
7341 9. Added ${quote_local_part.
7343 10. MIME decoding is now applied to the contents of Subject: header lines when
7346 11. Now that a reference to $sender_host_address automatically causes a reverse
7347 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7348 host lookup before query-style lookups in lists that might use this
7349 variable. This has therefore been abolished, and the "net-" prefix is no
7350 longer necessary for query-style lookups.
7352 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7353 have been a typo for LFLAGS, so it has been changed.
7355 13. The install script calls Exim with "-C /dev/null" in order to find the
7356 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7357 to be output. However, since Exim outputs its version number before the
7358 error, it didn't break the script. It just looked ugly. I fixed this by
7359 always allowing "-C /dev/null" if the caller is root.
7361 14. Ignore overlarge ACL variable number when reading spool file - insurance
7362 against a later release with more variables having written the file.
7364 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7365 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7366 this, as well as the form without the "IPv6" on the front.
7368 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7369 OS/Makefile-Darwin file.
7371 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7372 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7374 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7375 variable. However, some OS use ints for inodes. I've added cast to long int
7376 to get rid of the compiler warning.
7378 19. I had forgotten to lock out "/../" in configuration file names when
7379 ALT_CONFIG_PREFIX was set.
7381 20. Routers used for verification do not need to specify transports. However,
7382 if such a router generated a host list, and callout was configured, Exim
7383 crashed, because it could not find a port number from the (non-existent)
7384 transport. It now assumes port 25 in this circumstance.
7386 21. Added the -t option to exigrep.
7388 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7389 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7390 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7391 them is set, LOOKUP_LSEARCH is forced.
7393 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7395 24. Added sender and host information to the "rejected by local_scan()" log
7396 line; previously there was no indication of these.
7398 25. Added .include_if_exists.
7400 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7401 Linux. It turns out that not all file systems support this. Apparently some
7402 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7403 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7404 not supported on the file descriptor, is now ignored when Exim is trying to
7405 sync a directory. This applies only to Linux.
7407 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7409 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7410 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7413 29. Added support for maildirsize files from supplied patch (modified a bit).
7415 30. The use of :fail: followed by an empty string could lead Exim to respond to
7416 sender verification failures with (e.g.):
7418 550 Verification failed for <xxx>
7419 550 Sender verify failed
7421 where the first response line was missing the '-' that indicates it is not
7422 the final line of the response.
7424 31. The loop for finding the name of the user that called Exim had a hardwired
7425 limit of 10; it now uses the value of finduser_retries, which is used for
7426 all other user lookups.
7428 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7431 33. Exim was neglecting to zero errno before one call of strtol() when
7432 expanding a string and expecting an integer value. On some systems this
7433 resulted in spurious "integer overflow" errors. Also, it was casting the
7434 result into an int without checking.
7436 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7437 did not work. The code looks as if it has *never* worked, though it appears
7438 to have been documented since at least release 1.62. I have made it work.
7440 35. The "timeout_DNS" error in retry rules, also documented since at least
7441 1.62, also never worked. As it isn't clear exactly what this means, and
7442 clearly it isn't a major issue, I have abolished the feature by treating it
7443 as "timeout", and writing a warning to the main and panic logs.
7445 36. The display of retry rules for -brt wasn't always showing the error code
7448 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7449 timeout_connect_A, timeout_connect_MX.
7451 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7452 to the empty sender.
7454 39. The daemon was not analysing the content of -oX till after it had closed
7455 stderr and disconnected from the controlling terminal. This meant that any
7456 syntax errors were only noted on the panic log, and the return code from
7457 the command was 0. By re-arranging the code a little, I've made the
7458 decoding happen first, so such errors now appear on stderr, and the return
7459 code is 1. However, the actual setting up of the sockets still happens in
7460 the disconnected process, so errors there are still only recorded on the
7463 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7464 connections (as happens on some IP stacks) was logged at start up time as
7465 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7466 it from "IPv6 and IPv4", which means that two separate sockets are being
7469 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7470 says whether AF_INET or AF_INET6 was passed as an argument.
7472 42. Exiwhat output was messed up when time zones were included in log
7475 43. Exiwhat now gives more information about the daemon's listening ports,
7476 and whether -tls-on-connect was used.
7478 44. The "port" option of the smtp transport is now expanded.
7480 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7481 silently ignored. Now an error message is written to the main and panic
7484 46. There's a new ACL modifier called "logwrite" which writes to a log file
7485 as soon as it is encountered.
7487 47. Added $local_user_uid and $local_user_gid at routing time.
7489 48. Exim crashed when trying to verify a sender address that was being
7492 49. Exim was recognizing only a space character after ".include". It now also
7493 recognizes a tab character.
7495 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7496 extracting the relevant information from the specification. The man page no
7497 longer contains scrambled data for the -d option, and I've added a section
7498 at the front about calling Exim under different names.
7500 51. Added "extra_headers" argument to the "mail" command in filter files.
7502 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7505 53. Installed eximstats 1.29.
7507 54. Added transport_filter_timeout as a generic transport option.
7509 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7510 Cc: header lines. This was required by RFC 822, but it not required by RFC
7513 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7514 incoming messages that did not have them. Now it does so only if the
7515 message originates locally, that is, if there is no associated remote host
7516 address. When Resent- header lines are present, this applies to the Resent-
7517 lines rather than the non-Resent- lines.
7519 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7520 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7522 58. Messages for configuration errors now include the name of the main
7523 configuration file - useful now that there may be more than one file in a
7524 list (.included file names were always shown).
7526 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7527 for those rare installations that do not start the daemon as root or run it
7528 setuid root. I've cut out the call to initgroups() if the daemon is not
7531 60. The Exim user and group can now be bound into the binary as text strings
7532 that are looked up at the start of Exim's processing.
7534 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7536 62. Added $mailstore_basename variable.
7538 63. Installed patch to sieve.c from Michael Haardt.
7540 64. When Exim failed to open the panic log after failing to open the main log,
7541 the original message it was trying to log was written to stderr and debug
7542 output, but if they were not available (the usual case in production), it
7543 was lost. Now it is written to syslog before the two lines that record the
7544 failures to open the logs.
7546 65. Users' Exim filters run in subprocesses under the user's uid. It is
7547 possible for a "deliver" command or an alias in a "personal" command to
7548 provoke an address rewrite. If logging of address rewriting is configured,
7549 this fails because the process is not running as root or exim. There may be
7550 a better way of dealing with this, but for the moment (because 4.30 needs
7551 to be released), I have disabled address rewrite logging when running a
7552 filter in a non-root, non-exim process.
7558 1. The buildconfig auxiliary program wasn't quoting the value set for
7559 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7560 not defined. This bug was masked in 4.22 by the effect that was fixed in
7563 2. Some messages that were rejected after a message id was allocated were
7564 shown as "incomplete" by exigrep. It no longer does this for messages that
7565 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7567 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7568 have allow_domain_literals set, the ID did not get logged in the <= line.
7569 Domain literals are now always recognized in Message-ID: header lines.
7571 4. The first argument for a ${extract expansion item is the key name or field
7572 number. Leading and trailing spaces in this item were not being ignored,
7573 causing some misleading effects.
7575 5. When deliver_drop_privilege was set, single queue runner processes started
7576 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7577 same command in the process it spins off) were not dropping privilege.
7579 6. When the daemon running as "exim" started a queue runner, it always
7580 re-executed Exim in the spun-off process. This is a waste of effort when
7581 deliver_drop_privilege is set. The new process now just calls the
7582 queue-runner function directly.
7588 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7591 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7592 ignored. Though the use of -oP was forcing the writing of a pid file, it
7593 was always written to the default place.
7595 3. If the message "no IP address found for host xxxx" is generated during
7596 incoming verification, it is now followed by identification of the incoming
7597 connection (so you can more easily find what provoked it).
7599 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7601 5. Added some features to "harden" Exim a bit more against certain attacks:
7603 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7604 be put in Local/Makefile. This is like the never_users runtime option,
7605 but it cannot be overridden. The default setting is "root".
7607 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7608 prefix string with which any file named in a -C command line option
7611 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7612 is retained for -C and -D only if the caller of Exim is root. Without
7613 it, the exim user may also use -C and -D and retain privilege.
7615 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7616 command line option is disabled.
7618 6. Macro names set by the -D option must start with an upper case letter, just
7619 like macro names defined in the configuration file.
7621 7. Added "dereference=" facility to LDAP.
7623 8. Two instances of the typo "uknown" in the source files are fixed.
7625 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7626 the Configure-Makefile script screwed up while processing it.
7628 10. Incorporated PCRE 4.4.
7630 11. The SMTP synchronization check was not operating right at the start of an
7631 SMTP session. For example, it could not catch a HELO sent before the client
7632 waited for the greeting. There is now a check for outstanding input at the
7633 point when the greeting is written. Because of the duplex, asynchronous
7634 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7635 way, but not yet received, when the check is performed.
7637 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7638 on TCP/IP sockets, because this apparently causes some broken clients to
7641 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7642 unchanged) from the Cygwin maintainer.
7644 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7645 was supported instead of testing for mbx. Effectively a typo.
7647 15. The spa authenticator server code was not checking that the input it
7648 received was valid base64.
7650 16. The debug output line for the "set" modifier in ACLs was not showing the
7651 name of the variable that was being set.
7653 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7655 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7656 lookup on its own. Something else was needed to trigger the lookup. For
7657 example, a match in host_lookup or the need for a host name in a host list.
7658 Now, if $sender_host_name is referenced and the host name has not yet been
7659 looked up, a lookup is performed. If the lookup fails, the variable remains
7660 empty, and $host_lookup_failed is set to "1".
7662 19. Added "eqi" as a case-independent comparison operator.
7664 20. The saslauthd authentication condition could segfault if neither service
7665 nor realm was specified.
7667 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7668 error message that was logged was misleading, and incoming SMTP
7669 connections were dropped. The message is now more accurate, and temporary
7670 errors are given to SMTP connections.
7672 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7673 (except RSET) with a 421 error, until QUIT is received. However, it was
7674 failing to send a response to QUIT.
7676 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7677 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7678 ACL is now run after the helo_try_verify_hosts code.
7680 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7681 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7682 case-independent, so other case variants are also recognized). Apparently
7683 some systems use these upper case variants.
7685 25. If more than two messages were waiting for the same host, and a transport
7686 filter was specified for the transport, Exim sent two messages over the
7687 same TCP/IP connection, and then failed with "socket operation on non-
7688 socket" when it tried to send the third.
7690 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7692 27. The extern definition of crypt16() in expand.c was not being excluded when
7693 the OS had its own crypt16() function.
7695 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7696 as a preferred synonym for return_size_limit, both as an option and as an
7699 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7701 30. Changed the default configuration ACL to relax the local part checking rule
7702 for addresses that are not in any local domains. For these addresses,
7703 slashes and pipe symbols are allowed within local parts, but the sequence
7704 /../ is explicitly forbidden.
7706 31. SPA server authentication was not clearing the challenge buffer before
7709 32. log_message in a "warn" ACL statement was writing to the reject log as
7710 well as to the main log, which contradicts the documentation and doesn't
7711 seem right (because no rejection is happening). So I have stopped it.
7713 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7714 However, I am unable to do any testing of this.
7716 34. Fixed an infelicity in the appendfile transport. When checking directories
7717 for a mailbox, to see if any needed to be created, it was accidentally
7718 using path names with one or more superfluous leading slashes; tracing
7719 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7721 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7722 contents are added to the log line that is written for every discarded
7723 recipient. (Previously a log_message setting was ignored.)
7725 36. The ${quote: operator now quotes the string if it is empty.
7727 37. The install script runs exim in order to find its version number. If for
7728 some reason other than non-existence or emptiness, which it checks, it
7729 could not run './exim', it was installing it with an empty version number,
7730 i.e. as "exim-". This error state is now caught, and the installation is
7733 38. An argument was missing from the function that creates an error message
7734 when Exim fails to connect to the socket for saslauthd authentication.
7735 This could cause Exim to crash, or give a corrupted message.
7737 39. Added isip, isip4, and isip6 to ${if conditions.
7739 40. The ACL variables $acl_xx are now saved with the message, and can be
7740 accessed later in routers, transports, and filters.
7742 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7743 strings in the file are not string-expanded.
7745 42. If a MAIL command specified a SIZE value that was too large to fit into an
7746 int variable, the check against message_size_limit failed. Such values are
7747 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7748 day this will have to be increased, but I don't think I want to be around
7749 when emails are that large.
7756 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7757 iconv() is not standard in FreeBSD.
7759 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7760 IPv6 enabled. The observed symptom was a segmentation fault on return from
7761 the function os_common_find_running_interfaces() in src/os.c.
7763 3. In the check_special_case() function in daemon.c I had used "errno" as an
7764 argument name, which causes warnings on some systems. This was basically a
7765 typo, since it was named "eno" in the comments!
7767 4. The code that waits for the clock to tick (at a resolution of some fraction
7768 of a second) so as to ensure message-id uniqueness was always waiting for
7769 at least one whole tick, when it could have waited for less. [This is
7770 almost certainly not relevant at current processor speeds, where it is
7771 unlikely to ever wait at all. But we try to future-proof.]
7773 5. The function that sleeps for a time interval that includes fractions of a
7774 second contained a race. It did not block SIGALRM between setting the
7775 timer, and suspending (a couple of lines later). If the interval was short
7776 and the sigsuspend() was delayed until after it had expired, the suspension
7777 never ended. On busy systems this could lead to processes getting stuck for
7780 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7781 process, before it forks any delivery processes. The open lookup caching
7782 mechanism meant that the open file or database connection was passed into
7783 the delivery process. The problem was that delivery processes always tidy
7784 up cached lookup data. This could cause a problem for the next delivery
7785 process started by the queue runner, because the external queue runner
7786 process does not know about the closure. So the next delivery process
7787 still has data in the lookup cache. In the case of a file lookup, there was
7788 no problem because closing a file descriptor in a subprocess doesn't affect
7789 the parent. However, if the lookup was caching a connection to a database,
7790 the connection was closed, and the second delivery process was likely to
7791 see errors such as "PGSQL: query failed: server closed the connection
7792 unexpectedly". The problem has been fixed by closing all cached lookups
7793 in a queue runner before running a delivery process.
7795 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7796 seem to have the "const" qualifier which it has on other OS. I've
7799 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7800 *of the same type* (client or server) with the same public name that an
7801 error should be diagnosed.
7803 9. When Exim looked up a host name for an IP address, but failed to find the
7804 original IP address when looking up the host name (a safety check), it
7805 output the message "<ip address> does not match any IP for NULL", which was
7806 confusing, to say the least. The bug was that the host name should have
7807 appeared instead of "NULL".
7809 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7810 user that is built into the binary, and the -C or -D options is used, root
7811 privilege is dropped before the configuration file is read. In addition,
7812 logging is switched to stderr instead of the normal log files. If the
7813 configuration then re-defines the Exim user, the unprivileged environment
7814 is probably not what is expected, so Exim logs a panic warning message (but
7817 However, if deliver_drop_privilege is set, the unprivileged state may well
7818 be exactly what is intended, so the warning has been cut out in that case,
7819 and Exim is allowed to try to write to its normal log files.
7825 1. smtp_return_error_details was not giving details for temporary sender
7826 or receiver verification errors.
7828 2. Diagnose a configuration error if two authenticators have the same public
7831 3. Exim used not to create the message log file for a message until the first
7832 delivery attempt. This could be confusing when incoming messages were held
7833 for policy or load reasons. The message log file is now created at the time
7834 the message is received, and an initial "Received" line is written to it.
7836 4. The automatically generated man page for command line options had a minor
7837 bug that caused no ill effects; however, a more serious problem was that
7838 the procedure for building the man page automatically didn't always
7839 operate. Consequently, release 4.20 contains an out-of-date version. This
7840 shouldn't happen again.
7842 5. When building Exim with embedded Perl support, the script that builds the
7843 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7844 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7846 6. The freeze_tell option was not being used for messages that were frozen on
7847 arrival, either by an ACL or by local_scan().
7849 7. Added the smtp_incomplete_transaction log selector.
7851 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7852 was accepting AUTH without a new EHLO.
7854 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7855 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7858 10. Logging of TCP/IP connections (when configured) now happens in the main
7859 daemon process instead of the child process, so that the TCP/IP connection
7860 count is more accurate (but it can never be perfect).
7862 11. The use of "drop" in a nested ACL was not being handled correctly in the
7863 outer ACL. Now, if condition failure induced by the nested "drop" causes
7864 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7865 or "require"), the connection is dropped.
7867 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7868 that yield "discard" can now be used with an "accept" or a "discard" verb,
7869 but an error is generated for any others (because I can't see a useful way
7870 to define what should happen).
7872 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7873 are now processed in the same way as lines in the Exim configuration file.
7874 In particular, continuation lines are supported.
7876 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7878 15. Added -ti meaning -t -i.
7880 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7881 domains, and warn by logging if others are found.
7883 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7884 returned from GIFCONF when Exim is trying to find the list of interfaces on
7885 a host. The code in os.c has been modified to copy each ifreq to an aligned
7886 structure in all cases.
7888 Also, in some cases, the returned ifreq's were being copied to a 'struct
7889 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7890 means the last couple of bytes of an IPv6 address could be chopped if the
7891 ifreq contained only a normal sockaddr (14 bytes storage).
7893 18. Named domain lists were not supported in the hosts_treat_as_local option.
7894 An entry such as +xxxx was not recognized, and was treated as a literal
7897 19. Ensure that header lines added by a DATA ACL are included in the reject log
7898 if the ACL subsequently rejects the message.
7900 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7901 MD5 (which is deprecated).
7903 21. When testing a filter file using -bf, Exim was writing a message when it
7904 took the sender from a "From " line in the message, but it was not doing so
7905 when it took $return_path from a Return-Path: header line. It now does.
7907 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7908 with a valid header line field name (a series of printing characters
7909 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7911 23. Changed "disc" in the source to "disk" to conform to the documentation and
7912 the book and for uniformity.
7914 24. Ignore Sendmail's -Ooption=value command line item.
7916 25. When execve() failed while trying to run a command in a pipe transport,
7917 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7918 could be confused with a return value of 69 from the command itself. This
7919 has been changed to 127, the value the shell returns if it is asked to run
7920 a non-existent command. The wording for the related log line suggests a
7921 non-existent command as the problem.
7923 26. If received_header_text expands to an empty string, do not add a Received:
7924 header line to the message. (Well, it adds a token one on the spool, but
7925 marks it "old" so that it doesn't get used or transmitted.)
7927 27. Installed eximstats 1.28 (addition of -nt option).
7929 28. There was no check for failure on the call to getsockname() in the daemon
7930 code. This can fail if there is a shortage of resources on the system, with
7931 ENOMEM, for example. A temporary error is now given on failure.
7933 29. Contrary to the C standard, it seems that in some environments, the
7934 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7935 program. Exim now does this explicitly; it affects the formatting of
7936 timestamps using strftime().
7938 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7939 complaints. I've now initialized all the variables, to avoid this.
7941 32. Header lines added by a system filter were not being "seen" during
7942 transport-time rewrites.
7944 33. The info_callback() function passed to OpenSSL is set up with type void
7945 (*)(SSL *, int, int), as described somewhere. However, when calling the
7946 function (actually a macro) that sets it up, the type void(*)() is
7947 expected. I've put in a cast to prevent warnings from picky compilers.
7949 34. If a DNS black list lookup found a CNAME record, but there were no A
7950 records associated with the domain it pointed at, Exim crashed.
7952 35. If a DNS black list lookup returned more than one A record, Exim ignored
7953 all but the first. It now scans all returned addresses if a particular IP
7954 value is being sought. In this situation, the contents of the
7955 $dnslist_value variable are a list of all the addresses, separated by a
7958 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7959 to accept a host name and all its aliases if the forward lookup for any of
7960 them yielded the IP address of the incoming connection. Now it accepts only
7961 those names whose forward lookup yields the correct IP address. Any other
7962 names are discarded. This closes a loophole whereby a rogue DNS
7963 administrator could create reverse DNS records to break through a
7964 wildcarded host restriction in an ACL.
7966 37. If a user filter or a system filter that ran in a subprocess used any of
7967 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7968 the wrong values were passed to the pipe command ($thisaddress had the
7969 value of $0, $0 had the value of $1, etc). This bug was introduced by
7970 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7972 38. Improved the line breaking for long SMTP error messages from ACLs.
7973 Previously, if there was no break point between 40 and 75 characters, Exim
7974 left the rest of the message alone. Two changes have been made: (a) I've
7975 reduced the minimum length to 35 characters; (b) if it can't find a break
7976 point between 35 and 75 characters, it looks ahead and uses the first one
7977 that it finds. This may give the occasional overlong line, but at least the
7978 remaining text gets split now.
7980 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7981 file descriptors might be low, and that setting 1000 would always raise it.
7982 It turns out that in some environments, the limit is already over 1000 and
7983 that lowering it causes trouble. So now Exim takes care not to decrease it.
7985 40. When delivering a message, the value of $return_path is set to $sender_
7986 address at the start of routing (routers may change the value). By an
7987 oversight, this default was not being set up when an address was tested by
7988 -bt or -bv, which affected the outcome if any router or filter referred to
7991 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7992 log only when "message" or "log_message" are set. However, if one of the
7993 conditions was an address verification, or a call to a nested ACL, the
7994 messages generated by the underlying test were being passed through. This
7995 no longer happens. The underlying message is available in $acl_verify_
7996 message for both "message" and "log_message" expansions, so it can be
7997 passed through if needed.
7999 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8000 new expansion $bh_ to give the encoded byte string without charset
8001 translation. Translation happens only if iconv() is available; HAVE_ICONV
8002 indicates this at build time. HEADERS_CHARSET gives the charset to
8003 translate to; headers_charset can change it in the configuration, and
8004 "headers charset" can change it in an individual filter file.
8006 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8007 that creates RFC 2047 encoded "words" labels them as that charset instead
8008 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8009 expansion operator; (ii) when Exim creates a From: line for a local
8010 message; (iii) when a header line is rewritten to include a "phrase" part.
8012 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8013 buggy, causing it to skip the first lines of messages whose message ID
8014 ended in 'D'. This would not have bitten before Exim release 4.14, because
8015 message IDs were unlikely to end in 'D' before then. The effect was to have
8016 incorrect size information for certain domains.
8018 45. #include "config.h" was missing at the start of the crypt16.c module. This
8019 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8022 46. If there was a timeout during a "random" callout check, Exim treated it as
8023 a failure of the random address, and carried on sending RSET and the real
8024 address. If the delay was just some slowness somewhere, the response to the
8025 original RCPT would be taken as a response to RSET and so on, causing
8026 mayhem of various kinds.
8028 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8029 when I implemented it. It didn't allow for the fact that some option values
8030 may legitimately be negative (e.g. size_addition), and it didn't even do
8031 the right test for positive values.
8033 48. Domain names in DNS records are case-independent. Exim always looks them up
8034 in lower case. Some resolvers return domain names in exactly the case they
8035 appear in the zone file, that is, they may contain uppercase letters. Not
8036 all resolvers do this - some return always lower case. Exim was treating a
8037 change of case by a resolver as a change of domain, similar to a widening
8038 of a domain abbreviation. This triggered its re-routing code and so it was
8039 trying to route what was effectively the same domain again. This normally
8040 caused routing to fail (because the router wouldn't handle the domain
8041 twice). Now Exim checks for this case specially, and just changes the
8042 casing of the domain that it ultimately uses when it transmits the message
8045 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8048 50. If a filter generated a file delivery with a non-absolute name (possible if
8049 no home directory exists for the router), the forbid_file option was not
8052 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8053 the existing equality matching.
8055 52. Exim was using ints instead of ino_t variables in some places where it was
8056 dealing with inode numbers.
8058 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8059 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8060 called TMPDIR, and if it finds it is different, it changes its value.
8062 54. The smtp_printf() function is now made available to local_scan() so
8063 additional output lines can be written before returning. There is also an
8064 smtp_fflush() function to enable the detection of a dropped connection.
8065 The variables smtp_input and smtp_batched_input are exported to
8068 55. Changed the default runtime configuration: the message "Unknown user"
8069 has been removed from the ACL, and instead placed on the localuser router,
8070 using the cannot_route_message feature. This means that any verification
8071 failures that generate their own messages won't get overridden. Similarly,
8072 the "Unrouteable address" message that was in the ACL for unverifiable
8073 relay addresses has also been removed.
8075 56. Added hosts_avoid_esmtp to the smtp transport.
8077 57. The exicyclog script was not checking for the esoteric option
8078 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8079 will work only if exicyclog is run under the appropriate euid.
8081 58. Following a discussion on the list, the rules by which Exim recognises line
8082 endings on incoming messages have been changed. The -dropcr and drop_cr
8083 options are now no-ops, retained only for backwards compatibility. The
8084 following line terminators are recognized: LF CRLF CR. However, special
8085 processing applies to CR:
8087 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8088 nor a local message in the state where . is a terminator.
8090 (ii) If a bare CR is encountered in a header line, an extra space is added
8091 after the line terminator so as not to end the header. The reasoning
8092 behind this is that bare CRs in header lines are most likely either
8093 to be mistakes, or people trying to play silly games.
8095 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8096 was being incorrectly given as 18 bytes larger than it should have been.
8097 This is a VOB (very old bug).
8099 60. This may never have affected anything current, but just in case it has:
8100 When the local host is found other than at the start of a list of hosts,
8101 the local host, those with the same MX, and any that follow, are discarded.
8102 When the list in question was part of a longer list of hosts, the following
8103 hosts (not currently being processed) were also being discarded. This no
8104 longer happens. I'm not sure if this situation could ever has previously
8107 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8110 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8111 header. This is something that is recommended in a new Internet Draft, and
8112 is something that is documented as being done by Sendmail. There are two
8113 possible values. For messages generated by the autoreply transport, Exim
8116 Auto-Submitted: auto-replied
8118 whereas for all other generated messages (e.g. bounces) it adds
8120 Auto-Submitted: auto-generated
8122 63. The "personal" condition in filters now includes a test for the
8123 Auto-Submitted: header. If it contains the string "auto-" the message it
8124 not considered personal.
8126 64. Added rcpt_include_affixes as a generic transport option.
8128 65. Added queue_only_override (default true).
8130 66. Added the syslog_duplication option.
8132 67. If what should have been the first header line of a message consisted of
8133 a space followed by a colon, Exim was mis-interpreting it as a header line.
8134 It isn't of course - it is syntactically invalid and should therefore be
8135 treated as the start of the message body. The misbehaviour could have
8136 caused a number of strange effects, including loss of data in subsequent
8137 header lines, and spool format errors.
8139 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8140 client host had authenticated. This control can now be exercised by an ACL
8141 for more flexibility.
8143 69. By default, callouts do not happen when testing with -bh. There is now a
8144 variant, -bhc, which does actually run the callout code, including
8145 consulting and updating the callout cache.
8147 70. Added support for saslauthd authentication, courtesy of Alexander
8150 71. If statvfs() failed on the spool or log directories while checking their
8151 size for availability, Exim confusingly gave the error "space shortage".
8152 Furthermore, in debugging mode it crashed with a floating point exception.
8153 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8154 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8155 serious problem, Exim now writes to the main and panic logs when this
8156 happens, with details of the failure. It then refuses to accept the
8157 incoming message, giving the message "spool directory problem" or "log
8158 directory problem" with a 421 code for SMTP messages.
8160 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8161 0, 1, and 2 exist, because some OS complain for execs without them (see
8162 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8163 descriptors. However, the code omitted to check that the open succeeded,
8164 causing mysterious errors if for some reason the permissions on /dev/null
8165 got screwed. Now Exim writes a message to the main and panic logs, and
8166 bombs out if it can't open /dev/null.
8168 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8169 interact so that it is all more flexible. It is supposed to remain
8170 backwards compatible. Also added extra_local_interfaces.
8172 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8173 to bomb out with an assertion failure - to the client this appears as a
8174 connection drop. This problem occurs in the part of the code that was taken
8175 from the Samba project. Fortunately, the assertion is in a very simple
8176 function, so I have fixed this by reproducing the function inline in the
8177 one place where it is called, and arranging for authentication to fail
8178 instead of killing the process with assert().
8180 75. The SPA client code was not working when the server requested OEM rather
8181 than Unicode encoding.
8183 76. Added code to make require_files with a specific uid setting more usable in
8184 the case where statting the file as root fails - usually a non-root-mounted
8185 NFS file system. When this happens and the failure is EACCES, Exim now
8186 forks a subprocess and does the per-uid checking as the relevant uid.
8188 77. Added process_log_path.
8190 78. If log_file_path was not explicitly set, a setting of check_log_space or
8191 check_log_inodes was ignored.
8193 79. If a space check for the spool or log partitions fails, the incident is now
8194 logged. Of course, in the latter case the data may get lost...
8196 80. Added the %p formatting code to string_format() so that it can be used to
8197 print addresses in debug_print(). Adjusted all the address printing in the
8198 debugging in store.c to use %p rather than %d.
8200 81. There was a concern that a line of code in smtp_in.c could overflow a
8201 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8202 initially expressed, the concern was not well-founded, because trailing
8203 spaces are removed early. However, if the trailing spaces were followed by
8204 a NULL, they did not get removed, so the overflow was possible. Two fixes
8207 (a) I re-wrote the offending code in a cleaner fashion.
8208 (b) If an incoming SMTP command contains a NULL character, it is rejected
8211 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8212 runs initgroups(), so that if the Exim user is in any additional groups,
8213 they will be used during message reception.
8219 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob