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.
98 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
99 the call into OpenSSL to send a TLS Close was being repeated; this
100 resulted in the library waiting for the peer's Close. If that was never
101 sent we waited forever. Fix by tracking send calls.
107 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
108 related applications. Previously an "H" was used where available info
109 says that "M" should be, so change to match.
111 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
112 as arguments, so an implementation trying to copy these into a local
113 buffer was taking a taint-enforcement trap. Fix by using dynamically
114 created buffers. Similar fix for radius expansion condition.
116 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
117 reasonable, eg. to count headers. Fix by using dynamically created
118 buffers rather than a local. Do similar fixes for ACL actions "dcc",
119 "log_reject_target", "malware" and "spam"; the arguments are expanded
120 so could be handling tainted values.
122 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
123 broken the (no-op) support for this sendmail command. Restore it
124 to doing nothing, silently, and returning good status.
126 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
127 record path was given (or the default used) without a leading directory
128 path, an error occurred on trying to open it. Use the transport's working
131 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
132 transport. Previously it was the name on the DNS A-record; use instead
133 the head of the CNAME chain leading there (if there is one). This seems
134 to align better with RFC 6125.
136 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
137 smtp_accept_max_per_host allocated resources which were not released
138 when the limit was exceeded. This eventually crashed the daemon. Fix
139 by adding a release action in that path.
141 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
142 expanded; previously using tainted values was rejected. Fix by using
143 dynamically-created buffers.
145 JH/09 Relax restrictions on ACL verify condition needing access to message
146 headers. Previously they were only permitted in data and non-smtp ACLs;
147 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
148 not_blind, header_sender and header_names_ascii verification.
150 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
151 Previously a macro used one argument twice; when called with the
152 argument as an expression having side-effects, incorrect operation
153 resulted. Use an inlineable function.
155 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
156 held open for a verify callout. Previously this wan not accounted for
157 and a corrupt onward SMTP conversation resulted.
159 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
160 Previously they were embedded naively in the constructed address; when
161 needed, strip the quoting and quote the entire local_part.
162 Also make the inbound_srs expansion condition handle quoting.
164 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
165 excluded, not matching the documentation.
167 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
168 was given for the sqlite_dbfile a trap resulted.
170 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
171 "name" argument resulted in a trap. There is no reason to disallow such;
172 this was a coding error.
174 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
175 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
176 spent suspended, ignoring the POSIX definition. Previously we assumed
177 it did and a constant offset from real time could be used as a correction.
178 Change to using the same clock source for the start-of-message and the
179 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
180 exists, just to get a clock slightly more aligned to reality.
182 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
183 RFC says it is optional some validators care. The missing char was not
184 intended but triggered by a line-wrap alignment. Discovery and fix by
185 Guillaume Outters, hacked on by JH.
187 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
188 name being quoted was tainted a trap would be taken. Fix by using
189 dynamically created buffers. The routine could have been called by a
190 rewrite with the "h" flag, by using the "-F" command-line option, or
191 by using a "name=" option on a control=submission ACL modifier.
193 JH/19 SPF: change the Authentication-Results expansion component to give
194 smtp.helo when the sender domain is empty. Previously it gave
197 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
198 not in 127.0.0.0/8 to help in spotting list domains taken over by a
199 domain-parking registrar.
201 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
202 Previously when a whitespace character was specified it was not inserted
203 after removing the newline.
205 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
206 the domain part of the recipient address. This overrides any tls_sni
207 option set, which was previously used.
209 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
212 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
213 is_tainted() had an off-by-one error in the overenthusiastic direction.
214 Find and fix by Gavan. Although NetBSD is not a supported platform for
215 4.94 this bug could affect other platforms.
217 PP/01 Fix default prime selection to be consistent.
218 One path used ike23 still, instead of exim.dev.20160529.3; now both
219 execution flows will use the same DH primes (currently
220 exim.dev.20160529.3).
222 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
223 option in smtp transport, to match the documentation. Previously
224 verification was not being done in some cases where it should have been.
226 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
227 than one server was defined and depending on the platform memory layout
228 details, an internal consistency trap could be hit while walking the list
231 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
232 files. The value is used by the authresults expansion item. Previously
233 if this was used in a router or transport, a crash could result.
235 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
236 have a sufficient incidence of EINTR returns from select that an
237 interaction with other operations done by the main daemon loop exposed
238 a bug in the error-handling. This was benign apart from the log
241 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
242 for consistency with delivered "=>" and failed "**" lines. While we're
243 there, handle PRX and TFO.
245 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
246 applied. This resulted, if any header-line rewrite rules were configured,
247 in a panic-log triggerable by sending a message with a long address in
248 a header. Fix by increasing the arbitrary limit to larger than a single
249 (dewrapped) 5322 header line maximum size.
251 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
252 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
253 RFC 6648 which deprecates X- options in protocols as a general practice.
254 Changeover between the implementations is handled by the mechanisms
257 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
258 a remote which uses callout/hold. Previously the local was queued.
260 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
263 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
264 log line. The asterisk is now consistently appended to the remote IP
265 (and port, if given), and will also be provided on defer and fail log
266 lines. Previously it could be placed on the local IP if that was being
267 logged, and was only provided on delivery lines.
269 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
271 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
272 authenticator client_send option. Previously the next char, after a pair
273 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
274 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
275 documentation. There is still no way to get a leading ^ immediately
276 after a NUL (ie. for the password of a PLAIN method authenticator.
278 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
279 files. For bad sizes read, delete the record and whine to paniclog.
281 JH/38 When logging an AUTH failure, as server, do not include sensitive
282 information. Previously, the credentials would be included if given
283 as part of the AUTH command line and an ACL denied authentication.
285 JH/39 Bug 2691: fix $local_part_data. When the matching list element
286 referred to a file, bad data was returned. This likely also affected
289 JH/40 The gsasl authenticator now supports caching of the salted password
290 generated by the client-side implementation. This required the addition
291 of a new variable: $auth4.
293 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
294 left undeleted; the attempt to re-create it then failed - resulting in
295 the usual "SIGHUP tp have daemon reload configuration" to not work.
296 This affected any platform not supporting "abstract" Unix-domain
297 sockets (i.e. not Linux).
299 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
300 recipients" response to RCPT in a later response, with a 250. The
301 previous coding assumed this would not happen, and under PIPELINING
302 would result in both lost and duplicate recipients for a message.
304 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
305 Previously the weighting was incorrectly applied. Similar fix for socks
306 proxies. Found and fixed by Heiko Schlichting.
308 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
309 not handle sub-lists included using the +namedlist syntax. While
310 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
311 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
314 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
315 Previously the permanent pool was used, so the sore could not be freed.
316 This meant a connection with many messages would use continually-growing
319 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
320 per-pool so as not to waste too much space. Previously a constant size
321 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
322 DOS attacks harder. The cost is wasted memory use in the larger blocks.
324 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
325 same space repeatedly, and a smaller process footprint.
327 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
328 for GnuTLS. Previously the name started "%s" which, while not a bug,
329 looked as if if might be one.
331 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
332 "continued connection") make the $proxy_* variables available. Previously
333 the information was not passed across the exec() call for subsequent
334 transport executions. This also mean that the log lines for the
335 messages can show the proxy information.
337 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
338 receive time. With modern systems the difference is significant.
339 The historical behaviour can be restored by disabling (a new) log_selector
340 "queue_time_exclusive".
342 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
343 content) were specifically tested for. Now, also cover expansions
344 resulting in ACL names and inline ACL content.
346 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
347 making it unusable in complex expressions.
349 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
350 fail with a taint-check on the spoolfile name, and leave the message
353 HS/01 Enforce absolute PID file path name.
355 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
357 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
359 PP/02 Bug 2643: Correct TLS DH constants.
360 A missing NUL termination in our code-generation tool had led to some
361 incorrect Diffie-Hellman constants in the Exim source.
362 Reported by kylon94, code-gen tool fix by Simon Arlott.
364 PP/03 Impose security length checks on various command-line options.
365 Fixes CVE-2020-SPRSS reported by Qualys.
367 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
368 better. Reported by Qualys.
370 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
371 providing a particularly obnoxious sender full name.
374 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
376 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
379 PP/08 Change default for recipients_max from unlimited to 50,000.
381 PP/09 Fix security issue with too many recipients on a message (to remove a
382 known security problem if someone does set recipients_max to unlimited,
383 or if local additions add to the recipient list).
384 Fixes CVE-2020-RCPTL reported by Qualys.
386 PP/10 Fix security issue in SMTP verb option parsing
387 Fixes CVE-2020-EXOPT reported by Qualys.
389 PP/11 Fix security issue in BDAT state confusion.
390 Ensure we reset known-good where we know we need to not be reading BDAT
391 data, as a general case fix, and move the places where we switch to BDAT
392 mode until after various protocol state checks.
393 Fixes CVE-2020-BDATA reported by Qualys.
395 HS/03 Die on "/../" in msglog file names
397 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
398 the Exim runtime user are allowed to create files.
400 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
403 QS/03 When reading the output from interpreted forward files we do not
404 pass the pipe between the parent and the interpreting process to
405 executed child processes (if any).
407 QS/04 Always die if requested from internal logging, even is logging is
410 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
411 the API; compilation noo longer completes with DMARC support included.
412 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
413 been reported on other platforms.
415 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
417 JH/56 Make the majority of info read from config files readonly, for defence-in-
418 depth against exploits. Suggestion by Qualys.
419 Not supported on Solaris 10.
421 JH/57 Fix control=fakreject for a custom message containing tainted data.
422 Previously this resulted in a log complaint, due to a re-expansion present
423 since fakereject was originally introduced.
425 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
426 modifier was given, a loop resulted.
428 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
429 pipelined SMTP command followed the BDAT LAST then it would be
430 incorrectly treated as part of the message body, causing a verification
433 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
434 require looser validation rules than those for 821-level addresses,
435 which only permit IP addresses.
441 JH/01 Avoid costly startup code when not strictly needed. This reduces time
442 for some exim process initialisations. It does mean that the logging
443 of TLS configuration problems is only done for the daemon startup.
445 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
447 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
448 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
451 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
452 DKIM signing is being done. Previously a transport_filter would always
453 disable CHUNKING, falling back to traditional DATA.
455 JH/05 Regard command-line recipients as tainted.
457 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
459 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
460 PAM library frees one of the arguments given to it, despite the
461 documentation. Therefore a plain malloc must be used.
463 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
464 on-stack buffers were used, resulting in a taint trap when DSN information
465 copied from a received message was written into the buffer.
467 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
468 the ordering of its ARC headers. This caused a crash.
470 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
471 a new record was being constructed with information from the peer, a trap
474 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
475 installation would get error messages from DMARC verify, when it hit the
476 nonexistent file indicated by the default. Distros wanting DMARC enabled
477 should both provide the file and set the option.
478 Also enforce no DMARC verification for command-line sourced messages.
480 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
481 could, depending on the platform, hang at the STARTTLS response.
483 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
484 message on a connection. Previously if one message had ARC headers and
485 the following one did not, a crash could result when adding an
486 Authentication-Results: header.
488 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
489 Exim main code and Exim-related utities. The introduction of taint
490 tracking also did many adjustments to string handling. Since then, eximon
491 frequently terminated with an assert failure.
493 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
494 check for 452 responses. This slightly helps the inefficieny of doing
495 a large alias-expansion into a recipient-limited target. The max_rcpt
496 transport option still applies (and at the current default, will override
497 the new feature). The check is done for either cause of synch, and forces
498 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
499 connection. The new facility is not tunable at this time.
501 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
502 library live data was being used, so the results became garbage. Make
503 copies while it is still usable.
505 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
506 on delivery deferred (==) and failed (**) lines (if a delivery was
507 attemtped). Previously it was only on completion (=>) lines.
509 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
510 for the expansion of the server_scram_iter and server_scram_salt options.
512 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
513 are now specifically given a NO_DATA response without hitting the system
514 resolver. The library goes on to do the now-standard TXT lookup.
515 Use of dnsdb lookups is not affected.
517 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
518 only retrieve the errormessage once. Previously two calls to dlerror()
519 were used, and the second one (for mainlog/paniclog) retrieved null
522 JH/20 Taint checking: disallow use of tainted data for
523 - the appendfile transport file and directory options
524 - the pipe transport command
525 - the autoreply transport file, log and once options
526 - file names used by the redirect router (including filter files)
528 - paths used by single-key lookups
529 Previously this was permitted.
531 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
532 adjusted the size of a major service buffer; this failed because the
533 buffer was in use at the time. Change to a compile-time increase in the
534 buffer size, when this authenticator is compiled into exim.
536 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
537 previous fast-mode was untenable in the face of glibs using mmap to
538 support larger malloc requests.
540 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
541 New values supported, if defined on system where compiled:
542 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
543 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
545 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
546 running a limited number of proceses in parallel, a benefit is gained. The
547 amount varies with the platform hardware and load. The use of the option
548 queue_run_in_order means we cannot do this, as ordering becomes
551 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
552 had introduced a string-copy (for ensuring NUL-termination) which was not
553 appropriate for that case, which can include embedded NUL bytes in the
554 block of data. Investigation showed the copy to actually be needless, the
555 data being length-specified.
557 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
558 done during a receiving connection, and both used TLS, global info was
559 used rather than per-connection info for tracking the state of data
560 queued for transmission. This could result in a connection hang.
562 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
563 Previously, when delivering serveral messages down a single connection
564 only the first would provide a SIZE. This was due to the size information
565 not being properly tracked.
567 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
568 TAI (at 37 seconds currently), pretend to be in UTC for time-related
569 expansion and logging. Previously, spurious values such as a future
570 minute could be seen.
572 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
573 it could crash from a null-deref. This could also affect the
574 ${addresses: } operator and ${readsock } item.
576 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
577 message following a mime one, the variable was not reset.
579 JH/30 When an pipelined-connect fails at the first response, assume incorrect
580 cached capability (perhaps the peer reneged?) and immediately retry in
583 JH/31 Fix spurious detection of timeout while writing to transport filter.
585 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
586 an attempt to copy the string was made before checking it.
588 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
589 taint of the lookup key was maintained; we now regard the presence in the
590 filesystem as sufficient validation.
592 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
593 argument is supplied.
595 JH/35 The dsearch lookup now requires that the directory is an absolute path.
596 Previously this was not checked, and nonempty relative paths made an
597 access under Exim's current working directory.
599 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
600 Previously no event was raised.
602 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
603 parameter supplied by the sender MAIL FROM command. Previously it was
604 ignored, and only the check_spool_space option value for the required
607 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
608 the size of the signing public-key. Previously it was instead giving
609 the size of the signature hash.
611 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
612 the default. See the (new) dkim_verify_min_keysizes option.
614 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
615 and an ACL use a lookup for checking either the local_part or domain,
616 stale data could be accessed. Ensure that variable references are
617 dropped between messages.
619 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
620 by the client was not checked as pointing within response data before
621 being used. A malicious client could thus cause an out-of-bounds read and
622 possibly gain authentication. Fix by adding the check.
624 JH/42 Internationalisation: change the default for downconversion in the smtp
625 transport to be "if needed". Previously it was "as previously set" for
626 the message, which usually meant "if needed" for message-submission but
627 "no" for everything else. However, MTAs have been seen using SMTPUTF8
628 even when the envelope addresses did not need it, resulting in forwarding
629 failures to non-supporting MTAs. A downconvert in such cases will be
630 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
631 transport. The change does mean that addresses needing conversion will
632 be converted when previously a delivery failure would occur.
634 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
635 response was received it would be used unchecked in a fail-DSN, violating
636 standards on line-length limits. Truncate if needed.
638 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
639 doesn't work. (Additionally add a new main config option to configure the
646 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
647 side, to decode a TLS 1.3 packet capture.
649 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
650 Previously the default library behaviour applied, sending two, each in
653 JH/03 Debug output for ACL now gives the config file name and line number for
656 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
658 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
660 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
661 buffer overrun for (non-chunking) other transports.
663 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
664 TLS1.3, means that a server rejecting a client certificate is not visible
665 to the client until the first read of encrypted data (typically the
666 response to EHLO). Add detection for that case and treat it as a failed
667 TLS connection attempt, so that the normal retry-in-clear can work (if
668 suitably configured).
670 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
671 and/or domain. Found and fixed by Jason Betts.
673 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
674 configuration). If a CNAME target was not a wellformed name pattern, a
677 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
678 the OS reports them interleaved with other addresses.
680 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
681 used both for input and for a verify callout, both encrypted, SMTP
682 responses being sent by the server could be lost. This resulted in
683 dropped connections and sometimes bounces generated by a peer sending
686 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
687 string. Previously it was possible to cause undefined behaviour in a
688 library routine (usually a crash). Found by "zerons".
690 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
693 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
694 API was removed, so update to use the newer ones.
696 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
697 any timeout set, is taking a long time. Previously we would hang on to a
698 rotated logfile "forever" if the input was arriving with long gaps
699 (a previous attempt to fix addressed lack, for a long time, of initial
702 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
703 shared (NFS) environment. The length of the tempfile name is now
704 4 + 16 ("hdr.$message_exim_id") which might break on file
705 systems which restrict the file name length to lower values.
708 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
709 shared (NFS) environment.
711 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
712 did for all versions <4.90). Notably -M, -m, --invert, -I may be
715 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
716 on some platforms for bit 31.
718 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
719 to changes apparently associated with TLS1.3 handling some of the APIs
720 previously used were either nonfunctional or inappropriate. Strings
721 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
722 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
723 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
724 This affects log line X= elements, the $tls_{in,out}_cipher variables,
725 and the use of specific cipher names in the encrypted= ACL condition.
727 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
729 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
730 verification result was not updated unless hosts_require_ocsp applied.
732 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
733 queue_list_requires_admin set to false, non-admin users were denied the
736 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
737 directory-of-certs mode. Previously they were advertised despite the
740 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
741 A single TCP connection by a client will now hold a TLS connection open
742 for multiple message deliveries, by default. Previously the default was to
745 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
746 default. If built with the facility, DANE will be used. The facility
747 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
749 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
750 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
751 must be defined and you must still, unless you define DISABLE_TLS, manage
752 the the include-dir and library-file requirements that go with that
753 choice. Non-TLS builds are still supported.
755 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
758 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
759 default. If the platform supports and has the facility enabled, it will
760 be requested on all coneections.
762 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
763 controlled by the build-time option SUPPORT_PIPE_CONNECT.
765 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
767 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
768 success-DSN messages. Previously the From: header was always the default
769 one for these; the option was ignored.
771 JH/28 Fix the timeout on smtp response to apply to the whole response.
772 Previously it was reset for every read, so a teergrubing peer sending
773 single bytes within the time limit could extend the connection for a
774 long time. Credit to Qualsys Security Advisory Team for the discovery.
776 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
777 delivery address, which leaked information of the results of local
778 forwarding. Change to the original envelope recipient address, per
781 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
782 requested. Previously not bounce was generated and a log entry of
783 error ignored was made.
785 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
787 JH/32 Introduce a general tainting mechanism for values read from the input
788 channel, and values derived from them. Refuse to expand any tainted
789 values, to catch one form of exploit.
791 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
792 was unused and the unexpanded text used for the test. Found and
793 fixed by Ruben Jenster.
795 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
796 an attempt to use a TLS library read routine dereffed a nul pointer,
799 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
800 them in our smtp response.
802 JH/36 Have the general router option retry_use_local_part default to true when
803 any of the restrictive preconditions are set (to anything). Previously it
804 was only for check_local user. The change removes one item of manual
805 configuration which is required for proper retries when a remote router
806 handles a subset of addresses for a domain.
808 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
809 link count into consideration.
811 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
812 caused the extension of big_buffer, the following lines were ignored.
814 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
815 accordance with RFC 2308. Previously there was no expiry, so a longlived
816 receive process (eg. due to ACL delays) versus a short SOA value could
819 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
821 JH/39 Promote DMARC support to mainline.
823 JH/40 Bug 2452: Add a References: header to DSNs.
825 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
826 parameters. The relevant library call is documented as "Deprecated: This
827 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
828 3.6.0, DH parameters are negotiated following RFC7919."
830 HS/06 Change the default of dnssec_request_domains to "*"
832 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
833 carried on and emitted a BDAT command, even when PIPELINING was not
836 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
837 buffer was used for the filename, resulting in a trap when tainted
838 arguments (eg. $domain) were used.
840 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
841 recommended to avoid a possible server-load attack. The feature can be
842 re-enabled via the openssl_options main cofiguration option.
844 JH/45 local_scan API: documented the current smtp_printf() call. This changed
845 for version 4.90 - adding a "more data" boolean to the arguments.
846 Bumped the ABI version number also, this having been missed previously;
847 release versions 4.90 to 4.92.3 inclusive were effectively broken in
848 respect of usage of smtp_printf() by either local_scan code or libraries
849 accessed via the ${dlfunc } expansion item. Both will need coding
850 adjustment for any calls to smtp_printf() to match the new function
851 signature; a FALSE value for the new argument is always safe.
853 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
854 the file-offset (which the Linux syscall does, and exim expects); this
855 resulted in an indefinite loop.
857 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
858 to do ARC verification. The Authentication-Results: header line added
859 by the configuration then had no ARC item.
861 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
862 an incorrect mode was used for reading the result, resulting in it being
869 JH/01 Remove code calling the customisable local_scan function, unless a new
870 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
872 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
873 non-signal-safe functions being used.
875 JH/03 Bug 2269: When presented with a received message having a stupidly large
876 number of DKIM-Signature headers, disable DKIM verification to avoid
877 a resource-consumption attack. The limit is set at twenty.
879 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
880 report of oldest_pass in ${authres } in consequence, and separate out
881 some descriptions of reasons for verification fail.
883 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
884 files in the spool were present and unlocked. A queue-runner could spot
885 them, resulting in a duplicate delivery. Fix that by doing the unlock
886 after the unlink. Investigation by Tim Stewart. Take the opportunity to
887 add more error-checking on spoolfile handling while that code is being
890 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
891 No known attacks, no CVE, this is defensive hardening.
893 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
894 a queue-runner could start a delivery while other operations were ongoing.
895 Cutthrough delivery was a common victim, resulting in duplicate delivery.
896 Found and investigated by Tim Stewart. Fix by using the open message data
897 file handle rather than opening another, and not locally closing it (which
898 releases a lock) for that case, while creating the temporary .eml format
899 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
901 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
902 $sender_verify_failure/$recipient_verify_failure to "random".
904 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
907 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
908 Previously this would segfault.
910 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
913 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
914 like zero, since the resolver should be doing this for us, But we need one
915 as a CNAME but no MX presence gets the CNAME returned; we need to check
916 that doesn't point to an MX to declare it "no MX returned" rather than
917 "error, loop". A new main option is added so the older capability of
918 following some limited number of chain links is maintained.
920 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
922 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
923 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
924 now-deprecated earlier definitions used only specified the range up to TLS
925 1.2 (in the older-version library docs).
927 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
929 JH/15 Rework TLS client-side context management. Stop using a global, and
930 explicitly pass a context around. This enables future use of TLS for
931 connections to service-daemons (eg. malware scanning) while a client smtp
932 connection is using TLS; with cutthrough connections this is quite likely.
934 JH/16 Fix ARC verification to do AS checks in reverse order.
936 JH/17 Support a "tls" option on the ${readsocket } expansion item.
938 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
939 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
940 Previously the "utf8" would be re-prepended for every additional message.
942 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
943 Previously thery were accepted, resulting in issues when attempting to
944 forward messages to a non-supporting MTA.
946 PP/02 Let -n work with printing macros too, not just options.
948 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
949 one parent address was copied, and bogus data was used at delivery-logging
950 time. Either a crash (after delivery) or bogus log data could result.
951 Discovery and analysis by Tim Stewart.
953 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
954 Previously if the string ended mid-character, we did not insert the
955 promised '?' replacement.
957 PP/04 Documentation: current string operators work on bytes, not codepoints.
959 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
960 should pack well giving a smaller memory footprint so better caching and
961 therefore performance. Group the declarations where this can't be done so
962 that the byte-sized flag variables are not interspersed among pointer
963 variables, giving a better chance of good packing by the compiler.
965 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
966 non-null, to avoid issues with sites running BATV. Previously reports were
967 sent with an empty envelope sender so looked like bounces.
969 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
970 The ignore_error flag wasn't being returned from the filter subprocess so
971 was not set for later routers. Investigation and fix by Matthias Kurz.
973 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
974 and a msg:complete for the whole, when a message is manually removed using
975 -Mrm. Developement by Matthias Kurz, hacked on by JH.
977 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
978 a "Gnu special" function, asprintf() in the DB utility binary builds; I
979 hope that is portable enough.
981 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
982 requiring a known-CA anchor certificate; make it now rely entirely on the
983 TLSA as an anchor. Checking the name on the leaf cert against the name
984 on the A-record for the host is still done for TA (but not for EE mode).
986 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
987 included in delivery lines for non-proxied connections, when compiled with
988 SUPPORT_SOCKS and running with proxy logging enabled.
990 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
991 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
992 move the existing event to fire before the normal logging of message
993 failure so that custom logging is bracketed by normal logging.
995 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
996 msg:fail:internal event. Developement by Matthias Kurz.
998 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
999 far too small for todays use of crypto signatures stored there. Go all
1000 the way to the max DNS message size of 64kB, even though this might be
1001 overmuch for IOT constrained device use.
1003 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1004 copy a string over itself. The library routine is documented as not
1005 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1007 JH/32 For main options check_spool_space and check_inode_space, where the
1008 platform supports 64b integers, support more than the previous 2^31 kB
1009 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1010 the previous G, M, k.
1012 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1013 $authenticated_fail_id variable on authentication failure. Previously
1016 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1017 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1018 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1019 GNUTLS_SEC_PARAM_MEDIUM.
1021 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1022 side. Previously we would continue as if no SNI had been received.
1024 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1025 "<" character, which should be a list-separator specification, we walked
1026 off past the nul-terimation.
1028 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1029 causes) even when the retry time is not yet met. Previously they were
1030 not, meaning that when (say) an account was over-quota and temp-rejecting,
1031 and multiple senders' messages were queued, only one sender would get
1032 notified on each configured delay_warning cycle.
1034 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1036 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1037 cipher-suites, an error can be left on the stack even for a succeeding
1038 accept; this results in impossible error messages when a later operation
1041 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1042 return error codes indicating retry. Under TLS1.3 this becomes required.
1044 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1045 it only wrote the new authenticators, resulting in a lack of tracking of
1046 peer changes of ESMTP extensions until the next cache flush.
1048 JH/41 Fix the loop reading a message header line to check for integer overflow,
1049 and more-often against header_maxsize. Previously a crafted message could
1050 induce a crash of the recive process; now the message is cleanly rejected.
1052 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1053 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1059 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1060 When redis_servers is set to a list of > 1 element, and the Redis servers
1061 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1062 case of MOVED into a DEFER case instead, thus moving the query onto the
1063 next server in the list. For a cluster of N elements, all N servers must
1064 be defined in redis_servers.
1066 GF/02 Catch and remove uninitialized value warning in exiqsumm
1067 Check for existence of @ARGV before looking at $ARGV[0]
1069 JH/01 Replace the store_release() internal interface with store_newblock(),
1070 which internalises the check required to safely use the old one, plus
1071 the allocate and data copy operations duplicated in both (!) of the
1072 extant use locations.
1074 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1075 modifier. This matches the restriction on the commandline.
1077 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1078 Previously only the last row was returned.
1080 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1081 we assumed that tags in the header were well-formed, and parsed the
1082 element content after inspecting only the first char of the tag.
1083 Assumptions at that stage could crash the receive process on malformed
1086 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1087 While running the DKIM ACL we operate on the Permanent memory pool so that
1088 variables created with "set" persist to the DATA ACL. Also (at any time)
1089 DNS lookups that fail create cache records using the Permanent pool. But
1090 expansions release any allocations made on the current pool - so a dnsdb
1091 lookup expansion done in the DKIM ACL releases the memory used for the
1092 DNS negative-cache, and bad things result. Solution is to switch to the
1093 Main pool for expansions.
1094 While we're in that code, add checks on the DNS cache during store_reset,
1095 active in the testsuite.
1096 Problem spotted, and debugging aided, by Wolfgang Breyha.
1098 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1099 When none of the hosts presented to a transport match an already-open
1100 connection, close it and proceed with the list. Previously we would
1101 queue the message. Spotted by Lena with Yahoo, probably involving
1104 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1105 Previously a spurious "250 OK id=" response was appended to the proper
1108 JH/08 The "support for" informational output now, which built with Content
1109 Scanning support, has a line for the malware scanner interfaces compiled
1110 in. Interface can be individually included or not at build time.
1112 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1113 by the template makefile "src/EDITME". The "STREAM" support for an older
1114 ClamAV interface method is removed.
1116 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1117 rows affected is given instead).
1119 JH/11 The runtime Berkeley DB library version is now additionally output by
1120 "exim -d -bV". Previously only the compile-time version was shown.
1122 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1123 SMTP connection. Previously, when one had more recipients than the
1124 first, an abortive onward connection was made. Move to full support for
1125 multiple onward connections in sequence, handling cutthrough connection
1126 for all multi-message initiating connections.
1128 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1129 routers. Previously, a multi-recipient message would fail to match the
1130 onward-connection opened for the first recipient, and cause its closure.
1132 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1133 a timeout on read on a GnuTLS initiating connection, resulting in the
1134 initiating connection being dropped. This mattered most when the callout
1135 was marked defer_ok. Fix to keep the two timeout-detection methods
1138 JH/15 Relax results from ACL control request to enable cutthrough, in
1139 unsupported situations, from error to silently (except under debug)
1140 ignoring. This covers use with PRDR, frozen messages, queue-only and
1143 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1145 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1146 metadata, resulting in a crash in free().
1148 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1149 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1150 Broken also in d185889f4, with init system revamp.
1152 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1153 Previously we abruptly closed the connection after reading a malware-
1154 found indication; now we go on to read the "scan ok" response line,
1157 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1158 ACL. Previously, a crash would result.
1160 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1161 macro text after a replacement (previously it was only once per line) and
1162 by skipping builtin macros when searching for an uppercase lead character.
1164 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1165 for the build is renamed.
1167 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1168 was allocated for every new TLS startup, meaning one per message. Fix
1169 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1171 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1172 reported the original. Fix to report (as far as possible) the ACL
1173 result replacing the original.
1175 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1176 OpenSSL. Certificate information is loaded for every new TLS startup,
1177 and the resources needed to be freed.
1179 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1181 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1184 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1185 DATA response info to the (existing) per-recipient response info for
1186 the "C=" log element. It can have useful tracking info from the
1187 destination system. Patch from Simon Arlott.
1189 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1190 length value. Previously this would segfault.
1192 HS/02 Support Avast multiline protoocol, this allows passing flags to
1193 newer versions of the scanner.
1195 JH/28 Ensure that variables possibly set during message acceptance are marked
1196 dead before release of memory in the daemon loop. This stops complaints
1197 about them when the debug_store option is enabled. Discovered specifically
1198 for sender_rate_period, but applies to a whole set of variables.
1199 Do the same for the queue-runner and queue-list loops, for variables set
1200 from spool message files. Do the same for the SMTP per-message loop, for
1201 certain variables indirectly set in ACL operations.
1203 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1204 as a multi-recipient message from a mailinglist manager). The coding had
1205 an arbitrary cutoff number of characters while checking for more input;
1206 enforced by writing a NUL into the buffer. This corrupted long / fast
1207 input. The problem was exposed more widely when more pipelineing of SMTP
1208 responses was introduced, and one Exim system was feeding another.
1209 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1210 receiving system, and refused recipients seen by the sending system
1211 (propating to people being dropped from mailing lists).
1212 Discovered and pinpointed by David Carter.
1214 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1215 replaced by the ${authresults } expansion.
1217 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1219 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1220 allows proper process termination in container environments.
1222 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1223 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1225 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1226 and "err_perm", deprecated since 4.83 when the RFC-defined words
1227 "temperror" and "permerror" were introduced.
1229 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1230 transport-filters or DKIM-signing. The restriction was lost in the
1231 consolidation of verify-callout and delivery SMTP handling.
1232 Extend the restriction to also cover ARC-signing.
1234 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1235 in defer=pass mode supply a 450 to the initiator. Previously the message
1238 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1239 tls_require_ciphers is used as before.
1241 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1242 "pass_unscanned". Only tmpfails from the scanner are written to
1243 the paniclog, as they may require admin intervention (permission
1244 denied, license issues). Other scanner errors (like decompression
1245 bombs) do not cause a paniclog entry.
1247 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1248 Previously it was possible to log spurious information in receive log
1251 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1252 triggered odd behaviour from Outlook Express clients.
1254 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1257 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1258 since the IETF WG has not yet settled on that versus the original
1259 "bare" representation.
1261 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1262 Previously the millisecond value corrupted the output.
1263 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1264 corrupted the output.
1270 JH/01 Rework error string handling in TLS interface so that the caller in
1271 more cases is responsible for logging. This permits library-sourced
1272 string to be attached to addresses during delivery, and collapses
1273 pairs of long lines into single ones.
1275 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1276 during configuration. Wildcards are allowed and expanded.
1278 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1279 This permits better logging.
1281 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1282 do not close it down and have the child transport start it up again on
1283 the passed-on TCP connection. Instead, proxy the child (and any
1284 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1285 affected: the continued delivery log lines do not have any DNSSEC, TLS
1286 Certificate or OCSP information. TLS cipher information is still logged.
1288 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1289 identical IP addresses on different listening ports. Will also affect
1292 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1293 add noisy ifdef guards to special-case this sillyness.
1294 Patch from Bernd Kuhls.
1296 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1297 than 255 are no longer allowed.
1299 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1300 Disable the session-cache too, which might reduce our load. Since we
1301 currrectly use a new context for every connection, both as server and
1302 client, there is no benefit for these.
1303 GnuTLS appears to not support tickets server-side by default (we don't
1304 call gnutls_session_ticket_enable_server()) but client side is enabled
1305 by default on recent versions (3.1.3 +) unless the PFS priority string
1308 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1309 <https://reproducible-builds.org/specs/source-date-epoch/>.
1311 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1312 the check for any unsuccessful recipients did not notice the limit, and
1313 erroneously found still-pending ones.
1315 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1316 MSG_MORE. Only in-clear (not on TLS connections).
1318 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1319 transport-filter is involved we can buffer the headers in memory for
1320 creating the signature, and read the spool data file once for the
1321 signature and again for transmission.
1323 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1324 4.77 as the kernel support then wasn't solid, having issues in 64bit
1325 mode. Now, it's been long enough. Add support for FreeBSD also.
1327 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1328 case where the routing stage had gathered several addresses to send to
1329 a host before calling the transport for the first, we previously failed
1330 to close down TLS in the old transport process before passing the TCP
1331 connection to the new process. The new one sent a STARTTLS command
1332 which naturally failed, giving a failed delivery and bloating the retry
1333 database. Investigation and fix prototype from Wolfgang Breyha.
1335 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1336 false-negatives in the check that the sender had not preempted a response
1337 or prompt from Exim (running as a server), due to that code's lack of
1338 awareness of the SMTP input buffering.
1340 PP/04 Add commandline_checks_require_admin option.
1341 Exim drops privileges sanely, various checks such as -be aren't a
1342 security problem, as long as you trust local users with access to their
1343 own account. When invoked by services which pass untrusted data to
1344 Exim, this might be an issue. Set this option in main configuration
1345 AND make fixes to the calling application, such as using `--` to stop
1348 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1349 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1350 responses to those, into a single TLS record each way (this usually means
1351 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1354 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1355 affects you only if you're dancing at the edge of the param size limits.
1356 If you are, and this message makes sense to you, then: raise the
1357 configured limit or use OpenSSL 1.1. Nothing we can do for older
1360 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1361 cmdline element to get the documented default one. Previously it was
1364 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1365 are unsuable for non-message-specific reasons - eg. connection timeout,
1366 banner-time rejection.
1368 JH/16 Fix logging of delivery remote port, when specified by router, under
1371 PP/06 Repair manualroute's ability to take options in any order, even if one
1372 is the name of a transport.
1375 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1377 JH/17 Change the list-building routines interface to use the expanding-string
1378 triplet model, for better allocation and copying behaviour.
1380 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1381 Previously it was constructed the first time a possibly-matching string
1382 was met in the configuration file input during startup; now it is done
1385 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1386 compatible one, to avoid the (poorly documented) possibility of a config
1387 file in the working directory redirecting the DB files, possibly correpting
1388 some existing file. CVE-2017-10140 assigned for BDB.
1390 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1391 cache-hot. Previously, although the result was properly cached, the
1392 initial verify call returned a defer.
1394 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1395 the main verify for receipient in uncached-mode.
1397 JH/22 Retire historical build files to an "unsupported" subdir. These are
1398 defined as "ones for which we have no current evidence of testing".
1400 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1401 if present. Previously it was ignored.
1403 JH/24 Start using specified-initialisers in C structure init coding. This is
1404 a C99 feature (it's 2017, so now considered safe).
1406 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1407 if was a fixed-sized field and bitmask ops via macros; it is now more
1410 PP/07 GitHub PR 56: Apply MariaDB build fix.
1411 Patch provided by Jaroslav Škarvada.
1413 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1414 during Coverity cleanups [4.87 JH/47]
1415 Diagnosis and fix provided by Michael Fischer v. Mollard.
1417 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1418 the right size to place the terminating semicolon on its own folded
1419 line, the header hash was calculated to an incorrect value thanks to
1420 the (relaxed) space the fold became.
1422 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1423 and confused the parent.
1425 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1426 which could crash as a result. This could lead to undeliverable messages.
1428 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1431 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1432 matters on fast-turnover and PID-randomising systems, which were getting
1433 out-of-order delivery.
1435 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1436 a possibly-overlapping copy. The symptom was that "Remote host closed
1437 connection in response to HELO" was logged instead of the actual 4xx
1440 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1441 Previously only that bufferd was discarded, resulting in SYMTP command
1444 JH/32 DKIM: when a message has multiple signatures matching an identity given
1445 in dkim_verify_signers, run the dkim acl once for each. Previously only
1446 one run was done. Bug 2189.
1448 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1449 "panic the current process" to "deliberately defer". The panic log is
1450 still written with the problem list name; the mail and reject logs now
1451 get a temp-reject line for the message that was being handled, saying
1452 something like "domains check lookup or other defer". The SMTP 451
1453 message is still "Temporary local problem".
1455 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1456 A crafted sequence of BDAT commands could result in in-use memory beeing
1457 freed. CVE-2017-16943.
1459 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1460 from SMTP input. Previously it was always done; now only done for DATA
1461 and not BDAT commands. CVE-2017-16944.
1463 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1464 to the message (such as an overlong header line). Previously this was
1465 not done and we did not exit BDAT mode. Followon from the previous item
1466 though a different problem.
1472 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1473 than -2003 did; needs libidn2 in addition to libidn.
1475 JH/02 The path option on a pipe transport is now expanded before use.
1477 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1478 Patch provided by "Björn", documentation fix added too.
1480 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1481 missing a wire-to-host endian conversion.
1483 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1484 close after a BDAT command line could be taken as a following command,
1485 giving a synch failure. Fix by only checking for synch immediately
1486 before acknowledging the chunk.
1488 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1489 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1490 macro. Patches provided by Josh Soref.
1492 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1493 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1494 by IANA as a service extension. However, John Klensin suggests that we
1497 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1498 the dkim code may be unix-mode line endings rather than smtp wire-format
1499 CRLF, so prepend a CR to any bare LF.
1501 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1502 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1504 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1505 the first were themselves being wrongly included in the feed into dkim
1506 processing; with most chunk sizes in use this resulted in an incorrect
1507 body hash calculated value.
1509 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1510 DKIM signature block, for verification. Although advised against by
1511 standards it is specifically not ruled illegal.
1513 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1515 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1516 missing a body hash (the bh= tag).
1518 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1519 It seems that HAProxy sends the Proxy Protocol information in clear and
1520 only then does a TLS startup, so do the same.
1522 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1523 TCP connections (such as for Spamd) unless the daemon successfully set
1524 Fast Open mode on its listening sockets. This fixes breakage seen on
1525 too-old kernels or those not configured for Fast Open, at the cost of
1526 requiring both directions being enabled for TFO, and TFO never being used
1527 by non-daemon-related Exim processes.
1529 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1530 endings, at least on the first header line. Try to canonify any that get
1531 past that check, despite the cost.
1533 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1534 now limited to an arbitrary five deep, while parsing addresses with the
1535 strip_excess_angle_brackets option enabled.
1537 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1538 instead leave the unprompted TLS handshake in socket buffer for the
1539 TLS library to consume.
1541 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1543 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1545 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1546 region in one operation, for speed, and this leaves assigned pointers
1547 dangling. Add checks run only under the testsuite which checks all
1548 variables at a store-reset and panics on a dangling pointer; add code
1549 explicitly nulling out all the variables discovered. Fixes one known
1550 bug: a transport crash, where a dangling pointer for $sending_ip_address
1551 originally assigned in a verify callout, is re-used.
1553 PP/06 Drop '.' from @INC in various Perl scripts.
1555 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1557 PP/08 Reduce a number of compilation warnings under clang; building with
1558 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1559 should be warning-free.
1561 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1563 HS/01 Fix portability problems introduced by PP/08 for platforms where
1564 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1566 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1567 chunk. This allows us to accept broken chunked messages. We need a more
1568 general solution here.
1570 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1571 already-broken messages in the queue.
1573 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1575 JH/19 Fix reference counting bug in routing-generated-address tracking.
1581 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1582 supports it and a size is available (ie. the sending peer gave us one).
1584 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1585 years of being deprecated). The replacements are the ACLs
1586 acl_smtp_mime and acl_not_smtp_mime.
1588 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1589 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1590 or even an in-clear connection were permitted. Now, if the host lookup
1591 was dnssec and dane was requested then the host is only used if the
1592 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1593 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1594 if one fails this test.
1595 This means that a poorly-configured remote DNS will make it incommunicado;
1596 but it protects against a DNS-interception attack on it.
1598 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1599 non-noisy when a race steals the message being considered.
1601 JH/05 If main configuration option tls_certificate is unset, generate a
1602 self-signed certificate for inbound TLS connections.
1604 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1605 in rewrites and routers.
1607 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1608 and logged a warning sing 4.83; now they are a configuration file error.
1610 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1611 (lacking @domain). Apply the same qualification processing as RCPT.
1613 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1615 JH/10 Support ${sha256:} applied to a string (as well as the previous
1618 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1619 a cutthrough deliver is pending, as we always want to make a connection.
1620 This also avoids re-routing the message when later placing the cutthrough
1621 connection after a verify cache hit.
1622 Do not update it with the verify result either.
1624 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1625 when routing results in more than one destination address.
1627 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1628 signing (which inhibits the cutthrough capability). Previously only
1629 the presence of an option was tested; now an expansion evaluating as
1630 empty is permissible (obviously it should depend only on data available
1631 when the cutthrough connection is made).
1633 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1634 the relevant preceding SMTP command did not note the pipelining mode.
1636 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1637 Previously they were not counted.
1639 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1640 as one having no matching records. Previously we deferred the message
1641 that needed the lookup.
1643 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1644 distinguished as "(=".
1646 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1647 for missing MX records. Previously it only worked for missing A records.
1649 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1651 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1652 after the data-go-ahead and data-ack. Patch from Jason Betts.
1654 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1655 even for a "none" policy. Patch from Tony Meyer.
1657 JH/22 Fix continued use of a connection for further deliveries. If a port was
1658 specified by a router, it must also match for the delivery to be
1661 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1662 When one of the recipients of a message was unsuitable for the connection
1663 (has no matching addresses), we lost track of needing to mark it
1664 deferred. As a result mail would be lost.
1666 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1668 JH/25 Decoding ACL controls is now done using a binary search; the source code
1669 takes up less space and should be simpler to maintain. Merge the ACL
1670 condition decode tables also, with similar effect.
1672 JH/26 Fix problem with one_time used on a redirect router which returned the
1673 parent address unchanged. A retry would see the parent address marked as
1674 delivered, so not attempt the (identical) child. As a result mail would
1677 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1678 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1679 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1682 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1683 Main config options check_{log,spool}_{inodes,space} are now
1684 100 inodes, 10MB unless set otherwise in the configuration.
1686 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1687 Previously it only applied to the main-section connection policy
1690 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1692 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1693 by me. Added RFC7919 DH primes as an alternative.
1695 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1696 are not in the system include path.
1698 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1699 GnuTLS, when a session startup failed (eg because the client disconnected)
1700 Exim did stdio operations after fclose. This was exposed by a recent
1701 change which nulled out the file handle after the fclose.
1703 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1704 signed directly by the cert-signing cert, rather than an intermediate
1705 OCSP-signing cert. This is the model used by LetsEncrypt.
1707 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1709 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1710 an incoming connection.
1712 HS/02 Bug 1802: Do not half-close the connection after sending a request
1715 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1716 fallback to "prime256v1".
1718 JH/34 SECURITY: Use proper copy of DATA command in error message.
1719 Could leak key material. Remotely exploitable. CVE-2016-9963.
1725 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1726 and 3.4.4 - once the server is enabled to respond to an OCSP request
1727 it does even when not requested, resulting in a stapling non-aware
1728 client dropping the TLS connection.
1730 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1731 support variable-length bit vectors. No functional change.
1733 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1734 The I= interface field on outgoing lines is now after the H= remote
1735 host field, same as incoming lines. There is a separate
1736 outgoing_interface log selector which allows you to disable the
1739 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1740 If not running log_selector +smtp_connection the mainlog would be held
1741 open indefinitely after a "too many connections" event, including to a
1742 deleted file after a log rotate. Leave the per net connection logging
1743 leaving it open for efficiency as that will be quickly detected by the
1744 check on the next write.
1746 HS/01 Bug 1671: Fix post transport crash.
1747 Processing the wait-<transport> messages could crash the delivery
1748 process if the message IDs didn't exist for some reason. When
1749 using 'split_spool_directory=yes' the construction of the spool
1750 file name failed already, exposing the same netto behaviour.
1752 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1753 mime_regex ACL conditions.
1755 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1756 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1757 to HELO, local diagnostic string.
1759 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1760 log line to debug. Even when configured with a tls authenticator many
1761 client connections are expected to not authenticate in this way, so
1762 an authenticate fail is not an error.
1764 HS/02 Add the Exim version string to the process info. This way exiwhat
1765 gives some more detail about the running daemon.
1767 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1768 matter for fast-change records such as DNSBLs.
1770 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1771 retry record, even if constant. There may be multiple transports with
1772 different interface settings and the retry behaviour needs to be kept
1775 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1777 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1779 JH/10 Bug 840: fix log_defer_output option of pipe transport
1781 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1782 hosts_randomize. This matters a lot when combined with mua_wrapper.
1784 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1785 ${quote_pgsql:<string>} operator.
1787 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1789 JH/14 Change header file name for freeradius-client. Relevant if compiling
1790 with Radius support; from the Gentoo tree and checked under Fedora.
1792 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1794 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1795 option transport_filter, meaning no filtering.
1797 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1799 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1800 defaults to "*" (all hosts). The variable is now available when not built
1801 with TLS, default unset, mainly to enable keeping the testsuite sane.
1802 If a server certificate is not supplied (via tls_certificate) an error is
1803 logged, and clients will find TLS connections fail on startup. Presumably
1804 they will retry in-clear.
1805 Packagers of Exim are strongly encouraged to create a server certificate
1806 at installation time.
1808 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1809 with the $config_file variable.
1811 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1812 in transport context, after the attempt, and per-recipient. The latter type
1813 is per host attempted. The event data is the error message, and the errno
1814 information encodes the lookup type (A vs. MX) used for the (first) host,
1815 and the trailing two digits of the smtp 4xx response.
1817 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1818 to write to mainlog (or rejectlog, paniclog) in the window between file
1819 creation and permissions/ownership being changed. Particularly affects
1820 installations where exicyclog is run as root, rather than exim user;
1821 result is that the running daemon panics and dies.
1823 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1825 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1826 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1827 "pri" and "weight". Note that the previous implicit priority given by the
1828 list order is no longer honoured.
1830 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1831 for DKIM processing.
1833 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1834 by defining SUPPORT_SOCKS.
1836 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1837 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1838 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1839 variables are renamed to proxy_{local,external}_{address,port}.
1841 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1842 for a build by defining SUPPORT_I18N
1844 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1845 of the query string, and make ${quote_redis:} do that quoting.
1847 JH/29 Move Events support from Experimental to mainline, enabled by default
1848 and removable for a build by defining DISABLE_EVENT.
1850 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1852 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1853 cached by the daemon.
1855 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1856 by defining LOOKUP_REDIS. The libhiredis library is required.
1858 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1859 keys are given for lookup.
1861 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1862 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1863 only supported when built with TLS support. The PolarSSL SHA routines
1864 are still used when the TLS library is too old for convenient support.
1866 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1867 openssl_options), for security. OpenSSL forces this from version 1.1.0
1868 server-side so match that on older versions.
1870 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1871 allocation for $value could be released as the expansion processing
1872 concluded, but leaving the global pointer active for it.
1874 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1875 and to use the domains and local_parts ACL conditions.
1877 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1878 incorrectly not doubled on cutthrough transmission, hence seen as a
1879 body-termination at the receiving system - resulting in truncated mails.
1880 Commonly the sender saw a TCP-level error, and retransmitted the message
1881 via the normal store-and-forward channel. This could result in duplicates
1882 received - but deduplicating mailstores were liable to retain only the
1883 initial truncated version.
1885 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1887 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1889 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1890 we're in there, support oversigning also; bug 1309.
1892 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1894 HS/04 Add support for keep_environment and add_environment options.
1896 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1897 either intentional arithmetic overflow during PRNG, or testing config-
1900 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1901 delivery resulted in actual delivery. Cancel cutthrough before DATA
1904 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1905 rejects a recipient: pass the reject to the originator.
1907 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1908 Many were false-positives and ignorable, but it's worth fixing the
1911 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1912 for the new environment-manipulation done at startup. Move the routines
1913 from being local to tls.c to being global via the os.c file.
1915 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1916 an extract embedded as result-arg for a map, the first arg for extract
1917 is unavailable so we cannot tell if this is a numbered or keyed
1918 extraction. Accept either.
1924 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1927 JH/02 The smtp transport option "multi_domain" is now expanded.
1929 JH/03 The smtp transport now requests PRDR by default, if the server offers
1932 JH/04 Certificate name checking on server certificates, when exim is a client,
1933 is now done by default. The transport option tls_verify_cert_hostnames
1934 can be used to disable this per-host. The build option
1935 EXPERIMENTAL_CERTNAMES is withdrawn.
1937 JH/05 The value of the tls_verify_certificates smtp transport and main options
1938 default to the word "system" to access the system default CA bundle.
1939 For GnuTLS, only version 3.0.20 or later.
1941 JH/06 Verification of the server certificate for a TLS connection is now tried
1942 (but not required) by default. The verification status is now logged by
1943 default, for both outbound TLS and client-certificate supplying inbound
1946 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1949 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1950 Status Notification (bounce) messages are now MIME format per RFC 3464.
1951 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1952 under the control of the dsn_advertise_hosts option, and routers may
1953 have a dsn_lasthop option.
1955 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1956 default, modifiable by a malware= option. The list separator for
1957 the options can now be changed in the usual way. Bug 68.
1959 JH/10 The smtp_receive_timeout main option is now expanded before use.
1961 JH/11 The incoming_interface log option now also enables logging of the
1962 local interface on delivery outgoing connections.
1964 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1965 if the interface and destination host and port all match.
1967 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1970 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1971 Patch from Andrew Lewis.
1973 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1974 now supports optional time-restrictions, weighting, and priority
1975 modifiers per server. Patch originally by <rommer@active.by>.
1977 JH/16 The spamd_address main option now supports a mixed list of local
1978 and remote servers. Remote servers can be IPv6 addresses, and
1979 specify a port-range.
1981 JH/17 Bug 68: The spamd_address main option now supports an optional
1982 timeout value per server.
1984 JH/18 Bug 1581: Router and transport options headers_add/remove can
1985 now have the list separator specified.
1987 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1990 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1993 JH/21 Support for the A6 type of dns record is withdrawn.
1995 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1996 rather than the verbs used.
1998 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1999 from 255 to 1024 chars.
2001 JH/24 Verification callouts now attempt to use TLS by default.
2003 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2004 are generic router options now. The defaults didn't change.
2006 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2007 Original patch from Alexander Shikoff, worked over by JH.
2009 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2010 files if ZCAT_COMMAND is not executable.
2012 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2014 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2016 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2017 Normally benign, it bites when the pair was led to by a CNAME;
2018 modern usage is to not canonicalize the domain to a CNAME target
2019 (and we were inconsistent anyway for A-only vs AAAA+A).
2021 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2023 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2024 when evaluating $sender_host_dnssec.
2026 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2027 $sender_helo_dnssec variable.
2029 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2031 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2033 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2035 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2036 documented as working, but never had. Support all but $spam_report.
2038 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2039 added for tls authenticator.
2041 HS/03 Add perl_taintmode main config option
2047 TL/01 When running the test suite, the README says that variables such as
2048 no_msglog_check are global and can be placed anywhere in a specific
2049 test's script, however it was observed that placement needed to be near
2050 the beginning for it to behave that way. Changed the runtest perl
2051 script to read through the entire script once to detect and set these
2052 variables, reset to the beginning of the script, and then run through
2053 the script parsing/test process like normal.
2055 TL/02 The BSD's have an arc4random API. One of the functions to induce
2056 adding randomness was arc4random_stir(), but it has been removed in
2057 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2058 function when detected.
2060 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2061 cause callback expansion.
2063 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2064 syntax errors in an expansion can be treated as a string instead of
2065 logging or causing an error, due to the internal use of bool_lax
2066 instead of bool when processing it.
2068 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2069 server certificates when making smtp deliveries.
2071 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2073 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2075 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2077 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2078 Merged patch from Sebastian Wiedenroth.
2080 JH/05 Fix results-pipe from transport process. Several recipients, combined
2081 with certificate use, exposed issues where response data items split
2082 over buffer boundaries were not parsed properly. This eventually
2083 resulted in duplicates being sent. This issue only became common enough
2084 to notice due to the introduction of connection certificate information,
2085 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2087 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2088 size buffer was used, resulting in syntax errors when an expansion
2091 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2092 version 3.3.6 or later.
2094 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2095 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2096 both become "event_action", the variables become $event_name, $event_data
2097 and $event_defer_errno. There is a new variable $verify_mode, usable in
2098 routers, transports and related events. The tls:cert event is now also
2099 raised for inbound connections, if the main configuration event_action
2102 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2103 early OCSP support, but no stapling (appears to be less than 1.0.0).
2105 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2106 server certificate names available under the smtp transport option
2107 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2110 JH/10 Time-related extraction expansions from certificates now use the main
2111 option "timezone" setting for output formatting, and are consistent
2112 between OpenSSL and GnuTLS compilations. Bug 1541.
2114 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2115 encoded parameter in the incoming message. Bug 1558.
2117 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2118 include certificate info, eximon was claiming there were spoolfile
2121 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2123 JH/14 Log delivery-related information more consistently, using the sequence
2124 "H=<name> [<ip>]" wherever possible.
2126 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2127 are problematic for Debian distribution, omit them from the release
2130 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2132 JH/16 Fix string representation of time values on 64bit time_t architectures.
2135 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2136 output list separator was used.
2141 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2142 checkers that were complaining about end of non-void function with no
2145 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2146 This was a regression introduced in 4.83 by another bugfix.
2148 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2150 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2151 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2157 TF/01 Correctly close the server side of TLS when forking for delivery.
2159 When a message was received over SMTP with TLS, Exim failed to clear up
2160 the incoming connection properly after forking off the child process to
2161 deliver the message. In some situations the subsequent outgoing
2162 delivery connection happened to have the same fd number as the incoming
2163 connection previously had. Exim would try to use TLS and fail, logging
2164 a "Bad file descriptor" error.
2166 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2167 utilities have not been installed.
2169 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2170 temporary space as the ACL may create new global variables.
2172 TL/01 LDAP support uses per connection or global context settings, depending
2173 upon the detected version of the libraries at build time.
2175 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2176 to extract and use the src ip:port in logging and expansions as if it
2177 were a direct connection from the outside internet. PPv2 support was
2178 updated based on HAProxy spec change in May 2014.
2180 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2182 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2183 Properly escape header and check for NULL return.
2185 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2188 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2190 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2191 characters in header names, implemented as a verify condition.
2192 Contributed by Michael Fischer v. Mollard.
2194 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2195 results permerror and temperror. Previous values are deprecated but
2196 still accepted. In a future release, err_perm and err_temp will be
2197 completely removed, which will be a backward incompatibility if the
2198 ACL tests for either of these two old results. Patch contributed by
2199 user bes-internal on the mailing list.
2201 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2203 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2204 selectors, in both main and reject logs.
2206 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2209 JH/07 Add malware type "sock" for talking to simple daemon.
2211 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2213 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2214 routers/transports under cutthrough routing.
2216 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2217 numbers. Touch up "bool" conditional to keep the same definition.
2219 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2221 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2223 JH/12 Expand items in router/transport headers_add or headers_remove lists
2224 individually rather than the list as a whole. Bug 1452.
2226 Required for reasonable handling of multiple headers_ options when
2227 they may be empty; requires that headers_remove items with embedded
2228 colons must have them doubled (or the list-separator changed).
2230 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2231 view the policy declared in the DMARC record. Currently, $dmarc_status
2232 is a combined value of both the record presence and the result of the
2235 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2237 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2238 dnslookup router and the smtp transport (applying to the forward
2241 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2242 of ldap servers used for a specific lookup. Patch provided by Heiko
2245 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2246 New variable $lookup_dnssec_authenticated for observability.
2248 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2249 Patch submitted by Lars Timman.
2251 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2253 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2254 Requires trusted mode and valid format message id, aborts otherwise.
2255 Patch contributed by Heiko Schlichting.
2257 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2258 certextract with support for various fields. Bug 1358.
2260 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2261 is requested by default, modifiable by smtp transport option
2264 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2265 operate on certificate variables to give certificate fingerprints
2266 Also new ${sha256:cert_variable}.
2268 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2270 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2271 Christian Aistleitner.
2273 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2275 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2276 file. Patch from Wolfgang Breyha.
2278 JH/25 Expand the coverage of the delivery $host and $host_address to
2279 client authenticators run in verify callout. Bug 1476.
2281 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2282 align with daemon_smtp_ports. Bug 72.
2284 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2285 support and error reporting did not work properly.
2287 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2288 and is readable. Patch from Andrew Colin Kissa.
2290 TL/14 Enhance documentation of ${run expansion and how it parses the
2291 commandline after expansion, particularly in the case when an
2292 unquoted variable expansion results in an empty value.
2294 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2296 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2297 by renaming away from it. Addresses GH issue 15, reported by
2300 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2302 TL/15 SECURITY: prevent double expansion in math comparison functions
2303 (can expand unsanitized data). Not remotely exploitable.
2310 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2312 PP/02 Make -n do something, by making it not do something.
2313 When combined with -bP, the name of an option is not output.
2315 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2318 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2319 $sender_host_name and config options to manage this, and basic check
2322 PP/05 DSCP support for outbound connections and control modifier for inbound.
2324 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2325 (Only plugin which currently uses this is kerberos4, which nobody should
2326 be using, but we should make it available and other future plugins might
2327 conceivably use it, even though it would break NAT; stuff *should* be
2328 using channel bindings instead).
2330 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2331 name; added for Sendmail compatibility; requires admin caller.
2332 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2333 just ignore it); requires trusted caller.
2334 Also parse but ignore: -Ac -Am -X<logfile>
2337 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2339 TL/02 Add +smtp_confirmation as a default logging option.
2341 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2342 Patch by Magnus Holmgren from 2007-02-20.
2344 TL/04 Bugzilla 1281 - Spec typo.
2345 Bugzilla 1283 - Spec typo.
2346 Bugzilla 1290 - Spec grammar fixes.
2348 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2350 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2352 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2353 mailing list by by Dmitry Isaikin.
2355 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2357 JH/02 Support "G" suffix to numbers in ${if comparisons.
2359 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2361 NM/01 Bugzilla 1197 - Spec typo
2362 Bugzilla 1196 - Spec examples corrections
2364 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2366 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2367 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2370 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2371 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2373 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2374 "acl {{name}{arg}...}", and optional args on acl condition
2377 JH/05 Permit multiple router/transport headers_add/remove lines.
2379 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2381 JH/07 Avoid using a waiting database for a single-message-only transport.
2382 Performance patch from Paul Fisher. Bugzilla 1262.
2384 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2387 JH/09 Add $headers_added variable, with content from use of ACL modifier
2388 add_header (but not yet added to the message). Bugzilla 199.
2390 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2391 Pulled from Bugzilla 817 by Wolfgang Breyha.
2393 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2395 (nb: this is the same fix as in Exim 4.80.1)
2397 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2400 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2403 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2405 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2406 advertise SMTP AUTH mechanism to us, instead of a generic
2407 protocol violation error. Also, make Exim more robust to bad
2408 data from the Dovecot auth socket.
2410 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2412 When a queue runner is handling a message, Exim first routes the
2413 recipient addresses, during which it prunes them based on the retry
2414 hints database. After that it attempts to deliver the message to
2415 any remaining recipients. It then updates the hints database using
2418 So if a recipient address works intermittently, it can get repeatedly
2419 deferred at routing time. The retry hints record remains fresh so the
2420 address never reaches the final cutoff time.
2422 This is a fairly common occurrence when a user is bumping up against
2423 their storage quota. Exim had some logic in its local delivery code
2424 to deal with this. However it did not apply to per-recipient defers
2425 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2427 This change adds a proper retry rule check during routing so that the
2428 final cutoff time is checked against the message's age. We only do
2429 this check if there is an address retry record and there is not a
2430 domain retry record; this implies that previous attempts to handle
2431 the address had the retry_use_local_parts option turned on. We use
2432 this as an approximation for the destination being like a local
2433 delivery, as in LMTP.
2435 I suspect this new check makes the old local delivery cutoff check
2436 redundant, but I have not verified this so I left the code in place.
2438 TF/02 Correct gecos expansion when From: is a prefix of the username.
2440 Test 0254 submits a message to Exim with the header
2444 When I ran the test suite under the user fanf2, Exim expanded
2445 the header to contain my full name, whereas it should have added
2446 a Resent-Sender: header. It erroneously treats any prefix of the
2447 username as equal to the username.
2449 This change corrects that bug.
2451 GF/01 DCC debug and logging tidyup
2452 Error conditions log to paniclog rather than rejectlog.
2453 Debug lines prefixed by "DCC: " to remove any ambiguity.
2455 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2457 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2458 Bug spotted by Jeremy Harris; was flawed since initial commit.
2459 Would have resulted in OCSP responses post-SNI triggering an Exim
2460 NULL dereference and crash.
2462 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2464 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2465 Bug detection, analysis and fix by Samuel Thibault.
2466 Bugzilla 1331, Debian bug #698092.
2468 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2470 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2471 Server implementation by Todd Lyons, client by JH.
2472 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2473 config variable "prdr_enable" controls whether the server
2474 advertises the facility. If the client requests PRDR a new
2475 acl_data_smtp_prdr ACL is called once for each recipient, after
2476 the body content is received and before the acl_smtp_data ACL.
2477 The client is controlled by both of: a hosts_try_prdr option
2478 on the smtp transport, and the server advertisement.
2479 Default client logging of deliveries and rejections involving
2480 PRDR are flagged with the string "PRDR".
2482 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2483 fclose(). Diagnosis by Todd Lyons.
2485 PP/17 Update configure.default to handle IPv6 localhost better.
2486 Patch by Alain Williams (plus minor tweaks).
2489 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2490 This is now consistent with GnuTLS, and is now documented: the
2491 previous undocumented portable approach to treating the option as
2492 unset was to force an expansion failure. That still works, and
2493 an empty string is now equivalent.
2495 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2496 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2497 not performing validation itself.
2499 PP/20 Added force_command boolean option to pipe transport.
2500 Patch from Nick Koston, of cPanel Inc.
2502 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2505 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2507 PP/21 Fix eximon continuous updating with timestamped log-files.
2508 Broken in a format-string cleanup in 4.80, missed when I repaired the
2509 other false fix of the same issue.
2510 Report and fix from Heiko Schlichting.
2513 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2514 Report from Prashanth Katuri.
2516 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2517 It's SecureTransport, so affects any MacOS clients which use the
2518 system-integrated TLS libraries, including email clients.
2520 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2521 using a MIME ACL for non-SMTP local injection.
2522 Report and assistance in diagnosis by Warren Baker.
2524 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2526 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2528 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2529 last id that failed so it may be referenced in subsequent ACL's.
2531 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2534 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2535 ldap library initialization, allowing self-signed CA's to be
2536 used. Also properly sets require_cert option later in code by
2537 using NULL (global ldap config) instead of ldap handle (per
2538 session). Bug diagnosis and testing by alxgomz.
2540 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2541 the src/util/ subdirectory.
2543 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2544 renamed to Transport Post Delivery Action by Jeremy Harris, as
2547 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2548 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2549 redis_servers = needs to be configured which will be used by the redis
2550 lookup. Patch from Warren Baker, of The Packet Hub.
2552 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2554 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2555 hostname or reverse DNS when processing a host list. Used suggestions
2556 from multiple comments on this bug.
2558 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2560 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2561 Missed a few lines, added it to make the runtest require no keyboard
2564 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2565 contains upper case chars. Make router use caseful_local_part.
2567 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2568 support when GnuTLS has been built with p11-kit.
2574 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2576 This, or similar/improved, will also be change PP/11 of 4.82.
2582 PP/01 Handle short writes when writing local log-files.
2583 In practice, only affects FreeBSD (8 onwards).
2584 Bugzilla 1053, with thanks to Dmitry Isaikin.
2586 NM/01 Bugzilla 949 - Documentation tweak
2588 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2591 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2593 PP/02 Implemented gsasl authenticator.
2595 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2597 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2598 `pkg-config foo` for cflags/libs.
2600 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2601 with rest of GSASL and with heimdal_gssapi.
2603 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2604 `pkg-config foo` for cflags/libs for the TLS implementation.
2606 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2607 properties get this fed in as external SSF. A number of robustness
2608 and debugging improvements to the cyrus_sasl authenticator.
2610 PP/08 cyrus_sasl server now expands the server_realm option.
2612 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2613 Patch by Jeremy Harris.
2615 PP/10 Added dbmjz lookup type.
2617 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2619 PP/12 MAIL args handles TAB as well as SP, for better interop with
2620 non-compliant senders.
2621 Analysis and variant patch by Todd Lyons.
2623 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2624 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2625 Patch from Dirk Mueller <dmueller@suse.com>
2627 PP/13 tls_peerdn now print-escaped for spool files.
2628 Observed some $tls_peerdn in wild which contained \n, which resulted
2629 in spool file corruption.
2631 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2632 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2633 or write after TLS renegotiation, which otherwise led to messages
2636 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2637 as a tracking header (ie: a signed header comes before the signature).
2638 Patch from Wolfgang Breyha.
2640 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2641 comma-sep list; embedded commas doubled.
2643 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2645 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2647 Report and patch from Dmitry Banschikov.
2649 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2650 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2651 failures. We appear to now support TLS1.1+ with Exim.
2653 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2654 lets Exim select keys and certificates based upon TLS SNI from client.
2655 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2656 before an outbound SMTP session. New log_selector, +tls_sni.
2658 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2659 NULL dereference. Report and patch from Alun Jones.
2661 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2662 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2663 Not seeing resolver debug output on NetBSD, but suspect this is a
2664 resolver implementation change.
2666 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2667 Left warnings. Added "eximon gdb" invocation mode.
2669 PP/21 Defaulting "accept_8bitmime" to true, not false.
2671 PP/22 Added -bw for inetd wait mode support.
2673 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2674 locate the relevant includes and libraries. Made this the default.
2676 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2677 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2679 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2680 This may cause build issues on older platforms.
2682 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2683 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2684 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2685 Added SNI support via GnuTLS too.
2686 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2688 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2690 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2691 Applied second patch from Janne, implementing suggestion to default
2692 multiple-strings-in-record handling to match SPF spec.
2694 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2696 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2697 read-only, out of scope).
2698 Patch from Wolfgang Breyha, report from Stuart Northfield.
2700 PP/29 Fix three issues highlighted by clang analyser static analysis.
2701 Only crash-plausible issue would require the Cambridge-specific
2702 iplookup router and a misconfiguration.
2703 Report from Marcin Mirosław.
2705 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2707 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2708 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2709 As part of this, removing so much warning spew let me fix some minor
2710 real issues in debug logging.
2712 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2713 assignment on my part. Fixed.
2715 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2716 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2717 Janne Snabb (who went above and beyond: thank you).
2719 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2720 string otherwise requires a connection and a bunch more work and it's
2721 relatively easy to get wrong. Should also expose TLS library linkage
2724 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2725 64-bit ${eval} (JH/03).
2727 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2728 GNU libc to support some of the 64-bit stuff, should not lead to
2729 conflicts. Defined before os.h is pulled in, so if a given platform
2730 needs to override this, it can.
2732 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2733 protection layer was required, which is not implemented.
2734 Bugzilla 1254, patch from Wolfgang Breyha.
2736 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2737 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2738 tls_dhparam take prime identifiers. Also unbreak combination of
2739 OpenSSL+DH_params+TLSSNI.
2741 PP/39 Disable SSLv2 by default in OpenSSL support.
2747 PP/01 Solaris build fix for Oracle's LDAP libraries.
2748 Bugzilla 1109, patch from Stephen Usher.
2750 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2752 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2755 TF/02 Fix a couple more cases where we did not log the error message
2756 when unlink() failed. See also change 4.74-TF/03.
2758 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2759 lock up or crash if it happened to be inside a call to libc when it
2760 got a SIGUSR1 from exiwhat.
2762 The SIGUSR1 handler appends the current process status to the process
2763 log which is later printed by exiwhat. It used to use the general
2764 purpose logging code to do this, but several functions it calls are
2765 not safe for signals.
2767 The new output code in the SIGUSR1 handler is specific to the process
2768 log, and simple enough that it's easy to inspect for signal safety.
2769 Removing some special cases also simplifies the general logging code.
2770 Removing the spurious timestamps from the process log simplifies
2773 TF/04 Improved ratelimit ACL condition.
2775 The /noupdate option has been deprecated in favour of /readonly which
2776 has clearer semantics. The /leaky, /strict, and /readonly update modes
2777 are mutually exclusive. The update mode is no longer included in the
2778 database key; it just determines when the database is updated. (This
2779 means that when you upgrade Exim will forget old rate measurements.)
2781 Exim now checks that the per_* options are used with an update mode that
2782 makes sense for the current ACL. For example, when Exim is processing a
2783 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2784 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2785 must specify per_mail/readonly. If you omit the update mode it defaults to
2786 /leaky where that makes sense (as before) or /readonly where required.
2788 The /noupdate option is now undocumented but still supported for
2789 backwards compatibility. It is equivalent to /readonly except that in
2790 ACLs where /readonly is required you may specify /leaky/noupdate or
2791 /strict/noupdate which are treated the same as /readonly.
2793 A useful new feature is the /count= option. This is a generalization
2794 of the per_byte option, so that you can measure the throughput of other
2795 aggregate values. For example, the per_byte option is now equivalent
2796 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2798 The per_rcpt option has been generalized using the /count= mechanism
2799 (though it's more complicated than the per_byte equivalence). When it is
2800 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2801 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2802 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2803 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2804 Note that using per_rcpt with a non-readonly update mode in more than
2805 one ACL will cause the recipients to be double-counted. (The per_mail
2806 and per_byte options don't have this problem.)
2808 The handling of very low rates has changed slightly. If the computed rate
2809 is less than the event's count (usually one) then this event is the first
2810 after a long gap. In this case the rate is set to the same as this event's
2811 count, so that the first message of a spam run is counted properly.
2813 The major new feature is a mechanism for counting the rate of unique
2814 events. The new per_addr option counts the number of different
2815 recipients that someone has sent messages to in the last time period. It
2816 behaves like per_rcpt if all the recipient addresses are different, but
2817 duplicate recipient addresses do not increase the measured rate. Like
2818 the /count= option this is a general mechanism, so the per_addr option
2819 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2820 example, measure the rate that a client uses different sender addresses
2821 with the options per_mail/unique=$sender_address. There are further
2822 details in the main documentation.
2824 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2826 TF/06 Removed a few PCRE remnants.
2828 TF/07 Automatically extract Exim's version number from tags in the git
2829 repository when doing development or release builds.
2831 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2832 Bugzilla 879. Patch from Paul Fisher.
2834 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2835 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2838 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2840 PP/05 Variable $av_failed, true if the AV scanner deferred.
2841 Bugzilla 1078. Patch from John Horne.
2843 PP/06 Stop make process more reliably on build failure.
2844 Bugzilla 1087. Patch from Heiko Schlittermann.
2846 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2847 Bugzilla 1089. Patch from Heiko Schlittermann.
2849 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2850 Bugzilla 1131. Patch from Holger Weiß.
2852 PP/09 Handle IPv6 addresses with SPF.
2853 Bugzilla 860. Patch from Wolfgang Breyha.
2855 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2857 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2860 PP/11 match_* no longer expand right-hand-side by default.
2861 New compile-time build option, EXPAND_LISTMATCH_RHS.
2862 New expansion conditions, "inlist", "inlisti".
2864 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2866 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2868 PP/14 fix log_write() format string regression from TF/03.
2869 Bugzilla 1152. Patch from Dmitry Isaikin.
2875 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2877 PP/02 Harmonised TLS library version reporting; only show if debugging.
2878 Layout now matches that introduced for other libraries in 4.74 PP/03.
2880 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2882 PP/04 New "dns_use_edns0" global option.
2884 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2887 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2888 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2890 TK/01 Updated PolarSSL code to 0.14.2.
2891 Bugzilla 1097. Patch from Andreas Metzler.
2893 PP/07 Catch divide-by-zero in ${eval:...}.
2894 Fixes bugzilla 1102.
2896 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2899 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2900 format-string attack -- SECURITY: remote arbitrary code execution.
2902 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2903 time unintentionally subject to list matching rules, letting the header
2904 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2905 arbitrary string expansion). This allowed for information disclosure.
2907 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2908 INT_MIN/-1 -- value coerced to INT_MAX.
2914 NM/01 Workaround for PCRE version dependency in version reporting
2917 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2918 This fixes portability to compilers other than gcc, notably
2919 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2921 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2922 makefiles for portability to HP-UX and POSIX correctness.
2924 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2925 Also via indented variable definition in the Makefile.
2926 (Debugging by Oliver Heesakkers).
2928 PP/02 Restore caching of spamd results with expanded spamd_address.
2929 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2931 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2932 Improves build reliability. Fix from: Frank Elsner
2934 NM/02 Fix wide character breakage in the rfc2047 coding
2935 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2937 NM/03 Allow underscore in dnslist lookups
2938 Fixes bug 1026. Patch from Graeme Fowler
2940 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2941 Code patches from Adam Ciarcinski of NetBSD.
2943 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2946 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2947 is logged, to avoid truncation. Patch from John Horne.
2949 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2950 Patch from Jakob Hirsch.
2952 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2953 SQL string expansion failure details.
2954 Patch from Andrey Oktyabrski.
2956 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2957 Patch from Simon Arlott.
2959 PP/09 New lookups functionality failed to compile on old gcc which rejects
2960 extern declarations in function scope.
2961 Patch from Oliver Fleischmann
2963 PP/10 Use sig_atomic_t for flags set from signal handlers.
2964 Check getgroups() return and improve debugging.
2965 Fixed developed for diagnosis in bug 927 (which turned out to be
2968 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2969 Patch from Mark Zealey.
2971 PP/12 Bugzilla 1056: Improved spamd server selection.
2972 Patch from Mark Zealey.
2974 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2975 Based on patch from Heiko Schlittermann.
2977 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2978 Patch from Uwe Doering, sign-off by Michael Haardt.
2980 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2981 variable declaration deep within a block. Bug and patch from
2984 PP/15 lookups-Makefile IRIX compatibility coercion.
2986 PP/16 Make DISABLE_DKIM build knob functional.
2988 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2989 Patch by Simon Arlott
2991 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2992 variable argument macros. Our copy now differs from upstream.
2998 TF/01 Failure to get a lock on a hints database can have serious
2999 consequences so log it to the panic log.
3001 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3002 controlled using the smtp_confirmation log selector.
3004 TF/03 Include the error message when we fail to unlink a spool file.
3006 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3007 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3008 for maintaining out-of-tree patches for some time.
3010 PP/01 Bugzilla 139: Documentation and portability issues.
3011 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3012 Handle per-OS dynamic-module compilation flags.
3014 PP/02 Let /dev/null have normal permissions.
3015 The 4.73 fixes were a little too stringent and complained about the
3016 permissions on /dev/null. Exempt it from some checks.
3017 Reported by Andreas M. Kirchwitz.
3019 PP/03 Report version information for many libraries, including
3020 Exim version information for dynamically loaded libraries. Created
3021 version.h, now support a version extension string for distributors
3022 who patch heavily. Dynamic module ABI change.
3024 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3025 privilege escalation vulnerability whereby the Exim run-time user
3026 can cause root to append content of the attacker's choosing to
3029 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3032 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3033 If dropping privileges for untrusted macros, we disabled normal logging
3034 on the basis that it would fail; for the Exim run-time user, this is not
3035 the case, and it resulted in successful deliveries going unlogged.
3036 Fixed. Reported by Andreas Metzler.
3042 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3043 only prepend for the Resent-* case. Fixes regression introduced in
3044 Exim 4.70 by NM/22 for Bugzilla 607.
3046 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3047 increasing numbers of administrators be bitten by this.
3049 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3051 PP/03 Bugzilla 994: added openssl_options main configuration option.
3053 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3055 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3057 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3059 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3060 without a peer certificate, leading to a segfault because of an
3061 assumption that peers always have certificates. Be a little more
3062 paranoid. Problem reported by Martin Tscholak.
3064 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3065 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3066 NB: ClamAV planning to remove STREAM in "middle of 2010".
3067 CL also introduces -bmalware, various -d+acl logging additions and
3068 more caution in buffer sizes.
3070 PP/09 Implemented reverse_ip expansion operator.
3072 PP/10 Bugzilla 937: provide a "debug" ACL control.
3074 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3076 PP/12 Bugzilla 973: Implement --version.
3078 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3080 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3082 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3084 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3085 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3086 ignore trailing whitespace.
3088 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3090 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3093 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3094 Notification from Dr Andrew Aitchison.
3096 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3097 ExtendedDetectionInfo response format.
3098 Notification from John Horne.
3100 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3103 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3104 XSL and documented dependency on system catalogs, with examples of how
3107 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3110 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3111 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3112 configuration file which is writeable by the Exim user or group.
3114 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3115 of configuration files to cover files specified with the -C option if
3116 they are going to be used with root privileges, not just the default
3119 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3120 option (effectively making it always true).
3122 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3123 files to be used while preserving root privileges.
3125 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3126 that rogue child processes cannot use them.
3128 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3129 run-time user, instead of root.
3131 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3132 Exim run-time user without dropping privileges.
3134 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3135 result string, instead of calling string_vformat() twice with the same
3138 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3139 for other users. Others should always drop root privileges if they use
3140 -C on the command line, even for a whitelisted configure file.
3142 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3144 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3150 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3151 $data_path, and $header_path variables; fixed documentation bugs and
3154 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3155 exipick to access non-standard spools, including the "frozen" queue
3158 NM/01 Bugzilla 965: Support mysql stored procedures.
3159 Patch from Alain Williams
3161 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3163 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3164 Patch from Andreas Metzler
3166 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3167 Patch from Kirill Miazine
3169 NM/05 Bugzilla 671: Added umask to procmail example.
3171 JJ/03 installed exipick 20100323.0, fixing doc bug
3173 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3174 directory. Notification and patch from Dan Rosenberg.
3176 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3178 TK/02 Improve log output when DKIM signing operation fails.
3180 MH/01 Treat the transport option dkim_domain as a colon separated
3181 list, not as a single string, and sign the message with each element,
3182 omitting multiple occurences of the same signer.
3184 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3185 Bugzilla 985, 986. Patch by Simon Arlott
3187 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3188 Patch by Simon Arlott
3190 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3191 MBX locking. Notification from Dan Rosenberg.
3197 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3199 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3201 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3203 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3205 NM/04 Bugzilla 915: Flag broken perl installation during build.
3211 TK/01 Added patch by Johannes Berg that expands the main option
3212 "spamd_address" if it starts with a dollar sign.
3214 TK/02 Write list of recipients to X-Envelope-Sender header when building
3215 the mbox-format spool file for content scanning (suggested by Jakob
3218 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3219 (http://www.dcc-servers.net/) support via dccifd. Activated by
3220 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3222 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3223 by Mark Daniel Reidel <mr@df.eu>.
3225 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3226 When building exim an external PCRE library is now needed -
3227 PCRE is a system library on the majority of modern systems.
3228 See entry on PCRE_LIBS in EDITME file.
3230 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3231 conversation. Added nologin parameter to request.
3232 Patch contributed by Kirill Miazine.
3234 TF/01 Do not log submission mode rewrites if they do not change the address.
3236 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3238 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3239 log files in place. Contributed by Roberto Lima.
3241 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3243 TF/03 Bugzilla 615: When checking the local_parts router precondition
3244 after a local_part_suffix or local_part_prefix option, Exim now
3245 does not use the address's named list lookup cache, since this
3246 contains cached lookups for the whole local part.
3248 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3249 Robert Millan. Documentation is in experimental-spec.txt.
3251 TF/04 Bugzilla 668: Fix parallel build (make -j).
3253 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3255 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3256 Patch provided by Jan Srzednicki.
3258 TF/05 Leading white space used to be stripped from $spam_report which
3259 wrecked the formatting. Now it is preserved.
3261 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3262 that they are available at delivery time.
3264 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3266 TF/08 TLS error reporting now respects the incoming_interface and
3267 incoming_port log selectors.
3269 TF/09 Produce a more useful error message if an SMTP transport's hosts
3270 setting expands to an empty string.
3272 NM/06 Bugzilla 744: EXPN did not work under TLS.
3273 Patch provided by Phil Pennock.
3275 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3276 Patch provided by Richard Godbee.
3278 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3279 acl_smtp_notquit, added index entry.
3281 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3282 Patch provided by Eugene Bujak.
3284 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3285 accept(). Patch provided by Maxim Dounin.
3287 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3288 Patch provided by Phil Pennock.
3290 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3292 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3293 Patch provided by Brad "anomie" Jorsch.
3295 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3296 Patch provided by Dean Brooks.
3298 TK/05 Add native DKIM support (does not depend on external libraries).
3300 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3301 Patch provided by Graeme Fowler.
3303 NM/16 Bugzilla 851: Documentation example syntax fix.
3305 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3307 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3310 NM/19 Bugzilla 745: TLS version reporting.
3311 Patch provided by Phil Pennock.
3313 NM/20 Bugzilla 167: bool: condition support.
3314 Patch provided by Phil Pennock.
3316 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3317 clients. Patch provided by Phil Pennock.
3319 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3320 Patch provided by Brad "anomie" Jorsch.
3322 NM/23 Bugzilla 687: Fix misparses in eximstats.
3323 Patch provided by Heiko Schlittermann.
3325 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3326 Patch provided by Heiko Schlittermann.
3328 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3329 plus update to original patch.
3331 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3333 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3334 Patch provided by David Brownlee.
3336 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3338 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3340 NM/30 Bugzilla 888: TLS documentation bugfixes.
3342 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3344 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3345 Unlike the original bugzilla I have changed all shell scripts in src tree.
3347 NM/33 Bugzilla 898: Transport filter timeout fix.
3348 Patch by Todd Rinaldo.
3350 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3351 Patch by Serge Demonchaux.
3353 NM/35 Bugzilla 39: Base64 decode bug fixes.
3354 Patch by Jakob Hirsch.
3356 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3358 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3360 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3362 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3368 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3369 ALT-N's libdkim that I have put here:
3370 http://duncanthrax.net/exim-experimental/
3372 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3373 were called 'true' and it seems that C99 defines that as a reserved
3374 keyword to be used with 'bool' variable types. That means you could
3375 not include C99-style headers which use bools without triggering
3376 build errors in sieve.c.
3378 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3379 as mailq or other aliases. Changed the --help handling significantly
3380 to do whats expected. exim_usage() emits usage/help information.
3382 SC/01 Added the -bylocaldomain option to eximstats.
3384 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3386 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3388 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3394 PH/01 Another patch from the Sieve maintainer.
3396 PH/02 When an IPv6 address is converted to a string for single-key lookup
3397 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3398 dots are used instead of colons so that keys in lsearch files need not
3399 contain colons. This was done some time before quoting was made available
3400 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3401 (notated using the quote facility) so as to distinguish them from IPv4
3402 keys. This meant that lookups for IP addresses in host lists did not work
3403 for iplsearch lookups.
3405 This has been fixed by arranging for IPv6 addresses to be expressed with
3406 colons if the lookup type is iplsearch. This is not incompatible, because
3407 previously such lookups could never work.
3409 The situation is now rather anomalous, since one *can* have colons in
3410 ordinary lsearch keys. However, making the change in all cases is
3411 incompatible and would probably break a number of configurations.
3413 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3416 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3417 conversion specification without a maximum field width, thereby enabling
3418 a rogue spamd server to cause a buffer overflow. While nobody in their
3419 right mind would setup Exim to query an untrusted spamd server, an
3420 attacker that gains access to a server running spamd could potentially
3421 exploit this vulnerability to run arbitrary code as the Exim user.
3423 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3424 $primary_hostname instead of what libspf2 thinks the hosts name is.
3426 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3427 a directory entry by the name of the lookup key. Previously, if a
3428 symlink pointed to a non-existing file or a file in a directory that
3429 Exim lacked permissions to read, a lookup for a key matching that
3430 symlink would fail. Now it is enough that a matching directory entry
3431 exists, symlink or not. (Bugzilla 503.)
3433 PH/03 The body_linecount and body_zerocount variables are now exported in the
3436 PH/04 Added the $dnslist_matched variable.
3438 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3439 This means they are set thereafter only if the connection becomes
3442 PH/06 Added the client_condition to authenticators so that some can be skipped
3443 by clients under certain conditions.
3445 PH/07 The error message for a badly-placed control=no_multiline_responses left
3446 "_responses" off the end of the name.
3448 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3450 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3451 (without spaces) instead of just copying the configuration text.
3453 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3455 PH/11 Added $max_received_linelength.
3457 PH/12 Added +ignore_defer and +include_defer to host lists.
3459 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3460 way in which PCRE > 7.0 is built.
3462 PH/14 Implemented queue_only_load_latch.
3464 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3465 MAIL command. The effect was to mangle the value on 64-bit systems.
3467 PH/16 Another patch from the Sieve maintainer.
3469 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3471 PH/18 If a system quota error occurred while trying to create the file for
3472 a maildir delivery, the message "Mailbox is full" was not appended to the
3473 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3474 only to a quota excession during the actual writing of the file.
3476 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3477 characters?) which causes problems in log lines. The DN values are now
3478 passed through string_printing() before being added to log lines.
3480 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3481 and InterBase are left for another time.)
3483 PH/21 Added message_body_newlines option.
3485 PH/22 Guard against possible overflow in moan_check_errorcopy().
3487 PH/23 POSIX allows open() to be a macro; guard against that.
3489 PH/24 If the recipient of an error message contained an @ in the local part
3490 (suitably quoted, of course), incorrect values were put in $domain and
3491 $local_part during the evaluation of errors_copy.
3497 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3498 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3501 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3502 issue a MAIL command.
3504 PH/02 In an ACL statement such as
3506 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3508 if a client was not listed at all, or was listed with a value other than
3509 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3510 the condition was not true (as it should be), so access was not denied.
3511 The bug was that the ! inversion was incorrectly passed on to the second
3512 item. This has been fixed.
3514 PH/03 Added additional dnslists conditions == and =& which are different from
3515 = and & when the dns lookup returns more than one IP address.
3517 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3518 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3520 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3521 FSYNC, which compiles an option called disable_fsync that allows for
3522 bypassing fsync(). The documentation is heavily laced with warnings.
3524 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3526 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3527 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3528 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3529 including adding "make clean"; (3) Added -fPIC when compiling the test
3530 dynamically loaded module, to get rid of a warning.
3532 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3533 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3534 = 0s. The bug is otherwise harmless.
3536 PH/07 There was a bug in the dovecot authenticator such that the value of
3537 $auth1 could be overwritten, and so not correctly preserved, after a
3538 successful authentication. This usually meant that the value preserved by
3539 the server_setid option was incorrect.
3541 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3543 PH/09 Installed PCRE release 7.0.
3545 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3546 run for batched SMTP input. It is now run at the start of every message
3547 in the batch. While fixing this I discovered that the process information
3548 (output by running exiwhat) was not always getting set for -bs and -bS
3549 input. This is fixed, and it now also says "batched" for BSMTP.
3551 PH/11 Added control=no_pipelining.
3553 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3554 patch, slightly modified), and move the expansion of helo_data till after
3555 the connection is made in the smtp transport (so it can use these
3558 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3560 PH/14 Added log_selector = +pid.
3562 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3564 PH/16 Add ${if forany and ${if forall.
3566 PH/17 Added dsn_from option to vary the From: line in DSNs.
3568 PH/18 Flush SMTP output before performing a callout, unless control =
3569 no_callout_flush is set.
3571 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3572 was true (the default) a successful delivery failed to delete the retry
3573 item, thus causing premature timeout of the address. The bug is now
3576 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3578 PH/21 Long custom messages for fakedefer and fakereject are now split up
3579 into multiline responses in the same way that messages for "deny" and
3580 other ACL rejections are.
3582 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3583 with slight modification.
3585 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3586 draft, changing the syntax and factoring some duplicate code".
3588 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3589 for deliveries of the second and subsequent messages over the same SMTP
3592 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3593 ${reduce, with only minor "tidies".
3595 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3597 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3598 expansion side effects.
3600 PH/27 When a message times out after an over-quota error from an Exim-imposed
3601 quota, the bounce message says "mailbox is full". This message was not
3602 being given when it was a system quota that was exceeded. It now should
3605 MH/03 Made $recipients available in local_scan(). local_scan() already has
3606 better access to the recipient list through recipients_list[], but
3607 $recipients can be useful in postmaster-provided expansion strings.
3609 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3610 in the case of a MAIL command with additional options following the
3611 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3612 were accidentally chopped off.
3614 PH/29 SMTP synchronization checks are implemented when a command is read -
3615 there is a check that no more input is waiting when there shouldn't be
3616 any. However, for some commands, a delay in an ACL can mean that it is
3617 some time before the response is written. In this time, more input might
3618 arrive, invalidly. So now there are extra checks after an ACL has run for
3619 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3620 pipelining has not been advertised.
3622 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3624 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3625 returned in the SMTP response when smtp_return_error_details was false.
3626 This has been fixed.
3628 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3629 instead of the C I/O that was originally supplied, because problems were
3630 reported on Solaris.
3632 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3633 Exim which did not show up earlier: it was assuming that a call to
3634 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3635 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3636 was a macro that became an assignment, so it seemed to work. This has
3637 changed to a proper function call with a void return, hence the compile
3638 error. Exim's code has been fixed.
3640 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3643 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3645 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3647 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3648 because it turns out to be too big or there is a timeout) there is no
3649 "Completed" line in the log. When some messages of this type were
3650 selected by exigrep, they were listed as "not completed". Others were
3651 picked up by some special patterns. I have improved the selection
3652 criteria to be more general.
3654 PH/38 The host_find_failed option in the manualroute router can now be set
3655 to "ignore", to completely ignore a host whose IP address cannot be
3656 found. If all hosts are ignored, the behaviour is controlled by the new
3657 host_all_ignored option.
3659 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3660 homing or because of multiple MX records with /mx) generated more than
3661 one IP address, and the following item turned out to be the local host,
3662 all the secondary addresses of the first item were incorrectly removed
3663 from the list, along with the local host and any following hosts (which
3664 is what is supposed to happen).
3666 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3667 whoever called Exim into the -H file. In the case of the daemon it was
3668 behaving confusingly. When first started, it used values for whoever
3669 started the daemon, but after a SIGHUP it used the Exim user (because it
3670 calls itself on a restart). I have changed the code so that it now always
3673 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3674 message are rejected with the same error (e.g. no authentication or bad
3675 sender address), and a DATA command is nevertheless sent (as can happen
3676 with PIPELINING or a stupid MUA), the error message that was given to the
3677 RCPT commands is included in the rejection of the DATA command. This is
3678 intended to be helpful for MUAs that show only the final error to their
3681 PH/42 Another patch from the Sieve maintainer.
3683 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3684 Eximstats - Fixed some broken HTML links and added missing column headers
3686 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3687 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3689 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3691 PH/43 Yet another patch from the Sieve maintainer.
3693 PH/44 I found a way to check for a TCP/IP connection going away before sending
3694 the response to the final '.' that terminates a message, but only in the
3695 case where the client has not sent further data following the '.'
3696 (unfortunately, this is allowed). However, in many cases there won't be
3697 any further data because there won't be any more messages to send. A call
3698 to select() can be used: if it shows that the input is "ready", there is
3699 either input waiting, or the socket has been closed. An attempt to read
3700 the next input character can distinguish the two cases. Previously, Exim
3701 would have sent an OK response which the client would never have see.
3702 This could lead to message repetition. This fix should cure that, at
3703 least in a lot of common cases.
3705 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3706 advertised in response to EHLO.
3712 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3713 fixed by 4.65/MH/01 (is this a record?) are fixed:
3715 (i) An empty string was always treated as zero by the numeric comparison
3716 operators. This behaviour has been restored.
3718 (ii) It is documented that the numeric comparison operators always treat
3719 their arguments as decimal numbers. This was broken in that numbers
3720 starting with 0 were being interpreted as octal.
3722 While fixing these problems I realized that there was another issue that
3723 hadn't been noticed. Values of message_size_limit (both the global option
3724 and the transport option) were treated as octal if they started with 0.
3725 The documentation was vague. These values are now always treated as
3726 decimal, and I will make that clear in the documentation.
3732 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3733 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3736 MH/01 Don't check that the operands of numeric comparison operators are
3737 integers when their expansion is in "skipping" mode (fixes bug
3738 introduced by 4.64-PH/07).
3740 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3741 child addresses, Exim now panics and dies. Previously, because the count
3742 is held in a short int, deliveries were likely to be lost. As such a
3743 large number of recipients for a single message is ridiculous
3744 (performance will be very, very poor), I have chosen to impose a limit
3745 rather than extend the field.
3751 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3752 leftover -K file (the existence of which was triggered by #402).
3753 While we were at it, introduced process PID as part of the -K
3754 filename. This should rule out race conditions when creating
3757 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3758 processing considerably. Previous code took too long for large mails,
3759 triggering a timeout which in turn triggers #401.
3761 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3762 in the DK code in transports.c. sendfile() is not really portable,
3763 hence the _LINUX specificness.
3765 TF/01 In the add_headers option to the mail command in an Exim filter,
3766 there was a bug that Exim would claim a syntax error in any
3767 header after the first one which had an odd number of characters
3770 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3771 callout verification, Exim cached a "reject" for the entire domain. This
3772 is correct for most verifications, but it is not correct for a recipient
3773 verification with use_sender or use_postmaster set, because in that case
3774 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3775 case of MAIL FROM:<> rejection from other early rejections (e.g.
3776 rejection of HELO). When verifying a recipient using a non-null MAIL
3777 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3778 Whatever the result of the callout, the value of the domain cache is
3779 left unchanged (for any other kind of callout, getting as far as trying
3780 RCPT means that the domain itself is ok).
3782 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3785 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3786 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3787 session, the daemon ignores ECONNECT errors and logs others; it now
3788 ignores EPIPE as well.
3790 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3791 (quoted-printable decoding).
3793 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3794 later the small subsequent patch to fix an introduced bug.
3796 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3798 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3800 PH/08 An error is now given if message_size_limit is specified negative.
3802 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3803 to be given (somewhat) arbitrary names.
3805 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3808 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3809 miscellaneous code fixes
3811 PH/10 Added the log_reject_target ACL modifier to specify where to log
3814 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3815 hostname. This is wrong, because it relates to the incoming message (and
3816 probably the interface on which it is arriving) and not to the outgoing
3817 callout (which could be using a different interface). This has been
3818 changed to use the value of the helo_data option from the smtp transport
3819 instead - this is what is used when a message is actually being sent. If
3820 there is no remote transport (possible with a router that sets up host
3821 addresses), $smtp_active_hostname is used.
3823 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3824 tweaks were necessary in order to get it to work (see also 21 below):
3825 (a) The code assumed that strncpy() returns a negative number on buffer
3826 overflow, which isn't the case. Replaced with Exim's string_format()
3828 (b) There were several signed/unsigned issues. I just did the minimum
3829 hacking in of casts. There is scope for a larger refactoring.
3830 (c) The code used strcasecmp() which is not a standard C function.
3831 Replaced with Exim's strcmpic() function.
3832 (d) The code set only $1; it now sets $auth1 as well.
3833 (e) A simple test gave the error "authentication client didn't specify
3834 service in request". It would seem that Dovecot has changed its
3835 interface. Fortunately there's a specification; I followed it and
3836 changed what the client sends and it appears to be working now.
3838 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3841 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3842 address A is aliased to B and C, where B exists and C does not. Without
3843 -v the output is "A verified" because verification stops after a
3844 successful redirection if more than one address is generated. However,
3845 with -v the child addresses are also verified. Exim was outputting "A
3846 failed to verify" and then showing the successful verification for C,
3847 with its parentage. It now outputs "B failed to verify", showing B's
3848 parentage before showing the successful verification of C.
3850 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3851 look up a TXT record in a specific list after matching in a combined
3854 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3855 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3856 they consult the DNS. I had assumed they would set it the way they
3857 wanted; and indeed my experiments on Linux seem to show that in some
3858 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3859 To be on the safe side, however, I have now made the interface to
3860 host_find_byname() similar to host_find_bydns(), with an argument
3861 containing the DNS resolver options. The host_find_byname() function now
3862 sets these options at its start, just as host_find_bydns() does. The smtp
3863 transport options dns_qualify_single and dns_search_parents are passed to
3864 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3865 of host_find_byname() use the default settings of RES_DEFNAMES
3866 (qualify_single) but not RES_DNSRCH (search_parents).
3868 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3869 spool_read_header() do less string testing, by means of a preliminary
3870 switch on the second character of optional "-foo" lines. (This is
3871 overdue, caused by the large number of possibilities that now exist.
3872 Originally there were few.) While I was there, I also converted the
3873 str(n)cmp tests so they don't re-test the leading "-" and the first
3874 character, in the hope this might squeeze out yet more improvement.
3876 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3877 flag allowing group syntax was set by the header_syntax check but not
3878 turned off, possible causing trouble later; (2) The flag was not being
3879 set at all for the header_verify test, causing "group"-style headers to
3880 be rejected. I have now set it in this case, and also caused header_
3881 verify to ignore an empty address taken from a group. While doing this, I
3882 came across some other cases where the code for allowing group syntax
3883 while scanning a header line wasn't quite right (mostly, not resetting
3884 the flag correctly in the right place). These bugs could have caused
3885 trouble for malformed header lines. I hope it is now all correct.
3887 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3888 with the "reply" argument non-NULL. The code, however (which originally
3889 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3890 but it didn't always do it. This confused somebody who was copying the
3891 code for some other use. I have removed all the tests.
3893 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3894 feature that was used to support insecure browsers during the U.S. crypto
3895 embargo. It requires special client support, and Exim is probably the
3896 only MTA that supported it -- and would never use it because real RSA is
3897 always available. This code has been removed, because it had the bad
3898 effect of slowing Exim down by computing (never used) parameters for the
3899 RSA_EXPORT functionality.
3901 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3902 authenticator to fail if there's a tab character in the incoming data
3903 (there should never be unless someone is messing about, as it's supposed
3904 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3905 if the connection is using TLS or if the remote IP is the same as the
3906 local IP, and the "valid-client-cert option" if a client certificate has
3909 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3910 authenticators. This can be used for authorization after authentication
3911 succeeds. (In the case of plaintext, it servers for both authentication
3914 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3915 if any retry times were supplied.
3917 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3918 connection, where there is no client IP address to check. In this
3919 situation, the verify now always succeeds.
3921 PH/25 Applied John Jetmore's -Mset patch.
3923 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3925 PH/27 In a string expansion for a processed (not raw) header when multiple
3926 headers of the same name were present, leading whitespace was being
3927 removed from all of them, but trailing whitespace was being removed only
3928 from the last one. Now trailing whitespace is removed from each header
3929 before concatenation. Completely empty headers in a concatenation (as
3930 before) are ignored.
3932 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3933 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3935 PH/29 [Removed. This was a change that I later backed out, and forgot to
3936 correct the ChangeLog entry (that I had efficiently created) before
3937 committing the later change.]
3939 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3940 address errors (4xx response to RCPT) over the same connection as other
3941 messages routed to the same hosts. Such deliveries are always "forced",
3942 so retry times are not inspected. This resulted in far too many retries
3943 for the affected addresses. The effect occurred only when there were more
3944 hosts than the hosts_max_try setting in the smtp transport when it had
3945 the 4xx errors. Those hosts that it had tried were not added to the list
3946 of hosts for which the message was waiting, so if all were tried, there
3947 was no problem. Two fixes have been applied:
3949 (i) If there are any address or message errors in an SMTP delivery, none
3950 of the hosts (tried or untried) are now added to the list of hosts
3951 for which the message is waiting, so the message should not be a
3952 candidate for sending over the same connection that was used for a
3953 successful delivery of some other message. This seems entirely
3954 reasonable: after all the message is NOT "waiting for some host".
3955 This is so "obvious" that I'm not sure why it wasn't done
3956 previously. Hope I haven't missed anything, but it can't do any
3957 harm, as the worst effect is to miss an optimization.
3959 (ii) If, despite (i), such a delivery is accidentally attempted, the
3960 routing retry time is respected, so at least it doesn't keep
3961 hammering the server.
3963 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3964 in ${readsocket because some servers need this prod.
3966 PH/32 Added some extra debug output when updating a wait-xxx database.
3968 PH/33 The hint "could be header name not terminated by colon", which has been
3969 given for certain expansion errors for a long time, was not being given
3970 for the ${if def:h_colon_omitted{... case.
3972 PH/34 The spec says: "With one important exception, whenever a domain list is
3973 being scanned, $domain contains the subject domain." There was at least
3974 one case where this was not true.
3976 PH/35 The error "getsockname() failed: connection reset by peer" was being
3977 written to the panic log as well as the main log, but it isn't really
3978 panic-worthy as it just means the connection died rather early on. I have
3979 removed the panic log writing for the ECONNRESET error when getsockname()
3982 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3983 runs only) independently of the message's sender address. This meant
3984 that, if the 4xx error was in fact related to the sender, a different
3985 message to the same recipient with a different sender could confuse
3986 things. In particular, this can happen when sending to a greylisting
3987 server, but other circumstances could also provoke similar problems.
3988 I have changed the default so that the retry time for these errors is now
3989 based a combination of the sender and recipient addresses. This change
3990 can be overridden by setting address_retry_include_sender=false in the
3993 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3994 remote server are returned as part of bounce messages. This was not
3995 happening for LMTP over a pipe (the lmtp transport), but now it is the
3996 same for both kinds of LMTP.
3998 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3999 in header lines that were in fact CNAMEs. This is no longer the case.
4001 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4002 and queue runs started by the daemon processed all messages. This has
4003 been fixed so that -R and -S can now usefully be given with -q<time>.
4005 PH/40 Import PCRE release 6.7 (fixes some bugs).
4007 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4009 PH/42 Give an error if -q is specified more than once.
4011 PH/43 Renamed the variables $interface_address and $interface_port as
4012 $received_ip_address and $received_port, to make it clear that these
4013 values apply to message reception, and not to the outgoing interface when
4014 a message is delivered. (The old names remain recognized, of course.)
4016 PH/44 There was no timeout on the connect() call when using a Unix domain
4017 socket in the ${readsocket expansion. There now is.
4019 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4020 be meaningful with "accept".
4022 SC/01 Eximstats V1.43
4023 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4025 SC/02 Eximstats V1.44
4026 Use a glob alias rather than an array ref in the generated
4027 parser. This improves both readability and performance.
4029 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4030 Collect SpamAssassin and rejection statistics.
4031 Don't display local sender or destination tables unless
4032 there is data to show.
4033 Added average volumes into the top table text output.
4035 SC/04 Eximstats V1.46
4036 Collect data on the number of addresses (recipients)
4037 as well as the number of messages.
4039 SC/05 Eximstats V1.47
4040 Added 'Message too big' to the list of mail rejection
4041 reasons (thanks to Marco Gaiarin).
4043 SC/06 Eximstats V1.48
4044 Mainlog lines which have GMT offsets and are too short to
4045 have a flag are now skipped.
4047 SC/07 Eximstats V1.49 (Alain Williams)
4048 Added the -emptyok flag.
4050 SC/08 Eximstats V1.50
4051 Fixes for obtaining the IP address from reject messages.
4053 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4054 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4055 whitespace changes from 4.64-PH/27
4057 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4060 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4061 are found, allow negative numbers in numeric criteria)
4063 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4065 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4066 to match changes made in 4.64-PH/43
4068 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4070 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4071 "domain" is actually an IP address, and force a failure. This locks out
4072 those revolvers/nameservers that support "A-for-A" lookups, in
4073 contravention of the specifications.
4075 PH/48 When a host name was looked up from an IP address, and the subsequent
4076 forward lookup of the name timed out, the host name was left in
4077 $sender_host_name, contrary to the specification.
4079 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4080 restricted to single-key lookups, Exim was not diagnosing an error if
4081 * or *@ was used with a query-style lookup.
4083 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4085 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4086 long ago, but noone interested enough thought of it. Let's just say that
4087 the "1.1" means that there are some new functions that weren't there at
4088 some point in the past.
4090 PH/51 Error processing for expansion failure of helo_data from an smtp
4091 transport during callout processing was broken.
4093 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4094 tested/used via the -bh/-bhc/-bs options.
4096 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4097 bug, fixed in subsequent PCRE releases).
4099 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4100 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4102 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4108 SC/01 Use a glob alias rather than an array ref in eximstats generated
4109 parser. This improves both readability and performance.
4111 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4112 Don't display local sender or destination tables in eximstats unless
4113 there is data to show.
4114 Added average volumes into the eximstats top table text output.
4116 SC/03 Collect data on the number of addresses (recipients) as well
4117 as the number of messages in eximstats.
4119 TF/01 Correct an error in the documentation for the redirect router. Exim
4120 does (usually) call initgroups() when daemonizing.
4122 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4123 with consistent privilege compared to when running as a daemon.
4125 TF/03 Note in the spec that $authenticated_id is not set for local
4126 submissions from trusted users.
4128 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4129 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4131 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4132 by adding some example configuration directives to the default
4133 configuration file. A little bit of work is required to uncomment the
4134 directives and define how usernames and passwords are checked, but
4135 there is now a framework to start from.
4137 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4138 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4139 without this. I don't know how relevant this is to other LDAP libraries.
4141 PH/02 Add the verb name to the "unknown ACL verb" error.
4143 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4145 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4147 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4148 directory not expanded when it should be if an expanded home directory
4149 was set for the address (which is overridden by the transport).
4151 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4154 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4155 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4156 because it is too late at that time, and has no effect.
4158 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4159 security issue with \' (bugzilla #107). I could not use the
4160 PQescapeStringConn() function, because it needs a PGconn value as one of
4163 PH/08 When testing addresses using -bt, indicate those final addresses that
4164 are duplicates that would not cause an additional delivery. At least one
4165 person was confused, thinking that -bt output corresponded to deliveries.
4166 (Suppressing duplicates isn't a good idea as you lose the information
4167 about possibly different redirections that led to the duplicates.)
4169 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4170 systems where poll() doesn't work, in particular OS X.
4172 PH/10 Added more information to debugging output for retry time not reached.
4174 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4175 operations in malware.c.
4177 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4180 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4181 syslog_duplication set false, log lines that would normally be written
4182 both the the main log and to the reject log were not written to syslog at
4185 PH/14 In the default configuration, change the use of "message" in ACL warn
4186 statements to "add_header".
4188 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4189 not followed by a command (e.g. "seen endif").
4191 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4192 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4195 PH/17 Added extra conditions to the default value of delay_warning_condition
4199 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4200 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4201 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4204 The Auto-Submitted: and various List- headers are standardised, whereas I
4205 don't think Precedence: ever was.
4207 PH/18 Refactored debugging code in route_finduser() to show more information,
4208 in particular, the error code if getpwnam() issues one.
4210 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4211 This is apparently needed in addition to the PH/07 change above to avoid
4212 any possible encoding problems.
4214 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4215 but not after initializing Perl.
4217 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4218 output them only if debugging. By default they are written stderr,
4219 apparently, which is not desirable.
4221 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4224 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4227 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4229 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4230 authenticated or an ident call has been made. Suppress the default
4231 values for $authenticated_id and $authenticated_sender (but permit -oMai
4232 and -oMas) when testing with -bh.
4234 PH/24 Re-jigged the order of the tests in the default configuration so that the
4235 tests for valid domains and recipients precede the DNS black list and CSA
4236 tests, on the grounds that those ones are more expensive.
4238 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4239 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4240 This bug exists in every version of Exim that I still have, right back to
4243 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4244 However, an attempt to turn on case-sensitivity in a regex key by
4245 including (?-i) didn't work because the subject string was already
4246 lowercased, and the effects were non-intuitive. It turns out that a
4247 one-line patch can be used to allow (?-i) to work as expected.
4253 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4254 other effects) broke the use of negated acl sub-conditions.
4256 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4259 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4260 "Deny" causes Exim to reject the incoming connection with a 554 error.
4261 Unfortunately, if there is a major crisis, such as a disk failure,
4262 tcp-wrappers gives "deny", whereas what one would like would be some
4263 kind of temporary error. A kludge has been added to help with this.
4264 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4265 554 error is used if errno is still zero or contains ENOENT (which occurs
4266 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4269 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4271 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4272 errors. Otherwise a message that provokes a temporary error (when other
4273 messages do not) can cause a whole host to time out.
4275 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4276 addresses were routed directly to files or pipes from a redirect router.
4277 File deliveries just didn't batch; pipe deliveries might have suffered
4280 PH/06 A failure to get a lock for a hints database would erroneously always say
4281 "Failed to get write lock", even when it was really a read lock.
4283 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4284 of 0600. This has been changed to use the value of the lockfile_mode
4285 option (which defaults to 0600).
4287 PH/08 Applied small patch from the Sieve maintainer.
4289 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4290 folder from quota calculations, a direct delivery into this folder messed
4291 up the contents of the maildirsize file. This was because the regex was
4292 used only to exclude .Trash (or whatever) when the size of the mailbox
4293 was calculated. There was no check that a delivery was happening into an
4294 excluded directory. This bug has been fixed by ignoring all quota
4295 processing for deliveries into excluded directories.
4297 PH/10 Added the maildirfolder_create_regex option to appendfile.
4303 PH/01 The code for finding all the local interface addresses on a FreeBSD
4304 system running IPv6 was broken. This may well have applied to all BSD
4305 systems, as well as to others that have similar system calls. The broken
4306 code found IPv4 interfaces correctly, but gave incorrect values for the
4307 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4308 that it would not match correctly against @[] and not recognize the IPv6
4311 PH/02 The ipliteral router was not recognizing addresses of the form user@
4312 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4314 PH/03 Added disable_ipv6.
4316 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4317 decoded form, because it is most often used to construct To: headers
4318 lines in autoreplies, and the decoded form may well be syntactically
4319 invalid. However, $reply_address has leading white space removed, and all
4320 newlines turned into spaces so that the autoreply transport does not
4323 PH/05 If group was specified without a user on a router, and no group or user
4324 was specified on a transport, the group from the router was ignored.
4326 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4327 for visible compile-time settings that can be used to change these
4328 numbers, for those that want even more. Backwards compatibility with old
4329 spool files has been maintained. However, going back to a previous Exim
4330 release will lost any variables that are in spool files.
4332 PH/07 Two small changes when running in the test harness: increase delay when
4333 passing a TCP/IP connection to a new process, in case the original
4334 process has to generate a bounce, and remove special handling of
4335 127.0.0.2 (sic), which is no longer necessary.
4337 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4338 be the same on different OS.
4340 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4343 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4344 whether --show-vars was specified or not
4346 JJ/02 exipick: Added support for new ACL variable spool format introduced
4349 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4350 syntactically invalid From: or Reply-to: line, and a filter used this to
4351 generate an autoreply, and therefore failed to obtain an address for the
4352 autoreply, Exim could try to deliver to a non-existent relative file
4353 name, causing unrelated and misleading errors. What now happens is that
4354 it logs this as a hard delivery error, but does not attempt to create a
4357 PH/11 The exinext utility has a -C option for testing purposes, but although
4358 the given file was scanned by exinext itself; it wasn't being passed on
4359 when Exim was called.
4361 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4362 an end-of-file indication when reading a command response.
4364 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4365 compiled. In many other places in Exim, IPv6 addresses are always
4366 recognized, so I have changed this. It also means that IPv4 domain
4367 literals of the form [IPV4:n.n.n.n] are now always recognized.
4369 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4370 used if the router is not running as root, for example, when verifying at
4371 ACL time, or when using -bh. The debugging output from this situation was
4372 non-existent - all you got was a failure to exec. I have made two
4375 (a) Failures to set uid/gid, the current directory, or a process leader
4376 in a subprocess such as that created by queryprogram now generate
4377 suitable debugging output when -d is set.
4379 (b) The queryprogram router detects when it is not running as root,
4380 outputs suitable debugging information if -d is set, and then runs
4381 the subprocess without attempting to change uid/gid.
4383 PH/15 Minor change to Makefile for building test_host (undocumented testing
4386 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4387 additional section of a DNS packet that returns MX or SRV records.
4388 Instead, it always explicitly searches for A/AAAA records. This avoids
4389 major problems that occur when a DNS server includes only records of one
4390 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4391 fixed another bug: if SRV records were looked up and the corresponding
4392 address records were *not* found in the additional section, the port
4393 values from the SRV records were lost.
4395 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4396 using the correct key (the original address) when searching the retry
4397 rules in order to find which one to use for generating the retry hint.
4399 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4400 adding the default one. Similarly, if it contains a Reply-To: header, the
4401 errors_reply_to option, if set, is not used.
4403 PH/19 When calculating a retry time, Exim used to measure the "time since
4404 failure" by looking at the "first failed" field in the retry record. Now
4405 it does not use this if it is later than than the arrival time of the
4406 message. Instead it uses the arrival time. This makes for better
4407 behaviour in cases where some deliveries succeed, thus re-setting the
4408 "first failed" field. An example is a quota failure for a huge message
4409 when small messages continue to be delivered. Without this change, the
4410 "time since failure" will always be short, possible causing more frequent
4411 delivery attempts for the huge message than are intended.
4412 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4414 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4415 $1, $2, $3) because the numerical variables can be reset during some
4416 expansion items (e.g. "match"), thereby losing the authentication data.
4418 PH/21 Make -bV show the size of off_t variables so that the test suite can
4419 decide whether to run tests for quotas > 2G.
4421 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4422 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4423 filecount value is greater than 2G or if a quota value is greater than 2G
4424 on a system where the size of off_t is not greater than 4, a panic error
4427 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4428 never match. The debug and -bh output now contains an explicit error
4429 message indicating a malformed IPv4 address or mask.
4431 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4432 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4433 PH/23 above applies.
4435 PH/25 Do not write to syslog when running in the test harness. The only
4436 occasion when this arises is a failure to open the main or panic logs
4437 (for which there is an explicit test).
4439 PH/26 Added the /no_tell option to "control=freeze".
4441 PH/27 If a host name lookup failed very early in a connection, for example, if
4442 the IP address matched host_lookup and the reverse lookup yielded a name
4443 that did not have a forward lookup, an error message of the form "no IP
4444 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4445 could be logged. Now it outputs the IP address instead of "NULL".
4447 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4448 allows the sender and the authenticated sender to be set when
4449 submitting a message from within Exim. Since child_open_exim() is
4450 documented for local_scan(), the new function should be too.
4452 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4453 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4454 results in an empty string is now treated as unset.
4456 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4458 PH/31 Added sender_verify_fail logging option.
4460 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4461 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4462 tidied the source and removed it altogether.
4464 PH/33 When a queue run was abandoned because the load average was too high, a
4465 log line was always written; now it is written only if the queue_run log
4466 selector is set. In addition, the log line for abandonment now contains
4467 information about the queue run such as the pid. This is always present
4468 in "start" and "stop" lines but was omitted from the "abandon" line.
4470 PH/34 Omit spaces between a header name and the colon in the error message that
4471 is given when verify = headers_syntax fails (if there are lots of them,
4472 the message gets confusing).
4474 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4475 names, as there are now some PTR records that contain slashes. This check
4476 is only to protect against broken name servers that fall over on strange
4477 characters, so the fact that it applies to all lookups doesn't matter.
4479 PH/36 Now that the new test suite is complete, we can remove some of the
4480 special code in Exim that was needed for the old test suite. For example,
4481 sorting DNS records because real resolvers return them in an arbitrary
4482 order. The new test suite's fake resolver always returns records in the
4485 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4486 bounces) except when queue_only is set, to avoid logging races between
4487 the different processes.
4489 PH/38 Panic-die if .include specifies a non-absolute path.
4491 PH/39 A tweak to the "H" retry rule from its user.
4493 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4494 a label. They prevented compilation on older perls.
4496 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4497 a warning to be raised on newish perls.
4499 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4500 on queue. Changes to match documented behaviour of showing count of
4501 messages matching specified criteria.
4503 PH/40 Changed the default ident timeout from 30s to 5s.
4505 PH/41 Added support for the use of login_cap features, on those BSD systems
4506 that have them, for controlling the resources used by pipe deliveries.
4508 PH/42 The content-scanning code uses fopen() to create files in which to put
4509 message data. Previously it was not paying any attention to the mode of
4510 the files. Exim runs with umask(0) because the rest of the code creates
4511 files with open(), and sets the required mode explicitly. Thus, these
4512 files were ending up world-writeable. This was not a big issue, because,
4513 being within the spool directory, they were not world-accessible. I have
4514 created a function called modefopen, which takes an additional mode
4515 argument. It sets umask(777), creates the file, chmods it to the required
4516 mode, then resets the umask. All the relevant calls to fopen() in the
4517 content scanning code have been changed to use this function.
4519 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4520 to 24 hours. This avoids potential overflow problems when processing G
4521 and H retry rules. I suspect nobody ever tinkers with this value.
4523 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4525 PH/45 When the plaintext authenticator is running as a client, the server's
4526 challenges are checked to ensure they are valid base64 strings. By
4527 default, the authentication attempt is cancelled if an invalid string is
4528 received. Setting client_ignore_invalid_base64 true ignores these errors.
4529 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4530 they are received. Thus, the responses can be made to depend on the
4531 challenges. If an invalid string is ignored, an empty string is placed in
4534 PH/46 Messages that are created by the autoreply transport now contains a
4535 References: header, in accordance with RFCs 2822 and 3834.
4537 PH/47 Added authenticated_sender_force to the smtp transport.
4539 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4541 PH/49 Installed latest patch from the Sieve maintainer.
4543 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4544 was also set, the appendfile transport was unnecessarily scanning a
4545 directory of message files (e.g. for maildir delivery) to find the count
4546 of files (along with the size), even though it did not need this
4547 information. It now does the scan only if it needs to find either the
4548 size of the count of files.
4550 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4552 PH/52 Two bugs concerned with error handling when the smtp transport is
4555 (i) Exim was not creating retry information for temporary errors given
4556 for individual recipients after the DATA command when the smtp transport
4557 was used in LMTP mode. This meant that they could be retried too
4558 frequently, and not timed out correctly.
4560 (ii) Exim was setting the flag that allows error details to be returned
4561 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4562 recipients that were returned after the DATA command.
4564 PH/53 This is related to PH/52, but is more general: for any failing address,
4565 when detailed error information was permitted to be returned to the
4566 sender, but the error was temporary, then after the final timeout, only
4567 "retry timeout exceeded" was returned. Now it returns the full error as
4568 well as "retry timeout exceeded".
4570 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4571 do this, and (what is worse) MTAs that accept it.
4573 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4574 will now be deprecated.
4576 PH/56 New os.c-cygwin from the Cygwin maintainer.
4578 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4579 formats (previously only available in exim formats via -bpr, -bpru,
4580 and -bpra. Now also available in native and exiqgrep formats)
4582 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4583 with very large, slow to parse queues
4585 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4587 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4589 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4590 responses to authentication challenges, though it was showing the
4591 challenges; (ii) I've removed the CR characters from the debug output for
4594 PH/58 Allow for the insertion of a newline as well as a space when a string
4595 is turned into more than one encoded-word during RFC 2047 encoding. The
4596 Sieve code now uses this.
4598 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4599 data_4xx, lost_connection, tls_required.
4601 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4602 message was being sent as an SMTP response.
4604 PH/61 Add -l and -k options to exicyclog.
4606 PH/62 When verifying, if an address was redirected to one new address, so that
4607 verification continued, and the new address failed or deferred after
4608 having set something in $address_data, the value of $address_data was not
4609 passed back to the ACL. This was different to the case when no
4610 redirection occurred. The value is now passed back in both cases.
4612 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4613 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4614 use login_cap.h, so on its own it isn't the distinguishing feature. The
4615 new name refers directly to the setclassresources() function.
4617 PH/65 Added configuration files for NetBSD3.
4619 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4621 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4622 is preferred over IPv4.
4624 PH/68 The bounce_return_message and bounce_return_body options were not being
4625 honoured for bounces generated during the reception of non-SMTP messages.
4626 In particular, this applied to messages rejected by the ACL. This bug has
4627 been fixed. However, if bounce_return_message is true and bounce_return_
4628 body is false, the headers that are returned for a non-SMTP message
4629 include only those that have been read before the error was detected.
4630 (In the case of an ACL rejection, they have all been read.)
4632 PH/69 The HTML version of the specification is now built in a directory called
4633 spec_html instead of spec.html, because the latter looks like a path with
4634 a MIME-type, and this confuses some software.
4636 PH/70 Catch two compiler warnings in sieve.c.
4638 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4639 function verify_get_ident() calls ip_connect() to connect a socket, but
4640 if the "connect()" function timed out, ip_connect() used to close the
4641 socket. However, verify_get_ident() also closes the socket later, and in
4642 between Exim writes to the log, which may get opened at this point. When
4643 the socket was closed in ip_connect(), the log could get the same file
4644 descriptor number as the socket. This naturally causes chaos. The fix is
4645 not to close the socket in ip_connect(); the socket should be closed by
4646 the function that creates it. There was only one place in the code where
4647 this was missing, in the iplookup router, which I don't think anybody now
4648 uses, but I've fixed it anyway.
4650 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4651 well as to direct DNS lookups. Otherwise the handling of names in host
4652 lists is inconsistent and therefore confusing.
4658 PH/01 Two changes to the default runtime configuration:
4660 (1) Move the checks for relay_from_hosts and authenticated clients from
4661 after to before the (commented out) DNS black list checks.
4663 (2) Add control=submission to the relay_from_hosts and authenticated
4664 clients checks, on the grounds that messages accepted by these
4665 statements are most likely to be submissions.
4667 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4669 (1) Generate an error if the third argument for the ${prvs expansion is
4672 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4675 (3) Reset the variables that are obtained from the first argument of
4676 ${prvscheck and used in the second argument before leaving the code,
4677 because their memory is reclaimed, so using them afterwards may do
4680 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4681 one (it's much easier than Tom thought :-).
4683 (5) Because of (4), we can now allow for the use of $prvscheck_result
4684 inside the third argument.
4686 PH/03 For some reason, the default setting of PATH when running a command from
4687 a pipe transport was just "/usr/bin". I have changed it to
4690 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4691 anything to be listed in the output from -bV.
4693 PH/05 When a filter generated an autoreply, the entire To: header line was
4694 quoted in the delivery log line, like this:
4696 => >A.N.Other <ano@some.domain> <original@ddress> ...
4698 This has been changed so that it extracts the operative address. There
4699 may be more than one such address. If so, they are comma-separated, like
4702 => >ano@some.domain,ona@other.domain <original@ddress> ...
4704 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4705 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4706 not being looked up in the rDNS to get a host name, Exim was showing the
4707 IP address twice in Received: lines, even though the IP addresses were
4708 identical. For example:
4710 Received: from [1.2.3.4] (helo=[1.2.3.4])
4712 However, if the real host name was known, it was omitting the HELO data
4713 if it matched the actual IP address. This has been tidied up so that it
4714 doesn't show the same IP address twice.
4716 PH/07 When both +timestamp and +memory debugging was on, the value given by
4717 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4718 called by the debug printing, thereby overwriting the timestamp buffer.
4719 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4721 PH/08 When the original message was included in an autoreply transport, it
4722 always said "this is a copy of the message, including all the headers",
4723 even if body_only or headers_only was set. It now gives an appropriate
4726 PH/09 Applied a patch from the Sieve maintainer which:
4728 o fixes some comments
4729 o adds the (disabled) notify extension core
4730 o adds some debug output for the result of if/elsif tests
4731 o points to the current vacation draft in the documentation
4732 and documents the missing references header update
4736 o fixes a bug in processing the envelope test (when testing
4737 multiple envelope elements, the last element determined the
4740 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4741 Electronic Mail") by including:
4743 Auto-submitted: auto-generated
4745 in the messages that it generates (bounce messages and others, such as
4746 warnings). In the case of bounce messages for non-SMTP messages, there was
4747 also a typo: it was using "Auto_submitted" (underscore instead of
4748 hyphen). Since every message generated by Exim is necessarily in response
4749 to another message, thes have all been changed to:
4751 Auto-Submitted: auto-replied
4753 in accordance with these statements in the RFC:
4755 The auto-replied keyword:
4757 - SHOULD be used on messages sent in direct response to another
4758 message by an automatic process,
4760 - MUST NOT be used on manually-generated messages,
4762 - MAY be used on Delivery Status Notifications (DSNs) and Message
4763 Disposition Notifications (MDNs),
4765 - MUST NOT be used on messages generated by automatic or periodic
4766 processes, except for messages which are automatic responses to
4769 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4770 to the default Received: header definition.
4772 PH/12 Added log selector acl_warn_skipped (default on).
4774 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4775 variables because (a) they are in the wrong storage pool and (b) even if
4776 they were copied, it wouldn't work properly because of the caching.
4778 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4779 checking when decoding. Apparently there are clients that generate
4780 overlong encoded strings. Why am I not surprised?
4782 PH/15 If the first argument of "${if match_address" was not empty, but did not
4783 contain an "@" character, Exim crashed. Now it writes a panic log message
4784 and treats the condition as false.
4786 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4788 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4789 extension "envelope-auth". The code is finished and in agreement with
4790 other implementations, but there is no documentation so far and in fact,
4791 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4792 not changing the active code.
4794 Print executed "if" and "elsif" statements when debugging is used. This
4795 helps a great deal to understand what a filter does.
4797 Document more things not specified clearly in RFC3028. I had all this
4798 sorted out, when out of a sudden new issues came to my mind. Oops."
4800 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4803 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4804 canonical form (as documented). However, after a host name lookup from
4805 the IP address, check_host() was doing a simple string comparison with
4806 addresses acquired from the DNS when checking that the found name did
4807 have the original IP as one of its addresses. Since any found IPv6
4808 addresses are likely to be in abbreviated form, the comparison could
4809 fail. Luckily, there already exists a function for doing the comparison
4810 by converting both addresses to binary, so now that is used instead of
4811 the text comparison.
4813 PH/20 There was another similar case to PH/19, when a complete host name was
4814 given in a host list; looking up its IP address could give an abbreviated
4815 form, whereas the current host's name might or might not be abbreviated.
4816 The same fix has been applied.
4822 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4823 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4826 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4827 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4829 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4831 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4832 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4833 Also, for "Auto-Submitted", treat anything other than "no" as
4834 non-personal, in accordance with RFC 3834. (Previously it treated
4835 anything starting "auto-" as non-personal.)
4837 TF/01 The control=submission/name=... option had a problem with syntax
4838 errors if the name included a slash character. The /name= option
4839 now slurps the rest of the string, so it can include any characters
4840 but it must come last in the list of options (after /sender_retain
4843 PH/05 Some modifications to the interface to the fake nameserver for the new
4851 TK/01 Added the "success_on_redirect" address verification option. See
4852 NewStuff for rationale and an example.
4854 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4856 PH/02 Patch to exigrep to allow it to work on syslog lines.
4858 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4859 fread() to skip over the body file's header line, because in Cygwin the
4860 header line is locked and is inaccessible.
4862 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4863 co-exist for some time) to make it clear that it is the Exim ID that is
4864 referenced, not the Message-ID: header line.
4866 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4867 string_format() function, because snprintf() does not exist on all
4870 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4871 ${stat: expansion item.
4873 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4874 protocol synchronization error", to keep the pedants happy.
4876 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4877 well as for IRIX systems, when gcc is being used. See the host.c source
4880 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4882 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4885 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4886 transport and to the smtp transport in LMTP mode.
4888 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4890 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4891 run a filter in a subprocess. This could lead to confusion in subsequent
4892 lookups in the parent process. There should also be a search_tidyup() at
4893 the end of the subprocess.
4895 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4896 only if the host matched helo_try_verify_hosts, which caused the
4897 verification to occur when the EHLO/HELO command was issued. The ACL just
4898 tested the remembered result. Now, if a previous verification attempt has
4899 not happened, "verify = helo" does it there and then.
4901 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4903 TK/03 Fix log output including CR from clamd.
4905 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4906 exist provoked a memory error which could cause a segfault.
4908 PH/15 Installed PCRE 6.2
4910 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4912 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4913 of the problem. Specifically, suggested +O2 rather than +O1 for the
4916 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4918 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4919 record, the port was not correctly logged when the outgoing_port log
4920 selector was set (it logged the transort's default port).
4922 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4923 fallback_hosts, and "hosts" in the smtp transport.
4925 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4926 host errors such as "Connection refused".
4928 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4929 authentication with radiusclient 0.4.9:
4931 - Error returned from rc_read_config was caught wrongly
4932 - Username/password not passed on to radius server due to wrong length.
4934 The presumption is that some radiusclient API changes for 4.51/PH/17
4935 were not taken care of correctly. The code is still untested by me (my
4936 Linux distribution still has 0.3.2 of radiusclient), but it was
4937 contributed by a Radius user.
4939 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4940 expanding the "port" option of the smtp transport.
4942 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4943 while reading a MIME header. Thanks to Tom Hughes for a patch.
4945 PH/24 Include config.h inside local_scan.h so that configuration settings are
4948 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4949 that in an ACL for RCPT (for example), you can examine exactly what was
4952 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4953 commands, but it was not correctly comparing the address with the actual
4954 client host address. Thus, it would show the EHLO address in Received:
4955 header lines when this was not necessary.
4957 PH/27 Added the % operator to ${eval:}.
4959 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4960 it should be ignoring failures (because with -C, for example, it has lost
4961 privilege). It wasn't ignoring creation failures other than "already
4964 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4967 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4968 because an input line was too long, either on its own, or by virtue of
4969 too many continuations, the temporary file was not being removed, and the
4970 return code was incorrect.
4972 PH/31 Missing "BOOL" in function definition in filtertest.c.
4974 PH/32 Applied Sieve patches from the maintainer.
4976 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4978 PH/33 Added "verify = not_blind".
4980 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4981 Local/Makefile (with some defaults set). These are used in built scripts
4982 such as exicyclog, but they have never been used in the exim_install
4983 script (though there are many overriding facilities there). I have
4984 arranged that the exim_install script now takes note of these two
4987 PH/35 Installed configuration files for Dragonfly.
4989 PH/36 When a locally submitted message by a trusted user did not contain a
4990 From: header, and the sender address was obtained from -f or from an SMTP
4991 MAIL command, and the trusted user did not use -F to supply a sender
4992 name, $originator_name was incorrectly used when constructing a From:
4993 header. Furthermore, $originator_name was used for submission mode
4994 messages from external hosts without From: headers in a similar way,
4995 which is clearly wrong.
4997 PH/37 Added control=suppress_local_fixups.
4999 PH/38 When log_selector = +received_sender was set, and the addition of the
5000 sender made the log line's construction buffer exactly full, or one byte
5001 less than full, an overflow happened when the terminating "\n" was
5004 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5005 when the result of a list match is failure because a DNS lookup failed.
5007 PH/40 RM_COMMAND is now used in the building process.
5009 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5010 the "build-* directories that it finds.
5012 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5013 address in a domain literal was a prefix of an interface address.
5015 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5016 when verifying a sender address, unless rewrite_headers is false.
5018 PH/44 Wrote a long comment about why errors_to addresses are verified as
5019 recipients, not senders.
5021 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5022 the ratelimit ACL was added.
5024 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5026 PH/46 Added extra information about PostgreSQL errors to the error string.
5028 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5029 avoiding the need to install special zones in a real server. This is
5030 backwards compatible; if it can't find the fake resolver, it drops back.
5031 Thus, both old and new test suites can be run.
5033 TF/02 Added util/ratelimit.pl
5035 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5036 clock is set back in time.
5038 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5039 Candler <B.Candler@pobox.com>.
5041 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5042 OK for addresses that are the result of redirections.
5044 PH/48 A number of further additions for the benefit of the new test suite,
5045 including a fake gethostbyname() that interfaces to the fake DNS resolver
5048 TF/06 The fix for widen_domains has also been applied to qualify_single and
5049 search_parents which are the other dnslookup options that can cause
5052 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5055 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5057 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5058 service" if it encountered an MX record with an empty target hostname.
5059 The message is now "an MX or SRV record indicated no SMTP service".
5061 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5062 if the DNS of the sending site is misconfigured. This is quite a
5063 common situation. This change restores the behaviour of treating a
5064 helo verification defer as a failure.
5066 PH/51 If self=fail was set on a router, the bounce message did not include the
5067 actual error message.
5073 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5075 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5076 command itself ended in error, the underlying message about the transport
5077 filter timeout was being overwritten with the pipe command error. Now the
5078 underlying error message should be appended to the second error message.
5080 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5082 PH/02 Reduce the amount of output that "make" produces by default. Full output
5083 can still be requested.
5085 PH/03 The warning log line about a condition test deferring for a "warn" verb
5086 was being output only once per connection, rather than after each
5087 occurrence (because it was using the same function as for successful
5088 "warn" verbs). This seems wrong, so I have changed it.
5090 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5091 it should not have, which might have caused a crash in the right
5092 circumstances, but probably never did.
5094 PH/04 Installed a modified version of Tony Finch's patch to make submission
5095 mode fix the return path as well as the Sender: header line, and to
5096 add a /name= option so that you can make the user's friendly name appear
5099 TF/03 Added the control = fakedefer ACL modifier.
5101 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5102 Mark Lowes for thorough testing.
5104 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5106 TK/03 Merged latest SRS patch from Miles Wilton.
5108 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5109 with the definition in sysexits.h (which is #included earlier).
5110 Fortunately, Exim does not actually use EX_OK. The code used to try to
5111 preserve the sysexits.h value, by assuming that macro definitions were
5112 scanned for macro replacements. I have been disabused of this notion,
5113 so now the code just undefines EX_OK before #including unistd.h.
5115 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5116 in the smtp transport. When a block could not be written in a single
5117 write() function, the timeout was being re-applied to each part-write.
5118 This seems wrong - if the receiver was accepting one byte at a time it
5119 would take for ever. The timeout is now adjusted when this happens. It
5120 doesn't have to be particularly precise.
5122 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5123 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5125 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5126 without a domain if the check to <postmaster@domain> fails.
5128 SC/01 Eximstats: added -xls and the ability to specify output files
5129 (patch written by Frank Heydlauf).
5131 SC/02 Eximstats: use FileHandles for outputting results.
5133 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5135 SC/04 Eximstats: fixed display of large numbers with -nvr option
5137 SC/05 Eximstats: fixed merging of reports with empty tables.
5139 SC/06 Eximstats: added the -include_original_destination flag
5141 SC/07 Eximstats: removed tabs and trailing whitespace.
5143 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5145 TK/06 MBOX spool code: Add real "From " MBOX separator line
5146 so the .eml file is really in mbox format (even though
5147 most programs do not really care). Patch from Alex Miller.
5149 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5150 The latter is generated from $received_to and is only set if the
5151 message has one envelope recipient. SA can use these headers,
5152 obviously out-of-the-box. Patch from Alex Miller.
5154 PH/08 The ${def test on a variable was returning false if the variable's
5155 value was "0", contrary to what the specification has always said!
5156 The result should be true unless the variable is empty.
5158 PH/09 The syntax error of a character other than { following "${if
5159 def:variable_name" (after optional whitespace) was not being diagnosed.
5160 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5161 accidental colon was present, for example, could give incorrect results.
5163 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5164 result is used (not including appendfile, where other changes are about
5167 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5168 This involved changing a lot of size variables from int to off_t. It
5169 should work with maildirs and everything.
5171 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5172 spamd dying while we are connected to it.
5174 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5177 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5178 function for BDB 4.3.
5180 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5182 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5183 the sources for PCRE have been re-arranged and more files are now
5186 PH/15 The code I had for printing potentially long long variables in PH/11
5187 above was not the best (it lost precision). The length of off_t variables
5188 is now inspected at build time, and an appropriate printing format (%ld
5189 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5190 to be "long long int" or "long int". This is needed for the internal
5191 formatting function string_vformat().
5193 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5194 the configuration file to be ":syslog", then the script "guesses" where
5195 the logs files are, rather than using the compiled in default. In our
5196 case the guess is not the same as the compiled default, so the script
5197 suddenly stopped working when I started to use syslog. The patch checks
5198 to see if log_file_path is "". If so, it attempts to read it from exim
5199 with no configuration file to get the compiled in version, before it
5200 falls back to the previous guessing code."
5202 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5203 implementing BATV in an Exim configuration. See NewStuff for the gory
5206 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5207 Makefile that are specific to HP-UX.
5209 PH/18 If the "use_postmaster" option was set for a recipient callout together
5210 with the "random" option, the postmaster address was used as the MAIL
5211 FROM address for the random test, but not for the subsequent recipient
5212 test. It is now used for both.
5214 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5215 patch removes a few documentation additions to RFC 3028, because the
5216 latest draft now contains them. It adds the new en;ascii-case comparator
5217 and a new error check for 8bit text in MIME parts. Comparator and
5218 require names are now matched exactly. I enabled the subaddress
5219 extension, but it is not well tested yet (read: it works for me)."
5221 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5222 rework some of the code of TK/09 above to avoid the hardwired use of
5223 "%lld" and "long long". Replaced the call to snprintf() with a call to
5226 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5227 records point to non-existent hosts", "retry timeout exceeded", and
5228 "retry time not reached for any host after a long failure period".
5230 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5231 experimental DomainKeys support:
5233 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5234 (2) On an error such as an illegally used "control", the wrong name for
5235 the control was given.
5237 These problems did NOT occur unless DomainKeys support was compiled.
5239 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5241 PH/24 Added ${if match_ip condition.
5243 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5244 databases so that it will be absolutely obvious if a crash occurs in the
5245 DB library. This is a regular occurrence (often caused by mis-matched
5248 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5249 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5250 picked up on a user's system that detects such things. There doesn't seem
5251 to be a gcc warning option for this - only an attribute that has to be
5252 put on the function's prototype. It seems that in Fedora Core 4 they have
5253 set this on a number of new functions. No doubt there will be more in due
5256 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5257 specify a transport. However, if an address that was verified by such a
5258 router was the subject of a callout, Exim crashed because it tried to
5259 read the rcpt_include_affixes from the non-existent transport. Now it
5260 just assumes that the setting of that option is false. This bug was
5261 introduced by 4.51/PH/31.
5263 PH/28 Changed -d+all to exclude +memory, because that information is very
5264 rarely of interest, but it makes the output a lot bigger. People tend to
5265 do -d+all out of habit.
5267 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5268 code in os-type was giving problems when libc.so lives in lib64, like on
5271 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5272 aren't the modern standard, and it seems that some systems' include files
5273 don't always have them. Exim was already checking for some of the newer
5274 ones like T_AAAA, and defining it itself. I've added checks for all the
5275 record types that Exim uses.
5277 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5278 not automatically generating a new one, as it is supposed to. This
5279 prevented TLS from working. If the file did exist, but contained invalid
5280 data, a new version was generated, as expected. It was only the case of a
5281 non-existent file that was broken.
5283 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5284 with a change in libdomainkeys > 0.64.
5286 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5287 from DNS. If the selector record carries the flag, it now has
5288 precedence over the domain-wide flag.
5290 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5292 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5293 the use of a port name that isn't defined in /etc/services) occurred, the
5294 message was deferred as in a normal delivery, and thus remained on the
5295 spool, instead of being failed because of the mua_wrapper setting. This
5296 is now fixed, and I tidied up some of the mua_wrapper messages at the
5299 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5300 the messages in a hash of arrays rather than using individual hashes.
5301 This is a bit cleaner and results in dramatic memory savings, albeit
5302 at a slight CPU cost.
5304 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5305 as requested by Marc Sherman.
5307 SC/10 Eximstats: added histograms for user specified patterns as requested
5310 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5312 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5313 fopen() in the content-scanning modules that did not already have it.
5319 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5320 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5322 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5324 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5326 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5327 to test Sieve filters that use "vacation".
5329 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5330 that changes the way the GnuTLS parameters are stored in the cache file.
5331 The new format can be generated externally. For backward compatibility,
5332 if the data in the cache doesn't make sense, Exim assumes it has read an
5333 old-format file, and it generates new data and writes a new file. This
5334 means that you can't go back to an older release without removing the
5337 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5338 work if there are any delivery delays because "one_time" forces the
5339 parent to be marked "delivered", so its unseen clone is never tried
5340 again. For this reason, Exim now forbids the simultaneous setting of
5343 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5344 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5345 entry if you want to know the details. The fix, however, neglected to
5346 consider the case when local delivery batching is involved. The test for
5347 "previously delivered" was not happening when checking to see if an
5348 address could be batched with a previous (undelivered) one; under
5349 certain circumstances this could lead to multiple deliveries to the same
5352 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5353 in its include files, and this causes problems building Exim.
5355 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5356 header_syntax) but Exim was just ignoring anything given after a slash.
5357 In particular, this caused confusion with an attempt to use "verify =
5358 reverse_host_lookup/defer_ok". An error is now given when options are
5359 supplied for verify items that do not have them. (Maybe reverse_host_
5360 lookup should have a defer_ok option, but that's a different point.)
5362 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5363 defined by RFC 821) to 2048, because there were problems with some AUTH
5364 commands, and RFC 1869 says the size should be increased for extended
5365 SMTP commands that take arguments.
5367 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5370 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5371 "unknown" error; now it says that the functionality isn't in the binary.
5373 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5374 an address' error message when a string expansion fails (syntax or
5375 whatever). Otherwise the password may appear in the log. Following change
5376 PH/42 below, there is no longer a chance of it appearing in a bounce
5379 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5381 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5382 of its addresses was ever tried. (Bugzilla bug #2.)
5384 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5385 the result incorrectly in the debug output. (It correctly added a newline
5386 to what was transported.)
5388 TF/01 Added $received_time.
5390 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5391 commented out examples of how to interface to a virus scanner and to
5392 SpamAssassin. Also added commented examples of av_scanner and
5393 spamd_address settings.
5395 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5396 and controls are allowed in which ACLs. There were a couple of minor
5397 errors. Some of the entries in the conditions table (which is a table of
5398 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5399 negation of where the condition IS allowed.
5401 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5403 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5404 header file does not have a version number, so I've had to invent a new
5405 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5406 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5407 radiusclient), but it was contributed by a Radius user.
5409 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5410 files or directories, for OpenSSL.
5412 PH/19 When an Exim process that is running as root has to create an Exim log
5413 file, it does so in a subprocess that runs as exim:exim so as to get the
5414 ownership right at creation (otherwise, other Exim processes might see
5415 the file with the wrong ownership). There was no test for failure of this
5416 fork() call, which would lead to the process getting stuck as it waited
5417 for a non-existent subprocess. Forks do occasionally fail when resources
5418 run out. I reviewed all the other calls to fork(); they all seem to check
5421 PH/20 When checking for unexpected SMTP input at connect time (before writing
5422 the banner), Exim was not dealing correctly with a non-positive return
5423 from the read() function. If the client had disconnected by this time,
5424 the result was a log entry for a synchronization error with an empty
5425 string after "input=" when read() returned zero. If read() returned -1
5426 (an event I could not check), uninitialized data bytes were printed.
5427 There were reports of junk text (parts of files, etc) appearing after
5430 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5432 PH/22 Added support for macro redefinition, and (re)definition in between
5433 driver and ACL definitions.
5435 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5436 forgetting to use the resulting value; it was using the unexpanded value.
5438 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5439 hadn't been configured. The fix is from Juergen Kreileder, who
5440 understands it better than I do:
5442 "Here's what I see happening with three configured cyrus_sasl
5443 authenticators configured (plain, login, cram-md5):
5445 On startup auth_cyrus_sasl_init() gets called for each of these.
5446 This means three calls to sasl_listmech() without a specified mech_list.
5447 => SASL tests which mechs of all available mechs actually work
5448 => three warnings about OTP not working
5449 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5451 With the patch, sasl_listmech() also gets called three times. But now
5452 SASL's mech_list option is set to the server_mech specified in the the
5453 authenticator. Or in other words, the answer from sasl_listmech()
5454 gets limited to just the mech you're testing for (which is different
5456 => the return list contains just 'plain' or 'login', 'cram-md5' or
5457 nothing depending on the value of ob->server_mech.
5459 I've just tested the patch: Authentication still works fine,
5460 unavailable mechs specified in the exim configuration are still
5461 caught, and the auth.log warnings about OTP are gone."
5463 PH/25 When debugging is enabled, the contents of the command line are added
5464 to the debugging output, even when log_selector=+arguments is not
5467 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5468 answer is "GNU", and only if the return is "GNU/something" is the answer
5471 PH/27 $acl_verify_message is now set immediately after the failure of a
5472 verification in an ACL, and so is available in subsequent modifiers. In
5473 particular, the message can be preserved by coding like this:
5475 warn !verify = sender
5476 set acl_m0 = $acl_verify_message
5478 Previously, $acl_verify_message was set only while expanding "message"
5479 and "log_message" when a very denied access.
5481 PH/28 Modified OS/os.c-Linux with
5483 -#ifndef OS_LOAD_AVERAGE
5484 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5486 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5487 nomenclature these days.)
5489 PH/29 Installed patch from the Sieve maintainer that adds the options
5490 sieve_useraddress and sieve_subaddress to the redirect router.
5492 PH/30 In these circumstances:
5493 . Two addresses routed to the same list of hosts;
5494 . First host does not offer TLS;
5495 . First host accepts first address;
5496 . First host gives temporary error to second address;
5497 . Second host offers TLS and a TLS session is established;
5498 . Second host accepts second address.
5499 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5500 and peerdn, if requested) that were in fact used only for the second
5503 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5504 attention to any local part prefix or suffix that was matched by the
5505 router that accepted the address. It now behaves in the same way as it
5506 does for delivery: the affixes are removed from the local part unless
5507 rcpt_include_affixes is set on the transport.
5509 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5510 timeout during the DATA phase of an incoming message.
5512 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5513 applied a patch sanctioned by the Sieve maintainer.
5515 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5516 the uid or gid is negative. A case of a negative gid caused this to be
5517 noticed. The fix allows for either to be negative.
5519 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5520 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5523 PH/36 The change PH/12 above was broken. Fixed it.
5525 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5526 the grounds that routing the same address twice would always produce the
5527 same answer. This might have been true once, but it is certainly no
5528 longer true now. Routing a child address may depend on the previous
5529 routing that produced that child. Some complicated redirection strategies
5530 went wrong when messages had multiple recipients, and made Exim's
5531 behaviour dependent on the order in which the addresses were given.
5533 I have moved the duplicate checking until after the routing is complete.
5534 Exim scans the addresses that are assigned to local and remote
5535 transports, and removes any duplicates. This means that more work will be
5536 done, as duplicates will always all be routed, but duplicates are
5537 presumably rare, so I don't expect this is of any significance.
5539 For deliveries to pipes, files, and autoreplies, the duplicate checking
5540 still happens during the routing process, since they are not going to be
5543 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5544 It corrects a timeout issue with spamd. This is Ian's comment: "The
5545 background is that sometimes spamd either never reads data from a
5546 connection it has accepted, or it never writes response data. The exiscan
5547 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5548 blindly assumes that writes won't block so it may never time out."
5550 PH/39 Allow G after quota size as well as K and M.
5552 PH/40 The value set for $authenticated_id in an authenticator may not contain
5553 binary zeroes or newlines because the value is written to log lines and
5554 to spool files. There was no check on this. Now the value is run through
5555 the string_printing() function so that such characters are converted to
5556 printable escape sequences.
5558 PH/41 $message_linecount is a new variable that contains the total number of
5559 lines in the message. Compare $body_linecount, which is the count for the
5562 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5563 bounce and delay warning messages, except in certain special cases, which
5566 (a) An SMTP error message from a remote host;
5567 (b) A message specified in a :fail: redirection;
5568 (c) A message specified in a "fail" command in a system filter;
5569 (d) A message specified in a FAIL return from the queryprogram router;
5570 (e) A message specified by the cannot_route_message router option.
5572 In these cases only, Exim does include the error details in bounce and
5573 warning messages. There are also a few cases where bland messages such
5574 as "unrouteable address" or "local delivery error" are given.
5576 PH/43 $value is now also set for the "else" part of a ${run expansion.
5578 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5579 being worked on, but at least Exim now implements the latest version to
5582 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5583 process to complete was treated as a delivery failure, a timeout while
5584 writing the message to the pipe was logged, but erroneously treated as a
5585 successful delivery. Such timeouts include transport filter timeouts. For
5586 consistency with the overall process timeout, these timeouts are now
5587 treated as errors, giving rise to delivery failures by default. However,
5588 there is now a new Boolean option for the pipe transport called
5589 timeout_defer, which, if set TRUE, converts the failures into defers for
5590 both kinds of timeout. A transport filter timeout is now identified in
5593 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5594 systems where "make" and "gmake" are different, calling "gmake" at top
5595 level broke things. I've arranged for the value of $(MAKE) to be passed
5596 from the Makefile to this script so that it can call the same version of
5600 A note about Exim versions 4.44 and 4.50
5601 ----------------------------------------
5603 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5604 changes of various kinds. As a consequence, a big documentation update was
5605 needed. This delayed the release for rather longer than seemed good, especially
5606 in the light of a couple of (minor) security issues. Therefore, the changes
5607 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5608 release. So 4.44 and 4.50 are in effect two different branches that both start
5611 I have left the 4.50 change log unchanged; it contains all the changes since
5612 4.43. The change log for 4.44 is below; many of its items are identical to
5613 those for 4.50. This seems to be the most sensible way to preserve the
5614 historical information.
5620 1. Minor wording change to the doc/README.SIEVE file.
5622 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5623 computation of the current number of files was incorrect.
5625 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5626 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5629 4. Give more explanation in the error message when the command for a transport
5630 filter fails to execute.
5632 5. There are several places where Exim runs a non-Exim command in a
5633 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5634 was being done only for the command run by the queryprogram router. It is
5635 now done for all such subprocesses. The other cases are: ${run, transport
5636 filters, and the commands run by the lmtp and pipe transports.
5638 6. Added CONFIGURE_GROUP build-time option.
5640 7. Some older OS have a limit of 256 on the maximum number of file
5641 descriptors. Exim was using setrlimit() to set 1000 as a large value
5642 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5643 systems. I've change it so that if it can't get 1000, it tries for 256.
5645 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5646 was an oversight, and furthermore, ever since the addition of extra
5647 controls (e.g. 4.43/32), the checks on when to allow different forms of
5648 "control" were broken. There should now be diagnostics for all cases when a
5649 control that does not make sense is encountered.
5651 9. Added the /retain_sender option to "control=submission".
5653 10. $recipients is now available in the predata ACL (oversight).
5655 11. Tidy the search cache before the fork to do a delivery from a message
5656 received from the command line. Otherwise the child will trigger a lookup
5657 failure and thereby defer the delivery if it tries to use (for example) a
5658 cached ldap connection that the parent has called unbind on.
5660 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5661 of $address_data from the recipient verification was clobbered by the
5662 sender verification.
5664 13. The value of address_data from a sender verification is now available in
5665 $sender_address_data in subsequent conditions in the ACL statement.
5667 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5669 15. Added a new option "connect=<time>" to callout options, to set a different
5672 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5673 was its contents. (It was OK if the option was not defined at all.)
5675 17. A "Completed" log line is now written for messages that are removed from
5676 the spool by the -Mrm option.
5678 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5679 information about exactly what failed.
5681 19. Added -dd to debug only the daemon process.
5683 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5684 handles timeouts, both on the server side and network timeouts. Renamed the
5685 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5687 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5688 It is now set to "smtps".
5690 22. $host_address is now set to the target address during the checking of
5691 ignore_target_hosts.
5693 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5694 being passed; this would have caused $sender_host_name to have been used if
5695 matching the list had actually called for a host name (not very likely,
5696 since this list is usually IP addresses). A host name is now passed as
5699 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5700 code that specifies a non-restarting handler (typically sigaction() in
5701 modern systems) in an attempt to fix a rare and obscure crash bug.
5703 25. Narrowed the window for a race in the daemon that could cause it to ignore
5704 SIGCHLD signals. This is not a major problem, because they are used only to
5705 wake it up if nothing else does.
5707 26. A malformed maildirsize file could cause Exim to calculate negative values
5708 for the mailbox size or file count. Odd effects could occur as a result.
5709 The maildirsize information is now recalculated if the size or filecount
5712 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5713 support for a long time. Removed HAVE_SYS_VFS_H.
5715 28. Installed the latest version of exipick from John Jetmore.
5717 29. In an address list, if the pattern was not a regular expression, an empty
5718 subject address (from a bounce message) matched only if the pattern was an
5719 empty string. Non-empty patterns were not even tested. This was the wrong
5720 because it is perfectly reasonable to use an empty address as part of a
5721 database query. An empty address is now tested by patterns that are
5722 lookups. However, all the other forms of pattern expect the subject to
5723 contain a local part and a domain, and therefore, for them, an empty
5724 address still always fails if the pattern is not itself empty.
5726 30. Exim went into a mad DNS loop when attempting to do a callout where the
5727 host was specified on an smtp transport, and looking it up yielded more
5728 than one IP address.
5730 31. Re-factored the code for checking spool and log partition space into a
5731 function that finds that data and another that does the check. The former
5732 is then used to implement four new variables: $spool_space, $log_space,
5733 $spool_inodes, and $log_inodes.
5735 32. The RFC2047 encoding function was originally intended for short strings
5736 such as real names; it was not keeping to the 75-character limit for
5737 encoded words that the RFC imposes. It now respects the limit, and
5738 generates multiple encoded words if necessary. To be on the safe side, I
5739 have increased the buffer size for the ${rfc2047: expansion operator from
5742 33. It is now permitted to omit both strings after an "if" condition; if the
5743 condition is true, the result is "true". As before, when the second string
5744 is omitted, a false condition yields an empty string. This makes it less
5745 cumbersome to write custom ACL and router conditions.
5747 34. Failure to deliver a bounce message always caused it to be frozen, even if
5748 there was an errors_to setting on the router. The errors_to setting is now
5751 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5752 canonical form (fully expanded) before being placed in
5753 $sender_host_address.
5755 36. The table in the code that translates DNS record types into text (T_A to
5756 "A" for instance) was missing entries for NS and CNAME. It is just possible
5757 that this could have caused confusion if both these types were looked up
5758 for the same domain, because the text type is used as part of Exim's
5759 per-process caching. But the chance of anyone hitting this buglet seems
5762 37. The dnsdb lookup has been extended in a number of ways.
5764 (1) There is a new type, "zns", which walks up the domain tree until it
5765 finds some nameserver records. It should be used with care.
5767 (2) There is a new type, "mxh", which is like "mx" except that it returns
5768 just the host names, not the priorities.
5770 (3) It is now possible to give a list of domains (or IP addresses) to be
5771 looked up. The behaviour when one of the lookups defers can be
5772 controlled by a keyword.
5774 (4) It is now possible to specify the separator character for use when
5775 multiple records are returned.
5777 38. The dnslists ACL condition has been extended: it is now possible to supply
5778 a list of IP addresses and/or domains to be looked up in a particular DNS
5781 39. Added log_selector=+queue_time_overall.
5783 40. When running the queue in the test harness, wait just a tad after forking a
5784 delivery process, to get repeatability of debugging output.
5786 41. Include certificate and key file names in error message when GnuTLS fails
5787 to set them up, because the GnuTLS error message doesn't include the name
5788 of the failing file when there is a problem reading it.
5790 42. Allow both -bf and -bF in the same test run.
5792 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5794 44. The "Exiscan patch" is now merged into the mainline Exim source.
5796 45. Sometimes the final signoff response after QUIT could fail to get
5797 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5798 < 0 before doing a fflush(). This bug looks as though it goes back to the
5799 introduction of TLS in release 3.20, but "sometimes" must have been rare
5800 because the tests only now provoked it.
5802 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5803 (this can affect the format of dates).
5805 47. exim_tidydb, when checking for the continued existence of a message for
5806 which it has found a message-specific retry record, was not finding
5807 messages that were in split spool directories. Consequently, it was
5808 deleting retry records that should have stayed in existence.
5810 48. Steve fixed some bugs in eximstats.
5812 49. The SPA authentication driver was not abandoning authentication and moving
5813 on to the next authenticator when an expansion was forced to fail,
5814 contradicting the general specification for all authenticators. Instead it
5815 was generating a temporary error. It now behaves as specified.
5817 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5818 (the order specifies the preference for clients). The order is now AES256,
5819 AES128, 3DES, ARCFOUR128.
5821 51. Small patch to Sieve code - explicitly set From: when generating an
5824 52. Exim crashed if a remote delivery caused a very long error message to be
5825 recorded - for instance if somebody sent an entire SpamAssassin report back
5826 as a large number of 550 error lines. This bug was coincidentally fixed by
5827 increasing the size of one of Exim's internal buffers (big_buffer) that
5828 happened as part of the Exiscan merge. However, to be on the safe side, I
5829 have made the code more robust (and fixed the comments that describe what
5832 53. Now that there can be additional text after "Completed" in log lines (if
5833 the queue_time_overall log selector is set), a one-byte patch to exigrep
5834 was needed to allow it to recognize "Completed" as not the last thing in
5837 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5838 patch that reportedly fixes this has been added. I am not expert enough to
5839 create a test for it. This is what the patch creator wrote:
5841 "I found a little strange behaviour of ldap code when working with
5842 Windows 2003 AD Domain, where users was placed in more than one
5843 Organization Units. When I tried to give exim partial DN, the exit code
5844 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5845 But simultaneously result of request was absolutely normal ldap result,
5846 so I produce this patch..."
5848 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5849 REFERENCE, so I have modified the code to exclude the patch when that macro
5852 55. Some experimental protocols are using DNS PTR records for new purposes. The
5853 keys for these records are domain names, not reversed IP addresses. The
5854 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5855 leaves it alone. Component reversal etc. now happens only for IP addresses.
5858 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5860 57. Double the size of the debug message buffer (to 2048) so that more of very
5861 long debug lines gets shown.
5863 58. The exicyclog utility now does better if the number of log files to keep
5864 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5866 59. Two changes related to the smtp_active_hostname option:
5868 (1) $smtp_active_hostname is now available as a variable.
5869 (2) The default for smtp_banner uses $smtp_active_hostname instead
5870 of $primary_hostname.
5872 60. The host_aton() function is supposed to be passed a string that is known
5873 to be a valid IP address. However, in the case of IPv6 addresses, it was
5874 not checking this. This is a hostage to fortune. Exim now panics and dies
5875 if the condition is not met. A case was found where this could be provoked
5876 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5877 components; fortuitously, this particular loophole had already been fixed
5878 by change 4.50/55 above.
5880 If there are any other similar loopholes, the new check in host_aton()
5881 itself should stop them being exploited. The report I received stated that
5882 data on the command line could provoke the exploit when Exim was running as
5883 exim, but did not say which command line option was involved. All I could
5884 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5885 running as the user.
5888 61. There was a buffer overflow vulnerability in the SPA authentication code
5889 (which came originally from the Samba project). I have added a test to the
5890 spa_base64_to_bits() function which I hope fixes it.
5893 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5894 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5896 63. The daemon start-up calls getloadavg() while still root for those OS that
5897 need the first call to be done as root, but it missed one case: when
5898 deliver_queue_load_max is set with deliver_drop_privilege. This is
5899 necessary for the benefit of the queue runner, because there is no re-exec
5900 when deliver_drop_privilege is set.
5902 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5903 This has been fixed.
5905 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5906 was in use, was not putting the data itself into the right store pool;
5907 consequently, it could be overwritten for a subsequent message in the same
5908 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5911 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5913 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5914 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5915 the function were treating the return as a boolean value, which happened to
5916 work because 0=false and not-0=true, but is not correct code.
5918 68. The host_aton() function was not handling scoped IPv6 addresses (those
5919 with, for example, "%eth0" on the end) correctly.
5921 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5922 negated items (that is, ~something) in unsigned ints. Some compilers
5923 apparently mutter when there is no cast.
5925 70. If an address verification called from an ACL failed, and did not produce a
5926 user-specific message (i.e. there was only a "system" message), nothing was
5927 put in $acl_verify_message. In this situation, it now puts the system
5930 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5931 session; change 4.31/43 added the unwanted input to the log line - except
5932 that it did not do this in the start of session case. It now does.
5934 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5935 This is wrong and can cause the other end to generate a synchronization
5936 error if it is another Exim or anything else that does the synchronization
5937 check. A QUIT command is no longer sent after a timeout.
5939 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5940 during host lookups.
5942 74. The defer_ok option of callout verification was not working if it was used
5943 when verifying addresses in header lines, that is, for this case:
5945 verify = header_sender/callout=defer_ok
5947 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5948 those file descriptors could be used for SMTP connections. If anything
5949 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5950 could be sent to the SMTP client, causing chaos. The daemon now opens
5951 stdin, stdout, and stderr to /dev/null when it puts itself into the
5954 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5955 log by default. The user can override this with suitable Perl magic.
5957 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5958 the log message when discard triggers, was not working for the DATA ACL or
5959 for the non-SMTP ACL.
5961 78. Error message wording change in sieve.c.
5963 79. If smtp_accept_max_per_host was set, the number of connections could be
5964 restricted to fewer than expected, because the daemon was trying to set up
5965 a new connection before checking whether the processes handling previous
5966 connections had finished. The check for completed processes is now done
5967 earlier. On busy systems, this bug wouldn't be noticed because something
5968 else would have woken the daemon, and it would have reaped the completed
5971 80. If a message was submitted locally by a user whose login name contained one
5972 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5973 It caused a spool format error. I have fixed the spool reading code. A
5974 related problem was that the "from" clause in the Received: line became
5975 illegal because of the space(s). It is now covered by ${quote_local_part.
5977 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5980 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5982 83. Patch from Sieve maintainer to fix unterminated string problem in
5983 "vacation" handling.
5985 84. Some minor changes to the Linux configuration files to help with other
5986 OS variants using glibc.
5988 85. One more patch for Sieve to update vacation handling to latest spec.
5991 ----------------------------------------------------
5992 See the note above about the 4.44 and 4.50 releases.
5993 ----------------------------------------------------
5999 1. Change 4.43/35 introduced a bug that caused file counts to be
6000 incorrectly computed when quota_filecount was set in an appendfile
6003 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6004 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6007 3. Give more explanation in the error message when the command for a transport
6008 filter fails to execute.
6010 4. There are several places where Exim runs a non-Exim command in a
6011 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6012 was being done only for the command run by the queryprogram router. It is
6013 now done for all such subprocesses. The other cases are: ${run, transport
6014 filters, and the commands run by the lmtp and pipe transports.
6016 5. Some older OS have a limit of 256 on the maximum number of file
6017 descriptors. Exim was using setrlimit() to set 1000 as a large value
6018 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6019 systems. I've change it so that if it can't get 1000, it tries for 256.
6021 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6022 was an oversight, and furthermore, ever since the addition of extra
6023 controls (e.g. 4.43/32), the checks on when to allow different forms of
6024 "control" were broken. There should now be diagnostics for all cases when a
6025 control that does not make sense is encountered.
6027 7. $recipients is now available in the predata ACL (oversight).
6029 8. Tidy the search cache before the fork to do a delivery from a message
6030 received from the command line. Otherwise the child will trigger a lookup
6031 failure and thereby defer the delivery if it tries to use (for example) a
6032 cached ldap connection that the parent has called unbind on.
6034 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6035 of $address_data from the recipient verification was clobbered by the
6036 sender verification.
6038 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6039 was its contents. (It was OK if the option was not defined at all.)
6041 11. A "Completed" log line is now written for messages that are removed from
6042 the spool by the -Mrm option.
6044 12. $host_address is now set to the target address during the checking of
6045 ignore_target_hosts.
6047 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6048 being passed; this would have caused $sender_host_name to have been used if
6049 matching the list had actually called for a host name (not very likely,
6050 since this list is usually IP addresses). A host name is now passed as
6053 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6054 code that specifies a non-restarting handler (typically sigaction() in
6055 modern systems) in an attempt to fix a rare and obscure crash bug.
6057 15. Narrowed the window for a race in the daemon that could cause it to ignore
6058 SIGCHLD signals. This is not a major problem, because they are used only to
6059 wake it up if nothing else does.
6061 16. A malformed maildirsize file could cause Exim to calculate negative values
6062 for the mailbox size or file count. Odd effects could occur as a result.
6063 The maildirsize information is now recalculated if the size or filecount
6066 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6067 support for a long time. Removed HAVE_SYS_VFS_H.
6069 18. Updated exipick to current release from John Jetmore.
6071 19. Allow an empty sender to be matched against a lookup in an address list.
6072 Previously the only cases considered were a regular expression, or an
6075 20. Exim went into a mad DNS lookup loop when doing a callout where the
6076 host was specified on the transport, if the DNS lookup yielded more than
6079 21. The RFC2047 encoding function was originally intended for short strings
6080 such as real names; it was not keeping to the 75-character limit for
6081 encoded words that the RFC imposes. It now respects the limit, and
6082 generates multiple encoded words if necessary. To be on the safe side, I
6083 have increased the buffer size for the ${rfc2047: expansion operator from
6086 22. Failure to deliver a bounce message always caused it to be frozen, even if
6087 there was an errors_to setting on the router. The errors_to setting is now
6090 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6091 canonical form (fully expanded) before being placed in
6092 $sender_host_address.
6094 24. Updated eximstats to version 1.33
6096 25. Include certificate and key file names in error message when GnuTLS fails
6097 to set them up, because the GnuTLS error message doesn't include the name
6098 of the failing file when there is a problem reading it.
6100 26. Expand error message when OpenSSL has problems setting up cert/key files.
6103 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6104 (this can affect the format of dates).
6106 28. exim_tidydb, when checking for the continued existence of a message for
6107 which it has found a message-specific retry record, was not finding
6108 messages that were in split spool directories. Consequently, it was
6109 deleting retry records that should have stayed in existence.
6111 29. eximstats updated to version 1.35
6112 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6113 1.35 - bugfix such that pie charts by volume are generated correctly
6115 30. The SPA authentication driver was not abandoning authentication and moving
6116 on to the next authenticator when an expansion was forced to fail,
6117 contradicting the general specification for all authenticators. Instead it
6118 was generating a temporary error. It now behaves as specified.
6120 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6121 (the order specifies the preference for clients). The order is now AES256,
6122 AES128, 3DES, ARCFOUR128.
6124 31. Small patch to Sieve code - explicitly set From: when generating an
6127 32. Exim crashed if a remote delivery caused a very long error message to be
6128 recorded - for instance if somebody sent an entire SpamAssassin report back
6129 as a large number of 550 error lines. This bug was coincidentally fixed by
6130 increasing the size of one of Exim's internal buffers (big_buffer) that
6131 happened as part of the Exiscan merge. However, to be on the safe side, I
6132 have made the code more robust (and fixed the comments that describe what
6135 33. Some experimental protocols are using DNS PTR records for new purposes. The
6136 keys for these records are domain names, not reversed IP addresses. The
6137 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6138 leaves it alone. Component reversal etc. now happens only for IP addresses.
6141 34. The host_aton() function is supposed to be passed a string that is known
6142 to be a valid IP address. However, in the case of IPv6 addresses, it was
6143 not checking this. This is a hostage to fortune. Exim now panics and dies
6144 if the condition is not met. A case was found where this could be provoked
6145 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6146 components; fortuitously, this particular loophole had already been fixed
6147 by change 4.50/55 or 4.44/33 above.
6149 If there are any other similar loopholes, the new check in host_aton()
6150 itself should stop them being exploited. The report I received stated that
6151 data on the command line could provoke the exploit when Exim was running as
6152 exim, but did not say which command line option was involved. All I could
6153 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6154 running as the user.
6157 35. There was a buffer overflow vulnerability in the SPA authentication code
6158 (which came originally from the Samba project). I have added a test to the
6159 spa_base64_to_bits() function which I hope fixes it.
6162 36. The daemon start-up calls getloadavg() while still root for those OS that
6163 need the first call to be done as root, but it missed one case: when
6164 deliver_queue_load_max is set with deliver_drop_privilege. This is
6165 necessary for the benefit of the queue runner, because there is no re-exec
6166 when deliver_drop_privilege is set.
6168 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6169 was in use, was not putting the data itself into the right store pool;
6170 consequently, it could be overwritten for a subsequent message in the same
6171 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6174 38. Sometimes the final signoff response after QUIT could fail to get
6175 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6176 < 0 before doing a fflush(). This bug looks as though it goes back to the
6177 introduction of TLS in release 3.20, but "sometimes" must have been rare
6178 because the tests only now provoked it.
6184 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6185 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6186 Now it returns an integer. A number of calls were still expecting a T/F
6187 return. Fortuitously, in all cases, the tests worked in OK situations,
6188 which is the norm. However, things would have gone wrong on any write
6189 failures on the smtp file descriptor. This function is used when sending
6190 messages over SMTP and also when doing verify callouts.
6192 2. When Exim is called to do synchronous delivery of a locally submitted
6193 message (the -odf or -odi options), it no longer closes stderr before doing
6196 3. Implemented the mua_wrapper option.
6198 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6200 5. Implemented the functions header_remove(), header_testname(),
6201 header_add_at_position(), and receive_remove_recipient(), and exported them
6204 6. If an ACL "warn" statement specified the addition of headers, Exim already
6205 inserted X-ACL-Warn: at the start if there was no header name. However, it
6206 was not making this test for the second and subsequent header lines if
6207 there were newlines in the string. This meant that an invalid header could
6208 be inserted if Exim was badly configured.
6210 7. Allow an ACL "warn" statement to add header lines at the start or after all
6211 the Received: headers, as well as at the end.
6213 8. Added the rcpt_4xx retry error code.
6215 9. Added postmaster_mailfrom=xxx to callout verification option.
6217 10. Added mailfrom=xxxx to the callout verification option, for verify=
6220 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6221 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6223 12. Inserted some casts to stop certain compilers warning when using pointer
6224 differences as field lengths or precisions in printf-type calls (mostly
6225 affecting debugging statements).
6227 13. Added optional readline() support for -be (dynamically loaded).
6229 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6230 same clock tick as a message's arrival, so that its received time was the
6231 same as the "first fail" time on the retry record, and that message
6232 remained on the queue past the ultimate address timeout, every queue runner
6233 would try a delivery (because it was past the ultimate address timeout) but
6234 after another failure, the ultimate address timeout, which should have then
6235 bounced the address, did not kick in. This was a "< instead of <=" error;
6236 in most cases the first failure would have been in the next clock tick
6237 after the received time, and all would be well.
6239 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6240 being recognized when the domain list was tested by the match_domain
6241 condition in an expansion string.
6243 16. Added the ${str2b64: operator.
6245 17. Exim was always calling setrlimit() to set a large limit for the number of
6246 processes, without checking whether the existing limit was already
6247 adequate. (It did check for the limit on file descriptors.) Furthermore,
6248 errors from getrlimit() and setrlimit() were being ignored. Now they are
6249 logged to the main and panic logs, but Exim does carry on, to try to do its
6250 job under whatever limits there are.
6252 18. Imported PCRE 5.0.
6254 19. Trivial typo in log message " temporarily refused connection" (the leading
6257 20. If the log selector return_path_on_delivery was set and an address was
6258 redirected to /dev/null, the delivery process crashed because it assumed
6259 that a return path would always be set for a "successful" delivery. In this
6260 case, the whole delivery is bypassed as an optimization, and therefore no
6263 21. Internal re-arrangement: the function for sending a challenge and reading
6264 a response while authentication was assuming a zero-terminated challenge
6265 string. It's now changed to take a pointer and a length, to allow for
6266 binary data in such strings.
6268 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6270 23. Exim was not respecting finduser_retries when seeking the login of the
6271 uid under which it was called; it was always trying 10 times. (The default
6272 setting of finduser_retries is zero.) Also, it was sleeping after the final
6273 failure, which is pointless.
6275 24. Implemented tls_on_connect_ports.
6277 25. Implemented acl_smtp_predata.
6279 26. If the domain in control=submission is set empty, Exim assumes that the
6280 authenticated id is a complete email address when it generates From: or
6281 Sender: header lines.
6283 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6284 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6285 chown and chgrp in /bin and hostname in /usr/bin.
6287 28. Exim was keeping the "process log" file open after each use, just as it
6288 does for the main log. This opens the possibility of it remaining open for
6289 long periods when the USR1 signal hits a daemon. Occasional processlog
6290 errors were reported, that could have been caused by this. Anyway, it seems
6291 much more sensible not to leave this file open at all, so that is what now
6294 29. The long-running daemon process does not normally write to the log once it
6295 has entered its main loop, and it closes the log before doing so. This is
6296 so that log files can straightforwardly be renamed and moved. However,
6297 there are a couple of unusual error situations where the daemon does write
6298 log entries, and I had neglected to close the log afterwards.
6300 30. The text of an SMTP error response that was received during a remote
6301 delivery was being truncated at 512 bytes. This is too short for some of
6302 the long messages that one sometimes sees. I've increased the limit to
6305 31. It is now possible to make retry rules that apply only when a message has a
6306 specific sender, in particular, an empty sender.
6308 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6309 it possible to be selective about when SMTP synchronization is enforced.
6311 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6313 32. Implemented hosts_connection_nolog.
6315 33. Added an ACL for QUIT.
6317 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6320 35. Added mailbox_size and mailbox_filecount to appendfile.
6322 36. Added control = no_multiline_responses to ACLs.
6324 37. There was a bug in the logic of the code that waits for the clock to tick
6325 in the case where the clock went backwards by a substantial amount such
6326 that the microsecond fraction of "now" was more than the microsecond
6327 fraction of "then" (but the whole seconds number was less).
6329 38. Added support for the libradius Radius client library this is found on
6330 FreeBSD (previously only the radiusclient library was supported).
6336 1. When certain lookups returned multiple values in the form name=value, the
6337 quoting of the values was not always being done properly. Specifically:
6338 (a) If the value started with a double quote, but contained no whitespace,
6340 (b) If the value contained whitespace other than a space character (i.e.
6341 tabs or newlines or carriage returns) it was not quoted.
6342 This fix has been applied to the mysql and pgsql lookups by writing a
6343 separate quoting function and calling it from the lookup code. The fix
6344 should probably also be applied to nisplus, ibase and oracle lookups, but
6345 since I cannot test any of those, I have not disturbed their existing code.
6347 2. A hit in the callout cache for a specific address caused a log line with no
6348 reason for rejecting RCPT. Now it says "Previous (cached) callout
6349 verification failure".
6351 3. There was an off-by-one bug in the queryprogram router. An over-long
6352 return line was truncated at 256 instead of 255 characters, thereby
6353 overflowing its buffer with the terminating zero. As well as fixing this, I
6354 have increased the buffer size to 1024 (and made a note to document this).
6356 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6357 when Exim is waiting for an SMTP response from a remote server, Exim
6358 restarts its select() call on the socket, thereby resetting its timeout.
6359 This is not a problem when such interrupts are rare. Somebody set up a cron
6360 job to run exiwhat every 2 minutes, which is less than the normal select()
6361 timeout (5 or 10 minutes). This meant that the select() timeout never
6362 kicked in because it was always reset. I have fixed this by comparing the
6363 time when an interrupt arrives with the time at the start of the first call
6364 to select(). If more time than the timeout has elapsed, the interrupt is
6365 treated as a timeout.
6367 5. Some internal re-factoring in preparation for the addition of Sieve
6368 extensions (by MH). In particular, the "personal" test is moved to a
6369 separate function, and given an option for scanning Cc: and Bcc: (which is
6370 not set for Exim filters).
6372 6. When Exim created an email address using the login of the caller as the
6373 local part (e.g. when creating a From: or Sender: header line), it was not
6374 quoting the local part when it contained special characters such as @.
6376 7. Installed new OpenBSD configuration files.
6378 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6379 try to make them clearer.
6381 9. Callout options, other than the timeout value, were being ignored when
6382 verifying sender addresses in header lines. For example, when using
6384 verify = header_sender/callout=no_cache
6386 the cache was (incorrectly) being used.
6388 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6389 only the Cygwin environment.
6391 11. When return_path_on_delivery was set as a log selector, if different remote
6392 addresses in the same message used different return paths and parallel
6393 remote delivery occurred, the wrong values would sometimes be logged.
6394 (Whenever a remote delivery process finished, the return path value from
6395 the most recently started remote delivery process was logged.)
6397 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6398 lines when AUTH and/or TLS are in use. This is the "received protocol"
6399 field. Exim used to use "asmtp" for authenticated SMTP, without any
6400 indication (in the protocol name) for TLS use. Now it follows the RFC and
6401 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6402 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6403 names appear in log lines as well as in Received: header lines.
6405 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6406 extensions, and comparison tests, and to fix some bugs.
6408 14. Changes to the "personal" filter test:
6410 (1) The test was buggy in that it was just doing the equivalent of
6411 "contains" tests on header lines. For example, if a user's address was
6412 anne@some.where, the "personal" test would incorrectly be true for
6414 To: susanne@some.where
6416 This test is now done by extracting each address from the header in turn,
6417 and checking the entire address. Other tests that are part of "personal"
6418 are now done using regular expressions (for example, to check local parts
6419 of addresses in From: header lines).
6421 (2) The list of non-personal local parts in From: addresses has been
6422 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6423 taken from the Sieve specification recommendations.
6425 (3) If the message contains any header line starting with "List-" it is
6426 treated as non-personal.
6428 (4) The test for "circular" in the Subject: header line has been removed
6429 because it now seems ill-conceived.
6431 15. Minor typos in src/EDITME comments corrected.
6433 16. Installed latest exipick from John Jetmore.
6435 17. If headers_add on a router specified a text string that was too long for
6436 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6437 of string_sprintf() is now avoided.
6439 18. $message_body_size was not set (it was always zero) when running the DATA
6440 ACL and the local_scan() function.
6442 19. For the "mail" command in an Exim filter, no default was being set for
6443 the once_repeat time, causing a random time value to be used if "once" was
6444 specified. (If the value happened to be <= 0, no repeat happened.) The
6445 default is now 0s, meaning "never repeat". The "vacation" command was OK
6446 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6447 (I found it when inspecting the code).
6449 20. There is now an overall timeout for performing a callout verification. It
6450 defaults to 4 times the callout timeout, which applies to individual SMTP
6451 commands during the callout. The overall timeout applies when there is more
6452 than one host that can be tried. The timeout is checked before trying the
6453 next host. This prevents very long delays if there are a large number of
6454 hosts and all are timing out (e.g. when the network connections are timing
6455 out). The value of the overall timeout can be changed by specifying an
6456 additional sub-option for "callout", called "maxwait". For example:
6458 verify = sender/callout=5s,maxwait=20s
6460 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6461 to the end before writing, but this should make it even safer).
6463 22. Exim was forgetting that it had advertised PIPELINING for the second and
6464 subsequent messages on an SMTP connection. It was also not resetting its
6465 memory on STARTTLS and an internal HELO.
6467 23. When Exim logs an SMTP synchronization error within a session, it now
6468 records whether PIPELINING has been advertised or not.
6470 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6471 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6472 rather than long int.
6474 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6476 26. Added the never_mail option to autoreply.
6482 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6483 crash if the getsockname() call failed; this can happen if a connection is
6484 closed very soon after it is established. The problem was simply in the
6485 order in which certain operations were done, causing Exim to try to write
6486 to the SMTP stream before it had set up the file descriptor. The bug has
6487 been fixed by making things happen in the correct order.
6493 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6494 before the connection was closed, thus losing the rejection response.
6496 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6497 some early Solaris releases, but causes trouble in current releases where
6498 socklen_t is defined.
6500 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6503 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6506 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6507 that releasing the top of it at the end releases what was used for sub-
6508 expansions (unless the block got too big). However, discard this block if
6509 the first thing is a variable or header, so that we can use its block when
6510 it is dynamic (useful for very large $message_headers, for example).
6512 6. Lookups now cache *every* query, not just the most recent. A new, separate
6513 store pool is used for this. It can be recovered when all lookup caches are
6514 flushed. Lookups now release memory at the end of their result strings.
6515 This has involved some general refactoring of the lookup sources.
6517 7. Some code has been added to the store_xxx() functions to reduce the amount
6518 of flapping under certain conditions.
6520 8. log_incoming_interface used to affect only the <= reception log lines. Now
6521 it causes the local interface and port to be added to several more SMTP log
6522 lines, for example "SMTP connection from", and rejection lines.
6524 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6526 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6528 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6529 could be overwritten at the end of the current message (or the start of a
6530 new message if it was set in a HELO ACL). The value is now preserved for
6531 the duration of the SMTP connection.
6533 12. If a transport had a headers_rewrite setting, and a matching header line
6534 contained an unqualified address, that address was qualified, even if it
6535 did not match any rewriting rules. The underlying bug was that the values
6536 of the flags that permit the existence of unqualified sender and recipient
6537 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6538 non-local messages, and by -bnq for local messages) were not being
6539 preserved with the message after it was received.
6541 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6542 "next input=" as part of the text comprising the host identity instead of
6543 the correct text. The code was using the same buffer for two different
6544 strings. However, depending on which order the printing function evaluated
6545 its arguments, the bug did not always show up. Under Linux, for example, my
6546 test suite worked just fine.
6548 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6549 doesn't work with some older versions of Perl. It has been changed to "my",
6550 which in any case is probably the better facility to use.
6552 15. A really picky compiler found some instances of statements for creating
6553 error messages that either had too many or two few arguments for the format
6556 16. The size of the buffer for calls to the DNS resolver has been increased
6557 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6558 for addresses that have a lot of PTR records. This alleviates a problem; it
6559 does not fully solve it.
6561 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6562 buffer now truncates the list and logs the incident, which is the same
6563 action as happens when Exim is looking up a host name and its aliases.
6564 Previously in this situation something unpredictable would happen;
6565 sometimes it was "internal error: store_reset failed".
6567 18. If a server dropped the connection unexpectedly when an Exim client was
6568 using GnuTLS and trying to read a response, the client delivery process
6569 crashed while trying to generate an error log message.
6571 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6572 string, for example:
6574 warn message = H1: something\nH2: something
6576 the text was added as a single header line from Exim's point of view
6577 though it ended up OK in the delivered message. However, searching for the
6578 second and subsequent header lines using $h_h2: did not work. This has been
6579 fixed. Similarly, if a system filter added multiple headers in this way,
6580 the routers could not see them.
6582 20. Expanded the error message when iplsearch is called with an invalid key to
6583 suggest using net-iplsearch in a host list.
6585 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6586 ACLs are no longer actually imposed (and a message to that effect is
6589 22. If a "gecos" field in a passwd entry contained escaped characters, in
6590 particular, if it contained a \" sequence, Exim got it wrong when building
6591 a From: or a Sender: header from that name. A second bug also caused
6592 incorrect handling when an unquoted " was present following a character
6593 that needed quoting.
6595 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6596 was not being matched caselessly.
6598 24. Arranged for all hyphens in the exim.8 source to be escaped with
6601 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6602 properly. Recipient callout cache records were still being keyed to include
6603 the sender, even when use_sender was set false. This led to far more
6604 callouts that were necessary. The sender is no longer included in the key
6605 when use_sender is false.
6607 26. Added "control = submission" modifier to ACLs.
6609 27. Added the ${base62d: operator to decode base 62 numbers.
6611 28. dnsdb lookups can now access SRV records.
6613 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6614 the configuration file.
6616 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6617 (-v) mode. This makes the output for a verbose queue run more intelligible.
6619 31. Added a use_postmaster feature to recipient callouts.
6621 32. Added the $body_zerocount variable, containing the number of binary zero
6622 bytes in the message body.
6624 33. The time of last modification of the "new" subdirectory is now used as the
6625 "mailbox time last read" when there is a quota error for a maildir
6628 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6630 35. Added +ignore_unknown as a special item in host lists.
6632 36. Code for decoding IPv6 addresses in host lists is now included, even if
6633 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6634 address was recognized as an IP address, but was then not correctly decoded
6635 into binary, causing unexpected and incorrect effects when compared with
6642 1. Very minor rewording of debugging text in manualroute to say "list of
6643 hosts" instead of "hostlist".
6645 2. If verify=header_syntax was set, and a header line with an unqualified
6646 address (no domain) and a large number of spaces between the end of the
6647 name and the colon was received, the reception process suffered a buffer
6648 overflow, and (when I tested it) crashed. This was caused by some obsolete
6649 code that should have been removed. The fix is to remove it!
6651 3. When running in the test harness, delay a bit after writing a bounce
6652 message to get a bit more predictability in the log output.
6654 4. Added a call to search_tidyup() just before forking a reception process. In
6655 theory, someone could use a lookup in the expansion of smtp_accept_max_
6656 per_host which, without the tidyup, could leave open a database connection.
6658 5. Added the variables $recipient_data and $sender_data which get set from a
6659 lookup success in an ACL "recipients" or "senders" condition, or a router
6660 "senders" option, similar to $domain_data and $local_part_data.
6662 6. Moved the writing of debug_print from before to after the "senders" test
6665 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6666 problems for message scanning, either using a data ACL, or using
6667 local_scan() because the Received: header was not generated till after they
6668 were called (in order to set the time as the time of reception completion).
6669 I have revised the way this works. The header is now generated after the
6670 body is received, but before the ACL or local_scan() are called. After they
6671 are run, the timestamp in the header is updated.
6677 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6678 before starting a queue runner without re-exec. This happened only when
6679 deliver_drop_privilege was set or when the Exim user was set to root. The
6680 effect of the bug was that timeouts during subsequent deliveries caused
6681 crashes instead of being properly handled. The handler is now left at its
6682 default (and expected) setting.
6684 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6685 message, again when deliver_drop_privilege is set or Exim is run as root.
6686 The bug described in (1) was not present in this case, but the tidying up
6687 of the other signals was missing. I have made the two cases consistent.
6689 3. The ignore_target_hosts setting on a manualroute router was being ignored
6690 for hosts that were looked up using the /MX notation.
6692 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6695 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6696 operated on the sender address. After changing the $sender_address to <>
6697 for the sender address verify, Exim was re-instated it as the original
6698 (before rewriting) address, but remembering that it had rewritten it, so it
6699 wasn't rewriting it again. This bug also had the effect of breaking the
6700 sender address verification caching when the sender address was rewritten.
6702 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6703 This has been changed so that if the ip literal address matches
6704 ignore_target_hosts, the router declines.
6706 7. Added expansion conditions match_domain, match_address, and match_local_
6707 part (NOT match_host).
6709 8. The placeholder for the Received: header didn't have a length field set.
6711 9. Added code to Exim itself and to exim_lock to test for a specific race
6712 condition that could lead to file corruption when using MBX delivery. The
6713 issue is with the lockfile that is created in /tmp. If this file is removed
6714 after a process has opened it but before that process has acquired a lock,
6715 there is the potential for a second process to recreate the file and also
6716 acquire a lock. This could lead to two Exim processes writing to the file
6717 at the same time. The added code performs the same test as UW imapd; it
6718 checks after acquiring the lock that its file descriptor still refers to
6719 the same named file.
6721 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6722 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6725 11. Added the smtp_active_hostname option. If used, this will typically be made
6726 to depend on the incoming interface address. Because $interface_address is
6727 not set up until the daemon has forked a reception process, error responses
6728 that can happen earlier (such as "too many connections") no longer contain
6731 12. If an expansion in a condition on a "warn" statement fails because a lookup
6732 defers, the "warn" statement is abandoned, and the next ACL statement is
6733 processed. Previously this caused the whole ACL to be aborted.
6735 13. Added the iplsearch lookup type.
6737 14. Added ident_timeout as a log selector.
6739 15. Added tls_certificate_verified as a log selector.
6741 16. Added a global option tls_require_ciphers (compare the smtp transport
6742 option of the same name). This controls incoming TLS connections.
6744 17. I finally figured out how to make tls_require_ciphers do a similar thing
6745 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6746 before starting the TLS session.
6748 18. Tabs are now shown as \t in -bP output.
6750 19. If the log selector return_path_on_delivery was set, Exim crashed when
6751 bouncing a message because it had too many Received: header lines.
6753 20. If two routers both had headers_remove settings, and the first one included
6754 a superfluous trailing colon, the final name in the first list and the
6755 first name in the second list were incorrectly joined into one item (with a
6756 colon in the middle).
6762 1. Added -C and -D options to the exinext utility, mainly to make it easier
6763 to include in the automated testing, but these could be helpful when
6764 multiple configurations are in use.
6766 2. The exinext utility was not formatting the output nicely when there was
6767 an alternate port involved in the retry record key, nor when there was a
6768 message id as well (for retries that were specific to a specific message
6769 and a specific host). It was also confused by IPv6 addresses, because of
6770 the additional colons they contain. I have fixed the IPv4 problem, and
6771 patched it up to do a reasonable job for IPv6.
6773 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6774 delivery, the log line now contains "pipelined" if PIPELINING was used.
6776 4. An SMTP transport process used to panic and die if the bind() call to set
6777 an explicit outgoing interface failed. This has been changed; it is now
6778 treated in the same way as a connect() failure.
6780 5. A reference to $sender_host_name in the part of a conditional expansion
6781 that was being skipped was still causing a DNS lookup. This no longer
6784 6. The def: expansion condition was not recognizing references to header lines
6785 that used bh_ and bheader_.
6787 7. Added the _cache feature to named lists.
6789 8. The code for checking quota_filecount in the appendfile transport was
6790 allowing one more file than it should have been.
6792 9. For compatibility with Sendmail, the command line option
6800 and sets the incoming protocol and host name (for trusted callers). The
6801 host name and its colon can be omitted when only the protocol is to be set.
6802 Note the Exim already has two private options, -pd and -ps, that refer to
6803 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6804 "s", but I don't think that's a major issue.
6806 10. A number of refactoring changes to the code, none of which should affect
6809 (a) The number of logging options was getting close to filling up the
6810 32-bit word that was used as a bit map. I have split them into two classes:
6811 those that are passed in the argument to log_write(), and those that are
6812 only ever tested independently outside of that function. These are now in
6813 separate 32-bit words, so there is plenty of room for expansion again.
6814 There is no change in the user interface or the logging behaviour.
6816 (b) When building, for example, log lines, the code previously used a
6817 macro that called string_cat() twice, in order to add two strings. This is
6818 not really sufficiently general. Furthermore, there was one instance where
6819 it was actually wrong because one of the argument was used twice, and in
6820 one call a function was used. (As it happened, calling the function twice
6821 did not affect the overall behaviour.) The macro has been replaced by a
6822 function that can join an arbitrary number of extra strings onto a growing
6825 (c) The code for expansion conditions now uses a table and a binary chop
6826 instead of a serial search (which was left over from when there were very
6827 few conditions). Also, it now recognizes conditions like "pam" even when
6828 the relevant support is not compiled in: a suitably worded error message is
6829 given if an attempt is made to use such a condition.
6831 11. Added ${time_interval:xxxxx}.
6833 12. A bug was causing one of the ddress fields not to be passed back correctly
6834 from remote delivery subprocesses. The field in question was not being
6835 subsequently used, so this caused to problems in practice.
6837 13. Added new log selectors queue_time and deliver_time.
6839 14. Might have fixed a bug in maildirsizefile handling that threw up
6840 "unexpected character" debug warnings, and recalculated the data
6841 unnecessarily. In any case, I expanded the warning message to give more
6844 15. Added the message "Restricted characters in address" to the statements in
6845 the default ACL that block characters like @ and % in local parts.
6847 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6848 Three changes have been made:
6850 (a) There was a serious bug; a negative response to MAIL caused the whole
6851 recipient domain to be cached as invalid, thereby blocking all messages
6852 to all local parts at the same domain, from all senders. This bug has
6853 been fixed. The domain is no longer cached after a negative response to
6854 MAIL if the sender used is not empty.
6856 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6859 (c) A new callout option, "use_sender" has been added for people who want
6860 the modified behaviour.
6866 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6869 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6870 indeed breaks things for older releases.
6872 3. Added additional logging to the case where there is a problem reading data
6873 from a filter that is running in a subprocess using a pipe, in order to
6874 try to track down a specific problem.
6876 4. Testing facility fudge: when running in the test harness and attempting
6877 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6878 getting "No route to host". Convert this to a timeout.
6880 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6883 6. Some OS don't have socklen_t but use size_t instead. This affects the
6884 fifth argument of getsockopt() amongst other things. This is now
6885 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6886 can be set for individual OS. I have set it for SunOS5, OSF1, and
6887 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6888 some earlier ones do not.
6890 7. Change 4.30/15 was not doing the test caselessly.
6892 8. The standard form for an IPv6 address literal was being rejected by address
6893 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6894 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6895 this, as well as the form without the "IPv6" on the front (but only when
6896 address literals are enabled, of course).
6898 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6900 10. Exim crashed if a message with an empty sender address specified by -f
6901 encountered a router with an errors_to setting. This could be provoked only
6902 by a command such as
6906 where an empty string was supplied; "<>" did not hit this bug.
6908 11. Installed PCRE release 4.5.
6910 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6911 remained set. It is now erased.
6913 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6914 times from message ids (which are base 36 rather than the normal 62).
6916 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6917 were being counted as actual protocol errors, and logged if the log
6918 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6919 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6920 and DATA following a set of rejected RCPTs do not count as protocol errors.
6921 In other words, Exim assumes they were pipelined, though this may not
6922 actually be the case. Of course, in all cases the client gets an
6923 appropriate error code.
6925 15. If a lookup fails in an ACL condition, a message about the failure may
6926 be available; it is used if testing the ACL cannot continue, because most
6927 such messages specify what the cause of the deferral is. However, some
6928 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6929 that caused an old message to be retained and used if a later statement
6930 caused a defer, replacing the real cause of the deferral.
6932 16. If an IP address had so many PTR records that the DNS lookup buffer
6933 was not large enough to hold them, Exim could crash while trying to process
6934 the truncated data. It now detects and logs this case.
6936 17. Further to 4.21/58, another change has been made: if (and only if) the
6937 first line of a message (the first header line) ends with CRLF, a bare LF
6938 in a subsequent header line has a space inserted after it, so as not to
6939 terminate the header.
6941 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6942 unnecessarily, used atoi() instead of strtol(), and didn't check the
6943 termination when getting file sizes from file names by regex.
6945 19. Completely re-implemented the support for maildirsize files, in the light
6946 of a number of problems with the previous contributed implementation
6947 (4.30/29). In particular:
6949 . If the quota is zero, the maildirsize file is maintained, but no quota is
6952 . If the maildir directory does not exist, it is created before any attempt
6953 to write a maildirsize file.
6955 . The quota value in the file is just a cache; if the quota is changed in
6956 the transport, the new value overrides.
6958 . A regular expression is available for excluding directories from the
6961 20. The autoreply transport checks the characters in options that define the
6962 message's headers; it allows continued headers, but it was checking with
6963 isspace() after an embedded newline instead of explicitly looking for a
6966 21. If all the "regular" hosts to which an address was routed had passed their
6967 expiry times, and had not reached their retry times, the address was
6968 bounced, even if fallback hosts were defined. Now Exim should go on to try
6971 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6972 equivalent code in the SMTP transport. Some hosts send humungous responses
6973 to HELO/EHLO, more than 1024 it seems.
6975 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6976 gives compiler warnings in some environments. I've now done it "properly",
6979 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6980 (because of problems with the built-in one) was declared to return uschar *
6981 instead of char *, causing compiler failure.
6983 25. Fixed a file descriptor leak when processing alias/forward files.
6985 26. Fixed a minor format string issue in dbfn.c.
6987 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6989 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6990 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6993 29. When a sender address is verified, it is cached, to save repeating the test
6994 when there is more than one recipient in a message. However, when the
6995 verification involves a callout, it is possible for different callout
6996 options to be set for different recipients. It is too complicated to keep
6997 track of this in the cache, so now Exim always runs a verification when a
6998 callout is required, relying on the callout cache for the optimization.
6999 The overhead is duplication of the address routing, but this should not be
7002 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7003 to be verified with callout=postmaster, and the main callout worked but the
7004 postmaster check failed, the verification correctly failed. However, if a
7005 subsequent RCPT command asked for sender verification *without* the
7006 postmaster check, incorrect caching caused this verification also to fail,
7009 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7010 it was not caching the DNS options (qualify_single, search_parents) that
7011 were used when the lookup failed. A subsequent lookup with different
7012 options therefore always gave the same answer, though there were cases
7013 where it should not have. (Example: a "domains = !$mx_any" option on a
7014 dnslookup router: the "domains" option is always processed without any
7015 widening, but the router might have qualify_single set.) Now Exim uses the
7016 cached value only when the same options are set.
7018 32. Added John Jetmore's "exipick" utility to the distribution.
7020 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7021 than a timeout (e.g. a certificate is required, and is not provided), an
7022 Exim server now closes the connection immediately. Previously it waited for
7023 the client to close - but if the client is SSL, it seems that they each
7024 wait for each other, leading to a delay before one of them times out.
7026 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7027 maintained 0.8.x compatibility because I don't think many are using it, and
7028 it is clearly obsolete.
7030 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7033 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7034 client certificate was expired. A simple patch fixes this, though I don't
7035 understand the full logic of why the verify callback is called multiple
7038 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7039 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7040 which causes problems with some clients (such as the Certicom SSL Plus
7041 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7042 disables the coutermeasure allowing Eudora to connect."
7044 38. Exim was not checking that a write() to a log file succeeded. This could
7045 lead to Bad Things if a log got too big, in particular if it hit a file
7046 size limit. Exim now panics and dies if it cannot write to a log file, just
7047 as it does if it cannot open a log file.
7049 39. Modified OS/Makefile-Linux so that it now contains
7051 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7053 The two -D definitions ensure that Exim is compiled with large file
7054 support, which makes it possible to handle log files that are bigger than
7057 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7058 instance) a domain was checked against a named list that involved a lookup,
7059 causing $domain_data to be set, then another domain was checked against the
7060 same list, then the first domain was re-checked, the value of $domain_data
7061 after the final check could be wrong. In particular, if the second check
7062 failed, it could be set empty. This bug probably also applied to
7065 41. The strip_trailing_dot option was not being applied to the address given
7066 with the -f command-line option.
7068 42. The code for reading a message's header from the spool was incrementing
7069 $received_count, but never initializing it. This meant that the value was
7070 incorrect (doubled) while delivering a message in the same process in which
7071 it was received. In the most common configuration of Exim, this never
7072 happens - a fresh exec is done - but it can happen when
7073 deliver_drop_privilege is set.
7075 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7076 it now includes up to 150 characters of the unexpected data in the log
7079 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7080 and building data strings. The size of both of these buffers was 10 000
7081 bytes - far larger than anybody would *ever* want, thought I. Needless to
7082 say, somebody hit the limit. I have increased the maximum line length to
7083 20 000 and the maximum data length of concatenated lines to 100 000. I have
7084 also fixed two bugs, because there was no checking on these buffers. Tsk,
7085 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7086 buffer is too small.
7088 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7089 lsearch lookups. Now it does.
7091 46. When parsing a route_list item in a manualroute router, a fixed-length
7092 buffer was used for the list of hosts. I made this 1024 bytes long,
7093 thinking that nobody would ever have a list of hosts that long. Wrong.
7094 Somebody had a whole pile of complicated expansion conditions, and the
7095 string was silently truncated, leading to an expansion error. It turns out
7096 that it is easier to change to an unlimited length (owing to other changes
7097 that have happened since this code was originally written) than to build
7098 structure for giving a limitation error. The length of the item that
7099 expands into the list of hosts is now unlimited.
7101 47. The lsearch lookup could not handle data where the length of text line was
7102 more than 4095 characters. Such lines were truncated, leading to shortened
7103 data being returned. It should now handle lines of any length.
7105 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7106 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7109 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7110 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7111 no longer happens in comment lines. A list of replacements is now placed
7112 at the head of all of the source files, except those whose only change is
7113 to replace PERL_COMMAND in the very first #! line.
7115 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7116 messages on the queue, with a bottom-up merge sort, using code contributed
7117 by Michael Haardt. This should make operations like -bp somewhat faster on
7118 large queues. It won't affect queue runners, except when queue_run_in_order
7121 51. Installed eximstats 1.31 in the distribution.
7123 52. Added support for SRV lookups to the dnslookup router.
7125 53. If an ACL referred to $message_body or $message_body_end, the value was not
7126 reset for any messages that followed in the same SMTP session.
7128 54. The store-handling optimization for building very long strings was not
7129 differentiating between the different store pools. I don't think this
7130 actually made any difference in practice, but I've tidied it.
7132 55. While running the routers to verify a sender address, $sender_address
7133 was still set to the sender address. This is wrong, because when routing to
7134 send a bounce to the sender, it would be empty. Therefore, I have changed
7135 it so that, while verifying a sender address, $sender_address is set to <>.
7136 (There is no change to what happens when verifying a recipient address.)
7138 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7139 target A or AAAA records (if not already returned) without resetting the
7140 qualify_single or search_parents options of the DNS resolver. These are
7141 inappropriate in this case because the targets of MX and SRV records must
7142 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7143 target that, when qualified, matched something in the local domain. These
7144 two options are now turned off when doing these lookups.
7146 57. It seems that at least some releases of Reiserfs (which does not have the
7147 concept of a fixed number of inodes) returns zero and not -1 for the
7148 number of available inodes. This interacted badly with check_spool_inodes,
7149 which assumed that -1 was the "no such thing" setting. What I have done is
7150 to check that the total number of inodes is greater than zero before doing
7151 the test of how many are available.
7153 58. When a "warn" ACL statement has a log_message modifier, the message is
7154 remembered, and not repeated. This is to avoid a lot of repetition when a
7155 message has many recipients that cause the same warning to be written.
7156 However, Exim was preserving the list of already written lines for an
7157 entire SMTP session, which doesn't seem right. The memory is now reset if a
7158 new message is started.
7160 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7161 debugging output unless log_rewrite was also set.
7163 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7164 of (int)(handle) when we know that handle contains (void *)(-1).
7166 61. The Exim daemon panic-logs an error return when it closes the incoming
7167 connection. However "connection reset by peer" seems to be common, and
7168 isn't really an error worthy of noting specially, so that particular error
7171 62. When Exim is trying to find all the local interfaces, it used to panic and
7172 die if the ioctl to get the interface flags failed. However, it seems that
7173 on at least one OS (Solaris 9) it is possible to have an interface that is
7174 included in the list of interfaces, but for which you get a failure error
7175 for this call. This happens when the interface is not "plumbed" into a
7176 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7177 failure of the "get flags" call assumes that the interface is down.
7179 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7180 makes life easier for people who are doing arithmetic on fields extracted
7181 from dates, where you often get leading zeros that should not be
7182 interpreted as octal.
7184 64. Added qualify_domain to the redirect router, to override the global
7187 65. If a pathologically long header line contained very many addresses (the
7188 report of this problem mentioned 10 000) and each of them was rewritten,
7189 Exim could use up a very large amount of memory. (It kept on making new
7190 copies of the header line as it rewrote, and never released the old ones.)
7191 At the expense of a bit more processing, the header rewriting function has
7192 been changed so that it no longer eats memory in this way.
7194 66. The generation of the Received: header has been moved from the time that a
7195 message starts to be received, to the time that it finishes. The timestamp
7196 in the Received: header should now be very close to that of the <= log
7197 line. There are two side-effects of this change:
7199 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7200 logged header lines no longer include the local Received: line, because
7201 it has not yet been created. The same applies to a copy of the message
7202 that is returned to a non-SMTP sender when a message is rejected.
7204 (b) When a filter file is tested using -bf, no additional Received: header
7205 is added to the test message. After some thought, I decided that this
7208 This change does not affect the value of $received_for. It is still set
7209 after address rewriting, but before local_scan() is called.
7211 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7213 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7214 gave an unhelpful panic error message, and a defer error. I have managed to
7215 change this behaviour so that it now rejects any supplied certificate,
7216 which seems right, as the list of acceptable certificates is empty.
7218 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7219 gave an unhelpful defer error. I have not managed to make this reject any
7220 supplied certificates, but the error message it gives is "no certificate
7221 supplied", which is not helpful.
7223 70. exigrep's output now also includes lines that are not associated with any
7224 message, but which match the given pattern. Implemented by a patch from
7225 Martin Sluka, which also tidied up the Perl a bit.
7227 71. Recipient callout verification, like sender verification, was using <> in
7228 the MAIL FROM command. This isn't really the right thing, since the actual
7229 sender may affect whether the remote host accepts the recipient or not. I
7230 have changed it to use the actual sender in the callout; this means that
7231 the cache record is now keyed on a recipient/sender pair, not just the
7232 recipient address. There doesn't seem to be a real danger of callout loops,
7233 since a callout by the remote host to check the sender would use <>.
7234 [SEE ABOVE: changed after hitting problems.]
7236 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7237 temporary errors. However, in the case of such a code being given after
7238 the end of a data transmission (i.e. after ".") Exim was failing to write
7239 a retry record for the message. (Yes, there was some broken host that was
7240 actually sending 8xx at this point.)
7242 73. An unknown lookup type in a host list could cause Exim to panic-die when
7243 the list was checked. (An example that provoked this was putting <; in the
7244 middle of a list instead of at the start.) If this happened during a DATA
7245 ACL check, a -D file could be left lying around. This kind of configuration
7246 error no longer causes Exim to die; instead it causes a defer error. The
7247 incident is still logged to the main and panic logs.
7249 74. Buglet left over from Exim 3 conversion. The message "too many messages
7250 in one connection" was written to the rejectlog but not the mainlog, except
7251 when address rewriting (yes!) was being logged.
7253 75. Added write_rejectlog option.
7255 76. When a system filter was run not as root (that is, when system_filter_user
7256 was set), the values of the $n variables were not being returned to the
7257 main process; thus, they were not subsequently available in the $sn
7260 77. Added +return_path_on_delivery log selector.
7262 78. A connection timeout was being treated differently from recipients deferred
7263 when testing hosts_max_try with a message that was older than the host's
7264 retry timeout. (The host should not be counted, thus allowing all hosts to
7265 be tried at least once before bouncing.) This may have been the cause of an
7266 occasionally reported bug whereby a message would remain on the queue
7267 longer than the retry timeout, but would be bounced if a delivery was
7268 forced. I say "may" because I never totally pinned down the problem;
7269 setting up timeout/retry tests is difficult. See also the next item.
7271 79. The ultimate address timeout was not being applied to errors that involved
7272 a combination of host plus message (for example, a timeout on a MAIL
7273 command). When an address resolved to a number of possible hosts, and they
7274 were not all tried for each delivery (e.g. because of hosts_max_try), a
7275 message could remain on the queue longer than the retry timeout.
7277 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7280 81. Fixed an obscure SMTP outgoing bug which required at least the following
7281 conditions: (a) there was another message waiting for the same server;
7282 (b) the server returned 5xx to all RCPT commands in the first message so
7283 that the message was not completed; (c) the server dropped the connection
7284 or gave a negative response to the RSET that Exim sends to abort the
7285 transaction. The observed case was a dropped connection after DATA that had
7286 been sent in pipelining mode. That is, the server had advertised PIPELINING
7287 but was not implementing it correctly. The effect of the bug was incorrect
7288 behaviour, such as trying another host, and this could lead to a crash.
7294 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7295 and daemon.c were passed as pointers to ints; they should have been
7296 pointers to socklen_t variables (which are typically unsigned ints).
7298 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7301 3. Fixed a really odd bug that affected only the testing scheme; patching a
7302 certain fixed string in the binary changed the value of another string that
7303 happened to be identical to the end of the original first string.
7305 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7306 name", it returns that address as the IP address. On some operating
7307 systems (e.g. Solaris), it also passes back the IP address string as the
7308 "host name". However, on others (e.g. Linux), it passes back an empty
7309 string. Exim wasn't checking for this, and was changing the host name to an
7310 empty string, assuming it had been canonicalized.
7312 5. Although rare, it is permitted to have more than one PTR record for a given
7313 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7314 all the names associated with an address, because they do in Solaris.
7315 However, it seems that they do not in Linux for data that comes from the
7316 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7317 I found this out when I moved to a new Linux workstation and tried to run
7318 the Exim test suite.
7320 To get round this problem I have changed the code so that it now does its
7321 own call to the DNS to look up PTR records when searching for a host name.
7322 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7323 addresses that are only in /etc/hosts are still found.
7325 This behaviour is, however, controlled by an option called host_lookup_
7326 order, which defaults to "bydns:byaddr". If people want to use the other
7327 order, or indeed, just use one or the other means of lookup, they can
7328 specify it in this variable.
7330 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7331 some operating systems, this comes back from gethostbyaddr() as an empty
7332 string, and this is what Exim used to test for. However, it seems that in
7333 other systems, "." is yielded. Exim now tests for this case too.
7335 7. The values of check_spool_space and check_log_space are now held internally
7336 as a number of kilobytes instead of an absolute number of bytes. If a
7337 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7338 kilobyte. This means that much larger values can be stored.
7340 8. Exim monitor: an attempt to get the action menu when not actually pointing
7341 at a message produces an empty menu entitled "No message selected". This
7342 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7343 no entries in it ("Shell widget menu has zero width and/or height"). So I
7344 have added a single, blank menu entry in this case.
7346 9. Added ${quote_local_part.
7348 10. MIME decoding is now applied to the contents of Subject: header lines when
7351 11. Now that a reference to $sender_host_address automatically causes a reverse
7352 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7353 host lookup before query-style lookups in lists that might use this
7354 variable. This has therefore been abolished, and the "net-" prefix is no
7355 longer necessary for query-style lookups.
7357 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7358 have been a typo for LFLAGS, so it has been changed.
7360 13. The install script calls Exim with "-C /dev/null" in order to find the
7361 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7362 to be output. However, since Exim outputs its version number before the
7363 error, it didn't break the script. It just looked ugly. I fixed this by
7364 always allowing "-C /dev/null" if the caller is root.
7366 14. Ignore overlarge ACL variable number when reading spool file - insurance
7367 against a later release with more variables having written the file.
7369 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7370 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7371 this, as well as the form without the "IPv6" on the front.
7373 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7374 OS/Makefile-Darwin file.
7376 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7377 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7379 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7380 variable. However, some OS use ints for inodes. I've added cast to long int
7381 to get rid of the compiler warning.
7383 19. I had forgotten to lock out "/../" in configuration file names when
7384 ALT_CONFIG_PREFIX was set.
7386 20. Routers used for verification do not need to specify transports. However,
7387 if such a router generated a host list, and callout was configured, Exim
7388 crashed, because it could not find a port number from the (non-existent)
7389 transport. It now assumes port 25 in this circumstance.
7391 21. Added the -t option to exigrep.
7393 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7394 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7395 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7396 them is set, LOOKUP_LSEARCH is forced.
7398 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7400 24. Added sender and host information to the "rejected by local_scan()" log
7401 line; previously there was no indication of these.
7403 25. Added .include_if_exists.
7405 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7406 Linux. It turns out that not all file systems support this. Apparently some
7407 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7408 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7409 not supported on the file descriptor, is now ignored when Exim is trying to
7410 sync a directory. This applies only to Linux.
7412 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7414 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7415 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7418 29. Added support for maildirsize files from supplied patch (modified a bit).
7420 30. The use of :fail: followed by an empty string could lead Exim to respond to
7421 sender verification failures with (e.g.):
7423 550 Verification failed for <xxx>
7424 550 Sender verify failed
7426 where the first response line was missing the '-' that indicates it is not
7427 the final line of the response.
7429 31. The loop for finding the name of the user that called Exim had a hardwired
7430 limit of 10; it now uses the value of finduser_retries, which is used for
7431 all other user lookups.
7433 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7436 33. Exim was neglecting to zero errno before one call of strtol() when
7437 expanding a string and expecting an integer value. On some systems this
7438 resulted in spurious "integer overflow" errors. Also, it was casting the
7439 result into an int without checking.
7441 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7442 did not work. The code looks as if it has *never* worked, though it appears
7443 to have been documented since at least release 1.62. I have made it work.
7445 35. The "timeout_DNS" error in retry rules, also documented since at least
7446 1.62, also never worked. As it isn't clear exactly what this means, and
7447 clearly it isn't a major issue, I have abolished the feature by treating it
7448 as "timeout", and writing a warning to the main and panic logs.
7450 36. The display of retry rules for -brt wasn't always showing the error code
7453 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7454 timeout_connect_A, timeout_connect_MX.
7456 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7457 to the empty sender.
7459 39. The daemon was not analysing the content of -oX till after it had closed
7460 stderr and disconnected from the controlling terminal. This meant that any
7461 syntax errors were only noted on the panic log, and the return code from
7462 the command was 0. By re-arranging the code a little, I've made the
7463 decoding happen first, so such errors now appear on stderr, and the return
7464 code is 1. However, the actual setting up of the sockets still happens in
7465 the disconnected process, so errors there are still only recorded on the
7468 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7469 connections (as happens on some IP stacks) was logged at start up time as
7470 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7471 it from "IPv6 and IPv4", which means that two separate sockets are being
7474 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7475 says whether AF_INET or AF_INET6 was passed as an argument.
7477 42. Exiwhat output was messed up when time zones were included in log
7480 43. Exiwhat now gives more information about the daemon's listening ports,
7481 and whether -tls-on-connect was used.
7483 44. The "port" option of the smtp transport is now expanded.
7485 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7486 silently ignored. Now an error message is written to the main and panic
7489 46. There's a new ACL modifier called "logwrite" which writes to a log file
7490 as soon as it is encountered.
7492 47. Added $local_user_uid and $local_user_gid at routing time.
7494 48. Exim crashed when trying to verify a sender address that was being
7497 49. Exim was recognizing only a space character after ".include". It now also
7498 recognizes a tab character.
7500 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7501 extracting the relevant information from the specification. The man page no
7502 longer contains scrambled data for the -d option, and I've added a section
7503 at the front about calling Exim under different names.
7505 51. Added "extra_headers" argument to the "mail" command in filter files.
7507 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7510 53. Installed eximstats 1.29.
7512 54. Added transport_filter_timeout as a generic transport option.
7514 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7515 Cc: header lines. This was required by RFC 822, but it not required by RFC
7518 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7519 incoming messages that did not have them. Now it does so only if the
7520 message originates locally, that is, if there is no associated remote host
7521 address. When Resent- header lines are present, this applies to the Resent-
7522 lines rather than the non-Resent- lines.
7524 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7525 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7527 58. Messages for configuration errors now include the name of the main
7528 configuration file - useful now that there may be more than one file in a
7529 list (.included file names were always shown).
7531 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7532 for those rare installations that do not start the daemon as root or run it
7533 setuid root. I've cut out the call to initgroups() if the daemon is not
7536 60. The Exim user and group can now be bound into the binary as text strings
7537 that are looked up at the start of Exim's processing.
7539 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7541 62. Added $mailstore_basename variable.
7543 63. Installed patch to sieve.c from Michael Haardt.
7545 64. When Exim failed to open the panic log after failing to open the main log,
7546 the original message it was trying to log was written to stderr and debug
7547 output, but if they were not available (the usual case in production), it
7548 was lost. Now it is written to syslog before the two lines that record the
7549 failures to open the logs.
7551 65. Users' Exim filters run in subprocesses under the user's uid. It is
7552 possible for a "deliver" command or an alias in a "personal" command to
7553 provoke an address rewrite. If logging of address rewriting is configured,
7554 this fails because the process is not running as root or exim. There may be
7555 a better way of dealing with this, but for the moment (because 4.30 needs
7556 to be released), I have disabled address rewrite logging when running a
7557 filter in a non-root, non-exim process.
7563 1. The buildconfig auxiliary program wasn't quoting the value set for
7564 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7565 not defined. This bug was masked in 4.22 by the effect that was fixed in
7568 2. Some messages that were rejected after a message id was allocated were
7569 shown as "incomplete" by exigrep. It no longer does this for messages that
7570 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7572 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7573 have allow_domain_literals set, the ID did not get logged in the <= line.
7574 Domain literals are now always recognized in Message-ID: header lines.
7576 4. The first argument for a ${extract expansion item is the key name or field
7577 number. Leading and trailing spaces in this item were not being ignored,
7578 causing some misleading effects.
7580 5. When deliver_drop_privilege was set, single queue runner processes started
7581 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7582 same command in the process it spins off) were not dropping privilege.
7584 6. When the daemon running as "exim" started a queue runner, it always
7585 re-executed Exim in the spun-off process. This is a waste of effort when
7586 deliver_drop_privilege is set. The new process now just calls the
7587 queue-runner function directly.
7593 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7596 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7597 ignored. Though the use of -oP was forcing the writing of a pid file, it
7598 was always written to the default place.
7600 3. If the message "no IP address found for host xxxx" is generated during
7601 incoming verification, it is now followed by identification of the incoming
7602 connection (so you can more easily find what provoked it).
7604 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7606 5. Added some features to "harden" Exim a bit more against certain attacks:
7608 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7609 be put in Local/Makefile. This is like the never_users runtime option,
7610 but it cannot be overridden. The default setting is "root".
7612 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7613 prefix string with which any file named in a -C command line option
7616 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7617 is retained for -C and -D only if the caller of Exim is root. Without
7618 it, the exim user may also use -C and -D and retain privilege.
7620 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7621 command line option is disabled.
7623 6. Macro names set by the -D option must start with an upper case letter, just
7624 like macro names defined in the configuration file.
7626 7. Added "dereference=" facility to LDAP.
7628 8. Two instances of the typo "uknown" in the source files are fixed.
7630 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7631 the Configure-Makefile script screwed up while processing it.
7633 10. Incorporated PCRE 4.4.
7635 11. The SMTP synchronization check was not operating right at the start of an
7636 SMTP session. For example, it could not catch a HELO sent before the client
7637 waited for the greeting. There is now a check for outstanding input at the
7638 point when the greeting is written. Because of the duplex, asynchronous
7639 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7640 way, but not yet received, when the check is performed.
7642 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7643 on TCP/IP sockets, because this apparently causes some broken clients to
7646 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7647 unchanged) from the Cygwin maintainer.
7649 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7650 was supported instead of testing for mbx. Effectively a typo.
7652 15. The spa authenticator server code was not checking that the input it
7653 received was valid base64.
7655 16. The debug output line for the "set" modifier in ACLs was not showing the
7656 name of the variable that was being set.
7658 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7660 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7661 lookup on its own. Something else was needed to trigger the lookup. For
7662 example, a match in host_lookup or the need for a host name in a host list.
7663 Now, if $sender_host_name is referenced and the host name has not yet been
7664 looked up, a lookup is performed. If the lookup fails, the variable remains
7665 empty, and $host_lookup_failed is set to "1".
7667 19. Added "eqi" as a case-independent comparison operator.
7669 20. The saslauthd authentication condition could segfault if neither service
7670 nor realm was specified.
7672 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7673 error message that was logged was misleading, and incoming SMTP
7674 connections were dropped. The message is now more accurate, and temporary
7675 errors are given to SMTP connections.
7677 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7678 (except RSET) with a 421 error, until QUIT is received. However, it was
7679 failing to send a response to QUIT.
7681 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7682 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7683 ACL is now run after the helo_try_verify_hosts code.
7685 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7686 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7687 case-independent, so other case variants are also recognized). Apparently
7688 some systems use these upper case variants.
7690 25. If more than two messages were waiting for the same host, and a transport
7691 filter was specified for the transport, Exim sent two messages over the
7692 same TCP/IP connection, and then failed with "socket operation on non-
7693 socket" when it tried to send the third.
7695 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7697 27. The extern definition of crypt16() in expand.c was not being excluded when
7698 the OS had its own crypt16() function.
7700 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7701 as a preferred synonym for return_size_limit, both as an option and as an
7704 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7706 30. Changed the default configuration ACL to relax the local part checking rule
7707 for addresses that are not in any local domains. For these addresses,
7708 slashes and pipe symbols are allowed within local parts, but the sequence
7709 /../ is explicitly forbidden.
7711 31. SPA server authentication was not clearing the challenge buffer before
7714 32. log_message in a "warn" ACL statement was writing to the reject log as
7715 well as to the main log, which contradicts the documentation and doesn't
7716 seem right (because no rejection is happening). So I have stopped it.
7718 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7719 However, I am unable to do any testing of this.
7721 34. Fixed an infelicity in the appendfile transport. When checking directories
7722 for a mailbox, to see if any needed to be created, it was accidentally
7723 using path names with one or more superfluous leading slashes; tracing
7724 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7726 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7727 contents are added to the log line that is written for every discarded
7728 recipient. (Previously a log_message setting was ignored.)
7730 36. The ${quote: operator now quotes the string if it is empty.
7732 37. The install script runs exim in order to find its version number. If for
7733 some reason other than non-existence or emptiness, which it checks, it
7734 could not run './exim', it was installing it with an empty version number,
7735 i.e. as "exim-". This error state is now caught, and the installation is
7738 38. An argument was missing from the function that creates an error message
7739 when Exim fails to connect to the socket for saslauthd authentication.
7740 This could cause Exim to crash, or give a corrupted message.
7742 39. Added isip, isip4, and isip6 to ${if conditions.
7744 40. The ACL variables $acl_xx are now saved with the message, and can be
7745 accessed later in routers, transports, and filters.
7747 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7748 strings in the file are not string-expanded.
7750 42. If a MAIL command specified a SIZE value that was too large to fit into an
7751 int variable, the check against message_size_limit failed. Such values are
7752 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7753 day this will have to be increased, but I don't think I want to be around
7754 when emails are that large.
7761 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7762 iconv() is not standard in FreeBSD.
7764 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7765 IPv6 enabled. The observed symptom was a segmentation fault on return from
7766 the function os_common_find_running_interfaces() in src/os.c.
7768 3. In the check_special_case() function in daemon.c I had used "errno" as an
7769 argument name, which causes warnings on some systems. This was basically a
7770 typo, since it was named "eno" in the comments!
7772 4. The code that waits for the clock to tick (at a resolution of some fraction
7773 of a second) so as to ensure message-id uniqueness was always waiting for
7774 at least one whole tick, when it could have waited for less. [This is
7775 almost certainly not relevant at current processor speeds, where it is
7776 unlikely to ever wait at all. But we try to future-proof.]
7778 5. The function that sleeps for a time interval that includes fractions of a
7779 second contained a race. It did not block SIGALRM between setting the
7780 timer, and suspending (a couple of lines later). If the interval was short
7781 and the sigsuspend() was delayed until after it had expired, the suspension
7782 never ended. On busy systems this could lead to processes getting stuck for
7785 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7786 process, before it forks any delivery processes. The open lookup caching
7787 mechanism meant that the open file or database connection was passed into
7788 the delivery process. The problem was that delivery processes always tidy
7789 up cached lookup data. This could cause a problem for the next delivery
7790 process started by the queue runner, because the external queue runner
7791 process does not know about the closure. So the next delivery process
7792 still has data in the lookup cache. In the case of a file lookup, there was
7793 no problem because closing a file descriptor in a subprocess doesn't affect
7794 the parent. However, if the lookup was caching a connection to a database,
7795 the connection was closed, and the second delivery process was likely to
7796 see errors such as "PGSQL: query failed: server closed the connection
7797 unexpectedly". The problem has been fixed by closing all cached lookups
7798 in a queue runner before running a delivery process.
7800 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7801 seem to have the "const" qualifier which it has on other OS. I've
7804 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7805 *of the same type* (client or server) with the same public name that an
7806 error should be diagnosed.
7808 9. When Exim looked up a host name for an IP address, but failed to find the
7809 original IP address when looking up the host name (a safety check), it
7810 output the message "<ip address> does not match any IP for NULL", which was
7811 confusing, to say the least. The bug was that the host name should have
7812 appeared instead of "NULL".
7814 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7815 user that is built into the binary, and the -C or -D options is used, root
7816 privilege is dropped before the configuration file is read. In addition,
7817 logging is switched to stderr instead of the normal log files. If the
7818 configuration then re-defines the Exim user, the unprivileged environment
7819 is probably not what is expected, so Exim logs a panic warning message (but
7822 However, if deliver_drop_privilege is set, the unprivileged state may well
7823 be exactly what is intended, so the warning has been cut out in that case,
7824 and Exim is allowed to try to write to its normal log files.
7830 1. smtp_return_error_details was not giving details for temporary sender
7831 or receiver verification errors.
7833 2. Diagnose a configuration error if two authenticators have the same public
7836 3. Exim used not to create the message log file for a message until the first
7837 delivery attempt. This could be confusing when incoming messages were held
7838 for policy or load reasons. The message log file is now created at the time
7839 the message is received, and an initial "Received" line is written to it.
7841 4. The automatically generated man page for command line options had a minor
7842 bug that caused no ill effects; however, a more serious problem was that
7843 the procedure for building the man page automatically didn't always
7844 operate. Consequently, release 4.20 contains an out-of-date version. This
7845 shouldn't happen again.
7847 5. When building Exim with embedded Perl support, the script that builds the
7848 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7849 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7851 6. The freeze_tell option was not being used for messages that were frozen on
7852 arrival, either by an ACL or by local_scan().
7854 7. Added the smtp_incomplete_transaction log selector.
7856 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7857 was accepting AUTH without a new EHLO.
7859 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7860 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7863 10. Logging of TCP/IP connections (when configured) now happens in the main
7864 daemon process instead of the child process, so that the TCP/IP connection
7865 count is more accurate (but it can never be perfect).
7867 11. The use of "drop" in a nested ACL was not being handled correctly in the
7868 outer ACL. Now, if condition failure induced by the nested "drop" causes
7869 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7870 or "require"), the connection is dropped.
7872 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7873 that yield "discard" can now be used with an "accept" or a "discard" verb,
7874 but an error is generated for any others (because I can't see a useful way
7875 to define what should happen).
7877 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7878 are now processed in the same way as lines in the Exim configuration file.
7879 In particular, continuation lines are supported.
7881 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7883 15. Added -ti meaning -t -i.
7885 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7886 domains, and warn by logging if others are found.
7888 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7889 returned from GIFCONF when Exim is trying to find the list of interfaces on
7890 a host. The code in os.c has been modified to copy each ifreq to an aligned
7891 structure in all cases.
7893 Also, in some cases, the returned ifreq's were being copied to a 'struct
7894 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7895 means the last couple of bytes of an IPv6 address could be chopped if the
7896 ifreq contained only a normal sockaddr (14 bytes storage).
7898 18. Named domain lists were not supported in the hosts_treat_as_local option.
7899 An entry such as +xxxx was not recognized, and was treated as a literal
7902 19. Ensure that header lines added by a DATA ACL are included in the reject log
7903 if the ACL subsequently rejects the message.
7905 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7906 MD5 (which is deprecated).
7908 21. When testing a filter file using -bf, Exim was writing a message when it
7909 took the sender from a "From " line in the message, but it was not doing so
7910 when it took $return_path from a Return-Path: header line. It now does.
7912 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7913 with a valid header line field name (a series of printing characters
7914 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7916 23. Changed "disc" in the source to "disk" to conform to the documentation and
7917 the book and for uniformity.
7919 24. Ignore Sendmail's -Ooption=value command line item.
7921 25. When execve() failed while trying to run a command in a pipe transport,
7922 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7923 could be confused with a return value of 69 from the command itself. This
7924 has been changed to 127, the value the shell returns if it is asked to run
7925 a non-existent command. The wording for the related log line suggests a
7926 non-existent command as the problem.
7928 26. If received_header_text expands to an empty string, do not add a Received:
7929 header line to the message. (Well, it adds a token one on the spool, but
7930 marks it "old" so that it doesn't get used or transmitted.)
7932 27. Installed eximstats 1.28 (addition of -nt option).
7934 28. There was no check for failure on the call to getsockname() in the daemon
7935 code. This can fail if there is a shortage of resources on the system, with
7936 ENOMEM, for example. A temporary error is now given on failure.
7938 29. Contrary to the C standard, it seems that in some environments, the
7939 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7940 program. Exim now does this explicitly; it affects the formatting of
7941 timestamps using strftime().
7943 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7944 complaints. I've now initialized all the variables, to avoid this.
7946 32. Header lines added by a system filter were not being "seen" during
7947 transport-time rewrites.
7949 33. The info_callback() function passed to OpenSSL is set up with type void
7950 (*)(SSL *, int, int), as described somewhere. However, when calling the
7951 function (actually a macro) that sets it up, the type void(*)() is
7952 expected. I've put in a cast to prevent warnings from picky compilers.
7954 34. If a DNS black list lookup found a CNAME record, but there were no A
7955 records associated with the domain it pointed at, Exim crashed.
7957 35. If a DNS black list lookup returned more than one A record, Exim ignored
7958 all but the first. It now scans all returned addresses if a particular IP
7959 value is being sought. In this situation, the contents of the
7960 $dnslist_value variable are a list of all the addresses, separated by a
7963 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7964 to accept a host name and all its aliases if the forward lookup for any of
7965 them yielded the IP address of the incoming connection. Now it accepts only
7966 those names whose forward lookup yields the correct IP address. Any other
7967 names are discarded. This closes a loophole whereby a rogue DNS
7968 administrator could create reverse DNS records to break through a
7969 wildcarded host restriction in an ACL.
7971 37. If a user filter or a system filter that ran in a subprocess used any of
7972 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7973 the wrong values were passed to the pipe command ($thisaddress had the
7974 value of $0, $0 had the value of $1, etc). This bug was introduced by
7975 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7977 38. Improved the line breaking for long SMTP error messages from ACLs.
7978 Previously, if there was no break point between 40 and 75 characters, Exim
7979 left the rest of the message alone. Two changes have been made: (a) I've
7980 reduced the minimum length to 35 characters; (b) if it can't find a break
7981 point between 35 and 75 characters, it looks ahead and uses the first one
7982 that it finds. This may give the occasional overlong line, but at least the
7983 remaining text gets split now.
7985 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7986 file descriptors might be low, and that setting 1000 would always raise it.
7987 It turns out that in some environments, the limit is already over 1000 and
7988 that lowering it causes trouble. So now Exim takes care not to decrease it.
7990 40. When delivering a message, the value of $return_path is set to $sender_
7991 address at the start of routing (routers may change the value). By an
7992 oversight, this default was not being set up when an address was tested by
7993 -bt or -bv, which affected the outcome if any router or filter referred to
7996 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7997 log only when "message" or "log_message" are set. However, if one of the
7998 conditions was an address verification, or a call to a nested ACL, the
7999 messages generated by the underlying test were being passed through. This
8000 no longer happens. The underlying message is available in $acl_verify_
8001 message for both "message" and "log_message" expansions, so it can be
8002 passed through if needed.
8004 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8005 new expansion $bh_ to give the encoded byte string without charset
8006 translation. Translation happens only if iconv() is available; HAVE_ICONV
8007 indicates this at build time. HEADERS_CHARSET gives the charset to
8008 translate to; headers_charset can change it in the configuration, and
8009 "headers charset" can change it in an individual filter file.
8011 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8012 that creates RFC 2047 encoded "words" labels them as that charset instead
8013 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8014 expansion operator; (ii) when Exim creates a From: line for a local
8015 message; (iii) when a header line is rewritten to include a "phrase" part.
8017 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8018 buggy, causing it to skip the first lines of messages whose message ID
8019 ended in 'D'. This would not have bitten before Exim release 4.14, because
8020 message IDs were unlikely to end in 'D' before then. The effect was to have
8021 incorrect size information for certain domains.
8023 45. #include "config.h" was missing at the start of the crypt16.c module. This
8024 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8027 46. If there was a timeout during a "random" callout check, Exim treated it as
8028 a failure of the random address, and carried on sending RSET and the real
8029 address. If the delay was just some slowness somewhere, the response to the
8030 original RCPT would be taken as a response to RSET and so on, causing
8031 mayhem of various kinds.
8033 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8034 when I implemented it. It didn't allow for the fact that some option values
8035 may legitimately be negative (e.g. size_addition), and it didn't even do
8036 the right test for positive values.
8038 48. Domain names in DNS records are case-independent. Exim always looks them up
8039 in lower case. Some resolvers return domain names in exactly the case they
8040 appear in the zone file, that is, they may contain uppercase letters. Not
8041 all resolvers do this - some return always lower case. Exim was treating a
8042 change of case by a resolver as a change of domain, similar to a widening
8043 of a domain abbreviation. This triggered its re-routing code and so it was
8044 trying to route what was effectively the same domain again. This normally
8045 caused routing to fail (because the router wouldn't handle the domain
8046 twice). Now Exim checks for this case specially, and just changes the
8047 casing of the domain that it ultimately uses when it transmits the message
8050 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8053 50. If a filter generated a file delivery with a non-absolute name (possible if
8054 no home directory exists for the router), the forbid_file option was not
8057 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8058 the existing equality matching.
8060 52. Exim was using ints instead of ino_t variables in some places where it was
8061 dealing with inode numbers.
8063 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8064 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8065 called TMPDIR, and if it finds it is different, it changes its value.
8067 54. The smtp_printf() function is now made available to local_scan() so
8068 additional output lines can be written before returning. There is also an
8069 smtp_fflush() function to enable the detection of a dropped connection.
8070 The variables smtp_input and smtp_batched_input are exported to
8073 55. Changed the default runtime configuration: the message "Unknown user"
8074 has been removed from the ACL, and instead placed on the localuser router,
8075 using the cannot_route_message feature. This means that any verification
8076 failures that generate their own messages won't get overridden. Similarly,
8077 the "Unrouteable address" message that was in the ACL for unverifiable
8078 relay addresses has also been removed.
8080 56. Added hosts_avoid_esmtp to the smtp transport.
8082 57. The exicyclog script was not checking for the esoteric option
8083 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8084 will work only if exicyclog is run under the appropriate euid.
8086 58. Following a discussion on the list, the rules by which Exim recognises line
8087 endings on incoming messages have been changed. The -dropcr and drop_cr
8088 options are now no-ops, retained only for backwards compatibility. The
8089 following line terminators are recognized: LF CRLF CR. However, special
8090 processing applies to CR:
8092 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8093 nor a local message in the state where . is a terminator.
8095 (ii) If a bare CR is encountered in a header line, an extra space is added
8096 after the line terminator so as not to end the header. The reasoning
8097 behind this is that bare CRs in header lines are most likely either
8098 to be mistakes, or people trying to play silly games.
8100 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8101 was being incorrectly given as 18 bytes larger than it should have been.
8102 This is a VOB (very old bug).
8104 60. This may never have affected anything current, but just in case it has:
8105 When the local host is found other than at the start of a list of hosts,
8106 the local host, those with the same MX, and any that follow, are discarded.
8107 When the list in question was part of a longer list of hosts, the following
8108 hosts (not currently being processed) were also being discarded. This no
8109 longer happens. I'm not sure if this situation could ever has previously
8112 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8115 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8116 header. This is something that is recommended in a new Internet Draft, and
8117 is something that is documented as being done by Sendmail. There are two
8118 possible values. For messages generated by the autoreply transport, Exim
8121 Auto-Submitted: auto-replied
8123 whereas for all other generated messages (e.g. bounces) it adds
8125 Auto-Submitted: auto-generated
8127 63. The "personal" condition in filters now includes a test for the
8128 Auto-Submitted: header. If it contains the string "auto-" the message it
8129 not considered personal.
8131 64. Added rcpt_include_affixes as a generic transport option.
8133 65. Added queue_only_override (default true).
8135 66. Added the syslog_duplication option.
8137 67. If what should have been the first header line of a message consisted of
8138 a space followed by a colon, Exim was mis-interpreting it as a header line.
8139 It isn't of course - it is syntactically invalid and should therefore be
8140 treated as the start of the message body. The misbehaviour could have
8141 caused a number of strange effects, including loss of data in subsequent
8142 header lines, and spool format errors.
8144 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8145 client host had authenticated. This control can now be exercised by an ACL
8146 for more flexibility.
8148 69. By default, callouts do not happen when testing with -bh. There is now a
8149 variant, -bhc, which does actually run the callout code, including
8150 consulting and updating the callout cache.
8152 70. Added support for saslauthd authentication, courtesy of Alexander
8155 71. If statvfs() failed on the spool or log directories while checking their
8156 size for availability, Exim confusingly gave the error "space shortage".
8157 Furthermore, in debugging mode it crashed with a floating point exception.
8158 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8159 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8160 serious problem, Exim now writes to the main and panic logs when this
8161 happens, with details of the failure. It then refuses to accept the
8162 incoming message, giving the message "spool directory problem" or "log
8163 directory problem" with a 421 code for SMTP messages.
8165 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8166 0, 1, and 2 exist, because some OS complain for execs without them (see
8167 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8168 descriptors. However, the code omitted to check that the open succeeded,
8169 causing mysterious errors if for some reason the permissions on /dev/null
8170 got screwed. Now Exim writes a message to the main and panic logs, and
8171 bombs out if it can't open /dev/null.
8173 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8174 interact so that it is all more flexible. It is supposed to remain
8175 backwards compatible. Also added extra_local_interfaces.
8177 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8178 to bomb out with an assertion failure - to the client this appears as a
8179 connection drop. This problem occurs in the part of the code that was taken
8180 from the Samba project. Fortunately, the assertion is in a very simple
8181 function, so I have fixed this by reproducing the function inline in the
8182 one place where it is called, and arranging for authentication to fail
8183 instead of killing the process with assert().
8185 75. The SPA client code was not working when the server requested OEM rather
8186 than Unicode encoding.
8188 76. Added code to make require_files with a specific uid setting more usable in
8189 the case where statting the file as root fails - usually a non-root-mounted
8190 NFS file system. When this happens and the failure is EACCES, Exim now
8191 forks a subprocess and does the per-uid checking as the relevant uid.
8193 77. Added process_log_path.
8195 78. If log_file_path was not explicitly set, a setting of check_log_space or
8196 check_log_inodes was ignored.
8198 79. If a space check for the spool or log partitions fails, the incident is now
8199 logged. Of course, in the latter case the data may get lost...
8201 80. Added the %p formatting code to string_format() so that it can be used to
8202 print addresses in debug_print(). Adjusted all the address printing in the
8203 debugging in store.c to use %p rather than %d.
8205 81. There was a concern that a line of code in smtp_in.c could overflow a
8206 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8207 initially expressed, the concern was not well-founded, because trailing
8208 spaces are removed early. However, if the trailing spaces were followed by
8209 a NULL, they did not get removed, so the overflow was possible. Two fixes
8212 (a) I re-wrote the offending code in a cleaner fashion.
8213 (b) If an incoming SMTP command contains a NULL character, it is rejected
8216 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8217 runs initgroups(), so that if the Exim user is in any additional groups,
8218 they will be used during message reception.
8224 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob