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.
5 SINCE Exim version 4.96
8 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
18 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
19 after reception to before a subsequent reception. This should
20 mean slightly faster delivery, and also confirmation of reception
23 JH/02 Move from using the pcre library to pcre2. The former is no longer
24 being developed or supported (by the original developer).
26 JH/03 Constification work in the filters module required a major version
27 bump for the local-scan API. Specifically, the "headers_charset"
28 global which is visible via the API is now const and may therefore
29 not be modified by local-scan code.
31 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
32 sendfile() didi not account for the way the ClamAV driver code called it.
34 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
35 time check was being done for every character; replace that with one
38 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
39 sent was prefixed with a length byte.
41 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
42 RFC 5321. Previously Dovecot (at least) would log errors during
45 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
46 in 4.61 without a reason logged. Binaries will be bigger, which might
47 matter on diskspace-constrained systems, but debug is easier.
49 JH/09 Fix macro-definition during "-be" expansion testing. The move to
50 write-protected store for macros had not accounted for these runtime
51 additions; fix by removing this protection for "-be" mode.
53 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
54 handing out large-numbered file descriptors, violating the usual Unix
55 assumption (and required by Posix) that the lowest possible number will be
56 allocated by the kernel when a new one is needed. In the daemon, and any
57 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
58 are not useable for FD_SET() [and hence select()] and overwrite the stack.
59 Assorted crashes happen.
61 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
62 main-section options or in a connect ACL, the value from the first ever
63 connection was never replaced for subsequent connections. Found by
66 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
67 though only once PCRE2 was introduced: the memory accounting used under
68 debug offset allocations by an int, giving a hard trap in early startup.
69 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
71 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
72 with underbars is given. The write-protection of configuration introduced
73 in 4.95 trapped when normalisation was applied to an option not needing
76 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
78 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
79 failing to close the certificates directory, every hour or any time it
82 JH/16 Debugging initiated by an ACL control now continues through into routing
83 and transport processes. Previously debugging stopped any time Exim
84 re-execs, or for processing a queued message.
86 JH/17 The "expand" debug selector now gives more detail, specifically on the
87 result of expansion operators and items.
89 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
90 bad comparison between the option value and the name of the file to
91 be included was done, and a mismatch was wrongly identified.
92 4.88 to 4.95 are affected.
94 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
96 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
97 supplied as the db file-pair basename. Previously, if a directory
98 path was given, for example via the autoreply "once" option, the DB
99 file.pag and file.dir files would be created in that directory's
102 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
103 "taint" log_selector. These were previously deprecated.
105 JH/22 Fix static address-list lookups to properly return the matched item.
106 Previously only the domain part was returned.
108 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
109 the call into OpenSSL to send a TLS Close was being repeated; this
110 resulted in the library waiting for the peer's Close. If that was never
111 sent we waited forever. Fix by tracking send calls.
113 JH/24 The ${run} expansion item now expands its command string elements after
114 splitting. Previously it was before; the new ordering makes handling
115 zero-length arguments simpler. The old ordering can be obtained by
116 appending a new option "preexpand", after a comma, to the "run".
118 JH/25 Taint-check exec arguments for transport-initiated external processes.
119 Previously, tainted values could be used. This affects "pipe", "lmtp" and
120 "queryprogram" transport, transport-filter, and ETRN commands.
121 The ${run} expansion is also affected: in "preexpand" mode no part of
122 the command line may be tainted, in default mode the executable name
125 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
126 the the facility was not passed across execs, and only the first message
127 passed over a connection could use BDAT; any further ones using DATA.
129 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
130 uses $sending_ip_address and an interface is specified.
131 Previously any use of the local address in the EHLO name disabled
132 PIPECONNECT, the common case being to use the rDNS of it.
134 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
135 resumption. Previously verify failed because no certificate status is
136 passed on the wire for the restarted session. Fix by using the recorded
137 ocsp status of the stored session for the new connection.
139 JH/29 TLS resumption: the key for session lookup in the client now includes
140 more info that a server could potentially use in configuring a TLS
141 session, avoiding oferring mismatching sessions to such a server.
142 Previously only the server IP was used.
144 JH/30 Fix string_copyn() for limit greater than actual string length.
145 Previously the copied amount was the limit, which could result in a
146 overlapping memcpy for newly allocated destination soon after a
147 source string shorter than the limit. Found/investigated by KM.
149 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
150 close; it may be needed for a subsequent connection. This caused a
151 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
153 JH/32 Fix CHUNKING for a second message on a connection when the first was
154 rejected. Previously we did not reset the chunking-offered state, and
155 erroneously rejected the BDAT command. Investigation help from
162 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
163 related applications. Previously an "H" was used where available info
164 says that "M" should be, so change to match.
166 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
167 as arguments, so an implementation trying to copy these into a local
168 buffer was taking a taint-enforcement trap. Fix by using dynamically
169 created buffers. Similar fix for radius expansion condition.
171 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
172 reasonable, eg. to count headers. Fix by using dynamically created
173 buffers rather than a local. Do similar fixes for ACL actions "dcc",
174 "log_reject_target", "malware" and "spam"; the arguments are expanded
175 so could be handling tainted values.
177 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
178 broken the (no-op) support for this sendmail command. Restore it
179 to doing nothing, silently, and returning good status.
181 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
182 record path was given (or the default used) without a leading directory
183 path, an error occurred on trying to open it. Use the transport's working
186 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
187 transport. Previously it was the name on the DNS A-record; use instead
188 the head of the CNAME chain leading there (if there is one). This seems
189 to align better with RFC 6125.
191 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
192 smtp_accept_max_per_host allocated resources which were not released
193 when the limit was exceeded. This eventually crashed the daemon. Fix
194 by adding a release action in that path.
196 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
197 expanded; previously using tainted values was rejected. Fix by using
198 dynamically-created buffers.
200 JH/09 Relax restrictions on ACL verify condition needing access to message
201 headers. Previously they were only permitted in data and non-smtp ACLs;
202 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
203 not_blind, header_sender and header_names_ascii verification.
205 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
206 Previously a macro used one argument twice; when called with the
207 argument as an expression having side-effects, incorrect operation
208 resulted. Use an inlineable function.
210 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
211 held open for a verify callout. Previously this wan not accounted for
212 and a corrupt onward SMTP conversation resulted.
214 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
215 Previously they were embedded naively in the constructed address; when
216 needed, strip the quoting and quote the entire local_part.
217 Also make the inbound_srs expansion condition handle quoting.
219 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
220 excluded, not matching the documentation.
222 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
223 was given for the sqlite_dbfile a trap resulted.
225 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
226 "name" argument resulted in a trap. There is no reason to disallow such;
227 this was a coding error.
229 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
230 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
231 spent suspended, ignoring the POSIX definition. Previously we assumed
232 it did and a constant offset from real time could be used as a correction.
233 Change to using the same clock source for the start-of-message and the
234 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
235 exists, just to get a clock slightly more aligned to reality.
237 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
238 RFC says it is optional some validators care. The missing char was not
239 intended but triggered by a line-wrap alignment. Discovery and fix by
240 Guillaume Outters, hacked on by JH.
242 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
243 name being quoted was tainted a trap would be taken. Fix by using
244 dynamically created buffers. The routine could have been called by a
245 rewrite with the "h" flag, by using the "-F" command-line option, or
246 by using a "name=" option on a control=submission ACL modifier.
248 JH/19 SPF: change the Authentication-Results expansion component to give
249 smtp.helo when the sender domain is empty. Previously it gave
252 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
253 not in 127.0.0.0/8 to help in spotting list domains taken over by a
254 domain-parking registrar.
256 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
257 Previously when a whitespace character was specified it was not inserted
258 after removing the newline.
260 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
261 the domain part of the recipient address. This overrides any tls_sni
262 option set, which was previously used.
264 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
267 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
268 is_tainted() had an off-by-one error in the overenthusiastic direction.
269 Find and fix by Gavan. Although NetBSD is not a supported platform for
270 4.94 this bug could affect other platforms.
272 PP/01 Fix default prime selection to be consistent.
273 One path used ike23 still, instead of exim.dev.20160529.3; now both
274 execution flows will use the same DH primes (currently
275 exim.dev.20160529.3).
277 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
278 option in smtp transport, to match the documentation. Previously
279 verification was not being done in some cases where it should have been.
281 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
282 than one server was defined and depending on the platform memory layout
283 details, an internal consistency trap could be hit while walking the list
286 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
287 files. The value is used by the authresults expansion item. Previously
288 if this was used in a router or transport, a crash could result.
290 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
291 have a sufficient incidence of EINTR returns from select that an
292 interaction with other operations done by the main daemon loop exposed
293 a bug in the error-handling. This was benign apart from the log
296 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
297 for consistency with delivered "=>" and failed "**" lines. While we're
298 there, handle PRX and TFO.
300 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
301 applied. This resulted, if any header-line rewrite rules were configured,
302 in a panic-log triggerable by sending a message with a long address in
303 a header. Fix by increasing the arbitrary limit to larger than a single
304 (dewrapped) 5322 header line maximum size.
306 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
307 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
308 RFC 6648 which deprecates X- options in protocols as a general practice.
309 Changeover between the implementations is handled by the mechanisms
312 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
313 a remote which uses callout/hold. Previously the local was queued.
315 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
318 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
319 log line. The asterisk is now consistently appended to the remote IP
320 (and port, if given), and will also be provided on defer and fail log
321 lines. Previously it could be placed on the local IP if that was being
322 logged, and was only provided on delivery lines.
324 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
326 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
327 authenticator client_send option. Previously the next char, after a pair
328 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
329 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
330 documentation. There is still no way to get a leading ^ immediately
331 after a NUL (ie. for the password of a PLAIN method authenticator.
333 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
334 files. For bad sizes read, delete the record and whine to paniclog.
336 JH/38 When logging an AUTH failure, as server, do not include sensitive
337 information. Previously, the credentials would be included if given
338 as part of the AUTH command line and an ACL denied authentication.
340 JH/39 Bug 2691: fix $local_part_data. When the matching list element
341 referred to a file, bad data was returned. This likely also affected
344 JH/40 The gsasl authenticator now supports caching of the salted password
345 generated by the client-side implementation. This required the addition
346 of a new variable: $auth4.
348 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
349 left undeleted; the attempt to re-create it then failed - resulting in
350 the usual "SIGHUP tp have daemon reload configuration" to not work.
351 This affected any platform not supporting "abstract" Unix-domain
352 sockets (i.e. not Linux).
354 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
355 recipients" response to RCPT in a later response, with a 250. The
356 previous coding assumed this would not happen, and under PIPELINING
357 would result in both lost and duplicate recipients for a message.
359 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
360 Previously the weighting was incorrectly applied. Similar fix for socks
361 proxies. Found and fixed by Heiko Schlichting.
363 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
364 not handle sub-lists included using the +namedlist syntax. While
365 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
366 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
369 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
370 Previously the permanent pool was used, so the sore could not be freed.
371 This meant a connection with many messages would use continually-growing
374 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
375 per-pool so as not to waste too much space. Previously a constant size
376 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
377 DOS attacks harder. The cost is wasted memory use in the larger blocks.
379 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
380 same space repeatedly, and a smaller process footprint.
382 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
383 for GnuTLS. Previously the name started "%s" which, while not a bug,
384 looked as if if might be one.
386 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
387 "continued connection") make the $proxy_* variables available. Previously
388 the information was not passed across the exec() call for subsequent
389 transport executions. This also mean that the log lines for the
390 messages can show the proxy information.
392 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
393 receive time. With modern systems the difference is significant.
394 The historical behaviour can be restored by disabling (a new) log_selector
395 "queue_time_exclusive".
397 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
398 content) were specifically tested for. Now, also cover expansions
399 resulting in ACL names and inline ACL content.
401 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
402 making it unusable in complex expressions.
404 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
405 fail with a taint-check on the spoolfile name, and leave the message
408 HS/01 Enforce absolute PID file path name.
410 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
412 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
414 PP/02 Bug 2643: Correct TLS DH constants.
415 A missing NUL termination in our code-generation tool had led to some
416 incorrect Diffie-Hellman constants in the Exim source.
417 Reported by kylon94, code-gen tool fix by Simon Arlott.
419 PP/03 Impose security length checks on various command-line options.
420 Fixes CVE-2020-SPRSS reported by Qualys.
422 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
423 better. Reported by Qualys.
425 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
426 providing a particularly obnoxious sender full name.
429 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
431 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
434 PP/08 Change default for recipients_max from unlimited to 50,000.
436 PP/09 Fix security issue with too many recipients on a message (to remove a
437 known security problem if someone does set recipients_max to unlimited,
438 or if local additions add to the recipient list).
439 Fixes CVE-2020-RCPTL reported by Qualys.
441 PP/10 Fix security issue in SMTP verb option parsing
442 Fixes CVE-2020-EXOPT reported by Qualys.
444 PP/11 Fix security issue in BDAT state confusion.
445 Ensure we reset known-good where we know we need to not be reading BDAT
446 data, as a general case fix, and move the places where we switch to BDAT
447 mode until after various protocol state checks.
448 Fixes CVE-2020-BDATA reported by Qualys.
450 HS/03 Die on "/../" in msglog file names
452 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
453 the Exim runtime user are allowed to create files.
455 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
458 QS/03 When reading the output from interpreted forward files we do not
459 pass the pipe between the parent and the interpreting process to
460 executed child processes (if any).
462 QS/04 Always die if requested from internal logging, even is logging is
465 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
466 the API; compilation noo longer completes with DMARC support included.
467 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
468 been reported on other platforms.
470 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
472 JH/56 Make the majority of info read from config files readonly, for defence-in-
473 depth against exploits. Suggestion by Qualys.
474 Not supported on Solaris 10.
476 JH/57 Fix control=fakreject for a custom message containing tainted data.
477 Previously this resulted in a log complaint, due to a re-expansion present
478 since fakereject was originally introduced.
480 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
481 modifier was given, a loop resulted.
483 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
484 pipelined SMTP command followed the BDAT LAST then it would be
485 incorrectly treated as part of the message body, causing a verification
488 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
489 require looser validation rules than those for 821-level addresses,
490 which only permit IP addresses.
496 JH/01 Avoid costly startup code when not strictly needed. This reduces time
497 for some exim process initialisations. It does mean that the logging
498 of TLS configuration problems is only done for the daemon startup.
500 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
502 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
503 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
506 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
507 DKIM signing is being done. Previously a transport_filter would always
508 disable CHUNKING, falling back to traditional DATA.
510 JH/05 Regard command-line recipients as tainted.
512 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
514 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
515 PAM library frees one of the arguments given to it, despite the
516 documentation. Therefore a plain malloc must be used.
518 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
519 on-stack buffers were used, resulting in a taint trap when DSN information
520 copied from a received message was written into the buffer.
522 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
523 the ordering of its ARC headers. This caused a crash.
525 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
526 a new record was being constructed with information from the peer, a trap
529 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
530 installation would get error messages from DMARC verify, when it hit the
531 nonexistent file indicated by the default. Distros wanting DMARC enabled
532 should both provide the file and set the option.
533 Also enforce no DMARC verification for command-line sourced messages.
535 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
536 could, depending on the platform, hang at the STARTTLS response.
538 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
539 message on a connection. Previously if one message had ARC headers and
540 the following one did not, a crash could result when adding an
541 Authentication-Results: header.
543 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
544 Exim main code and Exim-related utities. The introduction of taint
545 tracking also did many adjustments to string handling. Since then, eximon
546 frequently terminated with an assert failure.
548 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
549 check for 452 responses. This slightly helps the inefficieny of doing
550 a large alias-expansion into a recipient-limited target. The max_rcpt
551 transport option still applies (and at the current default, will override
552 the new feature). The check is done for either cause of synch, and forces
553 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
554 connection. The new facility is not tunable at this time.
556 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
557 library live data was being used, so the results became garbage. Make
558 copies while it is still usable.
560 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
561 on delivery deferred (==) and failed (**) lines (if a delivery was
562 attemtped). Previously it was only on completion (=>) lines.
564 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
565 for the expansion of the server_scram_iter and server_scram_salt options.
567 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
568 are now specifically given a NO_DATA response without hitting the system
569 resolver. The library goes on to do the now-standard TXT lookup.
570 Use of dnsdb lookups is not affected.
572 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
573 only retrieve the errormessage once. Previously two calls to dlerror()
574 were used, and the second one (for mainlog/paniclog) retrieved null
577 JH/20 Taint checking: disallow use of tainted data for
578 - the appendfile transport file and directory options
579 - the pipe transport command
580 - the autoreply transport file, log and once options
581 - file names used by the redirect router (including filter files)
583 - paths used by single-key lookups
584 Previously this was permitted.
586 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
587 adjusted the size of a major service buffer; this failed because the
588 buffer was in use at the time. Change to a compile-time increase in the
589 buffer size, when this authenticator is compiled into exim.
591 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
592 previous fast-mode was untenable in the face of glibs using mmap to
593 support larger malloc requests.
595 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
596 New values supported, if defined on system where compiled:
597 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
598 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
600 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
601 running a limited number of proceses in parallel, a benefit is gained. The
602 amount varies with the platform hardware and load. The use of the option
603 queue_run_in_order means we cannot do this, as ordering becomes
606 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
607 had introduced a string-copy (for ensuring NUL-termination) which was not
608 appropriate for that case, which can include embedded NUL bytes in the
609 block of data. Investigation showed the copy to actually be needless, the
610 data being length-specified.
612 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
613 done during a receiving connection, and both used TLS, global info was
614 used rather than per-connection info for tracking the state of data
615 queued for transmission. This could result in a connection hang.
617 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
618 Previously, when delivering serveral messages down a single connection
619 only the first would provide a SIZE. This was due to the size information
620 not being properly tracked.
622 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
623 TAI (at 37 seconds currently), pretend to be in UTC for time-related
624 expansion and logging. Previously, spurious values such as a future
625 minute could be seen.
627 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
628 it could crash from a null-deref. This could also affect the
629 ${addresses: } operator and ${readsock } item.
631 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
632 message following a mime one, the variable was not reset.
634 JH/30 When an pipelined-connect fails at the first response, assume incorrect
635 cached capability (perhaps the peer reneged?) and immediately retry in
638 JH/31 Fix spurious detection of timeout while writing to transport filter.
640 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
641 an attempt to copy the string was made before checking it.
643 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
644 taint of the lookup key was maintained; we now regard the presence in the
645 filesystem as sufficient validation.
647 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
648 argument is supplied.
650 JH/35 The dsearch lookup now requires that the directory is an absolute path.
651 Previously this was not checked, and nonempty relative paths made an
652 access under Exim's current working directory.
654 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
655 Previously no event was raised.
657 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
658 parameter supplied by the sender MAIL FROM command. Previously it was
659 ignored, and only the check_spool_space option value for the required
662 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
663 the size of the signing public-key. Previously it was instead giving
664 the size of the signature hash.
666 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
667 the default. See the (new) dkim_verify_min_keysizes option.
669 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
670 and an ACL use a lookup for checking either the local_part or domain,
671 stale data could be accessed. Ensure that variable references are
672 dropped between messages.
674 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
675 by the client was not checked as pointing within response data before
676 being used. A malicious client could thus cause an out-of-bounds read and
677 possibly gain authentication. Fix by adding the check.
679 JH/42 Internationalisation: change the default for downconversion in the smtp
680 transport to be "if needed". Previously it was "as previously set" for
681 the message, which usually meant "if needed" for message-submission but
682 "no" for everything else. However, MTAs have been seen using SMTPUTF8
683 even when the envelope addresses did not need it, resulting in forwarding
684 failures to non-supporting MTAs. A downconvert in such cases will be
685 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
686 transport. The change does mean that addresses needing conversion will
687 be converted when previously a delivery failure would occur.
689 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
690 response was received it would be used unchecked in a fail-DSN, violating
691 standards on line-length limits. Truncate if needed.
693 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
694 doesn't work. (Additionally add a new main config option to configure the
701 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
702 side, to decode a TLS 1.3 packet capture.
704 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
705 Previously the default library behaviour applied, sending two, each in
708 JH/03 Debug output for ACL now gives the config file name and line number for
711 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
713 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
715 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
716 buffer overrun for (non-chunking) other transports.
718 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
719 TLS1.3, means that a server rejecting a client certificate is not visible
720 to the client until the first read of encrypted data (typically the
721 response to EHLO). Add detection for that case and treat it as a failed
722 TLS connection attempt, so that the normal retry-in-clear can work (if
723 suitably configured).
725 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
726 and/or domain. Found and fixed by Jason Betts.
728 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
729 configuration). If a CNAME target was not a wellformed name pattern, a
732 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
733 the OS reports them interleaved with other addresses.
735 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
736 used both for input and for a verify callout, both encrypted, SMTP
737 responses being sent by the server could be lost. This resulted in
738 dropped connections and sometimes bounces generated by a peer sending
741 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
742 string. Previously it was possible to cause undefined behaviour in a
743 library routine (usually a crash). Found by "zerons".
745 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
748 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
749 API was removed, so update to use the newer ones.
751 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
752 any timeout set, is taking a long time. Previously we would hang on to a
753 rotated logfile "forever" if the input was arriving with long gaps
754 (a previous attempt to fix addressed lack, for a long time, of initial
757 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
758 shared (NFS) environment. The length of the tempfile name is now
759 4 + 16 ("hdr.$message_exim_id") which might break on file
760 systems which restrict the file name length to lower values.
763 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
764 shared (NFS) environment.
766 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
767 did for all versions <4.90). Notably -M, -m, --invert, -I may be
770 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
771 on some platforms for bit 31.
773 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
774 to changes apparently associated with TLS1.3 handling some of the APIs
775 previously used were either nonfunctional or inappropriate. Strings
776 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
777 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
778 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
779 This affects log line X= elements, the $tls_{in,out}_cipher variables,
780 and the use of specific cipher names in the encrypted= ACL condition.
782 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
784 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
785 verification result was not updated unless hosts_require_ocsp applied.
787 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
788 queue_list_requires_admin set to false, non-admin users were denied the
791 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
792 directory-of-certs mode. Previously they were advertised despite the
795 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
796 A single TCP connection by a client will now hold a TLS connection open
797 for multiple message deliveries, by default. Previously the default was to
800 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
801 default. If built with the facility, DANE will be used. The facility
802 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
804 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
805 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
806 must be defined and you must still, unless you define DISABLE_TLS, manage
807 the the include-dir and library-file requirements that go with that
808 choice. Non-TLS builds are still supported.
810 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
813 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
814 default. If the platform supports and has the facility enabled, it will
815 be requested on all coneections.
817 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
818 controlled by the build-time option SUPPORT_PIPE_CONNECT.
820 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
822 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
823 success-DSN messages. Previously the From: header was always the default
824 one for these; the option was ignored.
826 JH/28 Fix the timeout on smtp response to apply to the whole response.
827 Previously it was reset for every read, so a teergrubing peer sending
828 single bytes within the time limit could extend the connection for a
829 long time. Credit to Qualsys Security Advisory Team for the discovery.
831 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
832 delivery address, which leaked information of the results of local
833 forwarding. Change to the original envelope recipient address, per
836 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
837 requested. Previously not bounce was generated and a log entry of
838 error ignored was made.
840 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
842 JH/32 Introduce a general tainting mechanism for values read from the input
843 channel, and values derived from them. Refuse to expand any tainted
844 values, to catch one form of exploit.
846 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
847 was unused and the unexpanded text used for the test. Found and
848 fixed by Ruben Jenster.
850 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
851 an attempt to use a TLS library read routine dereffed a nul pointer,
854 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
855 them in our smtp response.
857 JH/36 Have the general router option retry_use_local_part default to true when
858 any of the restrictive preconditions are set (to anything). Previously it
859 was only for check_local user. The change removes one item of manual
860 configuration which is required for proper retries when a remote router
861 handles a subset of addresses for a domain.
863 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
864 link count into consideration.
866 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
867 caused the extension of big_buffer, the following lines were ignored.
869 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
870 accordance with RFC 2308. Previously there was no expiry, so a longlived
871 receive process (eg. due to ACL delays) versus a short SOA value could
874 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
876 JH/39 Promote DMARC support to mainline.
878 JH/40 Bug 2452: Add a References: header to DSNs.
880 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
881 parameters. The relevant library call is documented as "Deprecated: This
882 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
883 3.6.0, DH parameters are negotiated following RFC7919."
885 HS/06 Change the default of dnssec_request_domains to "*"
887 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
888 carried on and emitted a BDAT command, even when PIPELINING was not
891 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
892 buffer was used for the filename, resulting in a trap when tainted
893 arguments (eg. $domain) were used.
895 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
896 recommended to avoid a possible server-load attack. The feature can be
897 re-enabled via the openssl_options main cofiguration option.
899 JH/45 local_scan API: documented the current smtp_printf() call. This changed
900 for version 4.90 - adding a "more data" boolean to the arguments.
901 Bumped the ABI version number also, this having been missed previously;
902 release versions 4.90 to 4.92.3 inclusive were effectively broken in
903 respect of usage of smtp_printf() by either local_scan code or libraries
904 accessed via the ${dlfunc } expansion item. Both will need coding
905 adjustment for any calls to smtp_printf() to match the new function
906 signature; a FALSE value for the new argument is always safe.
908 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
909 the file-offset (which the Linux syscall does, and exim expects); this
910 resulted in an indefinite loop.
912 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
913 to do ARC verification. The Authentication-Results: header line added
914 by the configuration then had no ARC item.
916 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
917 an incorrect mode was used for reading the result, resulting in it being
924 JH/01 Remove code calling the customisable local_scan function, unless a new
925 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
927 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
928 non-signal-safe functions being used.
930 JH/03 Bug 2269: When presented with a received message having a stupidly large
931 number of DKIM-Signature headers, disable DKIM verification to avoid
932 a resource-consumption attack. The limit is set at twenty.
934 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
935 report of oldest_pass in ${authres } in consequence, and separate out
936 some descriptions of reasons for verification fail.
938 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
939 files in the spool were present and unlocked. A queue-runner could spot
940 them, resulting in a duplicate delivery. Fix that by doing the unlock
941 after the unlink. Investigation by Tim Stewart. Take the opportunity to
942 add more error-checking on spoolfile handling while that code is being
945 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
946 No known attacks, no CVE, this is defensive hardening.
948 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
949 a queue-runner could start a delivery while other operations were ongoing.
950 Cutthrough delivery was a common victim, resulting in duplicate delivery.
951 Found and investigated by Tim Stewart. Fix by using the open message data
952 file handle rather than opening another, and not locally closing it (which
953 releases a lock) for that case, while creating the temporary .eml format
954 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
956 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
957 $sender_verify_failure/$recipient_verify_failure to "random".
959 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
962 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
963 Previously this would segfault.
965 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
968 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
969 like zero, since the resolver should be doing this for us, But we need one
970 as a CNAME but no MX presence gets the CNAME returned; we need to check
971 that doesn't point to an MX to declare it "no MX returned" rather than
972 "error, loop". A new main option is added so the older capability of
973 following some limited number of chain links is maintained.
975 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
977 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
978 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
979 now-deprecated earlier definitions used only specified the range up to TLS
980 1.2 (in the older-version library docs).
982 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
984 JH/15 Rework TLS client-side context management. Stop using a global, and
985 explicitly pass a context around. This enables future use of TLS for
986 connections to service-daemons (eg. malware scanning) while a client smtp
987 connection is using TLS; with cutthrough connections this is quite likely.
989 JH/16 Fix ARC verification to do AS checks in reverse order.
991 JH/17 Support a "tls" option on the ${readsocket } expansion item.
993 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
994 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
995 Previously the "utf8" would be re-prepended for every additional message.
997 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
998 Previously thery were accepted, resulting in issues when attempting to
999 forward messages to a non-supporting MTA.
1001 PP/02 Let -n work with printing macros too, not just options.
1003 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1004 one parent address was copied, and bogus data was used at delivery-logging
1005 time. Either a crash (after delivery) or bogus log data could result.
1006 Discovery and analysis by Tim Stewart.
1008 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1009 Previously if the string ended mid-character, we did not insert the
1010 promised '?' replacement.
1012 PP/04 Documentation: current string operators work on bytes, not codepoints.
1014 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1015 should pack well giving a smaller memory footprint so better caching and
1016 therefore performance. Group the declarations where this can't be done so
1017 that the byte-sized flag variables are not interspersed among pointer
1018 variables, giving a better chance of good packing by the compiler.
1020 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1021 non-null, to avoid issues with sites running BATV. Previously reports were
1022 sent with an empty envelope sender so looked like bounces.
1024 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1025 The ignore_error flag wasn't being returned from the filter subprocess so
1026 was not set for later routers. Investigation and fix by Matthias Kurz.
1028 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1029 and a msg:complete for the whole, when a message is manually removed using
1030 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1032 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1033 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1034 hope that is portable enough.
1036 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1037 requiring a known-CA anchor certificate; make it now rely entirely on the
1038 TLSA as an anchor. Checking the name on the leaf cert against the name
1039 on the A-record for the host is still done for TA (but not for EE mode).
1041 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1042 included in delivery lines for non-proxied connections, when compiled with
1043 SUPPORT_SOCKS and running with proxy logging enabled.
1045 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1046 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1047 move the existing event to fire before the normal logging of message
1048 failure so that custom logging is bracketed by normal logging.
1050 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1051 msg:fail:internal event. Developement by Matthias Kurz.
1053 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1054 far too small for todays use of crypto signatures stored there. Go all
1055 the way to the max DNS message size of 64kB, even though this might be
1056 overmuch for IOT constrained device use.
1058 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1059 copy a string over itself. The library routine is documented as not
1060 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1062 JH/32 For main options check_spool_space and check_inode_space, where the
1063 platform supports 64b integers, support more than the previous 2^31 kB
1064 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1065 the previous G, M, k.
1067 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1068 $authenticated_fail_id variable on authentication failure. Previously
1071 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1072 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1073 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1074 GNUTLS_SEC_PARAM_MEDIUM.
1076 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1077 side. Previously we would continue as if no SNI had been received.
1079 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1080 "<" character, which should be a list-separator specification, we walked
1081 off past the nul-terimation.
1083 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1084 causes) even when the retry time is not yet met. Previously they were
1085 not, meaning that when (say) an account was over-quota and temp-rejecting,
1086 and multiple senders' messages were queued, only one sender would get
1087 notified on each configured delay_warning cycle.
1089 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1091 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1092 cipher-suites, an error can be left on the stack even for a succeeding
1093 accept; this results in impossible error messages when a later operation
1096 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1097 return error codes indicating retry. Under TLS1.3 this becomes required.
1099 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1100 it only wrote the new authenticators, resulting in a lack of tracking of
1101 peer changes of ESMTP extensions until the next cache flush.
1103 JH/41 Fix the loop reading a message header line to check for integer overflow,
1104 and more-often against header_maxsize. Previously a crafted message could
1105 induce a crash of the recive process; now the message is cleanly rejected.
1107 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1108 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1114 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1115 When redis_servers is set to a list of > 1 element, and the Redis servers
1116 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1117 case of MOVED into a DEFER case instead, thus moving the query onto the
1118 next server in the list. For a cluster of N elements, all N servers must
1119 be defined in redis_servers.
1121 GF/02 Catch and remove uninitialized value warning in exiqsumm
1122 Check for existence of @ARGV before looking at $ARGV[0]
1124 JH/01 Replace the store_release() internal interface with store_newblock(),
1125 which internalises the check required to safely use the old one, plus
1126 the allocate and data copy operations duplicated in both (!) of the
1127 extant use locations.
1129 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1130 modifier. This matches the restriction on the commandline.
1132 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1133 Previously only the last row was returned.
1135 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1136 we assumed that tags in the header were well-formed, and parsed the
1137 element content after inspecting only the first char of the tag.
1138 Assumptions at that stage could crash the receive process on malformed
1141 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1142 While running the DKIM ACL we operate on the Permanent memory pool so that
1143 variables created with "set" persist to the DATA ACL. Also (at any time)
1144 DNS lookups that fail create cache records using the Permanent pool. But
1145 expansions release any allocations made on the current pool - so a dnsdb
1146 lookup expansion done in the DKIM ACL releases the memory used for the
1147 DNS negative-cache, and bad things result. Solution is to switch to the
1148 Main pool for expansions.
1149 While we're in that code, add checks on the DNS cache during store_reset,
1150 active in the testsuite.
1151 Problem spotted, and debugging aided, by Wolfgang Breyha.
1153 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1154 When none of the hosts presented to a transport match an already-open
1155 connection, close it and proceed with the list. Previously we would
1156 queue the message. Spotted by Lena with Yahoo, probably involving
1159 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1160 Previously a spurious "250 OK id=" response was appended to the proper
1163 JH/08 The "support for" informational output now, which built with Content
1164 Scanning support, has a line for the malware scanner interfaces compiled
1165 in. Interface can be individually included or not at build time.
1167 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1168 by the template makefile "src/EDITME". The "STREAM" support for an older
1169 ClamAV interface method is removed.
1171 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1172 rows affected is given instead).
1174 JH/11 The runtime Berkeley DB library version is now additionally output by
1175 "exim -d -bV". Previously only the compile-time version was shown.
1177 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1178 SMTP connection. Previously, when one had more recipients than the
1179 first, an abortive onward connection was made. Move to full support for
1180 multiple onward connections in sequence, handling cutthrough connection
1181 for all multi-message initiating connections.
1183 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1184 routers. Previously, a multi-recipient message would fail to match the
1185 onward-connection opened for the first recipient, and cause its closure.
1187 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1188 a timeout on read on a GnuTLS initiating connection, resulting in the
1189 initiating connection being dropped. This mattered most when the callout
1190 was marked defer_ok. Fix to keep the two timeout-detection methods
1193 JH/15 Relax results from ACL control request to enable cutthrough, in
1194 unsupported situations, from error to silently (except under debug)
1195 ignoring. This covers use with PRDR, frozen messages, queue-only and
1198 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1200 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1201 metadata, resulting in a crash in free().
1203 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1204 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1205 Broken also in d185889f4, with init system revamp.
1207 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1208 Previously we abruptly closed the connection after reading a malware-
1209 found indication; now we go on to read the "scan ok" response line,
1212 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1213 ACL. Previously, a crash would result.
1215 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1216 macro text after a replacement (previously it was only once per line) and
1217 by skipping builtin macros when searching for an uppercase lead character.
1219 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1220 for the build is renamed.
1222 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1223 was allocated for every new TLS startup, meaning one per message. Fix
1224 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1226 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1227 reported the original. Fix to report (as far as possible) the ACL
1228 result replacing the original.
1230 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1231 OpenSSL. Certificate information is loaded for every new TLS startup,
1232 and the resources needed to be freed.
1234 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1236 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1239 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1240 DATA response info to the (existing) per-recipient response info for
1241 the "C=" log element. It can have useful tracking info from the
1242 destination system. Patch from Simon Arlott.
1244 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1245 length value. Previously this would segfault.
1247 HS/02 Support Avast multiline protoocol, this allows passing flags to
1248 newer versions of the scanner.
1250 JH/28 Ensure that variables possibly set during message acceptance are marked
1251 dead before release of memory in the daemon loop. This stops complaints
1252 about them when the debug_store option is enabled. Discovered specifically
1253 for sender_rate_period, but applies to a whole set of variables.
1254 Do the same for the queue-runner and queue-list loops, for variables set
1255 from spool message files. Do the same for the SMTP per-message loop, for
1256 certain variables indirectly set in ACL operations.
1258 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1259 as a multi-recipient message from a mailinglist manager). The coding had
1260 an arbitrary cutoff number of characters while checking for more input;
1261 enforced by writing a NUL into the buffer. This corrupted long / fast
1262 input. The problem was exposed more widely when more pipelineing of SMTP
1263 responses was introduced, and one Exim system was feeding another.
1264 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1265 receiving system, and refused recipients seen by the sending system
1266 (propating to people being dropped from mailing lists).
1267 Discovered and pinpointed by David Carter.
1269 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1270 replaced by the ${authresults } expansion.
1272 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1274 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1275 allows proper process termination in container environments.
1277 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1278 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1280 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1281 and "err_perm", deprecated since 4.83 when the RFC-defined words
1282 "temperror" and "permerror" were introduced.
1284 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1285 transport-filters or DKIM-signing. The restriction was lost in the
1286 consolidation of verify-callout and delivery SMTP handling.
1287 Extend the restriction to also cover ARC-signing.
1289 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1290 in defer=pass mode supply a 450 to the initiator. Previously the message
1293 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1294 tls_require_ciphers is used as before.
1296 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1297 "pass_unscanned". Only tmpfails from the scanner are written to
1298 the paniclog, as they may require admin intervention (permission
1299 denied, license issues). Other scanner errors (like decompression
1300 bombs) do not cause a paniclog entry.
1302 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1303 Previously it was possible to log spurious information in receive log
1306 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1307 triggered odd behaviour from Outlook Express clients.
1309 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1312 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1313 since the IETF WG has not yet settled on that versus the original
1314 "bare" representation.
1316 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1317 Previously the millisecond value corrupted the output.
1318 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1319 corrupted the output.
1325 JH/01 Rework error string handling in TLS interface so that the caller in
1326 more cases is responsible for logging. This permits library-sourced
1327 string to be attached to addresses during delivery, and collapses
1328 pairs of long lines into single ones.
1330 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1331 during configuration. Wildcards are allowed and expanded.
1333 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1334 This permits better logging.
1336 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1337 do not close it down and have the child transport start it up again on
1338 the passed-on TCP connection. Instead, proxy the child (and any
1339 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1340 affected: the continued delivery log lines do not have any DNSSEC, TLS
1341 Certificate or OCSP information. TLS cipher information is still logged.
1343 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1344 identical IP addresses on different listening ports. Will also affect
1347 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1348 add noisy ifdef guards to special-case this sillyness.
1349 Patch from Bernd Kuhls.
1351 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1352 than 255 are no longer allowed.
1354 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1355 Disable the session-cache too, which might reduce our load. Since we
1356 currrectly use a new context for every connection, both as server and
1357 client, there is no benefit for these.
1358 GnuTLS appears to not support tickets server-side by default (we don't
1359 call gnutls_session_ticket_enable_server()) but client side is enabled
1360 by default on recent versions (3.1.3 +) unless the PFS priority string
1363 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1364 <https://reproducible-builds.org/specs/source-date-epoch/>.
1366 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1367 the check for any unsuccessful recipients did not notice the limit, and
1368 erroneously found still-pending ones.
1370 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1371 MSG_MORE. Only in-clear (not on TLS connections).
1373 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1374 transport-filter is involved we can buffer the headers in memory for
1375 creating the signature, and read the spool data file once for the
1376 signature and again for transmission.
1378 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1379 4.77 as the kernel support then wasn't solid, having issues in 64bit
1380 mode. Now, it's been long enough. Add support for FreeBSD also.
1382 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1383 case where the routing stage had gathered several addresses to send to
1384 a host before calling the transport for the first, we previously failed
1385 to close down TLS in the old transport process before passing the TCP
1386 connection to the new process. The new one sent a STARTTLS command
1387 which naturally failed, giving a failed delivery and bloating the retry
1388 database. Investigation and fix prototype from Wolfgang Breyha.
1390 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1391 false-negatives in the check that the sender had not preempted a response
1392 or prompt from Exim (running as a server), due to that code's lack of
1393 awareness of the SMTP input buffering.
1395 PP/04 Add commandline_checks_require_admin option.
1396 Exim drops privileges sanely, various checks such as -be aren't a
1397 security problem, as long as you trust local users with access to their
1398 own account. When invoked by services which pass untrusted data to
1399 Exim, this might be an issue. Set this option in main configuration
1400 AND make fixes to the calling application, such as using `--` to stop
1403 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1404 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1405 responses to those, into a single TLS record each way (this usually means
1406 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1409 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1410 affects you only if you're dancing at the edge of the param size limits.
1411 If you are, and this message makes sense to you, then: raise the
1412 configured limit or use OpenSSL 1.1. Nothing we can do for older
1415 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1416 cmdline element to get the documented default one. Previously it was
1419 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1420 are unsuable for non-message-specific reasons - eg. connection timeout,
1421 banner-time rejection.
1423 JH/16 Fix logging of delivery remote port, when specified by router, under
1426 PP/06 Repair manualroute's ability to take options in any order, even if one
1427 is the name of a transport.
1430 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1432 JH/17 Change the list-building routines interface to use the expanding-string
1433 triplet model, for better allocation and copying behaviour.
1435 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1436 Previously it was constructed the first time a possibly-matching string
1437 was met in the configuration file input during startup; now it is done
1440 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1441 compatible one, to avoid the (poorly documented) possibility of a config
1442 file in the working directory redirecting the DB files, possibly correpting
1443 some existing file. CVE-2017-10140 assigned for BDB.
1445 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1446 cache-hot. Previously, although the result was properly cached, the
1447 initial verify call returned a defer.
1449 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1450 the main verify for receipient in uncached-mode.
1452 JH/22 Retire historical build files to an "unsupported" subdir. These are
1453 defined as "ones for which we have no current evidence of testing".
1455 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1456 if present. Previously it was ignored.
1458 JH/24 Start using specified-initialisers in C structure init coding. This is
1459 a C99 feature (it's 2017, so now considered safe).
1461 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1462 if was a fixed-sized field and bitmask ops via macros; it is now more
1465 PP/07 GitHub PR 56: Apply MariaDB build fix.
1466 Patch provided by Jaroslav Škarvada.
1468 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1469 during Coverity cleanups [4.87 JH/47]
1470 Diagnosis and fix provided by Michael Fischer v. Mollard.
1472 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1473 the right size to place the terminating semicolon on its own folded
1474 line, the header hash was calculated to an incorrect value thanks to
1475 the (relaxed) space the fold became.
1477 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1478 and confused the parent.
1480 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1481 which could crash as a result. This could lead to undeliverable messages.
1483 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1486 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1487 matters on fast-turnover and PID-randomising systems, which were getting
1488 out-of-order delivery.
1490 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1491 a possibly-overlapping copy. The symptom was that "Remote host closed
1492 connection in response to HELO" was logged instead of the actual 4xx
1495 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1496 Previously only that bufferd was discarded, resulting in SYMTP command
1499 JH/32 DKIM: when a message has multiple signatures matching an identity given
1500 in dkim_verify_signers, run the dkim acl once for each. Previously only
1501 one run was done. Bug 2189.
1503 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1504 "panic the current process" to "deliberately defer". The panic log is
1505 still written with the problem list name; the mail and reject logs now
1506 get a temp-reject line for the message that was being handled, saying
1507 something like "domains check lookup or other defer". The SMTP 451
1508 message is still "Temporary local problem".
1510 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1511 A crafted sequence of BDAT commands could result in in-use memory beeing
1512 freed. CVE-2017-16943.
1514 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1515 from SMTP input. Previously it was always done; now only done for DATA
1516 and not BDAT commands. CVE-2017-16944.
1518 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1519 to the message (such as an overlong header line). Previously this was
1520 not done and we did not exit BDAT mode. Followon from the previous item
1521 though a different problem.
1527 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1528 than -2003 did; needs libidn2 in addition to libidn.
1530 JH/02 The path option on a pipe transport is now expanded before use.
1532 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1533 Patch provided by "Björn", documentation fix added too.
1535 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1536 missing a wire-to-host endian conversion.
1538 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1539 close after a BDAT command line could be taken as a following command,
1540 giving a synch failure. Fix by only checking for synch immediately
1541 before acknowledging the chunk.
1543 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1544 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1545 macro. Patches provided by Josh Soref.
1547 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1548 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1549 by IANA as a service extension. However, John Klensin suggests that we
1552 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1553 the dkim code may be unix-mode line endings rather than smtp wire-format
1554 CRLF, so prepend a CR to any bare LF.
1556 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1557 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1559 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1560 the first were themselves being wrongly included in the feed into dkim
1561 processing; with most chunk sizes in use this resulted in an incorrect
1562 body hash calculated value.
1564 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1565 DKIM signature block, for verification. Although advised against by
1566 standards it is specifically not ruled illegal.
1568 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1570 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1571 missing a body hash (the bh= tag).
1573 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1574 It seems that HAProxy sends the Proxy Protocol information in clear and
1575 only then does a TLS startup, so do the same.
1577 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1578 TCP connections (such as for Spamd) unless the daemon successfully set
1579 Fast Open mode on its listening sockets. This fixes breakage seen on
1580 too-old kernels or those not configured for Fast Open, at the cost of
1581 requiring both directions being enabled for TFO, and TFO never being used
1582 by non-daemon-related Exim processes.
1584 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1585 endings, at least on the first header line. Try to canonify any that get
1586 past that check, despite the cost.
1588 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1589 now limited to an arbitrary five deep, while parsing addresses with the
1590 strip_excess_angle_brackets option enabled.
1592 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1593 instead leave the unprompted TLS handshake in socket buffer for the
1594 TLS library to consume.
1596 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1598 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1600 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1601 region in one operation, for speed, and this leaves assigned pointers
1602 dangling. Add checks run only under the testsuite which checks all
1603 variables at a store-reset and panics on a dangling pointer; add code
1604 explicitly nulling out all the variables discovered. Fixes one known
1605 bug: a transport crash, where a dangling pointer for $sending_ip_address
1606 originally assigned in a verify callout, is re-used.
1608 PP/06 Drop '.' from @INC in various Perl scripts.
1610 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1612 PP/08 Reduce a number of compilation warnings under clang; building with
1613 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1614 should be warning-free.
1616 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1618 HS/01 Fix portability problems introduced by PP/08 for platforms where
1619 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1621 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1622 chunk. This allows us to accept broken chunked messages. We need a more
1623 general solution here.
1625 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1626 already-broken messages in the queue.
1628 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1630 JH/19 Fix reference counting bug in routing-generated-address tracking.
1636 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1637 supports it and a size is available (ie. the sending peer gave us one).
1639 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1640 years of being deprecated). The replacements are the ACLs
1641 acl_smtp_mime and acl_not_smtp_mime.
1643 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1644 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1645 or even an in-clear connection were permitted. Now, if the host lookup
1646 was dnssec and dane was requested then the host is only used if the
1647 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1648 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1649 if one fails this test.
1650 This means that a poorly-configured remote DNS will make it incommunicado;
1651 but it protects against a DNS-interception attack on it.
1653 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1654 non-noisy when a race steals the message being considered.
1656 JH/05 If main configuration option tls_certificate is unset, generate a
1657 self-signed certificate for inbound TLS connections.
1659 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1660 in rewrites and routers.
1662 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1663 and logged a warning sing 4.83; now they are a configuration file error.
1665 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1666 (lacking @domain). Apply the same qualification processing as RCPT.
1668 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1670 JH/10 Support ${sha256:} applied to a string (as well as the previous
1673 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1674 a cutthrough deliver is pending, as we always want to make a connection.
1675 This also avoids re-routing the message when later placing the cutthrough
1676 connection after a verify cache hit.
1677 Do not update it with the verify result either.
1679 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1680 when routing results in more than one destination address.
1682 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1683 signing (which inhibits the cutthrough capability). Previously only
1684 the presence of an option was tested; now an expansion evaluating as
1685 empty is permissible (obviously it should depend only on data available
1686 when the cutthrough connection is made).
1688 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1689 the relevant preceding SMTP command did not note the pipelining mode.
1691 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1692 Previously they were not counted.
1694 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1695 as one having no matching records. Previously we deferred the message
1696 that needed the lookup.
1698 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1699 distinguished as "(=".
1701 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1702 for missing MX records. Previously it only worked for missing A records.
1704 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1706 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1707 after the data-go-ahead and data-ack. Patch from Jason Betts.
1709 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1710 even for a "none" policy. Patch from Tony Meyer.
1712 JH/22 Fix continued use of a connection for further deliveries. If a port was
1713 specified by a router, it must also match for the delivery to be
1716 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1717 When one of the recipients of a message was unsuitable for the connection
1718 (has no matching addresses), we lost track of needing to mark it
1719 deferred. As a result mail would be lost.
1721 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1723 JH/25 Decoding ACL controls is now done using a binary search; the source code
1724 takes up less space and should be simpler to maintain. Merge the ACL
1725 condition decode tables also, with similar effect.
1727 JH/26 Fix problem with one_time used on a redirect router which returned the
1728 parent address unchanged. A retry would see the parent address marked as
1729 delivered, so not attempt the (identical) child. As a result mail would
1732 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1733 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1734 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1737 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1738 Main config options check_{log,spool}_{inodes,space} are now
1739 100 inodes, 10MB unless set otherwise in the configuration.
1741 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1742 Previously it only applied to the main-section connection policy
1745 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1747 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1748 by me. Added RFC7919 DH primes as an alternative.
1750 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1751 are not in the system include path.
1753 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1754 GnuTLS, when a session startup failed (eg because the client disconnected)
1755 Exim did stdio operations after fclose. This was exposed by a recent
1756 change which nulled out the file handle after the fclose.
1758 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1759 signed directly by the cert-signing cert, rather than an intermediate
1760 OCSP-signing cert. This is the model used by LetsEncrypt.
1762 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1764 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1765 an incoming connection.
1767 HS/02 Bug 1802: Do not half-close the connection after sending a request
1770 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1771 fallback to "prime256v1".
1773 JH/34 SECURITY: Use proper copy of DATA command in error message.
1774 Could leak key material. Remotely exploitable. CVE-2016-9963.
1780 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1781 and 3.4.4 - once the server is enabled to respond to an OCSP request
1782 it does even when not requested, resulting in a stapling non-aware
1783 client dropping the TLS connection.
1785 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1786 support variable-length bit vectors. No functional change.
1788 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1789 The I= interface field on outgoing lines is now after the H= remote
1790 host field, same as incoming lines. There is a separate
1791 outgoing_interface log selector which allows you to disable the
1794 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1795 If not running log_selector +smtp_connection the mainlog would be held
1796 open indefinitely after a "too many connections" event, including to a
1797 deleted file after a log rotate. Leave the per net connection logging
1798 leaving it open for efficiency as that will be quickly detected by the
1799 check on the next write.
1801 HS/01 Bug 1671: Fix post transport crash.
1802 Processing the wait-<transport> messages could crash the delivery
1803 process if the message IDs didn't exist for some reason. When
1804 using 'split_spool_directory=yes' the construction of the spool
1805 file name failed already, exposing the same netto behaviour.
1807 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1808 mime_regex ACL conditions.
1810 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1811 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1812 to HELO, local diagnostic string.
1814 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1815 log line to debug. Even when configured with a tls authenticator many
1816 client connections are expected to not authenticate in this way, so
1817 an authenticate fail is not an error.
1819 HS/02 Add the Exim version string to the process info. This way exiwhat
1820 gives some more detail about the running daemon.
1822 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1823 matter for fast-change records such as DNSBLs.
1825 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1826 retry record, even if constant. There may be multiple transports with
1827 different interface settings and the retry behaviour needs to be kept
1830 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1832 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1834 JH/10 Bug 840: fix log_defer_output option of pipe transport
1836 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1837 hosts_randomize. This matters a lot when combined with mua_wrapper.
1839 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1840 ${quote_pgsql:<string>} operator.
1842 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1844 JH/14 Change header file name for freeradius-client. Relevant if compiling
1845 with Radius support; from the Gentoo tree and checked under Fedora.
1847 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1849 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1850 option transport_filter, meaning no filtering.
1852 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1854 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1855 defaults to "*" (all hosts). The variable is now available when not built
1856 with TLS, default unset, mainly to enable keeping the testsuite sane.
1857 If a server certificate is not supplied (via tls_certificate) an error is
1858 logged, and clients will find TLS connections fail on startup. Presumably
1859 they will retry in-clear.
1860 Packagers of Exim are strongly encouraged to create a server certificate
1861 at installation time.
1863 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1864 with the $config_file variable.
1866 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1867 in transport context, after the attempt, and per-recipient. The latter type
1868 is per host attempted. The event data is the error message, and the errno
1869 information encodes the lookup type (A vs. MX) used for the (first) host,
1870 and the trailing two digits of the smtp 4xx response.
1872 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1873 to write to mainlog (or rejectlog, paniclog) in the window between file
1874 creation and permissions/ownership being changed. Particularly affects
1875 installations where exicyclog is run as root, rather than exim user;
1876 result is that the running daemon panics and dies.
1878 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1880 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1881 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1882 "pri" and "weight". Note that the previous implicit priority given by the
1883 list order is no longer honoured.
1885 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1886 for DKIM processing.
1888 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1889 by defining SUPPORT_SOCKS.
1891 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1892 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1893 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1894 variables are renamed to proxy_{local,external}_{address,port}.
1896 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1897 for a build by defining SUPPORT_I18N
1899 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1900 of the query string, and make ${quote_redis:} do that quoting.
1902 JH/29 Move Events support from Experimental to mainline, enabled by default
1903 and removable for a build by defining DISABLE_EVENT.
1905 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1907 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1908 cached by the daemon.
1910 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1911 by defining LOOKUP_REDIS. The libhiredis library is required.
1913 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1914 keys are given for lookup.
1916 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1917 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1918 only supported when built with TLS support. The PolarSSL SHA routines
1919 are still used when the TLS library is too old for convenient support.
1921 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1922 openssl_options), for security. OpenSSL forces this from version 1.1.0
1923 server-side so match that on older versions.
1925 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1926 allocation for $value could be released as the expansion processing
1927 concluded, but leaving the global pointer active for it.
1929 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1930 and to use the domains and local_parts ACL conditions.
1932 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1933 incorrectly not doubled on cutthrough transmission, hence seen as a
1934 body-termination at the receiving system - resulting in truncated mails.
1935 Commonly the sender saw a TCP-level error, and retransmitted the message
1936 via the normal store-and-forward channel. This could result in duplicates
1937 received - but deduplicating mailstores were liable to retain only the
1938 initial truncated version.
1940 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1942 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1944 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1945 we're in there, support oversigning also; bug 1309.
1947 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1949 HS/04 Add support for keep_environment and add_environment options.
1951 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1952 either intentional arithmetic overflow during PRNG, or testing config-
1955 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1956 delivery resulted in actual delivery. Cancel cutthrough before DATA
1959 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1960 rejects a recipient: pass the reject to the originator.
1962 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1963 Many were false-positives and ignorable, but it's worth fixing the
1966 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1967 for the new environment-manipulation done at startup. Move the routines
1968 from being local to tls.c to being global via the os.c file.
1970 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1971 an extract embedded as result-arg for a map, the first arg for extract
1972 is unavailable so we cannot tell if this is a numbered or keyed
1973 extraction. Accept either.
1979 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1982 JH/02 The smtp transport option "multi_domain" is now expanded.
1984 JH/03 The smtp transport now requests PRDR by default, if the server offers
1987 JH/04 Certificate name checking on server certificates, when exim is a client,
1988 is now done by default. The transport option tls_verify_cert_hostnames
1989 can be used to disable this per-host. The build option
1990 EXPERIMENTAL_CERTNAMES is withdrawn.
1992 JH/05 The value of the tls_verify_certificates smtp transport and main options
1993 default to the word "system" to access the system default CA bundle.
1994 For GnuTLS, only version 3.0.20 or later.
1996 JH/06 Verification of the server certificate for a TLS connection is now tried
1997 (but not required) by default. The verification status is now logged by
1998 default, for both outbound TLS and client-certificate supplying inbound
2001 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2004 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2005 Status Notification (bounce) messages are now MIME format per RFC 3464.
2006 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2007 under the control of the dsn_advertise_hosts option, and routers may
2008 have a dsn_lasthop option.
2010 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2011 default, modifiable by a malware= option. The list separator for
2012 the options can now be changed in the usual way. Bug 68.
2014 JH/10 The smtp_receive_timeout main option is now expanded before use.
2016 JH/11 The incoming_interface log option now also enables logging of the
2017 local interface on delivery outgoing connections.
2019 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2020 if the interface and destination host and port all match.
2022 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2025 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2026 Patch from Andrew Lewis.
2028 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2029 now supports optional time-restrictions, weighting, and priority
2030 modifiers per server. Patch originally by <rommer@active.by>.
2032 JH/16 The spamd_address main option now supports a mixed list of local
2033 and remote servers. Remote servers can be IPv6 addresses, and
2034 specify a port-range.
2036 JH/17 Bug 68: The spamd_address main option now supports an optional
2037 timeout value per server.
2039 JH/18 Bug 1581: Router and transport options headers_add/remove can
2040 now have the list separator specified.
2042 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2045 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2048 JH/21 Support for the A6 type of dns record is withdrawn.
2050 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2051 rather than the verbs used.
2053 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2054 from 255 to 1024 chars.
2056 JH/24 Verification callouts now attempt to use TLS by default.
2058 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2059 are generic router options now. The defaults didn't change.
2061 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2062 Original patch from Alexander Shikoff, worked over by JH.
2064 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2065 files if ZCAT_COMMAND is not executable.
2067 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2069 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2071 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2072 Normally benign, it bites when the pair was led to by a CNAME;
2073 modern usage is to not canonicalize the domain to a CNAME target
2074 (and we were inconsistent anyway for A-only vs AAAA+A).
2076 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2078 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2079 when evaluating $sender_host_dnssec.
2081 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2082 $sender_helo_dnssec variable.
2084 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2086 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2088 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2090 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2091 documented as working, but never had. Support all but $spam_report.
2093 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2094 added for tls authenticator.
2096 HS/03 Add perl_taintmode main config option
2102 TL/01 When running the test suite, the README says that variables such as
2103 no_msglog_check are global and can be placed anywhere in a specific
2104 test's script, however it was observed that placement needed to be near
2105 the beginning for it to behave that way. Changed the runtest perl
2106 script to read through the entire script once to detect and set these
2107 variables, reset to the beginning of the script, and then run through
2108 the script parsing/test process like normal.
2110 TL/02 The BSD's have an arc4random API. One of the functions to induce
2111 adding randomness was arc4random_stir(), but it has been removed in
2112 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2113 function when detected.
2115 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2116 cause callback expansion.
2118 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2119 syntax errors in an expansion can be treated as a string instead of
2120 logging or causing an error, due to the internal use of bool_lax
2121 instead of bool when processing it.
2123 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2124 server certificates when making smtp deliveries.
2126 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2128 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2130 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2132 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2133 Merged patch from Sebastian Wiedenroth.
2135 JH/05 Fix results-pipe from transport process. Several recipients, combined
2136 with certificate use, exposed issues where response data items split
2137 over buffer boundaries were not parsed properly. This eventually
2138 resulted in duplicates being sent. This issue only became common enough
2139 to notice due to the introduction of connection certificate information,
2140 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2142 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2143 size buffer was used, resulting in syntax errors when an expansion
2146 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2147 version 3.3.6 or later.
2149 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2150 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2151 both become "event_action", the variables become $event_name, $event_data
2152 and $event_defer_errno. There is a new variable $verify_mode, usable in
2153 routers, transports and related events. The tls:cert event is now also
2154 raised for inbound connections, if the main configuration event_action
2157 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2158 early OCSP support, but no stapling (appears to be less than 1.0.0).
2160 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2161 server certificate names available under the smtp transport option
2162 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2165 JH/10 Time-related extraction expansions from certificates now use the main
2166 option "timezone" setting for output formatting, and are consistent
2167 between OpenSSL and GnuTLS compilations. Bug 1541.
2169 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2170 encoded parameter in the incoming message. Bug 1558.
2172 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2173 include certificate info, eximon was claiming there were spoolfile
2176 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2178 JH/14 Log delivery-related information more consistently, using the sequence
2179 "H=<name> [<ip>]" wherever possible.
2181 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2182 are problematic for Debian distribution, omit them from the release
2185 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2187 JH/16 Fix string representation of time values on 64bit time_t architectures.
2190 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2191 output list separator was used.
2196 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2197 checkers that were complaining about end of non-void function with no
2200 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2201 This was a regression introduced in 4.83 by another bugfix.
2203 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2205 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2206 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2212 TF/01 Correctly close the server side of TLS when forking for delivery.
2214 When a message was received over SMTP with TLS, Exim failed to clear up
2215 the incoming connection properly after forking off the child process to
2216 deliver the message. In some situations the subsequent outgoing
2217 delivery connection happened to have the same fd number as the incoming
2218 connection previously had. Exim would try to use TLS and fail, logging
2219 a "Bad file descriptor" error.
2221 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2222 utilities have not been installed.
2224 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2225 temporary space as the ACL may create new global variables.
2227 TL/01 LDAP support uses per connection or global context settings, depending
2228 upon the detected version of the libraries at build time.
2230 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2231 to extract and use the src ip:port in logging and expansions as if it
2232 were a direct connection from the outside internet. PPv2 support was
2233 updated based on HAProxy spec change in May 2014.
2235 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2237 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2238 Properly escape header and check for NULL return.
2240 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2243 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2245 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2246 characters in header names, implemented as a verify condition.
2247 Contributed by Michael Fischer v. Mollard.
2249 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2250 results permerror and temperror. Previous values are deprecated but
2251 still accepted. In a future release, err_perm and err_temp will be
2252 completely removed, which will be a backward incompatibility if the
2253 ACL tests for either of these two old results. Patch contributed by
2254 user bes-internal on the mailing list.
2256 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2258 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2259 selectors, in both main and reject logs.
2261 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2264 JH/07 Add malware type "sock" for talking to simple daemon.
2266 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2268 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2269 routers/transports under cutthrough routing.
2271 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2272 numbers. Touch up "bool" conditional to keep the same definition.
2274 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2276 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2278 JH/12 Expand items in router/transport headers_add or headers_remove lists
2279 individually rather than the list as a whole. Bug 1452.
2281 Required for reasonable handling of multiple headers_ options when
2282 they may be empty; requires that headers_remove items with embedded
2283 colons must have them doubled (or the list-separator changed).
2285 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2286 view the policy declared in the DMARC record. Currently, $dmarc_status
2287 is a combined value of both the record presence and the result of the
2290 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2292 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2293 dnslookup router and the smtp transport (applying to the forward
2296 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2297 of ldap servers used for a specific lookup. Patch provided by Heiko
2300 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2301 New variable $lookup_dnssec_authenticated for observability.
2303 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2304 Patch submitted by Lars Timman.
2306 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2308 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2309 Requires trusted mode and valid format message id, aborts otherwise.
2310 Patch contributed by Heiko Schlichting.
2312 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2313 certextract with support for various fields. Bug 1358.
2315 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2316 is requested by default, modifiable by smtp transport option
2319 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2320 operate on certificate variables to give certificate fingerprints
2321 Also new ${sha256:cert_variable}.
2323 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2325 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2326 Christian Aistleitner.
2328 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2330 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2331 file. Patch from Wolfgang Breyha.
2333 JH/25 Expand the coverage of the delivery $host and $host_address to
2334 client authenticators run in verify callout. Bug 1476.
2336 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2337 align with daemon_smtp_ports. Bug 72.
2339 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2340 support and error reporting did not work properly.
2342 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2343 and is readable. Patch from Andrew Colin Kissa.
2345 TL/14 Enhance documentation of ${run expansion and how it parses the
2346 commandline after expansion, particularly in the case when an
2347 unquoted variable expansion results in an empty value.
2349 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2351 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2352 by renaming away from it. Addresses GH issue 15, reported by
2355 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2357 TL/15 SECURITY: prevent double expansion in math comparison functions
2358 (can expand unsanitized data). Not remotely exploitable.
2365 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2367 PP/02 Make -n do something, by making it not do something.
2368 When combined with -bP, the name of an option is not output.
2370 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2373 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2374 $sender_host_name and config options to manage this, and basic check
2377 PP/05 DSCP support for outbound connections and control modifier for inbound.
2379 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2380 (Only plugin which currently uses this is kerberos4, which nobody should
2381 be using, but we should make it available and other future plugins might
2382 conceivably use it, even though it would break NAT; stuff *should* be
2383 using channel bindings instead).
2385 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2386 name; added for Sendmail compatibility; requires admin caller.
2387 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2388 just ignore it); requires trusted caller.
2389 Also parse but ignore: -Ac -Am -X<logfile>
2392 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2394 TL/02 Add +smtp_confirmation as a default logging option.
2396 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2397 Patch by Magnus Holmgren from 2007-02-20.
2399 TL/04 Bugzilla 1281 - Spec typo.
2400 Bugzilla 1283 - Spec typo.
2401 Bugzilla 1290 - Spec grammar fixes.
2403 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2405 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2407 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2408 mailing list by by Dmitry Isaikin.
2410 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2412 JH/02 Support "G" suffix to numbers in ${if comparisons.
2414 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2416 NM/01 Bugzilla 1197 - Spec typo
2417 Bugzilla 1196 - Spec examples corrections
2419 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2421 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2422 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2425 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2426 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2428 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2429 "acl {{name}{arg}...}", and optional args on acl condition
2432 JH/05 Permit multiple router/transport headers_add/remove lines.
2434 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2436 JH/07 Avoid using a waiting database for a single-message-only transport.
2437 Performance patch from Paul Fisher. Bugzilla 1262.
2439 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2442 JH/09 Add $headers_added variable, with content from use of ACL modifier
2443 add_header (but not yet added to the message). Bugzilla 199.
2445 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2446 Pulled from Bugzilla 817 by Wolfgang Breyha.
2448 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2450 (nb: this is the same fix as in Exim 4.80.1)
2452 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2455 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2458 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2460 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2461 advertise SMTP AUTH mechanism to us, instead of a generic
2462 protocol violation error. Also, make Exim more robust to bad
2463 data from the Dovecot auth socket.
2465 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2467 When a queue runner is handling a message, Exim first routes the
2468 recipient addresses, during which it prunes them based on the retry
2469 hints database. After that it attempts to deliver the message to
2470 any remaining recipients. It then updates the hints database using
2473 So if a recipient address works intermittently, it can get repeatedly
2474 deferred at routing time. The retry hints record remains fresh so the
2475 address never reaches the final cutoff time.
2477 This is a fairly common occurrence when a user is bumping up against
2478 their storage quota. Exim had some logic in its local delivery code
2479 to deal with this. However it did not apply to per-recipient defers
2480 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2482 This change adds a proper retry rule check during routing so that the
2483 final cutoff time is checked against the message's age. We only do
2484 this check if there is an address retry record and there is not a
2485 domain retry record; this implies that previous attempts to handle
2486 the address had the retry_use_local_parts option turned on. We use
2487 this as an approximation for the destination being like a local
2488 delivery, as in LMTP.
2490 I suspect this new check makes the old local delivery cutoff check
2491 redundant, but I have not verified this so I left the code in place.
2493 TF/02 Correct gecos expansion when From: is a prefix of the username.
2495 Test 0254 submits a message to Exim with the header
2499 When I ran the test suite under the user fanf2, Exim expanded
2500 the header to contain my full name, whereas it should have added
2501 a Resent-Sender: header. It erroneously treats any prefix of the
2502 username as equal to the username.
2504 This change corrects that bug.
2506 GF/01 DCC debug and logging tidyup
2507 Error conditions log to paniclog rather than rejectlog.
2508 Debug lines prefixed by "DCC: " to remove any ambiguity.
2510 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2512 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2513 Bug spotted by Jeremy Harris; was flawed since initial commit.
2514 Would have resulted in OCSP responses post-SNI triggering an Exim
2515 NULL dereference and crash.
2517 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2519 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2520 Bug detection, analysis and fix by Samuel Thibault.
2521 Bugzilla 1331, Debian bug #698092.
2523 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2525 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2526 Server implementation by Todd Lyons, client by JH.
2527 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2528 config variable "prdr_enable" controls whether the server
2529 advertises the facility. If the client requests PRDR a new
2530 acl_data_smtp_prdr ACL is called once for each recipient, after
2531 the body content is received and before the acl_smtp_data ACL.
2532 The client is controlled by both of: a hosts_try_prdr option
2533 on the smtp transport, and the server advertisement.
2534 Default client logging of deliveries and rejections involving
2535 PRDR are flagged with the string "PRDR".
2537 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2538 fclose(). Diagnosis by Todd Lyons.
2540 PP/17 Update configure.default to handle IPv6 localhost better.
2541 Patch by Alain Williams (plus minor tweaks).
2544 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2545 This is now consistent with GnuTLS, and is now documented: the
2546 previous undocumented portable approach to treating the option as
2547 unset was to force an expansion failure. That still works, and
2548 an empty string is now equivalent.
2550 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2551 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2552 not performing validation itself.
2554 PP/20 Added force_command boolean option to pipe transport.
2555 Patch from Nick Koston, of cPanel Inc.
2557 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2560 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2562 PP/21 Fix eximon continuous updating with timestamped log-files.
2563 Broken in a format-string cleanup in 4.80, missed when I repaired the
2564 other false fix of the same issue.
2565 Report and fix from Heiko Schlichting.
2568 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2569 Report from Prashanth Katuri.
2571 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2572 It's SecureTransport, so affects any MacOS clients which use the
2573 system-integrated TLS libraries, including email clients.
2575 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2576 using a MIME ACL for non-SMTP local injection.
2577 Report and assistance in diagnosis by Warren Baker.
2579 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2581 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2583 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2584 last id that failed so it may be referenced in subsequent ACL's.
2586 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2589 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2590 ldap library initialization, allowing self-signed CA's to be
2591 used. Also properly sets require_cert option later in code by
2592 using NULL (global ldap config) instead of ldap handle (per
2593 session). Bug diagnosis and testing by alxgomz.
2595 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2596 the src/util/ subdirectory.
2598 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2599 renamed to Transport Post Delivery Action by Jeremy Harris, as
2602 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2603 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2604 redis_servers = needs to be configured which will be used by the redis
2605 lookup. Patch from Warren Baker, of The Packet Hub.
2607 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2609 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2610 hostname or reverse DNS when processing a host list. Used suggestions
2611 from multiple comments on this bug.
2613 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2615 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2616 Missed a few lines, added it to make the runtest require no keyboard
2619 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2620 contains upper case chars. Make router use caseful_local_part.
2622 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2623 support when GnuTLS has been built with p11-kit.
2629 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2631 This, or similar/improved, will also be change PP/11 of 4.82.
2637 PP/01 Handle short writes when writing local log-files.
2638 In practice, only affects FreeBSD (8 onwards).
2639 Bugzilla 1053, with thanks to Dmitry Isaikin.
2641 NM/01 Bugzilla 949 - Documentation tweak
2643 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2646 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2648 PP/02 Implemented gsasl authenticator.
2650 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2652 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2653 `pkg-config foo` for cflags/libs.
2655 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2656 with rest of GSASL and with heimdal_gssapi.
2658 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2659 `pkg-config foo` for cflags/libs for the TLS implementation.
2661 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2662 properties get this fed in as external SSF. A number of robustness
2663 and debugging improvements to the cyrus_sasl authenticator.
2665 PP/08 cyrus_sasl server now expands the server_realm option.
2667 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2668 Patch by Jeremy Harris.
2670 PP/10 Added dbmjz lookup type.
2672 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2674 PP/12 MAIL args handles TAB as well as SP, for better interop with
2675 non-compliant senders.
2676 Analysis and variant patch by Todd Lyons.
2678 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2679 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2680 Patch from Dirk Mueller <dmueller@suse.com>
2682 PP/13 tls_peerdn now print-escaped for spool files.
2683 Observed some $tls_peerdn in wild which contained \n, which resulted
2684 in spool file corruption.
2686 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2687 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2688 or write after TLS renegotiation, which otherwise led to messages
2691 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2692 as a tracking header (ie: a signed header comes before the signature).
2693 Patch from Wolfgang Breyha.
2695 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2696 comma-sep list; embedded commas doubled.
2698 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2700 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2702 Report and patch from Dmitry Banschikov.
2704 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2705 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2706 failures. We appear to now support TLS1.1+ with Exim.
2708 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2709 lets Exim select keys and certificates based upon TLS SNI from client.
2710 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2711 before an outbound SMTP session. New log_selector, +tls_sni.
2713 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2714 NULL dereference. Report and patch from Alun Jones.
2716 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2717 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2718 Not seeing resolver debug output on NetBSD, but suspect this is a
2719 resolver implementation change.
2721 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2722 Left warnings. Added "eximon gdb" invocation mode.
2724 PP/21 Defaulting "accept_8bitmime" to true, not false.
2726 PP/22 Added -bw for inetd wait mode support.
2728 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2729 locate the relevant includes and libraries. Made this the default.
2731 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2732 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2734 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2735 This may cause build issues on older platforms.
2737 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2738 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2739 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2740 Added SNI support via GnuTLS too.
2741 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2743 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2745 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2746 Applied second patch from Janne, implementing suggestion to default
2747 multiple-strings-in-record handling to match SPF spec.
2749 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2751 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2752 read-only, out of scope).
2753 Patch from Wolfgang Breyha, report from Stuart Northfield.
2755 PP/29 Fix three issues highlighted by clang analyser static analysis.
2756 Only crash-plausible issue would require the Cambridge-specific
2757 iplookup router and a misconfiguration.
2758 Report from Marcin Mirosław.
2760 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2762 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2763 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2764 As part of this, removing so much warning spew let me fix some minor
2765 real issues in debug logging.
2767 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2768 assignment on my part. Fixed.
2770 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2771 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2772 Janne Snabb (who went above and beyond: thank you).
2774 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2775 string otherwise requires a connection and a bunch more work and it's
2776 relatively easy to get wrong. Should also expose TLS library linkage
2779 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2780 64-bit ${eval} (JH/03).
2782 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2783 GNU libc to support some of the 64-bit stuff, should not lead to
2784 conflicts. Defined before os.h is pulled in, so if a given platform
2785 needs to override this, it can.
2787 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2788 protection layer was required, which is not implemented.
2789 Bugzilla 1254, patch from Wolfgang Breyha.
2791 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2792 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2793 tls_dhparam take prime identifiers. Also unbreak combination of
2794 OpenSSL+DH_params+TLSSNI.
2796 PP/39 Disable SSLv2 by default in OpenSSL support.
2802 PP/01 Solaris build fix for Oracle's LDAP libraries.
2803 Bugzilla 1109, patch from Stephen Usher.
2805 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2807 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2810 TF/02 Fix a couple more cases where we did not log the error message
2811 when unlink() failed. See also change 4.74-TF/03.
2813 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2814 lock up or crash if it happened to be inside a call to libc when it
2815 got a SIGUSR1 from exiwhat.
2817 The SIGUSR1 handler appends the current process status to the process
2818 log which is later printed by exiwhat. It used to use the general
2819 purpose logging code to do this, but several functions it calls are
2820 not safe for signals.
2822 The new output code in the SIGUSR1 handler is specific to the process
2823 log, and simple enough that it's easy to inspect for signal safety.
2824 Removing some special cases also simplifies the general logging code.
2825 Removing the spurious timestamps from the process log simplifies
2828 TF/04 Improved ratelimit ACL condition.
2830 The /noupdate option has been deprecated in favour of /readonly which
2831 has clearer semantics. The /leaky, /strict, and /readonly update modes
2832 are mutually exclusive. The update mode is no longer included in the
2833 database key; it just determines when the database is updated. (This
2834 means that when you upgrade Exim will forget old rate measurements.)
2836 Exim now checks that the per_* options are used with an update mode that
2837 makes sense for the current ACL. For example, when Exim is processing a
2838 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2839 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2840 must specify per_mail/readonly. If you omit the update mode it defaults to
2841 /leaky where that makes sense (as before) or /readonly where required.
2843 The /noupdate option is now undocumented but still supported for
2844 backwards compatibility. It is equivalent to /readonly except that in
2845 ACLs where /readonly is required you may specify /leaky/noupdate or
2846 /strict/noupdate which are treated the same as /readonly.
2848 A useful new feature is the /count= option. This is a generalization
2849 of the per_byte option, so that you can measure the throughput of other
2850 aggregate values. For example, the per_byte option is now equivalent
2851 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2853 The per_rcpt option has been generalized using the /count= mechanism
2854 (though it's more complicated than the per_byte equivalence). When it is
2855 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2856 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2857 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2858 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2859 Note that using per_rcpt with a non-readonly update mode in more than
2860 one ACL will cause the recipients to be double-counted. (The per_mail
2861 and per_byte options don't have this problem.)
2863 The handling of very low rates has changed slightly. If the computed rate
2864 is less than the event's count (usually one) then this event is the first
2865 after a long gap. In this case the rate is set to the same as this event's
2866 count, so that the first message of a spam run is counted properly.
2868 The major new feature is a mechanism for counting the rate of unique
2869 events. The new per_addr option counts the number of different
2870 recipients that someone has sent messages to in the last time period. It
2871 behaves like per_rcpt if all the recipient addresses are different, but
2872 duplicate recipient addresses do not increase the measured rate. Like
2873 the /count= option this is a general mechanism, so the per_addr option
2874 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2875 example, measure the rate that a client uses different sender addresses
2876 with the options per_mail/unique=$sender_address. There are further
2877 details in the main documentation.
2879 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2881 TF/06 Removed a few PCRE remnants.
2883 TF/07 Automatically extract Exim's version number from tags in the git
2884 repository when doing development or release builds.
2886 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2887 Bugzilla 879. Patch from Paul Fisher.
2889 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2890 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2893 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2895 PP/05 Variable $av_failed, true if the AV scanner deferred.
2896 Bugzilla 1078. Patch from John Horne.
2898 PP/06 Stop make process more reliably on build failure.
2899 Bugzilla 1087. Patch from Heiko Schlittermann.
2901 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2902 Bugzilla 1089. Patch from Heiko Schlittermann.
2904 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2905 Bugzilla 1131. Patch from Holger Weiß.
2907 PP/09 Handle IPv6 addresses with SPF.
2908 Bugzilla 860. Patch from Wolfgang Breyha.
2910 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2912 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2915 PP/11 match_* no longer expand right-hand-side by default.
2916 New compile-time build option, EXPAND_LISTMATCH_RHS.
2917 New expansion conditions, "inlist", "inlisti".
2919 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2921 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2923 PP/14 fix log_write() format string regression from TF/03.
2924 Bugzilla 1152. Patch from Dmitry Isaikin.
2930 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2932 PP/02 Harmonised TLS library version reporting; only show if debugging.
2933 Layout now matches that introduced for other libraries in 4.74 PP/03.
2935 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2937 PP/04 New "dns_use_edns0" global option.
2939 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2942 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2943 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2945 TK/01 Updated PolarSSL code to 0.14.2.
2946 Bugzilla 1097. Patch from Andreas Metzler.
2948 PP/07 Catch divide-by-zero in ${eval:...}.
2949 Fixes bugzilla 1102.
2951 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2954 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2955 format-string attack -- SECURITY: remote arbitrary code execution.
2957 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2958 time unintentionally subject to list matching rules, letting the header
2959 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2960 arbitrary string expansion). This allowed for information disclosure.
2962 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2963 INT_MIN/-1 -- value coerced to INT_MAX.
2969 NM/01 Workaround for PCRE version dependency in version reporting
2972 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2973 This fixes portability to compilers other than gcc, notably
2974 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2976 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2977 makefiles for portability to HP-UX and POSIX correctness.
2979 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2980 Also via indented variable definition in the Makefile.
2981 (Debugging by Oliver Heesakkers).
2983 PP/02 Restore caching of spamd results with expanded spamd_address.
2984 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2986 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2987 Improves build reliability. Fix from: Frank Elsner
2989 NM/02 Fix wide character breakage in the rfc2047 coding
2990 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2992 NM/03 Allow underscore in dnslist lookups
2993 Fixes bug 1026. Patch from Graeme Fowler
2995 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2996 Code patches from Adam Ciarcinski of NetBSD.
2998 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3001 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3002 is logged, to avoid truncation. Patch from John Horne.
3004 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3005 Patch from Jakob Hirsch.
3007 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3008 SQL string expansion failure details.
3009 Patch from Andrey Oktyabrski.
3011 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3012 Patch from Simon Arlott.
3014 PP/09 New lookups functionality failed to compile on old gcc which rejects
3015 extern declarations in function scope.
3016 Patch from Oliver Fleischmann
3018 PP/10 Use sig_atomic_t for flags set from signal handlers.
3019 Check getgroups() return and improve debugging.
3020 Fixed developed for diagnosis in bug 927 (which turned out to be
3023 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3024 Patch from Mark Zealey.
3026 PP/12 Bugzilla 1056: Improved spamd server selection.
3027 Patch from Mark Zealey.
3029 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3030 Based on patch from Heiko Schlittermann.
3032 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3033 Patch from Uwe Doering, sign-off by Michael Haardt.
3035 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3036 variable declaration deep within a block. Bug and patch from
3039 PP/15 lookups-Makefile IRIX compatibility coercion.
3041 PP/16 Make DISABLE_DKIM build knob functional.
3043 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3044 Patch by Simon Arlott
3046 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3047 variable argument macros. Our copy now differs from upstream.
3053 TF/01 Failure to get a lock on a hints database can have serious
3054 consequences so log it to the panic log.
3056 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3057 controlled using the smtp_confirmation log selector.
3059 TF/03 Include the error message when we fail to unlink a spool file.
3061 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3062 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3063 for maintaining out-of-tree patches for some time.
3065 PP/01 Bugzilla 139: Documentation and portability issues.
3066 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3067 Handle per-OS dynamic-module compilation flags.
3069 PP/02 Let /dev/null have normal permissions.
3070 The 4.73 fixes were a little too stringent and complained about the
3071 permissions on /dev/null. Exempt it from some checks.
3072 Reported by Andreas M. Kirchwitz.
3074 PP/03 Report version information for many libraries, including
3075 Exim version information for dynamically loaded libraries. Created
3076 version.h, now support a version extension string for distributors
3077 who patch heavily. Dynamic module ABI change.
3079 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3080 privilege escalation vulnerability whereby the Exim run-time user
3081 can cause root to append content of the attacker's choosing to
3084 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3087 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3088 If dropping privileges for untrusted macros, we disabled normal logging
3089 on the basis that it would fail; for the Exim run-time user, this is not
3090 the case, and it resulted in successful deliveries going unlogged.
3091 Fixed. Reported by Andreas Metzler.
3097 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3098 only prepend for the Resent-* case. Fixes regression introduced in
3099 Exim 4.70 by NM/22 for Bugzilla 607.
3101 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3102 increasing numbers of administrators be bitten by this.
3104 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3106 PP/03 Bugzilla 994: added openssl_options main configuration option.
3108 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3110 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3112 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3114 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3115 without a peer certificate, leading to a segfault because of an
3116 assumption that peers always have certificates. Be a little more
3117 paranoid. Problem reported by Martin Tscholak.
3119 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3120 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3121 NB: ClamAV planning to remove STREAM in "middle of 2010".
3122 CL also introduces -bmalware, various -d+acl logging additions and
3123 more caution in buffer sizes.
3125 PP/09 Implemented reverse_ip expansion operator.
3127 PP/10 Bugzilla 937: provide a "debug" ACL control.
3129 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3131 PP/12 Bugzilla 973: Implement --version.
3133 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3135 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3137 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3139 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3140 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3141 ignore trailing whitespace.
3143 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3145 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3148 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3149 Notification from Dr Andrew Aitchison.
3151 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3152 ExtendedDetectionInfo response format.
3153 Notification from John Horne.
3155 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3158 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3159 XSL and documented dependency on system catalogs, with examples of how
3162 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3165 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3166 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3167 configuration file which is writeable by the Exim user or group.
3169 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3170 of configuration files to cover files specified with the -C option if
3171 they are going to be used with root privileges, not just the default
3174 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3175 option (effectively making it always true).
3177 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3178 files to be used while preserving root privileges.
3180 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3181 that rogue child processes cannot use them.
3183 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3184 run-time user, instead of root.
3186 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3187 Exim run-time user without dropping privileges.
3189 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3190 result string, instead of calling string_vformat() twice with the same
3193 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3194 for other users. Others should always drop root privileges if they use
3195 -C on the command line, even for a whitelisted configure file.
3197 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3199 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3205 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3206 $data_path, and $header_path variables; fixed documentation bugs and
3209 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3210 exipick to access non-standard spools, including the "frozen" queue
3213 NM/01 Bugzilla 965: Support mysql stored procedures.
3214 Patch from Alain Williams
3216 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3218 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3219 Patch from Andreas Metzler
3221 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3222 Patch from Kirill Miazine
3224 NM/05 Bugzilla 671: Added umask to procmail example.
3226 JJ/03 installed exipick 20100323.0, fixing doc bug
3228 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3229 directory. Notification and patch from Dan Rosenberg.
3231 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3233 TK/02 Improve log output when DKIM signing operation fails.
3235 MH/01 Treat the transport option dkim_domain as a colon separated
3236 list, not as a single string, and sign the message with each element,
3237 omitting multiple occurences of the same signer.
3239 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3240 Bugzilla 985, 986. Patch by Simon Arlott
3242 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3243 Patch by Simon Arlott
3245 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3246 MBX locking. Notification from Dan Rosenberg.
3252 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3254 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3256 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3258 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3260 NM/04 Bugzilla 915: Flag broken perl installation during build.
3266 TK/01 Added patch by Johannes Berg that expands the main option
3267 "spamd_address" if it starts with a dollar sign.
3269 TK/02 Write list of recipients to X-Envelope-Sender header when building
3270 the mbox-format spool file for content scanning (suggested by Jakob
3273 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3274 (http://www.dcc-servers.net/) support via dccifd. Activated by
3275 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3277 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3278 by Mark Daniel Reidel <mr@df.eu>.
3280 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3281 When building exim an external PCRE library is now needed -
3282 PCRE is a system library on the majority of modern systems.
3283 See entry on PCRE_LIBS in EDITME file.
3285 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3286 conversation. Added nologin parameter to request.
3287 Patch contributed by Kirill Miazine.
3289 TF/01 Do not log submission mode rewrites if they do not change the address.
3291 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3293 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3294 log files in place. Contributed by Roberto Lima.
3296 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3298 TF/03 Bugzilla 615: When checking the local_parts router precondition
3299 after a local_part_suffix or local_part_prefix option, Exim now
3300 does not use the address's named list lookup cache, since this
3301 contains cached lookups for the whole local part.
3303 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3304 Robert Millan. Documentation is in experimental-spec.txt.
3306 TF/04 Bugzilla 668: Fix parallel build (make -j).
3308 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3310 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3311 Patch provided by Jan Srzednicki.
3313 TF/05 Leading white space used to be stripped from $spam_report which
3314 wrecked the formatting. Now it is preserved.
3316 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3317 that they are available at delivery time.
3319 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3321 TF/08 TLS error reporting now respects the incoming_interface and
3322 incoming_port log selectors.
3324 TF/09 Produce a more useful error message if an SMTP transport's hosts
3325 setting expands to an empty string.
3327 NM/06 Bugzilla 744: EXPN did not work under TLS.
3328 Patch provided by Phil Pennock.
3330 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3331 Patch provided by Richard Godbee.
3333 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3334 acl_smtp_notquit, added index entry.
3336 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3337 Patch provided by Eugene Bujak.
3339 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3340 accept(). Patch provided by Maxim Dounin.
3342 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3343 Patch provided by Phil Pennock.
3345 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3347 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3348 Patch provided by Brad "anomie" Jorsch.
3350 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3351 Patch provided by Dean Brooks.
3353 TK/05 Add native DKIM support (does not depend on external libraries).
3355 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3356 Patch provided by Graeme Fowler.
3358 NM/16 Bugzilla 851: Documentation example syntax fix.
3360 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3362 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3365 NM/19 Bugzilla 745: TLS version reporting.
3366 Patch provided by Phil Pennock.
3368 NM/20 Bugzilla 167: bool: condition support.
3369 Patch provided by Phil Pennock.
3371 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3372 clients. Patch provided by Phil Pennock.
3374 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3375 Patch provided by Brad "anomie" Jorsch.
3377 NM/23 Bugzilla 687: Fix misparses in eximstats.
3378 Patch provided by Heiko Schlittermann.
3380 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3381 Patch provided by Heiko Schlittermann.
3383 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3384 plus update to original patch.
3386 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3388 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3389 Patch provided by David Brownlee.
3391 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3393 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3395 NM/30 Bugzilla 888: TLS documentation bugfixes.
3397 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3399 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3400 Unlike the original bugzilla I have changed all shell scripts in src tree.
3402 NM/33 Bugzilla 898: Transport filter timeout fix.
3403 Patch by Todd Rinaldo.
3405 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3406 Patch by Serge Demonchaux.
3408 NM/35 Bugzilla 39: Base64 decode bug fixes.
3409 Patch by Jakob Hirsch.
3411 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3413 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3415 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3417 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3423 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3424 ALT-N's libdkim that I have put here:
3425 http://duncanthrax.net/exim-experimental/
3427 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3428 were called 'true' and it seems that C99 defines that as a reserved
3429 keyword to be used with 'bool' variable types. That means you could
3430 not include C99-style headers which use bools without triggering
3431 build errors in sieve.c.
3433 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3434 as mailq or other aliases. Changed the --help handling significantly
3435 to do whats expected. exim_usage() emits usage/help information.
3437 SC/01 Added the -bylocaldomain option to eximstats.
3439 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3441 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3443 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3449 PH/01 Another patch from the Sieve maintainer.
3451 PH/02 When an IPv6 address is converted to a string for single-key lookup
3452 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3453 dots are used instead of colons so that keys in lsearch files need not
3454 contain colons. This was done some time before quoting was made available
3455 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3456 (notated using the quote facility) so as to distinguish them from IPv4
3457 keys. This meant that lookups for IP addresses in host lists did not work
3458 for iplsearch lookups.
3460 This has been fixed by arranging for IPv6 addresses to be expressed with
3461 colons if the lookup type is iplsearch. This is not incompatible, because
3462 previously such lookups could never work.
3464 The situation is now rather anomalous, since one *can* have colons in
3465 ordinary lsearch keys. However, making the change in all cases is
3466 incompatible and would probably break a number of configurations.
3468 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3471 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3472 conversion specification without a maximum field width, thereby enabling
3473 a rogue spamd server to cause a buffer overflow. While nobody in their
3474 right mind would setup Exim to query an untrusted spamd server, an
3475 attacker that gains access to a server running spamd could potentially
3476 exploit this vulnerability to run arbitrary code as the Exim user.
3478 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3479 $primary_hostname instead of what libspf2 thinks the hosts name is.
3481 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3482 a directory entry by the name of the lookup key. Previously, if a
3483 symlink pointed to a non-existing file or a file in a directory that
3484 Exim lacked permissions to read, a lookup for a key matching that
3485 symlink would fail. Now it is enough that a matching directory entry
3486 exists, symlink or not. (Bugzilla 503.)
3488 PH/03 The body_linecount and body_zerocount variables are now exported in the
3491 PH/04 Added the $dnslist_matched variable.
3493 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3494 This means they are set thereafter only if the connection becomes
3497 PH/06 Added the client_condition to authenticators so that some can be skipped
3498 by clients under certain conditions.
3500 PH/07 The error message for a badly-placed control=no_multiline_responses left
3501 "_responses" off the end of the name.
3503 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3505 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3506 (without spaces) instead of just copying the configuration text.
3508 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3510 PH/11 Added $max_received_linelength.
3512 PH/12 Added +ignore_defer and +include_defer to host lists.
3514 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3515 way in which PCRE > 7.0 is built.
3517 PH/14 Implemented queue_only_load_latch.
3519 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3520 MAIL command. The effect was to mangle the value on 64-bit systems.
3522 PH/16 Another patch from the Sieve maintainer.
3524 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3526 PH/18 If a system quota error occurred while trying to create the file for
3527 a maildir delivery, the message "Mailbox is full" was not appended to the
3528 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3529 only to a quota excession during the actual writing of the file.
3531 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3532 characters?) which causes problems in log lines. The DN values are now
3533 passed through string_printing() before being added to log lines.
3535 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3536 and InterBase are left for another time.)
3538 PH/21 Added message_body_newlines option.
3540 PH/22 Guard against possible overflow in moan_check_errorcopy().
3542 PH/23 POSIX allows open() to be a macro; guard against that.
3544 PH/24 If the recipient of an error message contained an @ in the local part
3545 (suitably quoted, of course), incorrect values were put in $domain and
3546 $local_part during the evaluation of errors_copy.
3552 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3553 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3556 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3557 issue a MAIL command.
3559 PH/02 In an ACL statement such as
3561 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3563 if a client was not listed at all, or was listed with a value other than
3564 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3565 the condition was not true (as it should be), so access was not denied.
3566 The bug was that the ! inversion was incorrectly passed on to the second
3567 item. This has been fixed.
3569 PH/03 Added additional dnslists conditions == and =& which are different from
3570 = and & when the dns lookup returns more than one IP address.
3572 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3573 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3575 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3576 FSYNC, which compiles an option called disable_fsync that allows for
3577 bypassing fsync(). The documentation is heavily laced with warnings.
3579 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3581 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3582 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3583 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3584 including adding "make clean"; (3) Added -fPIC when compiling the test
3585 dynamically loaded module, to get rid of a warning.
3587 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3588 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3589 = 0s. The bug is otherwise harmless.
3591 PH/07 There was a bug in the dovecot authenticator such that the value of
3592 $auth1 could be overwritten, and so not correctly preserved, after a
3593 successful authentication. This usually meant that the value preserved by
3594 the server_setid option was incorrect.
3596 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3598 PH/09 Installed PCRE release 7.0.
3600 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3601 run for batched SMTP input. It is now run at the start of every message
3602 in the batch. While fixing this I discovered that the process information
3603 (output by running exiwhat) was not always getting set for -bs and -bS
3604 input. This is fixed, and it now also says "batched" for BSMTP.
3606 PH/11 Added control=no_pipelining.
3608 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3609 patch, slightly modified), and move the expansion of helo_data till after
3610 the connection is made in the smtp transport (so it can use these
3613 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3615 PH/14 Added log_selector = +pid.
3617 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3619 PH/16 Add ${if forany and ${if forall.
3621 PH/17 Added dsn_from option to vary the From: line in DSNs.
3623 PH/18 Flush SMTP output before performing a callout, unless control =
3624 no_callout_flush is set.
3626 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3627 was true (the default) a successful delivery failed to delete the retry
3628 item, thus causing premature timeout of the address. The bug is now
3631 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3633 PH/21 Long custom messages for fakedefer and fakereject are now split up
3634 into multiline responses in the same way that messages for "deny" and
3635 other ACL rejections are.
3637 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3638 with slight modification.
3640 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3641 draft, changing the syntax and factoring some duplicate code".
3643 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3644 for deliveries of the second and subsequent messages over the same SMTP
3647 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3648 ${reduce, with only minor "tidies".
3650 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3652 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3653 expansion side effects.
3655 PH/27 When a message times out after an over-quota error from an Exim-imposed
3656 quota, the bounce message says "mailbox is full". This message was not
3657 being given when it was a system quota that was exceeded. It now should
3660 MH/03 Made $recipients available in local_scan(). local_scan() already has
3661 better access to the recipient list through recipients_list[], but
3662 $recipients can be useful in postmaster-provided expansion strings.
3664 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3665 in the case of a MAIL command with additional options following the
3666 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3667 were accidentally chopped off.
3669 PH/29 SMTP synchronization checks are implemented when a command is read -
3670 there is a check that no more input is waiting when there shouldn't be
3671 any. However, for some commands, a delay in an ACL can mean that it is
3672 some time before the response is written. In this time, more input might
3673 arrive, invalidly. So now there are extra checks after an ACL has run for
3674 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3675 pipelining has not been advertised.
3677 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3679 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3680 returned in the SMTP response when smtp_return_error_details was false.
3681 This has been fixed.
3683 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3684 instead of the C I/O that was originally supplied, because problems were
3685 reported on Solaris.
3687 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3688 Exim which did not show up earlier: it was assuming that a call to
3689 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3690 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3691 was a macro that became an assignment, so it seemed to work. This has
3692 changed to a proper function call with a void return, hence the compile
3693 error. Exim's code has been fixed.
3695 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3698 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3700 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3702 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3703 because it turns out to be too big or there is a timeout) there is no
3704 "Completed" line in the log. When some messages of this type were
3705 selected by exigrep, they were listed as "not completed". Others were
3706 picked up by some special patterns. I have improved the selection
3707 criteria to be more general.
3709 PH/38 The host_find_failed option in the manualroute router can now be set
3710 to "ignore", to completely ignore a host whose IP address cannot be
3711 found. If all hosts are ignored, the behaviour is controlled by the new
3712 host_all_ignored option.
3714 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3715 homing or because of multiple MX records with /mx) generated more than
3716 one IP address, and the following item turned out to be the local host,
3717 all the secondary addresses of the first item were incorrectly removed
3718 from the list, along with the local host and any following hosts (which
3719 is what is supposed to happen).
3721 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3722 whoever called Exim into the -H file. In the case of the daemon it was
3723 behaving confusingly. When first started, it used values for whoever
3724 started the daemon, but after a SIGHUP it used the Exim user (because it
3725 calls itself on a restart). I have changed the code so that it now always
3728 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3729 message are rejected with the same error (e.g. no authentication or bad
3730 sender address), and a DATA command is nevertheless sent (as can happen
3731 with PIPELINING or a stupid MUA), the error message that was given to the
3732 RCPT commands is included in the rejection of the DATA command. This is
3733 intended to be helpful for MUAs that show only the final error to their
3736 PH/42 Another patch from the Sieve maintainer.
3738 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3739 Eximstats - Fixed some broken HTML links and added missing column headers
3741 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3742 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3744 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3746 PH/43 Yet another patch from the Sieve maintainer.
3748 PH/44 I found a way to check for a TCP/IP connection going away before sending
3749 the response to the final '.' that terminates a message, but only in the
3750 case where the client has not sent further data following the '.'
3751 (unfortunately, this is allowed). However, in many cases there won't be
3752 any further data because there won't be any more messages to send. A call
3753 to select() can be used: if it shows that the input is "ready", there is
3754 either input waiting, or the socket has been closed. An attempt to read
3755 the next input character can distinguish the two cases. Previously, Exim
3756 would have sent an OK response which the client would never have see.
3757 This could lead to message repetition. This fix should cure that, at
3758 least in a lot of common cases.
3760 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3761 advertised in response to EHLO.
3767 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3768 fixed by 4.65/MH/01 (is this a record?) are fixed:
3770 (i) An empty string was always treated as zero by the numeric comparison
3771 operators. This behaviour has been restored.
3773 (ii) It is documented that the numeric comparison operators always treat
3774 their arguments as decimal numbers. This was broken in that numbers
3775 starting with 0 were being interpreted as octal.
3777 While fixing these problems I realized that there was another issue that
3778 hadn't been noticed. Values of message_size_limit (both the global option
3779 and the transport option) were treated as octal if they started with 0.
3780 The documentation was vague. These values are now always treated as
3781 decimal, and I will make that clear in the documentation.
3787 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3788 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3791 MH/01 Don't check that the operands of numeric comparison operators are
3792 integers when their expansion is in "skipping" mode (fixes bug
3793 introduced by 4.64-PH/07).
3795 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3796 child addresses, Exim now panics and dies. Previously, because the count
3797 is held in a short int, deliveries were likely to be lost. As such a
3798 large number of recipients for a single message is ridiculous
3799 (performance will be very, very poor), I have chosen to impose a limit
3800 rather than extend the field.
3806 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3807 leftover -K file (the existence of which was triggered by #402).
3808 While we were at it, introduced process PID as part of the -K
3809 filename. This should rule out race conditions when creating
3812 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3813 processing considerably. Previous code took too long for large mails,
3814 triggering a timeout which in turn triggers #401.
3816 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3817 in the DK code in transports.c. sendfile() is not really portable,
3818 hence the _LINUX specificness.
3820 TF/01 In the add_headers option to the mail command in an Exim filter,
3821 there was a bug that Exim would claim a syntax error in any
3822 header after the first one which had an odd number of characters
3825 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3826 callout verification, Exim cached a "reject" for the entire domain. This
3827 is correct for most verifications, but it is not correct for a recipient
3828 verification with use_sender or use_postmaster set, because in that case
3829 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3830 case of MAIL FROM:<> rejection from other early rejections (e.g.
3831 rejection of HELO). When verifying a recipient using a non-null MAIL
3832 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3833 Whatever the result of the callout, the value of the domain cache is
3834 left unchanged (for any other kind of callout, getting as far as trying
3835 RCPT means that the domain itself is ok).
3837 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3840 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3841 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3842 session, the daemon ignores ECONNECT errors and logs others; it now
3843 ignores EPIPE as well.
3845 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3846 (quoted-printable decoding).
3848 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3849 later the small subsequent patch to fix an introduced bug.
3851 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3853 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3855 PH/08 An error is now given if message_size_limit is specified negative.
3857 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3858 to be given (somewhat) arbitrary names.
3860 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3863 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3864 miscellaneous code fixes
3866 PH/10 Added the log_reject_target ACL modifier to specify where to log
3869 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3870 hostname. This is wrong, because it relates to the incoming message (and
3871 probably the interface on which it is arriving) and not to the outgoing
3872 callout (which could be using a different interface). This has been
3873 changed to use the value of the helo_data option from the smtp transport
3874 instead - this is what is used when a message is actually being sent. If
3875 there is no remote transport (possible with a router that sets up host
3876 addresses), $smtp_active_hostname is used.
3878 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3879 tweaks were necessary in order to get it to work (see also 21 below):
3880 (a) The code assumed that strncpy() returns a negative number on buffer
3881 overflow, which isn't the case. Replaced with Exim's string_format()
3883 (b) There were several signed/unsigned issues. I just did the minimum
3884 hacking in of casts. There is scope for a larger refactoring.
3885 (c) The code used strcasecmp() which is not a standard C function.
3886 Replaced with Exim's strcmpic() function.
3887 (d) The code set only $1; it now sets $auth1 as well.
3888 (e) A simple test gave the error "authentication client didn't specify
3889 service in request". It would seem that Dovecot has changed its
3890 interface. Fortunately there's a specification; I followed it and
3891 changed what the client sends and it appears to be working now.
3893 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3896 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3897 address A is aliased to B and C, where B exists and C does not. Without
3898 -v the output is "A verified" because verification stops after a
3899 successful redirection if more than one address is generated. However,
3900 with -v the child addresses are also verified. Exim was outputting "A
3901 failed to verify" and then showing the successful verification for C,
3902 with its parentage. It now outputs "B failed to verify", showing B's
3903 parentage before showing the successful verification of C.
3905 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3906 look up a TXT record in a specific list after matching in a combined
3909 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3910 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3911 they consult the DNS. I had assumed they would set it the way they
3912 wanted; and indeed my experiments on Linux seem to show that in some
3913 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3914 To be on the safe side, however, I have now made the interface to
3915 host_find_byname() similar to host_find_bydns(), with an argument
3916 containing the DNS resolver options. The host_find_byname() function now
3917 sets these options at its start, just as host_find_bydns() does. The smtp
3918 transport options dns_qualify_single and dns_search_parents are passed to
3919 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3920 of host_find_byname() use the default settings of RES_DEFNAMES
3921 (qualify_single) but not RES_DNSRCH (search_parents).
3923 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3924 spool_read_header() do less string testing, by means of a preliminary
3925 switch on the second character of optional "-foo" lines. (This is
3926 overdue, caused by the large number of possibilities that now exist.
3927 Originally there were few.) While I was there, I also converted the
3928 str(n)cmp tests so they don't re-test the leading "-" and the first
3929 character, in the hope this might squeeze out yet more improvement.
3931 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3932 flag allowing group syntax was set by the header_syntax check but not
3933 turned off, possible causing trouble later; (2) The flag was not being
3934 set at all for the header_verify test, causing "group"-style headers to
3935 be rejected. I have now set it in this case, and also caused header_
3936 verify to ignore an empty address taken from a group. While doing this, I
3937 came across some other cases where the code for allowing group syntax
3938 while scanning a header line wasn't quite right (mostly, not resetting
3939 the flag correctly in the right place). These bugs could have caused
3940 trouble for malformed header lines. I hope it is now all correct.
3942 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3943 with the "reply" argument non-NULL. The code, however (which originally
3944 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3945 but it didn't always do it. This confused somebody who was copying the
3946 code for some other use. I have removed all the tests.
3948 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3949 feature that was used to support insecure browsers during the U.S. crypto
3950 embargo. It requires special client support, and Exim is probably the
3951 only MTA that supported it -- and would never use it because real RSA is
3952 always available. This code has been removed, because it had the bad
3953 effect of slowing Exim down by computing (never used) parameters for the
3954 RSA_EXPORT functionality.
3956 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3957 authenticator to fail if there's a tab character in the incoming data
3958 (there should never be unless someone is messing about, as it's supposed
3959 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3960 if the connection is using TLS or if the remote IP is the same as the
3961 local IP, and the "valid-client-cert option" if a client certificate has
3964 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3965 authenticators. This can be used for authorization after authentication
3966 succeeds. (In the case of plaintext, it servers for both authentication
3969 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3970 if any retry times were supplied.
3972 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3973 connection, where there is no client IP address to check. In this
3974 situation, the verify now always succeeds.
3976 PH/25 Applied John Jetmore's -Mset patch.
3978 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3980 PH/27 In a string expansion for a processed (not raw) header when multiple
3981 headers of the same name were present, leading whitespace was being
3982 removed from all of them, but trailing whitespace was being removed only
3983 from the last one. Now trailing whitespace is removed from each header
3984 before concatenation. Completely empty headers in a concatenation (as
3985 before) are ignored.
3987 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3988 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3990 PH/29 [Removed. This was a change that I later backed out, and forgot to
3991 correct the ChangeLog entry (that I had efficiently created) before
3992 committing the later change.]
3994 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3995 address errors (4xx response to RCPT) over the same connection as other
3996 messages routed to the same hosts. Such deliveries are always "forced",
3997 so retry times are not inspected. This resulted in far too many retries
3998 for the affected addresses. The effect occurred only when there were more
3999 hosts than the hosts_max_try setting in the smtp transport when it had
4000 the 4xx errors. Those hosts that it had tried were not added to the list
4001 of hosts for which the message was waiting, so if all were tried, there
4002 was no problem. Two fixes have been applied:
4004 (i) If there are any address or message errors in an SMTP delivery, none
4005 of the hosts (tried or untried) are now added to the list of hosts
4006 for which the message is waiting, so the message should not be a
4007 candidate for sending over the same connection that was used for a
4008 successful delivery of some other message. This seems entirely
4009 reasonable: after all the message is NOT "waiting for some host".
4010 This is so "obvious" that I'm not sure why it wasn't done
4011 previously. Hope I haven't missed anything, but it can't do any
4012 harm, as the worst effect is to miss an optimization.
4014 (ii) If, despite (i), such a delivery is accidentally attempted, the
4015 routing retry time is respected, so at least it doesn't keep
4016 hammering the server.
4018 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4019 in ${readsocket because some servers need this prod.
4021 PH/32 Added some extra debug output when updating a wait-xxx database.
4023 PH/33 The hint "could be header name not terminated by colon", which has been
4024 given for certain expansion errors for a long time, was not being given
4025 for the ${if def:h_colon_omitted{... case.
4027 PH/34 The spec says: "With one important exception, whenever a domain list is
4028 being scanned, $domain contains the subject domain." There was at least
4029 one case where this was not true.
4031 PH/35 The error "getsockname() failed: connection reset by peer" was being
4032 written to the panic log as well as the main log, but it isn't really
4033 panic-worthy as it just means the connection died rather early on. I have
4034 removed the panic log writing for the ECONNRESET error when getsockname()
4037 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4038 runs only) independently of the message's sender address. This meant
4039 that, if the 4xx error was in fact related to the sender, a different
4040 message to the same recipient with a different sender could confuse
4041 things. In particular, this can happen when sending to a greylisting
4042 server, but other circumstances could also provoke similar problems.
4043 I have changed the default so that the retry time for these errors is now
4044 based a combination of the sender and recipient addresses. This change
4045 can be overridden by setting address_retry_include_sender=false in the
4048 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4049 remote server are returned as part of bounce messages. This was not
4050 happening for LMTP over a pipe (the lmtp transport), but now it is the
4051 same for both kinds of LMTP.
4053 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4054 in header lines that were in fact CNAMEs. This is no longer the case.
4056 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4057 and queue runs started by the daemon processed all messages. This has
4058 been fixed so that -R and -S can now usefully be given with -q<time>.
4060 PH/40 Import PCRE release 6.7 (fixes some bugs).
4062 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4064 PH/42 Give an error if -q is specified more than once.
4066 PH/43 Renamed the variables $interface_address and $interface_port as
4067 $received_ip_address and $received_port, to make it clear that these
4068 values apply to message reception, and not to the outgoing interface when
4069 a message is delivered. (The old names remain recognized, of course.)
4071 PH/44 There was no timeout on the connect() call when using a Unix domain
4072 socket in the ${readsocket expansion. There now is.
4074 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4075 be meaningful with "accept".
4077 SC/01 Eximstats V1.43
4078 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4080 SC/02 Eximstats V1.44
4081 Use a glob alias rather than an array ref in the generated
4082 parser. This improves both readability and performance.
4084 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4085 Collect SpamAssassin and rejection statistics.
4086 Don't display local sender or destination tables unless
4087 there is data to show.
4088 Added average volumes into the top table text output.
4090 SC/04 Eximstats V1.46
4091 Collect data on the number of addresses (recipients)
4092 as well as the number of messages.
4094 SC/05 Eximstats V1.47
4095 Added 'Message too big' to the list of mail rejection
4096 reasons (thanks to Marco Gaiarin).
4098 SC/06 Eximstats V1.48
4099 Mainlog lines which have GMT offsets and are too short to
4100 have a flag are now skipped.
4102 SC/07 Eximstats V1.49 (Alain Williams)
4103 Added the -emptyok flag.
4105 SC/08 Eximstats V1.50
4106 Fixes for obtaining the IP address from reject messages.
4108 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4109 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4110 whitespace changes from 4.64-PH/27
4112 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4115 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4116 are found, allow negative numbers in numeric criteria)
4118 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4120 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4121 to match changes made in 4.64-PH/43
4123 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4125 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4126 "domain" is actually an IP address, and force a failure. This locks out
4127 those revolvers/nameservers that support "A-for-A" lookups, in
4128 contravention of the specifications.
4130 PH/48 When a host name was looked up from an IP address, and the subsequent
4131 forward lookup of the name timed out, the host name was left in
4132 $sender_host_name, contrary to the specification.
4134 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4135 restricted to single-key lookups, Exim was not diagnosing an error if
4136 * or *@ was used with a query-style lookup.
4138 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4140 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4141 long ago, but noone interested enough thought of it. Let's just say that
4142 the "1.1" means that there are some new functions that weren't there at
4143 some point in the past.
4145 PH/51 Error processing for expansion failure of helo_data from an smtp
4146 transport during callout processing was broken.
4148 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4149 tested/used via the -bh/-bhc/-bs options.
4151 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4152 bug, fixed in subsequent PCRE releases).
4154 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4155 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4157 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4163 SC/01 Use a glob alias rather than an array ref in eximstats generated
4164 parser. This improves both readability and performance.
4166 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4167 Don't display local sender or destination tables in eximstats unless
4168 there is data to show.
4169 Added average volumes into the eximstats top table text output.
4171 SC/03 Collect data on the number of addresses (recipients) as well
4172 as the number of messages in eximstats.
4174 TF/01 Correct an error in the documentation for the redirect router. Exim
4175 does (usually) call initgroups() when daemonizing.
4177 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4178 with consistent privilege compared to when running as a daemon.
4180 TF/03 Note in the spec that $authenticated_id is not set for local
4181 submissions from trusted users.
4183 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4184 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4186 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4187 by adding some example configuration directives to the default
4188 configuration file. A little bit of work is required to uncomment the
4189 directives and define how usernames and passwords are checked, but
4190 there is now a framework to start from.
4192 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4193 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4194 without this. I don't know how relevant this is to other LDAP libraries.
4196 PH/02 Add the verb name to the "unknown ACL verb" error.
4198 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4200 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4202 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4203 directory not expanded when it should be if an expanded home directory
4204 was set for the address (which is overridden by the transport).
4206 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4209 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4210 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4211 because it is too late at that time, and has no effect.
4213 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4214 security issue with \' (bugzilla #107). I could not use the
4215 PQescapeStringConn() function, because it needs a PGconn value as one of
4218 PH/08 When testing addresses using -bt, indicate those final addresses that
4219 are duplicates that would not cause an additional delivery. At least one
4220 person was confused, thinking that -bt output corresponded to deliveries.
4221 (Suppressing duplicates isn't a good idea as you lose the information
4222 about possibly different redirections that led to the duplicates.)
4224 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4225 systems where poll() doesn't work, in particular OS X.
4227 PH/10 Added more information to debugging output for retry time not reached.
4229 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4230 operations in malware.c.
4232 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4235 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4236 syslog_duplication set false, log lines that would normally be written
4237 both the the main log and to the reject log were not written to syslog at
4240 PH/14 In the default configuration, change the use of "message" in ACL warn
4241 statements to "add_header".
4243 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4244 not followed by a command (e.g. "seen endif").
4246 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4247 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4250 PH/17 Added extra conditions to the default value of delay_warning_condition
4254 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4255 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4256 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4259 The Auto-Submitted: and various List- headers are standardised, whereas I
4260 don't think Precedence: ever was.
4262 PH/18 Refactored debugging code in route_finduser() to show more information,
4263 in particular, the error code if getpwnam() issues one.
4265 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4266 This is apparently needed in addition to the PH/07 change above to avoid
4267 any possible encoding problems.
4269 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4270 but not after initializing Perl.
4272 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4273 output them only if debugging. By default they are written stderr,
4274 apparently, which is not desirable.
4276 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4279 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4282 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4284 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4285 authenticated or an ident call has been made. Suppress the default
4286 values for $authenticated_id and $authenticated_sender (but permit -oMai
4287 and -oMas) when testing with -bh.
4289 PH/24 Re-jigged the order of the tests in the default configuration so that the
4290 tests for valid domains and recipients precede the DNS black list and CSA
4291 tests, on the grounds that those ones are more expensive.
4293 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4294 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4295 This bug exists in every version of Exim that I still have, right back to
4298 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4299 However, an attempt to turn on case-sensitivity in a regex key by
4300 including (?-i) didn't work because the subject string was already
4301 lowercased, and the effects were non-intuitive. It turns out that a
4302 one-line patch can be used to allow (?-i) to work as expected.
4308 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4309 other effects) broke the use of negated acl sub-conditions.
4311 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4314 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4315 "Deny" causes Exim to reject the incoming connection with a 554 error.
4316 Unfortunately, if there is a major crisis, such as a disk failure,
4317 tcp-wrappers gives "deny", whereas what one would like would be some
4318 kind of temporary error. A kludge has been added to help with this.
4319 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4320 554 error is used if errno is still zero or contains ENOENT (which occurs
4321 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4324 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4326 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4327 errors. Otherwise a message that provokes a temporary error (when other
4328 messages do not) can cause a whole host to time out.
4330 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4331 addresses were routed directly to files or pipes from a redirect router.
4332 File deliveries just didn't batch; pipe deliveries might have suffered
4335 PH/06 A failure to get a lock for a hints database would erroneously always say
4336 "Failed to get write lock", even when it was really a read lock.
4338 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4339 of 0600. This has been changed to use the value of the lockfile_mode
4340 option (which defaults to 0600).
4342 PH/08 Applied small patch from the Sieve maintainer.
4344 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4345 folder from quota calculations, a direct delivery into this folder messed
4346 up the contents of the maildirsize file. This was because the regex was
4347 used only to exclude .Trash (or whatever) when the size of the mailbox
4348 was calculated. There was no check that a delivery was happening into an
4349 excluded directory. This bug has been fixed by ignoring all quota
4350 processing for deliveries into excluded directories.
4352 PH/10 Added the maildirfolder_create_regex option to appendfile.
4358 PH/01 The code for finding all the local interface addresses on a FreeBSD
4359 system running IPv6 was broken. This may well have applied to all BSD
4360 systems, as well as to others that have similar system calls. The broken
4361 code found IPv4 interfaces correctly, but gave incorrect values for the
4362 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4363 that it would not match correctly against @[] and not recognize the IPv6
4366 PH/02 The ipliteral router was not recognizing addresses of the form user@
4367 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4369 PH/03 Added disable_ipv6.
4371 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4372 decoded form, because it is most often used to construct To: headers
4373 lines in autoreplies, and the decoded form may well be syntactically
4374 invalid. However, $reply_address has leading white space removed, and all
4375 newlines turned into spaces so that the autoreply transport does not
4378 PH/05 If group was specified without a user on a router, and no group or user
4379 was specified on a transport, the group from the router was ignored.
4381 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4382 for visible compile-time settings that can be used to change these
4383 numbers, for those that want even more. Backwards compatibility with old
4384 spool files has been maintained. However, going back to a previous Exim
4385 release will lost any variables that are in spool files.
4387 PH/07 Two small changes when running in the test harness: increase delay when
4388 passing a TCP/IP connection to a new process, in case the original
4389 process has to generate a bounce, and remove special handling of
4390 127.0.0.2 (sic), which is no longer necessary.
4392 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4393 be the same on different OS.
4395 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4398 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4399 whether --show-vars was specified or not
4401 JJ/02 exipick: Added support for new ACL variable spool format introduced
4404 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4405 syntactically invalid From: or Reply-to: line, and a filter used this to
4406 generate an autoreply, and therefore failed to obtain an address for the
4407 autoreply, Exim could try to deliver to a non-existent relative file
4408 name, causing unrelated and misleading errors. What now happens is that
4409 it logs this as a hard delivery error, but does not attempt to create a
4412 PH/11 The exinext utility has a -C option for testing purposes, but although
4413 the given file was scanned by exinext itself; it wasn't being passed on
4414 when Exim was called.
4416 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4417 an end-of-file indication when reading a command response.
4419 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4420 compiled. In many other places in Exim, IPv6 addresses are always
4421 recognized, so I have changed this. It also means that IPv4 domain
4422 literals of the form [IPV4:n.n.n.n] are now always recognized.
4424 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4425 used if the router is not running as root, for example, when verifying at
4426 ACL time, or when using -bh. The debugging output from this situation was
4427 non-existent - all you got was a failure to exec. I have made two
4430 (a) Failures to set uid/gid, the current directory, or a process leader
4431 in a subprocess such as that created by queryprogram now generate
4432 suitable debugging output when -d is set.
4434 (b) The queryprogram router detects when it is not running as root,
4435 outputs suitable debugging information if -d is set, and then runs
4436 the subprocess without attempting to change uid/gid.
4438 PH/15 Minor change to Makefile for building test_host (undocumented testing
4441 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4442 additional section of a DNS packet that returns MX or SRV records.
4443 Instead, it always explicitly searches for A/AAAA records. This avoids
4444 major problems that occur when a DNS server includes only records of one
4445 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4446 fixed another bug: if SRV records were looked up and the corresponding
4447 address records were *not* found in the additional section, the port
4448 values from the SRV records were lost.
4450 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4451 using the correct key (the original address) when searching the retry
4452 rules in order to find which one to use for generating the retry hint.
4454 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4455 adding the default one. Similarly, if it contains a Reply-To: header, the
4456 errors_reply_to option, if set, is not used.
4458 PH/19 When calculating a retry time, Exim used to measure the "time since
4459 failure" by looking at the "first failed" field in the retry record. Now
4460 it does not use this if it is later than than the arrival time of the
4461 message. Instead it uses the arrival time. This makes for better
4462 behaviour in cases where some deliveries succeed, thus re-setting the
4463 "first failed" field. An example is a quota failure for a huge message
4464 when small messages continue to be delivered. Without this change, the
4465 "time since failure" will always be short, possible causing more frequent
4466 delivery attempts for the huge message than are intended.
4467 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4469 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4470 $1, $2, $3) because the numerical variables can be reset during some
4471 expansion items (e.g. "match"), thereby losing the authentication data.
4473 PH/21 Make -bV show the size of off_t variables so that the test suite can
4474 decide whether to run tests for quotas > 2G.
4476 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4477 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4478 filecount value is greater than 2G or if a quota value is greater than 2G
4479 on a system where the size of off_t is not greater than 4, a panic error
4482 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4483 never match. The debug and -bh output now contains an explicit error
4484 message indicating a malformed IPv4 address or mask.
4486 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4487 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4488 PH/23 above applies.
4490 PH/25 Do not write to syslog when running in the test harness. The only
4491 occasion when this arises is a failure to open the main or panic logs
4492 (for which there is an explicit test).
4494 PH/26 Added the /no_tell option to "control=freeze".
4496 PH/27 If a host name lookup failed very early in a connection, for example, if
4497 the IP address matched host_lookup and the reverse lookup yielded a name
4498 that did not have a forward lookup, an error message of the form "no IP
4499 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4500 could be logged. Now it outputs the IP address instead of "NULL".
4502 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4503 allows the sender and the authenticated sender to be set when
4504 submitting a message from within Exim. Since child_open_exim() is
4505 documented for local_scan(), the new function should be too.
4507 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4508 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4509 results in an empty string is now treated as unset.
4511 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4513 PH/31 Added sender_verify_fail logging option.
4515 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4516 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4517 tidied the source and removed it altogether.
4519 PH/33 When a queue run was abandoned because the load average was too high, a
4520 log line was always written; now it is written only if the queue_run log
4521 selector is set. In addition, the log line for abandonment now contains
4522 information about the queue run such as the pid. This is always present
4523 in "start" and "stop" lines but was omitted from the "abandon" line.
4525 PH/34 Omit spaces between a header name and the colon in the error message that
4526 is given when verify = headers_syntax fails (if there are lots of them,
4527 the message gets confusing).
4529 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4530 names, as there are now some PTR records that contain slashes. This check
4531 is only to protect against broken name servers that fall over on strange
4532 characters, so the fact that it applies to all lookups doesn't matter.
4534 PH/36 Now that the new test suite is complete, we can remove some of the
4535 special code in Exim that was needed for the old test suite. For example,
4536 sorting DNS records because real resolvers return them in an arbitrary
4537 order. The new test suite's fake resolver always returns records in the
4540 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4541 bounces) except when queue_only is set, to avoid logging races between
4542 the different processes.
4544 PH/38 Panic-die if .include specifies a non-absolute path.
4546 PH/39 A tweak to the "H" retry rule from its user.
4548 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4549 a label. They prevented compilation on older perls.
4551 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4552 a warning to be raised on newish perls.
4554 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4555 on queue. Changes to match documented behaviour of showing count of
4556 messages matching specified criteria.
4558 PH/40 Changed the default ident timeout from 30s to 5s.
4560 PH/41 Added support for the use of login_cap features, on those BSD systems
4561 that have them, for controlling the resources used by pipe deliveries.
4563 PH/42 The content-scanning code uses fopen() to create files in which to put
4564 message data. Previously it was not paying any attention to the mode of
4565 the files. Exim runs with umask(0) because the rest of the code creates
4566 files with open(), and sets the required mode explicitly. Thus, these
4567 files were ending up world-writeable. This was not a big issue, because,
4568 being within the spool directory, they were not world-accessible. I have
4569 created a function called modefopen, which takes an additional mode
4570 argument. It sets umask(777), creates the file, chmods it to the required
4571 mode, then resets the umask. All the relevant calls to fopen() in the
4572 content scanning code have been changed to use this function.
4574 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4575 to 24 hours. This avoids potential overflow problems when processing G
4576 and H retry rules. I suspect nobody ever tinkers with this value.
4578 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4580 PH/45 When the plaintext authenticator is running as a client, the server's
4581 challenges are checked to ensure they are valid base64 strings. By
4582 default, the authentication attempt is cancelled if an invalid string is
4583 received. Setting client_ignore_invalid_base64 true ignores these errors.
4584 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4585 they are received. Thus, the responses can be made to depend on the
4586 challenges. If an invalid string is ignored, an empty string is placed in
4589 PH/46 Messages that are created by the autoreply transport now contains a
4590 References: header, in accordance with RFCs 2822 and 3834.
4592 PH/47 Added authenticated_sender_force to the smtp transport.
4594 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4596 PH/49 Installed latest patch from the Sieve maintainer.
4598 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4599 was also set, the appendfile transport was unnecessarily scanning a
4600 directory of message files (e.g. for maildir delivery) to find the count
4601 of files (along with the size), even though it did not need this
4602 information. It now does the scan only if it needs to find either the
4603 size of the count of files.
4605 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4607 PH/52 Two bugs concerned with error handling when the smtp transport is
4610 (i) Exim was not creating retry information for temporary errors given
4611 for individual recipients after the DATA command when the smtp transport
4612 was used in LMTP mode. This meant that they could be retried too
4613 frequently, and not timed out correctly.
4615 (ii) Exim was setting the flag that allows error details to be returned
4616 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4617 recipients that were returned after the DATA command.
4619 PH/53 This is related to PH/52, but is more general: for any failing address,
4620 when detailed error information was permitted to be returned to the
4621 sender, but the error was temporary, then after the final timeout, only
4622 "retry timeout exceeded" was returned. Now it returns the full error as
4623 well as "retry timeout exceeded".
4625 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4626 do this, and (what is worse) MTAs that accept it.
4628 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4629 will now be deprecated.
4631 PH/56 New os.c-cygwin from the Cygwin maintainer.
4633 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4634 formats (previously only available in exim formats via -bpr, -bpru,
4635 and -bpra. Now also available in native and exiqgrep formats)
4637 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4638 with very large, slow to parse queues
4640 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4642 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4644 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4645 responses to authentication challenges, though it was showing the
4646 challenges; (ii) I've removed the CR characters from the debug output for
4649 PH/58 Allow for the insertion of a newline as well as a space when a string
4650 is turned into more than one encoded-word during RFC 2047 encoding. The
4651 Sieve code now uses this.
4653 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4654 data_4xx, lost_connection, tls_required.
4656 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4657 message was being sent as an SMTP response.
4659 PH/61 Add -l and -k options to exicyclog.
4661 PH/62 When verifying, if an address was redirected to one new address, so that
4662 verification continued, and the new address failed or deferred after
4663 having set something in $address_data, the value of $address_data was not
4664 passed back to the ACL. This was different to the case when no
4665 redirection occurred. The value is now passed back in both cases.
4667 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4668 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4669 use login_cap.h, so on its own it isn't the distinguishing feature. The
4670 new name refers directly to the setclassresources() function.
4672 PH/65 Added configuration files for NetBSD3.
4674 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4676 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4677 is preferred over IPv4.
4679 PH/68 The bounce_return_message and bounce_return_body options were not being
4680 honoured for bounces generated during the reception of non-SMTP messages.
4681 In particular, this applied to messages rejected by the ACL. This bug has
4682 been fixed. However, if bounce_return_message is true and bounce_return_
4683 body is false, the headers that are returned for a non-SMTP message
4684 include only those that have been read before the error was detected.
4685 (In the case of an ACL rejection, they have all been read.)
4687 PH/69 The HTML version of the specification is now built in a directory called
4688 spec_html instead of spec.html, because the latter looks like a path with
4689 a MIME-type, and this confuses some software.
4691 PH/70 Catch two compiler warnings in sieve.c.
4693 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4694 function verify_get_ident() calls ip_connect() to connect a socket, but
4695 if the "connect()" function timed out, ip_connect() used to close the
4696 socket. However, verify_get_ident() also closes the socket later, and in
4697 between Exim writes to the log, which may get opened at this point. When
4698 the socket was closed in ip_connect(), the log could get the same file
4699 descriptor number as the socket. This naturally causes chaos. The fix is
4700 not to close the socket in ip_connect(); the socket should be closed by
4701 the function that creates it. There was only one place in the code where
4702 this was missing, in the iplookup router, which I don't think anybody now
4703 uses, but I've fixed it anyway.
4705 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4706 well as to direct DNS lookups. Otherwise the handling of names in host
4707 lists is inconsistent and therefore confusing.
4713 PH/01 Two changes to the default runtime configuration:
4715 (1) Move the checks for relay_from_hosts and authenticated clients from
4716 after to before the (commented out) DNS black list checks.
4718 (2) Add control=submission to the relay_from_hosts and authenticated
4719 clients checks, on the grounds that messages accepted by these
4720 statements are most likely to be submissions.
4722 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4724 (1) Generate an error if the third argument for the ${prvs expansion is
4727 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4730 (3) Reset the variables that are obtained from the first argument of
4731 ${prvscheck and used in the second argument before leaving the code,
4732 because their memory is reclaimed, so using them afterwards may do
4735 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4736 one (it's much easier than Tom thought :-).
4738 (5) Because of (4), we can now allow for the use of $prvscheck_result
4739 inside the third argument.
4741 PH/03 For some reason, the default setting of PATH when running a command from
4742 a pipe transport was just "/usr/bin". I have changed it to
4745 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4746 anything to be listed in the output from -bV.
4748 PH/05 When a filter generated an autoreply, the entire To: header line was
4749 quoted in the delivery log line, like this:
4751 => >A.N.Other <ano@some.domain> <original@ddress> ...
4753 This has been changed so that it extracts the operative address. There
4754 may be more than one such address. If so, they are comma-separated, like
4757 => >ano@some.domain,ona@other.domain <original@ddress> ...
4759 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4760 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4761 not being looked up in the rDNS to get a host name, Exim was showing the
4762 IP address twice in Received: lines, even though the IP addresses were
4763 identical. For example:
4765 Received: from [1.2.3.4] (helo=[1.2.3.4])
4767 However, if the real host name was known, it was omitting the HELO data
4768 if it matched the actual IP address. This has been tidied up so that it
4769 doesn't show the same IP address twice.
4771 PH/07 When both +timestamp and +memory debugging was on, the value given by
4772 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4773 called by the debug printing, thereby overwriting the timestamp buffer.
4774 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4776 PH/08 When the original message was included in an autoreply transport, it
4777 always said "this is a copy of the message, including all the headers",
4778 even if body_only or headers_only was set. It now gives an appropriate
4781 PH/09 Applied a patch from the Sieve maintainer which:
4783 o fixes some comments
4784 o adds the (disabled) notify extension core
4785 o adds some debug output for the result of if/elsif tests
4786 o points to the current vacation draft in the documentation
4787 and documents the missing references header update
4791 o fixes a bug in processing the envelope test (when testing
4792 multiple envelope elements, the last element determined the
4795 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4796 Electronic Mail") by including:
4798 Auto-submitted: auto-generated
4800 in the messages that it generates (bounce messages and others, such as
4801 warnings). In the case of bounce messages for non-SMTP messages, there was
4802 also a typo: it was using "Auto_submitted" (underscore instead of
4803 hyphen). Since every message generated by Exim is necessarily in response
4804 to another message, thes have all been changed to:
4806 Auto-Submitted: auto-replied
4808 in accordance with these statements in the RFC:
4810 The auto-replied keyword:
4812 - SHOULD be used on messages sent in direct response to another
4813 message by an automatic process,
4815 - MUST NOT be used on manually-generated messages,
4817 - MAY be used on Delivery Status Notifications (DSNs) and Message
4818 Disposition Notifications (MDNs),
4820 - MUST NOT be used on messages generated by automatic or periodic
4821 processes, except for messages which are automatic responses to
4824 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4825 to the default Received: header definition.
4827 PH/12 Added log selector acl_warn_skipped (default on).
4829 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4830 variables because (a) they are in the wrong storage pool and (b) even if
4831 they were copied, it wouldn't work properly because of the caching.
4833 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4834 checking when decoding. Apparently there are clients that generate
4835 overlong encoded strings. Why am I not surprised?
4837 PH/15 If the first argument of "${if match_address" was not empty, but did not
4838 contain an "@" character, Exim crashed. Now it writes a panic log message
4839 and treats the condition as false.
4841 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4843 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4844 extension "envelope-auth". The code is finished and in agreement with
4845 other implementations, but there is no documentation so far and in fact,
4846 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4847 not changing the active code.
4849 Print executed "if" and "elsif" statements when debugging is used. This
4850 helps a great deal to understand what a filter does.
4852 Document more things not specified clearly in RFC3028. I had all this
4853 sorted out, when out of a sudden new issues came to my mind. Oops."
4855 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4858 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4859 canonical form (as documented). However, after a host name lookup from
4860 the IP address, check_host() was doing a simple string comparison with
4861 addresses acquired from the DNS when checking that the found name did
4862 have the original IP as one of its addresses. Since any found IPv6
4863 addresses are likely to be in abbreviated form, the comparison could
4864 fail. Luckily, there already exists a function for doing the comparison
4865 by converting both addresses to binary, so now that is used instead of
4866 the text comparison.
4868 PH/20 There was another similar case to PH/19, when a complete host name was
4869 given in a host list; looking up its IP address could give an abbreviated
4870 form, whereas the current host's name might or might not be abbreviated.
4871 The same fix has been applied.
4877 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4878 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4881 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4882 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4884 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4886 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4887 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4888 Also, for "Auto-Submitted", treat anything other than "no" as
4889 non-personal, in accordance with RFC 3834. (Previously it treated
4890 anything starting "auto-" as non-personal.)
4892 TF/01 The control=submission/name=... option had a problem with syntax
4893 errors if the name included a slash character. The /name= option
4894 now slurps the rest of the string, so it can include any characters
4895 but it must come last in the list of options (after /sender_retain
4898 PH/05 Some modifications to the interface to the fake nameserver for the new
4906 TK/01 Added the "success_on_redirect" address verification option. See
4907 NewStuff for rationale and an example.
4909 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4911 PH/02 Patch to exigrep to allow it to work on syslog lines.
4913 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4914 fread() to skip over the body file's header line, because in Cygwin the
4915 header line is locked and is inaccessible.
4917 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4918 co-exist for some time) to make it clear that it is the Exim ID that is
4919 referenced, not the Message-ID: header line.
4921 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4922 string_format() function, because snprintf() does not exist on all
4925 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4926 ${stat: expansion item.
4928 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4929 protocol synchronization error", to keep the pedants happy.
4931 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4932 well as for IRIX systems, when gcc is being used. See the host.c source
4935 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4937 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4940 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4941 transport and to the smtp transport in LMTP mode.
4943 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4945 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4946 run a filter in a subprocess. This could lead to confusion in subsequent
4947 lookups in the parent process. There should also be a search_tidyup() at
4948 the end of the subprocess.
4950 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4951 only if the host matched helo_try_verify_hosts, which caused the
4952 verification to occur when the EHLO/HELO command was issued. The ACL just
4953 tested the remembered result. Now, if a previous verification attempt has
4954 not happened, "verify = helo" does it there and then.
4956 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4958 TK/03 Fix log output including CR from clamd.
4960 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4961 exist provoked a memory error which could cause a segfault.
4963 PH/15 Installed PCRE 6.2
4965 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4967 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4968 of the problem. Specifically, suggested +O2 rather than +O1 for the
4971 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4973 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4974 record, the port was not correctly logged when the outgoing_port log
4975 selector was set (it logged the transort's default port).
4977 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4978 fallback_hosts, and "hosts" in the smtp transport.
4980 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4981 host errors such as "Connection refused".
4983 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4984 authentication with radiusclient 0.4.9:
4986 - Error returned from rc_read_config was caught wrongly
4987 - Username/password not passed on to radius server due to wrong length.
4989 The presumption is that some radiusclient API changes for 4.51/PH/17
4990 were not taken care of correctly. The code is still untested by me (my
4991 Linux distribution still has 0.3.2 of radiusclient), but it was
4992 contributed by a Radius user.
4994 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4995 expanding the "port" option of the smtp transport.
4997 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4998 while reading a MIME header. Thanks to Tom Hughes for a patch.
5000 PH/24 Include config.h inside local_scan.h so that configuration settings are
5003 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5004 that in an ACL for RCPT (for example), you can examine exactly what was
5007 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5008 commands, but it was not correctly comparing the address with the actual
5009 client host address. Thus, it would show the EHLO address in Received:
5010 header lines when this was not necessary.
5012 PH/27 Added the % operator to ${eval:}.
5014 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5015 it should be ignoring failures (because with -C, for example, it has lost
5016 privilege). It wasn't ignoring creation failures other than "already
5019 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5022 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5023 because an input line was too long, either on its own, or by virtue of
5024 too many continuations, the temporary file was not being removed, and the
5025 return code was incorrect.
5027 PH/31 Missing "BOOL" in function definition in filtertest.c.
5029 PH/32 Applied Sieve patches from the maintainer.
5031 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5033 PH/33 Added "verify = not_blind".
5035 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5036 Local/Makefile (with some defaults set). These are used in built scripts
5037 such as exicyclog, but they have never been used in the exim_install
5038 script (though there are many overriding facilities there). I have
5039 arranged that the exim_install script now takes note of these two
5042 PH/35 Installed configuration files for Dragonfly.
5044 PH/36 When a locally submitted message by a trusted user did not contain a
5045 From: header, and the sender address was obtained from -f or from an SMTP
5046 MAIL command, and the trusted user did not use -F to supply a sender
5047 name, $originator_name was incorrectly used when constructing a From:
5048 header. Furthermore, $originator_name was used for submission mode
5049 messages from external hosts without From: headers in a similar way,
5050 which is clearly wrong.
5052 PH/37 Added control=suppress_local_fixups.
5054 PH/38 When log_selector = +received_sender was set, and the addition of the
5055 sender made the log line's construction buffer exactly full, or one byte
5056 less than full, an overflow happened when the terminating "\n" was
5059 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5060 when the result of a list match is failure because a DNS lookup failed.
5062 PH/40 RM_COMMAND is now used in the building process.
5064 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5065 the "build-* directories that it finds.
5067 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5068 address in a domain literal was a prefix of an interface address.
5070 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5071 when verifying a sender address, unless rewrite_headers is false.
5073 PH/44 Wrote a long comment about why errors_to addresses are verified as
5074 recipients, not senders.
5076 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5077 the ratelimit ACL was added.
5079 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5081 PH/46 Added extra information about PostgreSQL errors to the error string.
5083 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5084 avoiding the need to install special zones in a real server. This is
5085 backwards compatible; if it can't find the fake resolver, it drops back.
5086 Thus, both old and new test suites can be run.
5088 TF/02 Added util/ratelimit.pl
5090 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5091 clock is set back in time.
5093 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5094 Candler <B.Candler@pobox.com>.
5096 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5097 OK for addresses that are the result of redirections.
5099 PH/48 A number of further additions for the benefit of the new test suite,
5100 including a fake gethostbyname() that interfaces to the fake DNS resolver
5103 TF/06 The fix for widen_domains has also been applied to qualify_single and
5104 search_parents which are the other dnslookup options that can cause
5107 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5110 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5112 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5113 service" if it encountered an MX record with an empty target hostname.
5114 The message is now "an MX or SRV record indicated no SMTP service".
5116 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5117 if the DNS of the sending site is misconfigured. This is quite a
5118 common situation. This change restores the behaviour of treating a
5119 helo verification defer as a failure.
5121 PH/51 If self=fail was set on a router, the bounce message did not include the
5122 actual error message.
5128 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5130 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5131 command itself ended in error, the underlying message about the transport
5132 filter timeout was being overwritten with the pipe command error. Now the
5133 underlying error message should be appended to the second error message.
5135 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5137 PH/02 Reduce the amount of output that "make" produces by default. Full output
5138 can still be requested.
5140 PH/03 The warning log line about a condition test deferring for a "warn" verb
5141 was being output only once per connection, rather than after each
5142 occurrence (because it was using the same function as for successful
5143 "warn" verbs). This seems wrong, so I have changed it.
5145 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5146 it should not have, which might have caused a crash in the right
5147 circumstances, but probably never did.
5149 PH/04 Installed a modified version of Tony Finch's patch to make submission
5150 mode fix the return path as well as the Sender: header line, and to
5151 add a /name= option so that you can make the user's friendly name appear
5154 TF/03 Added the control = fakedefer ACL modifier.
5156 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5157 Mark Lowes for thorough testing.
5159 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5161 TK/03 Merged latest SRS patch from Miles Wilton.
5163 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5164 with the definition in sysexits.h (which is #included earlier).
5165 Fortunately, Exim does not actually use EX_OK. The code used to try to
5166 preserve the sysexits.h value, by assuming that macro definitions were
5167 scanned for macro replacements. I have been disabused of this notion,
5168 so now the code just undefines EX_OK before #including unistd.h.
5170 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5171 in the smtp transport. When a block could not be written in a single
5172 write() function, the timeout was being re-applied to each part-write.
5173 This seems wrong - if the receiver was accepting one byte at a time it
5174 would take for ever. The timeout is now adjusted when this happens. It
5175 doesn't have to be particularly precise.
5177 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5178 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5180 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5181 without a domain if the check to <postmaster@domain> fails.
5183 SC/01 Eximstats: added -xls and the ability to specify output files
5184 (patch written by Frank Heydlauf).
5186 SC/02 Eximstats: use FileHandles for outputting results.
5188 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5190 SC/04 Eximstats: fixed display of large numbers with -nvr option
5192 SC/05 Eximstats: fixed merging of reports with empty tables.
5194 SC/06 Eximstats: added the -include_original_destination flag
5196 SC/07 Eximstats: removed tabs and trailing whitespace.
5198 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5200 TK/06 MBOX spool code: Add real "From " MBOX separator line
5201 so the .eml file is really in mbox format (even though
5202 most programs do not really care). Patch from Alex Miller.
5204 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5205 The latter is generated from $received_to and is only set if the
5206 message has one envelope recipient. SA can use these headers,
5207 obviously out-of-the-box. Patch from Alex Miller.
5209 PH/08 The ${def test on a variable was returning false if the variable's
5210 value was "0", contrary to what the specification has always said!
5211 The result should be true unless the variable is empty.
5213 PH/09 The syntax error of a character other than { following "${if
5214 def:variable_name" (after optional whitespace) was not being diagnosed.
5215 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5216 accidental colon was present, for example, could give incorrect results.
5218 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5219 result is used (not including appendfile, where other changes are about
5222 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5223 This involved changing a lot of size variables from int to off_t. It
5224 should work with maildirs and everything.
5226 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5227 spamd dying while we are connected to it.
5229 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5232 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5233 function for BDB 4.3.
5235 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5237 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5238 the sources for PCRE have been re-arranged and more files are now
5241 PH/15 The code I had for printing potentially long long variables in PH/11
5242 above was not the best (it lost precision). The length of off_t variables
5243 is now inspected at build time, and an appropriate printing format (%ld
5244 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5245 to be "long long int" or "long int". This is needed for the internal
5246 formatting function string_vformat().
5248 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5249 the configuration file to be ":syslog", then the script "guesses" where
5250 the logs files are, rather than using the compiled in default. In our
5251 case the guess is not the same as the compiled default, so the script
5252 suddenly stopped working when I started to use syslog. The patch checks
5253 to see if log_file_path is "". If so, it attempts to read it from exim
5254 with no configuration file to get the compiled in version, before it
5255 falls back to the previous guessing code."
5257 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5258 implementing BATV in an Exim configuration. See NewStuff for the gory
5261 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5262 Makefile that are specific to HP-UX.
5264 PH/18 If the "use_postmaster" option was set for a recipient callout together
5265 with the "random" option, the postmaster address was used as the MAIL
5266 FROM address for the random test, but not for the subsequent recipient
5267 test. It is now used for both.
5269 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5270 patch removes a few documentation additions to RFC 3028, because the
5271 latest draft now contains them. It adds the new en;ascii-case comparator
5272 and a new error check for 8bit text in MIME parts. Comparator and
5273 require names are now matched exactly. I enabled the subaddress
5274 extension, but it is not well tested yet (read: it works for me)."
5276 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5277 rework some of the code of TK/09 above to avoid the hardwired use of
5278 "%lld" and "long long". Replaced the call to snprintf() with a call to
5281 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5282 records point to non-existent hosts", "retry timeout exceeded", and
5283 "retry time not reached for any host after a long failure period".
5285 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5286 experimental DomainKeys support:
5288 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5289 (2) On an error such as an illegally used "control", the wrong name for
5290 the control was given.
5292 These problems did NOT occur unless DomainKeys support was compiled.
5294 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5296 PH/24 Added ${if match_ip condition.
5298 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5299 databases so that it will be absolutely obvious if a crash occurs in the
5300 DB library. This is a regular occurrence (often caused by mis-matched
5303 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5304 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5305 picked up on a user's system that detects such things. There doesn't seem
5306 to be a gcc warning option for this - only an attribute that has to be
5307 put on the function's prototype. It seems that in Fedora Core 4 they have
5308 set this on a number of new functions. No doubt there will be more in due
5311 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5312 specify a transport. However, if an address that was verified by such a
5313 router was the subject of a callout, Exim crashed because it tried to
5314 read the rcpt_include_affixes from the non-existent transport. Now it
5315 just assumes that the setting of that option is false. This bug was
5316 introduced by 4.51/PH/31.
5318 PH/28 Changed -d+all to exclude +memory, because that information is very
5319 rarely of interest, but it makes the output a lot bigger. People tend to
5320 do -d+all out of habit.
5322 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5323 code in os-type was giving problems when libc.so lives in lib64, like on
5326 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5327 aren't the modern standard, and it seems that some systems' include files
5328 don't always have them. Exim was already checking for some of the newer
5329 ones like T_AAAA, and defining it itself. I've added checks for all the
5330 record types that Exim uses.
5332 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5333 not automatically generating a new one, as it is supposed to. This
5334 prevented TLS from working. If the file did exist, but contained invalid
5335 data, a new version was generated, as expected. It was only the case of a
5336 non-existent file that was broken.
5338 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5339 with a change in libdomainkeys > 0.64.
5341 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5342 from DNS. If the selector record carries the flag, it now has
5343 precedence over the domain-wide flag.
5345 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5347 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5348 the use of a port name that isn't defined in /etc/services) occurred, the
5349 message was deferred as in a normal delivery, and thus remained on the
5350 spool, instead of being failed because of the mua_wrapper setting. This
5351 is now fixed, and I tidied up some of the mua_wrapper messages at the
5354 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5355 the messages in a hash of arrays rather than using individual hashes.
5356 This is a bit cleaner and results in dramatic memory savings, albeit
5357 at a slight CPU cost.
5359 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5360 as requested by Marc Sherman.
5362 SC/10 Eximstats: added histograms for user specified patterns as requested
5365 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5367 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5368 fopen() in the content-scanning modules that did not already have it.
5374 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5375 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5377 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5379 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5381 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5382 to test Sieve filters that use "vacation".
5384 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5385 that changes the way the GnuTLS parameters are stored in the cache file.
5386 The new format can be generated externally. For backward compatibility,
5387 if the data in the cache doesn't make sense, Exim assumes it has read an
5388 old-format file, and it generates new data and writes a new file. This
5389 means that you can't go back to an older release without removing the
5392 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5393 work if there are any delivery delays because "one_time" forces the
5394 parent to be marked "delivered", so its unseen clone is never tried
5395 again. For this reason, Exim now forbids the simultaneous setting of
5398 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5399 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5400 entry if you want to know the details. The fix, however, neglected to
5401 consider the case when local delivery batching is involved. The test for
5402 "previously delivered" was not happening when checking to see if an
5403 address could be batched with a previous (undelivered) one; under
5404 certain circumstances this could lead to multiple deliveries to the same
5407 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5408 in its include files, and this causes problems building Exim.
5410 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5411 header_syntax) but Exim was just ignoring anything given after a slash.
5412 In particular, this caused confusion with an attempt to use "verify =
5413 reverse_host_lookup/defer_ok". An error is now given when options are
5414 supplied for verify items that do not have them. (Maybe reverse_host_
5415 lookup should have a defer_ok option, but that's a different point.)
5417 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5418 defined by RFC 821) to 2048, because there were problems with some AUTH
5419 commands, and RFC 1869 says the size should be increased for extended
5420 SMTP commands that take arguments.
5422 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5425 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5426 "unknown" error; now it says that the functionality isn't in the binary.
5428 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5429 an address' error message when a string expansion fails (syntax or
5430 whatever). Otherwise the password may appear in the log. Following change
5431 PH/42 below, there is no longer a chance of it appearing in a bounce
5434 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5436 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5437 of its addresses was ever tried. (Bugzilla bug #2.)
5439 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5440 the result incorrectly in the debug output. (It correctly added a newline
5441 to what was transported.)
5443 TF/01 Added $received_time.
5445 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5446 commented out examples of how to interface to a virus scanner and to
5447 SpamAssassin. Also added commented examples of av_scanner and
5448 spamd_address settings.
5450 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5451 and controls are allowed in which ACLs. There were a couple of minor
5452 errors. Some of the entries in the conditions table (which is a table of
5453 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5454 negation of where the condition IS allowed.
5456 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5458 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5459 header file does not have a version number, so I've had to invent a new
5460 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5461 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5462 radiusclient), but it was contributed by a Radius user.
5464 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5465 files or directories, for OpenSSL.
5467 PH/19 When an Exim process that is running as root has to create an Exim log
5468 file, it does so in a subprocess that runs as exim:exim so as to get the
5469 ownership right at creation (otherwise, other Exim processes might see
5470 the file with the wrong ownership). There was no test for failure of this
5471 fork() call, which would lead to the process getting stuck as it waited
5472 for a non-existent subprocess. Forks do occasionally fail when resources
5473 run out. I reviewed all the other calls to fork(); they all seem to check
5476 PH/20 When checking for unexpected SMTP input at connect time (before writing
5477 the banner), Exim was not dealing correctly with a non-positive return
5478 from the read() function. If the client had disconnected by this time,
5479 the result was a log entry for a synchronization error with an empty
5480 string after "input=" when read() returned zero. If read() returned -1
5481 (an event I could not check), uninitialized data bytes were printed.
5482 There were reports of junk text (parts of files, etc) appearing after
5485 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5487 PH/22 Added support for macro redefinition, and (re)definition in between
5488 driver and ACL definitions.
5490 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5491 forgetting to use the resulting value; it was using the unexpanded value.
5493 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5494 hadn't been configured. The fix is from Juergen Kreileder, who
5495 understands it better than I do:
5497 "Here's what I see happening with three configured cyrus_sasl
5498 authenticators configured (plain, login, cram-md5):
5500 On startup auth_cyrus_sasl_init() gets called for each of these.
5501 This means three calls to sasl_listmech() without a specified mech_list.
5502 => SASL tests which mechs of all available mechs actually work
5503 => three warnings about OTP not working
5504 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5506 With the patch, sasl_listmech() also gets called three times. But now
5507 SASL's mech_list option is set to the server_mech specified in the the
5508 authenticator. Or in other words, the answer from sasl_listmech()
5509 gets limited to just the mech you're testing for (which is different
5511 => the return list contains just 'plain' or 'login', 'cram-md5' or
5512 nothing depending on the value of ob->server_mech.
5514 I've just tested the patch: Authentication still works fine,
5515 unavailable mechs specified in the exim configuration are still
5516 caught, and the auth.log warnings about OTP are gone."
5518 PH/25 When debugging is enabled, the contents of the command line are added
5519 to the debugging output, even when log_selector=+arguments is not
5522 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5523 answer is "GNU", and only if the return is "GNU/something" is the answer
5526 PH/27 $acl_verify_message is now set immediately after the failure of a
5527 verification in an ACL, and so is available in subsequent modifiers. In
5528 particular, the message can be preserved by coding like this:
5530 warn !verify = sender
5531 set acl_m0 = $acl_verify_message
5533 Previously, $acl_verify_message was set only while expanding "message"
5534 and "log_message" when a very denied access.
5536 PH/28 Modified OS/os.c-Linux with
5538 -#ifndef OS_LOAD_AVERAGE
5539 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5541 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5542 nomenclature these days.)
5544 PH/29 Installed patch from the Sieve maintainer that adds the options
5545 sieve_useraddress and sieve_subaddress to the redirect router.
5547 PH/30 In these circumstances:
5548 . Two addresses routed to the same list of hosts;
5549 . First host does not offer TLS;
5550 . First host accepts first address;
5551 . First host gives temporary error to second address;
5552 . Second host offers TLS and a TLS session is established;
5553 . Second host accepts second address.
5554 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5555 and peerdn, if requested) that were in fact used only for the second
5558 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5559 attention to any local part prefix or suffix that was matched by the
5560 router that accepted the address. It now behaves in the same way as it
5561 does for delivery: the affixes are removed from the local part unless
5562 rcpt_include_affixes is set on the transport.
5564 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5565 timeout during the DATA phase of an incoming message.
5567 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5568 applied a patch sanctioned by the Sieve maintainer.
5570 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5571 the uid or gid is negative. A case of a negative gid caused this to be
5572 noticed. The fix allows for either to be negative.
5574 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5575 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5578 PH/36 The change PH/12 above was broken. Fixed it.
5580 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5581 the grounds that routing the same address twice would always produce the
5582 same answer. This might have been true once, but it is certainly no
5583 longer true now. Routing a child address may depend on the previous
5584 routing that produced that child. Some complicated redirection strategies
5585 went wrong when messages had multiple recipients, and made Exim's
5586 behaviour dependent on the order in which the addresses were given.
5588 I have moved the duplicate checking until after the routing is complete.
5589 Exim scans the addresses that are assigned to local and remote
5590 transports, and removes any duplicates. This means that more work will be
5591 done, as duplicates will always all be routed, but duplicates are
5592 presumably rare, so I don't expect this is of any significance.
5594 For deliveries to pipes, files, and autoreplies, the duplicate checking
5595 still happens during the routing process, since they are not going to be
5598 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5599 It corrects a timeout issue with spamd. This is Ian's comment: "The
5600 background is that sometimes spamd either never reads data from a
5601 connection it has accepted, or it never writes response data. The exiscan
5602 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5603 blindly assumes that writes won't block so it may never time out."
5605 PH/39 Allow G after quota size as well as K and M.
5607 PH/40 The value set for $authenticated_id in an authenticator may not contain
5608 binary zeroes or newlines because the value is written to log lines and
5609 to spool files. There was no check on this. Now the value is run through
5610 the string_printing() function so that such characters are converted to
5611 printable escape sequences.
5613 PH/41 $message_linecount is a new variable that contains the total number of
5614 lines in the message. Compare $body_linecount, which is the count for the
5617 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5618 bounce and delay warning messages, except in certain special cases, which
5621 (a) An SMTP error message from a remote host;
5622 (b) A message specified in a :fail: redirection;
5623 (c) A message specified in a "fail" command in a system filter;
5624 (d) A message specified in a FAIL return from the queryprogram router;
5625 (e) A message specified by the cannot_route_message router option.
5627 In these cases only, Exim does include the error details in bounce and
5628 warning messages. There are also a few cases where bland messages such
5629 as "unrouteable address" or "local delivery error" are given.
5631 PH/43 $value is now also set for the "else" part of a ${run expansion.
5633 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5634 being worked on, but at least Exim now implements the latest version to
5637 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5638 process to complete was treated as a delivery failure, a timeout while
5639 writing the message to the pipe was logged, but erroneously treated as a
5640 successful delivery. Such timeouts include transport filter timeouts. For
5641 consistency with the overall process timeout, these timeouts are now
5642 treated as errors, giving rise to delivery failures by default. However,
5643 there is now a new Boolean option for the pipe transport called
5644 timeout_defer, which, if set TRUE, converts the failures into defers for
5645 both kinds of timeout. A transport filter timeout is now identified in
5648 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5649 systems where "make" and "gmake" are different, calling "gmake" at top
5650 level broke things. I've arranged for the value of $(MAKE) to be passed
5651 from the Makefile to this script so that it can call the same version of
5655 A note about Exim versions 4.44 and 4.50
5656 ----------------------------------------
5658 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5659 changes of various kinds. As a consequence, a big documentation update was
5660 needed. This delayed the release for rather longer than seemed good, especially
5661 in the light of a couple of (minor) security issues. Therefore, the changes
5662 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5663 release. So 4.44 and 4.50 are in effect two different branches that both start
5666 I have left the 4.50 change log unchanged; it contains all the changes since
5667 4.43. The change log for 4.44 is below; many of its items are identical to
5668 those for 4.50. This seems to be the most sensible way to preserve the
5669 historical information.
5675 1. Minor wording change to the doc/README.SIEVE file.
5677 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5678 computation of the current number of files was incorrect.
5680 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5681 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5684 4. Give more explanation in the error message when the command for a transport
5685 filter fails to execute.
5687 5. There are several places where Exim runs a non-Exim command in a
5688 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5689 was being done only for the command run by the queryprogram router. It is
5690 now done for all such subprocesses. The other cases are: ${run, transport
5691 filters, and the commands run by the lmtp and pipe transports.
5693 6. Added CONFIGURE_GROUP build-time option.
5695 7. Some older OS have a limit of 256 on the maximum number of file
5696 descriptors. Exim was using setrlimit() to set 1000 as a large value
5697 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5698 systems. I've change it so that if it can't get 1000, it tries for 256.
5700 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5701 was an oversight, and furthermore, ever since the addition of extra
5702 controls (e.g. 4.43/32), the checks on when to allow different forms of
5703 "control" were broken. There should now be diagnostics for all cases when a
5704 control that does not make sense is encountered.
5706 9. Added the /retain_sender option to "control=submission".
5708 10. $recipients is now available in the predata ACL (oversight).
5710 11. Tidy the search cache before the fork to do a delivery from a message
5711 received from the command line. Otherwise the child will trigger a lookup
5712 failure and thereby defer the delivery if it tries to use (for example) a
5713 cached ldap connection that the parent has called unbind on.
5715 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5716 of $address_data from the recipient verification was clobbered by the
5717 sender verification.
5719 13. The value of address_data from a sender verification is now available in
5720 $sender_address_data in subsequent conditions in the ACL statement.
5722 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5724 15. Added a new option "connect=<time>" to callout options, to set a different
5727 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5728 was its contents. (It was OK if the option was not defined at all.)
5730 17. A "Completed" log line is now written for messages that are removed from
5731 the spool by the -Mrm option.
5733 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5734 information about exactly what failed.
5736 19. Added -dd to debug only the daemon process.
5738 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5739 handles timeouts, both on the server side and network timeouts. Renamed the
5740 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5742 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5743 It is now set to "smtps".
5745 22. $host_address is now set to the target address during the checking of
5746 ignore_target_hosts.
5748 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5749 being passed; this would have caused $sender_host_name to have been used if
5750 matching the list had actually called for a host name (not very likely,
5751 since this list is usually IP addresses). A host name is now passed as
5754 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5755 code that specifies a non-restarting handler (typically sigaction() in
5756 modern systems) in an attempt to fix a rare and obscure crash bug.
5758 25. Narrowed the window for a race in the daemon that could cause it to ignore
5759 SIGCHLD signals. This is not a major problem, because they are used only to
5760 wake it up if nothing else does.
5762 26. A malformed maildirsize file could cause Exim to calculate negative values
5763 for the mailbox size or file count. Odd effects could occur as a result.
5764 The maildirsize information is now recalculated if the size or filecount
5767 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5768 support for a long time. Removed HAVE_SYS_VFS_H.
5770 28. Installed the latest version of exipick from John Jetmore.
5772 29. In an address list, if the pattern was not a regular expression, an empty
5773 subject address (from a bounce message) matched only if the pattern was an
5774 empty string. Non-empty patterns were not even tested. This was the wrong
5775 because it is perfectly reasonable to use an empty address as part of a
5776 database query. An empty address is now tested by patterns that are
5777 lookups. However, all the other forms of pattern expect the subject to
5778 contain a local part and a domain, and therefore, for them, an empty
5779 address still always fails if the pattern is not itself empty.
5781 30. Exim went into a mad DNS loop when attempting to do a callout where the
5782 host was specified on an smtp transport, and looking it up yielded more
5783 than one IP address.
5785 31. Re-factored the code for checking spool and log partition space into a
5786 function that finds that data and another that does the check. The former
5787 is then used to implement four new variables: $spool_space, $log_space,
5788 $spool_inodes, and $log_inodes.
5790 32. The RFC2047 encoding function was originally intended for short strings
5791 such as real names; it was not keeping to the 75-character limit for
5792 encoded words that the RFC imposes. It now respects the limit, and
5793 generates multiple encoded words if necessary. To be on the safe side, I
5794 have increased the buffer size for the ${rfc2047: expansion operator from
5797 33. It is now permitted to omit both strings after an "if" condition; if the
5798 condition is true, the result is "true". As before, when the second string
5799 is omitted, a false condition yields an empty string. This makes it less
5800 cumbersome to write custom ACL and router conditions.
5802 34. Failure to deliver a bounce message always caused it to be frozen, even if
5803 there was an errors_to setting on the router. The errors_to setting is now
5806 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5807 canonical form (fully expanded) before being placed in
5808 $sender_host_address.
5810 36. The table in the code that translates DNS record types into text (T_A to
5811 "A" for instance) was missing entries for NS and CNAME. It is just possible
5812 that this could have caused confusion if both these types were looked up
5813 for the same domain, because the text type is used as part of Exim's
5814 per-process caching. But the chance of anyone hitting this buglet seems
5817 37. The dnsdb lookup has been extended in a number of ways.
5819 (1) There is a new type, "zns", which walks up the domain tree until it
5820 finds some nameserver records. It should be used with care.
5822 (2) There is a new type, "mxh", which is like "mx" except that it returns
5823 just the host names, not the priorities.
5825 (3) It is now possible to give a list of domains (or IP addresses) to be
5826 looked up. The behaviour when one of the lookups defers can be
5827 controlled by a keyword.
5829 (4) It is now possible to specify the separator character for use when
5830 multiple records are returned.
5832 38. The dnslists ACL condition has been extended: it is now possible to supply
5833 a list of IP addresses and/or domains to be looked up in a particular DNS
5836 39. Added log_selector=+queue_time_overall.
5838 40. When running the queue in the test harness, wait just a tad after forking a
5839 delivery process, to get repeatability of debugging output.
5841 41. Include certificate and key file names in error message when GnuTLS fails
5842 to set them up, because the GnuTLS error message doesn't include the name
5843 of the failing file when there is a problem reading it.
5845 42. Allow both -bf and -bF in the same test run.
5847 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5849 44. The "Exiscan patch" is now merged into the mainline Exim source.
5851 45. Sometimes the final signoff response after QUIT could fail to get
5852 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5853 < 0 before doing a fflush(). This bug looks as though it goes back to the
5854 introduction of TLS in release 3.20, but "sometimes" must have been rare
5855 because the tests only now provoked it.
5857 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5858 (this can affect the format of dates).
5860 47. exim_tidydb, when checking for the continued existence of a message for
5861 which it has found a message-specific retry record, was not finding
5862 messages that were in split spool directories. Consequently, it was
5863 deleting retry records that should have stayed in existence.
5865 48. Steve fixed some bugs in eximstats.
5867 49. The SPA authentication driver was not abandoning authentication and moving
5868 on to the next authenticator when an expansion was forced to fail,
5869 contradicting the general specification for all authenticators. Instead it
5870 was generating a temporary error. It now behaves as specified.
5872 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5873 (the order specifies the preference for clients). The order is now AES256,
5874 AES128, 3DES, ARCFOUR128.
5876 51. Small patch to Sieve code - explicitly set From: when generating an
5879 52. Exim crashed if a remote delivery caused a very long error message to be
5880 recorded - for instance if somebody sent an entire SpamAssassin report back
5881 as a large number of 550 error lines. This bug was coincidentally fixed by
5882 increasing the size of one of Exim's internal buffers (big_buffer) that
5883 happened as part of the Exiscan merge. However, to be on the safe side, I
5884 have made the code more robust (and fixed the comments that describe what
5887 53. Now that there can be additional text after "Completed" in log lines (if
5888 the queue_time_overall log selector is set), a one-byte patch to exigrep
5889 was needed to allow it to recognize "Completed" as not the last thing in
5892 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5893 patch that reportedly fixes this has been added. I am not expert enough to
5894 create a test for it. This is what the patch creator wrote:
5896 "I found a little strange behaviour of ldap code when working with
5897 Windows 2003 AD Domain, where users was placed in more than one
5898 Organization Units. When I tried to give exim partial DN, the exit code
5899 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5900 But simultaneously result of request was absolutely normal ldap result,
5901 so I produce this patch..."
5903 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5904 REFERENCE, so I have modified the code to exclude the patch when that macro
5907 55. Some experimental protocols are using DNS PTR records for new purposes. The
5908 keys for these records are domain names, not reversed IP addresses. The
5909 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5910 leaves it alone. Component reversal etc. now happens only for IP addresses.
5913 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5915 57. Double the size of the debug message buffer (to 2048) so that more of very
5916 long debug lines gets shown.
5918 58. The exicyclog utility now does better if the number of log files to keep
5919 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5921 59. Two changes related to the smtp_active_hostname option:
5923 (1) $smtp_active_hostname is now available as a variable.
5924 (2) The default for smtp_banner uses $smtp_active_hostname instead
5925 of $primary_hostname.
5927 60. The host_aton() function is supposed to be passed a string that is known
5928 to be a valid IP address. However, in the case of IPv6 addresses, it was
5929 not checking this. This is a hostage to fortune. Exim now panics and dies
5930 if the condition is not met. A case was found where this could be provoked
5931 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5932 components; fortuitously, this particular loophole had already been fixed
5933 by change 4.50/55 above.
5935 If there are any other similar loopholes, the new check in host_aton()
5936 itself should stop them being exploited. The report I received stated that
5937 data on the command line could provoke the exploit when Exim was running as
5938 exim, but did not say which command line option was involved. All I could
5939 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5940 running as the user.
5943 61. There was a buffer overflow vulnerability in the SPA authentication code
5944 (which came originally from the Samba project). I have added a test to the
5945 spa_base64_to_bits() function which I hope fixes it.
5948 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5949 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5951 63. The daemon start-up calls getloadavg() while still root for those OS that
5952 need the first call to be done as root, but it missed one case: when
5953 deliver_queue_load_max is set with deliver_drop_privilege. This is
5954 necessary for the benefit of the queue runner, because there is no re-exec
5955 when deliver_drop_privilege is set.
5957 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5958 This has been fixed.
5960 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5961 was in use, was not putting the data itself into the right store pool;
5962 consequently, it could be overwritten for a subsequent message in the same
5963 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5966 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5968 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5969 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5970 the function were treating the return as a boolean value, which happened to
5971 work because 0=false and not-0=true, but is not correct code.
5973 68. The host_aton() function was not handling scoped IPv6 addresses (those
5974 with, for example, "%eth0" on the end) correctly.
5976 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5977 negated items (that is, ~something) in unsigned ints. Some compilers
5978 apparently mutter when there is no cast.
5980 70. If an address verification called from an ACL failed, and did not produce a
5981 user-specific message (i.e. there was only a "system" message), nothing was
5982 put in $acl_verify_message. In this situation, it now puts the system
5985 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5986 session; change 4.31/43 added the unwanted input to the log line - except
5987 that it did not do this in the start of session case. It now does.
5989 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5990 This is wrong and can cause the other end to generate a synchronization
5991 error if it is another Exim or anything else that does the synchronization
5992 check. A QUIT command is no longer sent after a timeout.
5994 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5995 during host lookups.
5997 74. The defer_ok option of callout verification was not working if it was used
5998 when verifying addresses in header lines, that is, for this case:
6000 verify = header_sender/callout=defer_ok
6002 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6003 those file descriptors could be used for SMTP connections. If anything
6004 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6005 could be sent to the SMTP client, causing chaos. The daemon now opens
6006 stdin, stdout, and stderr to /dev/null when it puts itself into the
6009 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6010 log by default. The user can override this with suitable Perl magic.
6012 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6013 the log message when discard triggers, was not working for the DATA ACL or
6014 for the non-SMTP ACL.
6016 78. Error message wording change in sieve.c.
6018 79. If smtp_accept_max_per_host was set, the number of connections could be
6019 restricted to fewer than expected, because the daemon was trying to set up
6020 a new connection before checking whether the processes handling previous
6021 connections had finished. The check for completed processes is now done
6022 earlier. On busy systems, this bug wouldn't be noticed because something
6023 else would have woken the daemon, and it would have reaped the completed
6026 80. If a message was submitted locally by a user whose login name contained one
6027 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6028 It caused a spool format error. I have fixed the spool reading code. A
6029 related problem was that the "from" clause in the Received: line became
6030 illegal because of the space(s). It is now covered by ${quote_local_part.
6032 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6035 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6037 83. Patch from Sieve maintainer to fix unterminated string problem in
6038 "vacation" handling.
6040 84. Some minor changes to the Linux configuration files to help with other
6041 OS variants using glibc.
6043 85. One more patch for Sieve to update vacation handling to latest spec.
6046 ----------------------------------------------------
6047 See the note above about the 4.44 and 4.50 releases.
6048 ----------------------------------------------------
6054 1. Change 4.43/35 introduced a bug that caused file counts to be
6055 incorrectly computed when quota_filecount was set in an appendfile
6058 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6059 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6062 3. Give more explanation in the error message when the command for a transport
6063 filter fails to execute.
6065 4. There are several places where Exim runs a non-Exim command in a
6066 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6067 was being done only for the command run by the queryprogram router. It is
6068 now done for all such subprocesses. The other cases are: ${run, transport
6069 filters, and the commands run by the lmtp and pipe transports.
6071 5. Some older OS have a limit of 256 on the maximum number of file
6072 descriptors. Exim was using setrlimit() to set 1000 as a large value
6073 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6074 systems. I've change it so that if it can't get 1000, it tries for 256.
6076 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6077 was an oversight, and furthermore, ever since the addition of extra
6078 controls (e.g. 4.43/32), the checks on when to allow different forms of
6079 "control" were broken. There should now be diagnostics for all cases when a
6080 control that does not make sense is encountered.
6082 7. $recipients is now available in the predata ACL (oversight).
6084 8. Tidy the search cache before the fork to do a delivery from a message
6085 received from the command line. Otherwise the child will trigger a lookup
6086 failure and thereby defer the delivery if it tries to use (for example) a
6087 cached ldap connection that the parent has called unbind on.
6089 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6090 of $address_data from the recipient verification was clobbered by the
6091 sender verification.
6093 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6094 was its contents. (It was OK if the option was not defined at all.)
6096 11. A "Completed" log line is now written for messages that are removed from
6097 the spool by the -Mrm option.
6099 12. $host_address is now set to the target address during the checking of
6100 ignore_target_hosts.
6102 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6103 being passed; this would have caused $sender_host_name to have been used if
6104 matching the list had actually called for a host name (not very likely,
6105 since this list is usually IP addresses). A host name is now passed as
6108 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6109 code that specifies a non-restarting handler (typically sigaction() in
6110 modern systems) in an attempt to fix a rare and obscure crash bug.
6112 15. Narrowed the window for a race in the daemon that could cause it to ignore
6113 SIGCHLD signals. This is not a major problem, because they are used only to
6114 wake it up if nothing else does.
6116 16. A malformed maildirsize file could cause Exim to calculate negative values
6117 for the mailbox size or file count. Odd effects could occur as a result.
6118 The maildirsize information is now recalculated if the size or filecount
6121 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6122 support for a long time. Removed HAVE_SYS_VFS_H.
6124 18. Updated exipick to current release from John Jetmore.
6126 19. Allow an empty sender to be matched against a lookup in an address list.
6127 Previously the only cases considered were a regular expression, or an
6130 20. Exim went into a mad DNS lookup loop when doing a callout where the
6131 host was specified on the transport, if the DNS lookup yielded more than
6134 21. The RFC2047 encoding function was originally intended for short strings
6135 such as real names; it was not keeping to the 75-character limit for
6136 encoded words that the RFC imposes. It now respects the limit, and
6137 generates multiple encoded words if necessary. To be on the safe side, I
6138 have increased the buffer size for the ${rfc2047: expansion operator from
6141 22. Failure to deliver a bounce message always caused it to be frozen, even if
6142 there was an errors_to setting on the router. The errors_to setting is now
6145 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6146 canonical form (fully expanded) before being placed in
6147 $sender_host_address.
6149 24. Updated eximstats to version 1.33
6151 25. Include certificate and key file names in error message when GnuTLS fails
6152 to set them up, because the GnuTLS error message doesn't include the name
6153 of the failing file when there is a problem reading it.
6155 26. Expand error message when OpenSSL has problems setting up cert/key files.
6158 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6159 (this can affect the format of dates).
6161 28. exim_tidydb, when checking for the continued existence of a message for
6162 which it has found a message-specific retry record, was not finding
6163 messages that were in split spool directories. Consequently, it was
6164 deleting retry records that should have stayed in existence.
6166 29. eximstats updated to version 1.35
6167 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6168 1.35 - bugfix such that pie charts by volume are generated correctly
6170 30. The SPA authentication driver was not abandoning authentication and moving
6171 on to the next authenticator when an expansion was forced to fail,
6172 contradicting the general specification for all authenticators. Instead it
6173 was generating a temporary error. It now behaves as specified.
6175 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6176 (the order specifies the preference for clients). The order is now AES256,
6177 AES128, 3DES, ARCFOUR128.
6179 31. Small patch to Sieve code - explicitly set From: when generating an
6182 32. Exim crashed if a remote delivery caused a very long error message to be
6183 recorded - for instance if somebody sent an entire SpamAssassin report back
6184 as a large number of 550 error lines. This bug was coincidentally fixed by
6185 increasing the size of one of Exim's internal buffers (big_buffer) that
6186 happened as part of the Exiscan merge. However, to be on the safe side, I
6187 have made the code more robust (and fixed the comments that describe what
6190 33. Some experimental protocols are using DNS PTR records for new purposes. The
6191 keys for these records are domain names, not reversed IP addresses. The
6192 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6193 leaves it alone. Component reversal etc. now happens only for IP addresses.
6196 34. The host_aton() function is supposed to be passed a string that is known
6197 to be a valid IP address. However, in the case of IPv6 addresses, it was
6198 not checking this. This is a hostage to fortune. Exim now panics and dies
6199 if the condition is not met. A case was found where this could be provoked
6200 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6201 components; fortuitously, this particular loophole had already been fixed
6202 by change 4.50/55 or 4.44/33 above.
6204 If there are any other similar loopholes, the new check in host_aton()
6205 itself should stop them being exploited. The report I received stated that
6206 data on the command line could provoke the exploit when Exim was running as
6207 exim, but did not say which command line option was involved. All I could
6208 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6209 running as the user.
6212 35. There was a buffer overflow vulnerability in the SPA authentication code
6213 (which came originally from the Samba project). I have added a test to the
6214 spa_base64_to_bits() function which I hope fixes it.
6217 36. The daemon start-up calls getloadavg() while still root for those OS that
6218 need the first call to be done as root, but it missed one case: when
6219 deliver_queue_load_max is set with deliver_drop_privilege. This is
6220 necessary for the benefit of the queue runner, because there is no re-exec
6221 when deliver_drop_privilege is set.
6223 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6224 was in use, was not putting the data itself into the right store pool;
6225 consequently, it could be overwritten for a subsequent message in the same
6226 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6229 38. Sometimes the final signoff response after QUIT could fail to get
6230 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6231 < 0 before doing a fflush(). This bug looks as though it goes back to the
6232 introduction of TLS in release 3.20, but "sometimes" must have been rare
6233 because the tests only now provoked it.
6239 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6240 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6241 Now it returns an integer. A number of calls were still expecting a T/F
6242 return. Fortuitously, in all cases, the tests worked in OK situations,
6243 which is the norm. However, things would have gone wrong on any write
6244 failures on the smtp file descriptor. This function is used when sending
6245 messages over SMTP and also when doing verify callouts.
6247 2. When Exim is called to do synchronous delivery of a locally submitted
6248 message (the -odf or -odi options), it no longer closes stderr before doing
6251 3. Implemented the mua_wrapper option.
6253 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6255 5. Implemented the functions header_remove(), header_testname(),
6256 header_add_at_position(), and receive_remove_recipient(), and exported them
6259 6. If an ACL "warn" statement specified the addition of headers, Exim already
6260 inserted X-ACL-Warn: at the start if there was no header name. However, it
6261 was not making this test for the second and subsequent header lines if
6262 there were newlines in the string. This meant that an invalid header could
6263 be inserted if Exim was badly configured.
6265 7. Allow an ACL "warn" statement to add header lines at the start or after all
6266 the Received: headers, as well as at the end.
6268 8. Added the rcpt_4xx retry error code.
6270 9. Added postmaster_mailfrom=xxx to callout verification option.
6272 10. Added mailfrom=xxxx to the callout verification option, for verify=
6275 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6276 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6278 12. Inserted some casts to stop certain compilers warning when using pointer
6279 differences as field lengths or precisions in printf-type calls (mostly
6280 affecting debugging statements).
6282 13. Added optional readline() support for -be (dynamically loaded).
6284 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6285 same clock tick as a message's arrival, so that its received time was the
6286 same as the "first fail" time on the retry record, and that message
6287 remained on the queue past the ultimate address timeout, every queue runner
6288 would try a delivery (because it was past the ultimate address timeout) but
6289 after another failure, the ultimate address timeout, which should have then
6290 bounced the address, did not kick in. This was a "< instead of <=" error;
6291 in most cases the first failure would have been in the next clock tick
6292 after the received time, and all would be well.
6294 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6295 being recognized when the domain list was tested by the match_domain
6296 condition in an expansion string.
6298 16. Added the ${str2b64: operator.
6300 17. Exim was always calling setrlimit() to set a large limit for the number of
6301 processes, without checking whether the existing limit was already
6302 adequate. (It did check for the limit on file descriptors.) Furthermore,
6303 errors from getrlimit() and setrlimit() were being ignored. Now they are
6304 logged to the main and panic logs, but Exim does carry on, to try to do its
6305 job under whatever limits there are.
6307 18. Imported PCRE 5.0.
6309 19. Trivial typo in log message " temporarily refused connection" (the leading
6312 20. If the log selector return_path_on_delivery was set and an address was
6313 redirected to /dev/null, the delivery process crashed because it assumed
6314 that a return path would always be set for a "successful" delivery. In this
6315 case, the whole delivery is bypassed as an optimization, and therefore no
6318 21. Internal re-arrangement: the function for sending a challenge and reading
6319 a response while authentication was assuming a zero-terminated challenge
6320 string. It's now changed to take a pointer and a length, to allow for
6321 binary data in such strings.
6323 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6325 23. Exim was not respecting finduser_retries when seeking the login of the
6326 uid under which it was called; it was always trying 10 times. (The default
6327 setting of finduser_retries is zero.) Also, it was sleeping after the final
6328 failure, which is pointless.
6330 24. Implemented tls_on_connect_ports.
6332 25. Implemented acl_smtp_predata.
6334 26. If the domain in control=submission is set empty, Exim assumes that the
6335 authenticated id is a complete email address when it generates From: or
6336 Sender: header lines.
6338 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6339 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6340 chown and chgrp in /bin and hostname in /usr/bin.
6342 28. Exim was keeping the "process log" file open after each use, just as it
6343 does for the main log. This opens the possibility of it remaining open for
6344 long periods when the USR1 signal hits a daemon. Occasional processlog
6345 errors were reported, that could have been caused by this. Anyway, it seems
6346 much more sensible not to leave this file open at all, so that is what now
6349 29. The long-running daemon process does not normally write to the log once it
6350 has entered its main loop, and it closes the log before doing so. This is
6351 so that log files can straightforwardly be renamed and moved. However,
6352 there are a couple of unusual error situations where the daemon does write
6353 log entries, and I had neglected to close the log afterwards.
6355 30. The text of an SMTP error response that was received during a remote
6356 delivery was being truncated at 512 bytes. This is too short for some of
6357 the long messages that one sometimes sees. I've increased the limit to
6360 31. It is now possible to make retry rules that apply only when a message has a
6361 specific sender, in particular, an empty sender.
6363 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6364 it possible to be selective about when SMTP synchronization is enforced.
6366 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6368 32. Implemented hosts_connection_nolog.
6370 33. Added an ACL for QUIT.
6372 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6375 35. Added mailbox_size and mailbox_filecount to appendfile.
6377 36. Added control = no_multiline_responses to ACLs.
6379 37. There was a bug in the logic of the code that waits for the clock to tick
6380 in the case where the clock went backwards by a substantial amount such
6381 that the microsecond fraction of "now" was more than the microsecond
6382 fraction of "then" (but the whole seconds number was less).
6384 38. Added support for the libradius Radius client library this is found on
6385 FreeBSD (previously only the radiusclient library was supported).
6391 1. When certain lookups returned multiple values in the form name=value, the
6392 quoting of the values was not always being done properly. Specifically:
6393 (a) If the value started with a double quote, but contained no whitespace,
6395 (b) If the value contained whitespace other than a space character (i.e.
6396 tabs or newlines or carriage returns) it was not quoted.
6397 This fix has been applied to the mysql and pgsql lookups by writing a
6398 separate quoting function and calling it from the lookup code. The fix
6399 should probably also be applied to nisplus, ibase and oracle lookups, but
6400 since I cannot test any of those, I have not disturbed their existing code.
6402 2. A hit in the callout cache for a specific address caused a log line with no
6403 reason for rejecting RCPT. Now it says "Previous (cached) callout
6404 verification failure".
6406 3. There was an off-by-one bug in the queryprogram router. An over-long
6407 return line was truncated at 256 instead of 255 characters, thereby
6408 overflowing its buffer with the terminating zero. As well as fixing this, I
6409 have increased the buffer size to 1024 (and made a note to document this).
6411 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6412 when Exim is waiting for an SMTP response from a remote server, Exim
6413 restarts its select() call on the socket, thereby resetting its timeout.
6414 This is not a problem when such interrupts are rare. Somebody set up a cron
6415 job to run exiwhat every 2 minutes, which is less than the normal select()
6416 timeout (5 or 10 minutes). This meant that the select() timeout never
6417 kicked in because it was always reset. I have fixed this by comparing the
6418 time when an interrupt arrives with the time at the start of the first call
6419 to select(). If more time than the timeout has elapsed, the interrupt is
6420 treated as a timeout.
6422 5. Some internal re-factoring in preparation for the addition of Sieve
6423 extensions (by MH). In particular, the "personal" test is moved to a
6424 separate function, and given an option for scanning Cc: and Bcc: (which is
6425 not set for Exim filters).
6427 6. When Exim created an email address using the login of the caller as the
6428 local part (e.g. when creating a From: or Sender: header line), it was not
6429 quoting the local part when it contained special characters such as @.
6431 7. Installed new OpenBSD configuration files.
6433 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6434 try to make them clearer.
6436 9. Callout options, other than the timeout value, were being ignored when
6437 verifying sender addresses in header lines. For example, when using
6439 verify = header_sender/callout=no_cache
6441 the cache was (incorrectly) being used.
6443 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6444 only the Cygwin environment.
6446 11. When return_path_on_delivery was set as a log selector, if different remote
6447 addresses in the same message used different return paths and parallel
6448 remote delivery occurred, the wrong values would sometimes be logged.
6449 (Whenever a remote delivery process finished, the return path value from
6450 the most recently started remote delivery process was logged.)
6452 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6453 lines when AUTH and/or TLS are in use. This is the "received protocol"
6454 field. Exim used to use "asmtp" for authenticated SMTP, without any
6455 indication (in the protocol name) for TLS use. Now it follows the RFC and
6456 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6457 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6458 names appear in log lines as well as in Received: header lines.
6460 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6461 extensions, and comparison tests, and to fix some bugs.
6463 14. Changes to the "personal" filter test:
6465 (1) The test was buggy in that it was just doing the equivalent of
6466 "contains" tests on header lines. For example, if a user's address was
6467 anne@some.where, the "personal" test would incorrectly be true for
6469 To: susanne@some.where
6471 This test is now done by extracting each address from the header in turn,
6472 and checking the entire address. Other tests that are part of "personal"
6473 are now done using regular expressions (for example, to check local parts
6474 of addresses in From: header lines).
6476 (2) The list of non-personal local parts in From: addresses has been
6477 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6478 taken from the Sieve specification recommendations.
6480 (3) If the message contains any header line starting with "List-" it is
6481 treated as non-personal.
6483 (4) The test for "circular" in the Subject: header line has been removed
6484 because it now seems ill-conceived.
6486 15. Minor typos in src/EDITME comments corrected.
6488 16. Installed latest exipick from John Jetmore.
6490 17. If headers_add on a router specified a text string that was too long for
6491 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6492 of string_sprintf() is now avoided.
6494 18. $message_body_size was not set (it was always zero) when running the DATA
6495 ACL and the local_scan() function.
6497 19. For the "mail" command in an Exim filter, no default was being set for
6498 the once_repeat time, causing a random time value to be used if "once" was
6499 specified. (If the value happened to be <= 0, no repeat happened.) The
6500 default is now 0s, meaning "never repeat". The "vacation" command was OK
6501 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6502 (I found it when inspecting the code).
6504 20. There is now an overall timeout for performing a callout verification. It
6505 defaults to 4 times the callout timeout, which applies to individual SMTP
6506 commands during the callout. The overall timeout applies when there is more
6507 than one host that can be tried. The timeout is checked before trying the
6508 next host. This prevents very long delays if there are a large number of
6509 hosts and all are timing out (e.g. when the network connections are timing
6510 out). The value of the overall timeout can be changed by specifying an
6511 additional sub-option for "callout", called "maxwait". For example:
6513 verify = sender/callout=5s,maxwait=20s
6515 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6516 to the end before writing, but this should make it even safer).
6518 22. Exim was forgetting that it had advertised PIPELINING for the second and
6519 subsequent messages on an SMTP connection. It was also not resetting its
6520 memory on STARTTLS and an internal HELO.
6522 23. When Exim logs an SMTP synchronization error within a session, it now
6523 records whether PIPELINING has been advertised or not.
6525 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6526 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6527 rather than long int.
6529 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6531 26. Added the never_mail option to autoreply.
6537 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6538 crash if the getsockname() call failed; this can happen if a connection is
6539 closed very soon after it is established. The problem was simply in the
6540 order in which certain operations were done, causing Exim to try to write
6541 to the SMTP stream before it had set up the file descriptor. The bug has
6542 been fixed by making things happen in the correct order.
6548 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6549 before the connection was closed, thus losing the rejection response.
6551 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6552 some early Solaris releases, but causes trouble in current releases where
6553 socklen_t is defined.
6555 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6558 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6561 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6562 that releasing the top of it at the end releases what was used for sub-
6563 expansions (unless the block got too big). However, discard this block if
6564 the first thing is a variable or header, so that we can use its block when
6565 it is dynamic (useful for very large $message_headers, for example).
6567 6. Lookups now cache *every* query, not just the most recent. A new, separate
6568 store pool is used for this. It can be recovered when all lookup caches are
6569 flushed. Lookups now release memory at the end of their result strings.
6570 This has involved some general refactoring of the lookup sources.
6572 7. Some code has been added to the store_xxx() functions to reduce the amount
6573 of flapping under certain conditions.
6575 8. log_incoming_interface used to affect only the <= reception log lines. Now
6576 it causes the local interface and port to be added to several more SMTP log
6577 lines, for example "SMTP connection from", and rejection lines.
6579 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6581 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6583 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6584 could be overwritten at the end of the current message (or the start of a
6585 new message if it was set in a HELO ACL). The value is now preserved for
6586 the duration of the SMTP connection.
6588 12. If a transport had a headers_rewrite setting, and a matching header line
6589 contained an unqualified address, that address was qualified, even if it
6590 did not match any rewriting rules. The underlying bug was that the values
6591 of the flags that permit the existence of unqualified sender and recipient
6592 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6593 non-local messages, and by -bnq for local messages) were not being
6594 preserved with the message after it was received.
6596 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6597 "next input=" as part of the text comprising the host identity instead of
6598 the correct text. The code was using the same buffer for two different
6599 strings. However, depending on which order the printing function evaluated
6600 its arguments, the bug did not always show up. Under Linux, for example, my
6601 test suite worked just fine.
6603 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6604 doesn't work with some older versions of Perl. It has been changed to "my",
6605 which in any case is probably the better facility to use.
6607 15. A really picky compiler found some instances of statements for creating
6608 error messages that either had too many or two few arguments for the format
6611 16. The size of the buffer for calls to the DNS resolver has been increased
6612 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6613 for addresses that have a lot of PTR records. This alleviates a problem; it
6614 does not fully solve it.
6616 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6617 buffer now truncates the list and logs the incident, which is the same
6618 action as happens when Exim is looking up a host name and its aliases.
6619 Previously in this situation something unpredictable would happen;
6620 sometimes it was "internal error: store_reset failed".
6622 18. If a server dropped the connection unexpectedly when an Exim client was
6623 using GnuTLS and trying to read a response, the client delivery process
6624 crashed while trying to generate an error log message.
6626 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6627 string, for example:
6629 warn message = H1: something\nH2: something
6631 the text was added as a single header line from Exim's point of view
6632 though it ended up OK in the delivered message. However, searching for the
6633 second and subsequent header lines using $h_h2: did not work. This has been
6634 fixed. Similarly, if a system filter added multiple headers in this way,
6635 the routers could not see them.
6637 20. Expanded the error message when iplsearch is called with an invalid key to
6638 suggest using net-iplsearch in a host list.
6640 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6641 ACLs are no longer actually imposed (and a message to that effect is
6644 22. If a "gecos" field in a passwd entry contained escaped characters, in
6645 particular, if it contained a \" sequence, Exim got it wrong when building
6646 a From: or a Sender: header from that name. A second bug also caused
6647 incorrect handling when an unquoted " was present following a character
6648 that needed quoting.
6650 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6651 was not being matched caselessly.
6653 24. Arranged for all hyphens in the exim.8 source to be escaped with
6656 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6657 properly. Recipient callout cache records were still being keyed to include
6658 the sender, even when use_sender was set false. This led to far more
6659 callouts that were necessary. The sender is no longer included in the key
6660 when use_sender is false.
6662 26. Added "control = submission" modifier to ACLs.
6664 27. Added the ${base62d: operator to decode base 62 numbers.
6666 28. dnsdb lookups can now access SRV records.
6668 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6669 the configuration file.
6671 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6672 (-v) mode. This makes the output for a verbose queue run more intelligible.
6674 31. Added a use_postmaster feature to recipient callouts.
6676 32. Added the $body_zerocount variable, containing the number of binary zero
6677 bytes in the message body.
6679 33. The time of last modification of the "new" subdirectory is now used as the
6680 "mailbox time last read" when there is a quota error for a maildir
6683 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6685 35. Added +ignore_unknown as a special item in host lists.
6687 36. Code for decoding IPv6 addresses in host lists is now included, even if
6688 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6689 address was recognized as an IP address, but was then not correctly decoded
6690 into binary, causing unexpected and incorrect effects when compared with
6697 1. Very minor rewording of debugging text in manualroute to say "list of
6698 hosts" instead of "hostlist".
6700 2. If verify=header_syntax was set, and a header line with an unqualified
6701 address (no domain) and a large number of spaces between the end of the
6702 name and the colon was received, the reception process suffered a buffer
6703 overflow, and (when I tested it) crashed. This was caused by some obsolete
6704 code that should have been removed. The fix is to remove it!
6706 3. When running in the test harness, delay a bit after writing a bounce
6707 message to get a bit more predictability in the log output.
6709 4. Added a call to search_tidyup() just before forking a reception process. In
6710 theory, someone could use a lookup in the expansion of smtp_accept_max_
6711 per_host which, without the tidyup, could leave open a database connection.
6713 5. Added the variables $recipient_data and $sender_data which get set from a
6714 lookup success in an ACL "recipients" or "senders" condition, or a router
6715 "senders" option, similar to $domain_data and $local_part_data.
6717 6. Moved the writing of debug_print from before to after the "senders" test
6720 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6721 problems for message scanning, either using a data ACL, or using
6722 local_scan() because the Received: header was not generated till after they
6723 were called (in order to set the time as the time of reception completion).
6724 I have revised the way this works. The header is now generated after the
6725 body is received, but before the ACL or local_scan() are called. After they
6726 are run, the timestamp in the header is updated.
6732 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6733 before starting a queue runner without re-exec. This happened only when
6734 deliver_drop_privilege was set or when the Exim user was set to root. The
6735 effect of the bug was that timeouts during subsequent deliveries caused
6736 crashes instead of being properly handled. The handler is now left at its
6737 default (and expected) setting.
6739 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6740 message, again when deliver_drop_privilege is set or Exim is run as root.
6741 The bug described in (1) was not present in this case, but the tidying up
6742 of the other signals was missing. I have made the two cases consistent.
6744 3. The ignore_target_hosts setting on a manualroute router was being ignored
6745 for hosts that were looked up using the /MX notation.
6747 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6750 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6751 operated on the sender address. After changing the $sender_address to <>
6752 for the sender address verify, Exim was re-instated it as the original
6753 (before rewriting) address, but remembering that it had rewritten it, so it
6754 wasn't rewriting it again. This bug also had the effect of breaking the
6755 sender address verification caching when the sender address was rewritten.
6757 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6758 This has been changed so that if the ip literal address matches
6759 ignore_target_hosts, the router declines.
6761 7. Added expansion conditions match_domain, match_address, and match_local_
6762 part (NOT match_host).
6764 8. The placeholder for the Received: header didn't have a length field set.
6766 9. Added code to Exim itself and to exim_lock to test for a specific race
6767 condition that could lead to file corruption when using MBX delivery. The
6768 issue is with the lockfile that is created in /tmp. If this file is removed
6769 after a process has opened it but before that process has acquired a lock,
6770 there is the potential for a second process to recreate the file and also
6771 acquire a lock. This could lead to two Exim processes writing to the file
6772 at the same time. The added code performs the same test as UW imapd; it
6773 checks after acquiring the lock that its file descriptor still refers to
6774 the same named file.
6776 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6777 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6780 11. Added the smtp_active_hostname option. If used, this will typically be made
6781 to depend on the incoming interface address. Because $interface_address is
6782 not set up until the daemon has forked a reception process, error responses
6783 that can happen earlier (such as "too many connections") no longer contain
6786 12. If an expansion in a condition on a "warn" statement fails because a lookup
6787 defers, the "warn" statement is abandoned, and the next ACL statement is
6788 processed. Previously this caused the whole ACL to be aborted.
6790 13. Added the iplsearch lookup type.
6792 14. Added ident_timeout as a log selector.
6794 15. Added tls_certificate_verified as a log selector.
6796 16. Added a global option tls_require_ciphers (compare the smtp transport
6797 option of the same name). This controls incoming TLS connections.
6799 17. I finally figured out how to make tls_require_ciphers do a similar thing
6800 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6801 before starting the TLS session.
6803 18. Tabs are now shown as \t in -bP output.
6805 19. If the log selector return_path_on_delivery was set, Exim crashed when
6806 bouncing a message because it had too many Received: header lines.
6808 20. If two routers both had headers_remove settings, and the first one included
6809 a superfluous trailing colon, the final name in the first list and the
6810 first name in the second list were incorrectly joined into one item (with a
6811 colon in the middle).
6817 1. Added -C and -D options to the exinext utility, mainly to make it easier
6818 to include in the automated testing, but these could be helpful when
6819 multiple configurations are in use.
6821 2. The exinext utility was not formatting the output nicely when there was
6822 an alternate port involved in the retry record key, nor when there was a
6823 message id as well (for retries that were specific to a specific message
6824 and a specific host). It was also confused by IPv6 addresses, because of
6825 the additional colons they contain. I have fixed the IPv4 problem, and
6826 patched it up to do a reasonable job for IPv6.
6828 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6829 delivery, the log line now contains "pipelined" if PIPELINING was used.
6831 4. An SMTP transport process used to panic and die if the bind() call to set
6832 an explicit outgoing interface failed. This has been changed; it is now
6833 treated in the same way as a connect() failure.
6835 5. A reference to $sender_host_name in the part of a conditional expansion
6836 that was being skipped was still causing a DNS lookup. This no longer
6839 6. The def: expansion condition was not recognizing references to header lines
6840 that used bh_ and bheader_.
6842 7. Added the _cache feature to named lists.
6844 8. The code for checking quota_filecount in the appendfile transport was
6845 allowing one more file than it should have been.
6847 9. For compatibility with Sendmail, the command line option
6855 and sets the incoming protocol and host name (for trusted callers). The
6856 host name and its colon can be omitted when only the protocol is to be set.
6857 Note the Exim already has two private options, -pd and -ps, that refer to
6858 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6859 "s", but I don't think that's a major issue.
6861 10. A number of refactoring changes to the code, none of which should affect
6864 (a) The number of logging options was getting close to filling up the
6865 32-bit word that was used as a bit map. I have split them into two classes:
6866 those that are passed in the argument to log_write(), and those that are
6867 only ever tested independently outside of that function. These are now in
6868 separate 32-bit words, so there is plenty of room for expansion again.
6869 There is no change in the user interface or the logging behaviour.
6871 (b) When building, for example, log lines, the code previously used a
6872 macro that called string_cat() twice, in order to add two strings. This is
6873 not really sufficiently general. Furthermore, there was one instance where
6874 it was actually wrong because one of the argument was used twice, and in
6875 one call a function was used. (As it happened, calling the function twice
6876 did not affect the overall behaviour.) The macro has been replaced by a
6877 function that can join an arbitrary number of extra strings onto a growing
6880 (c) The code for expansion conditions now uses a table and a binary chop
6881 instead of a serial search (which was left over from when there were very
6882 few conditions). Also, it now recognizes conditions like "pam" even when
6883 the relevant support is not compiled in: a suitably worded error message is
6884 given if an attempt is made to use such a condition.
6886 11. Added ${time_interval:xxxxx}.
6888 12. A bug was causing one of the ddress fields not to be passed back correctly
6889 from remote delivery subprocesses. The field in question was not being
6890 subsequently used, so this caused to problems in practice.
6892 13. Added new log selectors queue_time and deliver_time.
6894 14. Might have fixed a bug in maildirsizefile handling that threw up
6895 "unexpected character" debug warnings, and recalculated the data
6896 unnecessarily. In any case, I expanded the warning message to give more
6899 15. Added the message "Restricted characters in address" to the statements in
6900 the default ACL that block characters like @ and % in local parts.
6902 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6903 Three changes have been made:
6905 (a) There was a serious bug; a negative response to MAIL caused the whole
6906 recipient domain to be cached as invalid, thereby blocking all messages
6907 to all local parts at the same domain, from all senders. This bug has
6908 been fixed. The domain is no longer cached after a negative response to
6909 MAIL if the sender used is not empty.
6911 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6914 (c) A new callout option, "use_sender" has been added for people who want
6915 the modified behaviour.
6921 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6924 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6925 indeed breaks things for older releases.
6927 3. Added additional logging to the case where there is a problem reading data
6928 from a filter that is running in a subprocess using a pipe, in order to
6929 try to track down a specific problem.
6931 4. Testing facility fudge: when running in the test harness and attempting
6932 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6933 getting "No route to host". Convert this to a timeout.
6935 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6938 6. Some OS don't have socklen_t but use size_t instead. This affects the
6939 fifth argument of getsockopt() amongst other things. This is now
6940 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6941 can be set for individual OS. I have set it for SunOS5, OSF1, and
6942 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6943 some earlier ones do not.
6945 7. Change 4.30/15 was not doing the test caselessly.
6947 8. The standard form for an IPv6 address literal was being rejected by address
6948 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6949 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6950 this, as well as the form without the "IPv6" on the front (but only when
6951 address literals are enabled, of course).
6953 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6955 10. Exim crashed if a message with an empty sender address specified by -f
6956 encountered a router with an errors_to setting. This could be provoked only
6957 by a command such as
6961 where an empty string was supplied; "<>" did not hit this bug.
6963 11. Installed PCRE release 4.5.
6965 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6966 remained set. It is now erased.
6968 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6969 times from message ids (which are base 36 rather than the normal 62).
6971 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6972 were being counted as actual protocol errors, and logged if the log
6973 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6974 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6975 and DATA following a set of rejected RCPTs do not count as protocol errors.
6976 In other words, Exim assumes they were pipelined, though this may not
6977 actually be the case. Of course, in all cases the client gets an
6978 appropriate error code.
6980 15. If a lookup fails in an ACL condition, a message about the failure may
6981 be available; it is used if testing the ACL cannot continue, because most
6982 such messages specify what the cause of the deferral is. However, some
6983 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6984 that caused an old message to be retained and used if a later statement
6985 caused a defer, replacing the real cause of the deferral.
6987 16. If an IP address had so many PTR records that the DNS lookup buffer
6988 was not large enough to hold them, Exim could crash while trying to process
6989 the truncated data. It now detects and logs this case.
6991 17. Further to 4.21/58, another change has been made: if (and only if) the
6992 first line of a message (the first header line) ends with CRLF, a bare LF
6993 in a subsequent header line has a space inserted after it, so as not to
6994 terminate the header.
6996 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6997 unnecessarily, used atoi() instead of strtol(), and didn't check the
6998 termination when getting file sizes from file names by regex.
7000 19. Completely re-implemented the support for maildirsize files, in the light
7001 of a number of problems with the previous contributed implementation
7002 (4.30/29). In particular:
7004 . If the quota is zero, the maildirsize file is maintained, but no quota is
7007 . If the maildir directory does not exist, it is created before any attempt
7008 to write a maildirsize file.
7010 . The quota value in the file is just a cache; if the quota is changed in
7011 the transport, the new value overrides.
7013 . A regular expression is available for excluding directories from the
7016 20. The autoreply transport checks the characters in options that define the
7017 message's headers; it allows continued headers, but it was checking with
7018 isspace() after an embedded newline instead of explicitly looking for a
7021 21. If all the "regular" hosts to which an address was routed had passed their
7022 expiry times, and had not reached their retry times, the address was
7023 bounced, even if fallback hosts were defined. Now Exim should go on to try
7026 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7027 equivalent code in the SMTP transport. Some hosts send humungous responses
7028 to HELO/EHLO, more than 1024 it seems.
7030 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7031 gives compiler warnings in some environments. I've now done it "properly",
7034 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7035 (because of problems with the built-in one) was declared to return uschar *
7036 instead of char *, causing compiler failure.
7038 25. Fixed a file descriptor leak when processing alias/forward files.
7040 26. Fixed a minor format string issue in dbfn.c.
7042 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7044 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7045 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7048 29. When a sender address is verified, it is cached, to save repeating the test
7049 when there is more than one recipient in a message. However, when the
7050 verification involves a callout, it is possible for different callout
7051 options to be set for different recipients. It is too complicated to keep
7052 track of this in the cache, so now Exim always runs a verification when a
7053 callout is required, relying on the callout cache for the optimization.
7054 The overhead is duplication of the address routing, but this should not be
7057 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7058 to be verified with callout=postmaster, and the main callout worked but the
7059 postmaster check failed, the verification correctly failed. However, if a
7060 subsequent RCPT command asked for sender verification *without* the
7061 postmaster check, incorrect caching caused this verification also to fail,
7064 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7065 it was not caching the DNS options (qualify_single, search_parents) that
7066 were used when the lookup failed. A subsequent lookup with different
7067 options therefore always gave the same answer, though there were cases
7068 where it should not have. (Example: a "domains = !$mx_any" option on a
7069 dnslookup router: the "domains" option is always processed without any
7070 widening, but the router might have qualify_single set.) Now Exim uses the
7071 cached value only when the same options are set.
7073 32. Added John Jetmore's "exipick" utility to the distribution.
7075 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7076 than a timeout (e.g. a certificate is required, and is not provided), an
7077 Exim server now closes the connection immediately. Previously it waited for
7078 the client to close - but if the client is SSL, it seems that they each
7079 wait for each other, leading to a delay before one of them times out.
7081 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7082 maintained 0.8.x compatibility because I don't think many are using it, and
7083 it is clearly obsolete.
7085 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7088 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7089 client certificate was expired. A simple patch fixes this, though I don't
7090 understand the full logic of why the verify callback is called multiple
7093 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7094 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7095 which causes problems with some clients (such as the Certicom SSL Plus
7096 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7097 disables the coutermeasure allowing Eudora to connect."
7099 38. Exim was not checking that a write() to a log file succeeded. This could
7100 lead to Bad Things if a log got too big, in particular if it hit a file
7101 size limit. Exim now panics and dies if it cannot write to a log file, just
7102 as it does if it cannot open a log file.
7104 39. Modified OS/Makefile-Linux so that it now contains
7106 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7108 The two -D definitions ensure that Exim is compiled with large file
7109 support, which makes it possible to handle log files that are bigger than
7112 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7113 instance) a domain was checked against a named list that involved a lookup,
7114 causing $domain_data to be set, then another domain was checked against the
7115 same list, then the first domain was re-checked, the value of $domain_data
7116 after the final check could be wrong. In particular, if the second check
7117 failed, it could be set empty. This bug probably also applied to
7120 41. The strip_trailing_dot option was not being applied to the address given
7121 with the -f command-line option.
7123 42. The code for reading a message's header from the spool was incrementing
7124 $received_count, but never initializing it. This meant that the value was
7125 incorrect (doubled) while delivering a message in the same process in which
7126 it was received. In the most common configuration of Exim, this never
7127 happens - a fresh exec is done - but it can happen when
7128 deliver_drop_privilege is set.
7130 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7131 it now includes up to 150 characters of the unexpected data in the log
7134 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7135 and building data strings. The size of both of these buffers was 10 000
7136 bytes - far larger than anybody would *ever* want, thought I. Needless to
7137 say, somebody hit the limit. I have increased the maximum line length to
7138 20 000 and the maximum data length of concatenated lines to 100 000. I have
7139 also fixed two bugs, because there was no checking on these buffers. Tsk,
7140 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7141 buffer is too small.
7143 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7144 lsearch lookups. Now it does.
7146 46. When parsing a route_list item in a manualroute router, a fixed-length
7147 buffer was used for the list of hosts. I made this 1024 bytes long,
7148 thinking that nobody would ever have a list of hosts that long. Wrong.
7149 Somebody had a whole pile of complicated expansion conditions, and the
7150 string was silently truncated, leading to an expansion error. It turns out
7151 that it is easier to change to an unlimited length (owing to other changes
7152 that have happened since this code was originally written) than to build
7153 structure for giving a limitation error. The length of the item that
7154 expands into the list of hosts is now unlimited.
7156 47. The lsearch lookup could not handle data where the length of text line was
7157 more than 4095 characters. Such lines were truncated, leading to shortened
7158 data being returned. It should now handle lines of any length.
7160 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7161 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7164 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7165 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7166 no longer happens in comment lines. A list of replacements is now placed
7167 at the head of all of the source files, except those whose only change is
7168 to replace PERL_COMMAND in the very first #! line.
7170 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7171 messages on the queue, with a bottom-up merge sort, using code contributed
7172 by Michael Haardt. This should make operations like -bp somewhat faster on
7173 large queues. It won't affect queue runners, except when queue_run_in_order
7176 51. Installed eximstats 1.31 in the distribution.
7178 52. Added support for SRV lookups to the dnslookup router.
7180 53. If an ACL referred to $message_body or $message_body_end, the value was not
7181 reset for any messages that followed in the same SMTP session.
7183 54. The store-handling optimization for building very long strings was not
7184 differentiating between the different store pools. I don't think this
7185 actually made any difference in practice, but I've tidied it.
7187 55. While running the routers to verify a sender address, $sender_address
7188 was still set to the sender address. This is wrong, because when routing to
7189 send a bounce to the sender, it would be empty. Therefore, I have changed
7190 it so that, while verifying a sender address, $sender_address is set to <>.
7191 (There is no change to what happens when verifying a recipient address.)
7193 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7194 target A or AAAA records (if not already returned) without resetting the
7195 qualify_single or search_parents options of the DNS resolver. These are
7196 inappropriate in this case because the targets of MX and SRV records must
7197 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7198 target that, when qualified, matched something in the local domain. These
7199 two options are now turned off when doing these lookups.
7201 57. It seems that at least some releases of Reiserfs (which does not have the
7202 concept of a fixed number of inodes) returns zero and not -1 for the
7203 number of available inodes. This interacted badly with check_spool_inodes,
7204 which assumed that -1 was the "no such thing" setting. What I have done is
7205 to check that the total number of inodes is greater than zero before doing
7206 the test of how many are available.
7208 58. When a "warn" ACL statement has a log_message modifier, the message is
7209 remembered, and not repeated. This is to avoid a lot of repetition when a
7210 message has many recipients that cause the same warning to be written.
7211 However, Exim was preserving the list of already written lines for an
7212 entire SMTP session, which doesn't seem right. The memory is now reset if a
7213 new message is started.
7215 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7216 debugging output unless log_rewrite was also set.
7218 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7219 of (int)(handle) when we know that handle contains (void *)(-1).
7221 61. The Exim daemon panic-logs an error return when it closes the incoming
7222 connection. However "connection reset by peer" seems to be common, and
7223 isn't really an error worthy of noting specially, so that particular error
7226 62. When Exim is trying to find all the local interfaces, it used to panic and
7227 die if the ioctl to get the interface flags failed. However, it seems that
7228 on at least one OS (Solaris 9) it is possible to have an interface that is
7229 included in the list of interfaces, but for which you get a failure error
7230 for this call. This happens when the interface is not "plumbed" into a
7231 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7232 failure of the "get flags" call assumes that the interface is down.
7234 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7235 makes life easier for people who are doing arithmetic on fields extracted
7236 from dates, where you often get leading zeros that should not be
7237 interpreted as octal.
7239 64. Added qualify_domain to the redirect router, to override the global
7242 65. If a pathologically long header line contained very many addresses (the
7243 report of this problem mentioned 10 000) and each of them was rewritten,
7244 Exim could use up a very large amount of memory. (It kept on making new
7245 copies of the header line as it rewrote, and never released the old ones.)
7246 At the expense of a bit more processing, the header rewriting function has
7247 been changed so that it no longer eats memory in this way.
7249 66. The generation of the Received: header has been moved from the time that a
7250 message starts to be received, to the time that it finishes. The timestamp
7251 in the Received: header should now be very close to that of the <= log
7252 line. There are two side-effects of this change:
7254 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7255 logged header lines no longer include the local Received: line, because
7256 it has not yet been created. The same applies to a copy of the message
7257 that is returned to a non-SMTP sender when a message is rejected.
7259 (b) When a filter file is tested using -bf, no additional Received: header
7260 is added to the test message. After some thought, I decided that this
7263 This change does not affect the value of $received_for. It is still set
7264 after address rewriting, but before local_scan() is called.
7266 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7268 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7269 gave an unhelpful panic error message, and a defer error. I have managed to
7270 change this behaviour so that it now rejects any supplied certificate,
7271 which seems right, as the list of acceptable certificates is empty.
7273 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7274 gave an unhelpful defer error. I have not managed to make this reject any
7275 supplied certificates, but the error message it gives is "no certificate
7276 supplied", which is not helpful.
7278 70. exigrep's output now also includes lines that are not associated with any
7279 message, but which match the given pattern. Implemented by a patch from
7280 Martin Sluka, which also tidied up the Perl a bit.
7282 71. Recipient callout verification, like sender verification, was using <> in
7283 the MAIL FROM command. This isn't really the right thing, since the actual
7284 sender may affect whether the remote host accepts the recipient or not. I
7285 have changed it to use the actual sender in the callout; this means that
7286 the cache record is now keyed on a recipient/sender pair, not just the
7287 recipient address. There doesn't seem to be a real danger of callout loops,
7288 since a callout by the remote host to check the sender would use <>.
7289 [SEE ABOVE: changed after hitting problems.]
7291 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7292 temporary errors. However, in the case of such a code being given after
7293 the end of a data transmission (i.e. after ".") Exim was failing to write
7294 a retry record for the message. (Yes, there was some broken host that was
7295 actually sending 8xx at this point.)
7297 73. An unknown lookup type in a host list could cause Exim to panic-die when
7298 the list was checked. (An example that provoked this was putting <; in the
7299 middle of a list instead of at the start.) If this happened during a DATA
7300 ACL check, a -D file could be left lying around. This kind of configuration
7301 error no longer causes Exim to die; instead it causes a defer error. The
7302 incident is still logged to the main and panic logs.
7304 74. Buglet left over from Exim 3 conversion. The message "too many messages
7305 in one connection" was written to the rejectlog but not the mainlog, except
7306 when address rewriting (yes!) was being logged.
7308 75. Added write_rejectlog option.
7310 76. When a system filter was run not as root (that is, when system_filter_user
7311 was set), the values of the $n variables were not being returned to the
7312 main process; thus, they were not subsequently available in the $sn
7315 77. Added +return_path_on_delivery log selector.
7317 78. A connection timeout was being treated differently from recipients deferred
7318 when testing hosts_max_try with a message that was older than the host's
7319 retry timeout. (The host should not be counted, thus allowing all hosts to
7320 be tried at least once before bouncing.) This may have been the cause of an
7321 occasionally reported bug whereby a message would remain on the queue
7322 longer than the retry timeout, but would be bounced if a delivery was
7323 forced. I say "may" because I never totally pinned down the problem;
7324 setting up timeout/retry tests is difficult. See also the next item.
7326 79. The ultimate address timeout was not being applied to errors that involved
7327 a combination of host plus message (for example, a timeout on a MAIL
7328 command). When an address resolved to a number of possible hosts, and they
7329 were not all tried for each delivery (e.g. because of hosts_max_try), a
7330 message could remain on the queue longer than the retry timeout.
7332 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7335 81. Fixed an obscure SMTP outgoing bug which required at least the following
7336 conditions: (a) there was another message waiting for the same server;
7337 (b) the server returned 5xx to all RCPT commands in the first message so
7338 that the message was not completed; (c) the server dropped the connection
7339 or gave a negative response to the RSET that Exim sends to abort the
7340 transaction. The observed case was a dropped connection after DATA that had
7341 been sent in pipelining mode. That is, the server had advertised PIPELINING
7342 but was not implementing it correctly. The effect of the bug was incorrect
7343 behaviour, such as trying another host, and this could lead to a crash.
7349 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7350 and daemon.c were passed as pointers to ints; they should have been
7351 pointers to socklen_t variables (which are typically unsigned ints).
7353 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7356 3. Fixed a really odd bug that affected only the testing scheme; patching a
7357 certain fixed string in the binary changed the value of another string that
7358 happened to be identical to the end of the original first string.
7360 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7361 name", it returns that address as the IP address. On some operating
7362 systems (e.g. Solaris), it also passes back the IP address string as the
7363 "host name". However, on others (e.g. Linux), it passes back an empty
7364 string. Exim wasn't checking for this, and was changing the host name to an
7365 empty string, assuming it had been canonicalized.
7367 5. Although rare, it is permitted to have more than one PTR record for a given
7368 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7369 all the names associated with an address, because they do in Solaris.
7370 However, it seems that they do not in Linux for data that comes from the
7371 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7372 I found this out when I moved to a new Linux workstation and tried to run
7373 the Exim test suite.
7375 To get round this problem I have changed the code so that it now does its
7376 own call to the DNS to look up PTR records when searching for a host name.
7377 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7378 addresses that are only in /etc/hosts are still found.
7380 This behaviour is, however, controlled by an option called host_lookup_
7381 order, which defaults to "bydns:byaddr". If people want to use the other
7382 order, or indeed, just use one or the other means of lookup, they can
7383 specify it in this variable.
7385 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7386 some operating systems, this comes back from gethostbyaddr() as an empty
7387 string, and this is what Exim used to test for. However, it seems that in
7388 other systems, "." is yielded. Exim now tests for this case too.
7390 7. The values of check_spool_space and check_log_space are now held internally
7391 as a number of kilobytes instead of an absolute number of bytes. If a
7392 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7393 kilobyte. This means that much larger values can be stored.
7395 8. Exim monitor: an attempt to get the action menu when not actually pointing
7396 at a message produces an empty menu entitled "No message selected". This
7397 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7398 no entries in it ("Shell widget menu has zero width and/or height"). So I
7399 have added a single, blank menu entry in this case.
7401 9. Added ${quote_local_part.
7403 10. MIME decoding is now applied to the contents of Subject: header lines when
7406 11. Now that a reference to $sender_host_address automatically causes a reverse
7407 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7408 host lookup before query-style lookups in lists that might use this
7409 variable. This has therefore been abolished, and the "net-" prefix is no
7410 longer necessary for query-style lookups.
7412 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7413 have been a typo for LFLAGS, so it has been changed.
7415 13. The install script calls Exim with "-C /dev/null" in order to find the
7416 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7417 to be output. However, since Exim outputs its version number before the
7418 error, it didn't break the script. It just looked ugly. I fixed this by
7419 always allowing "-C /dev/null" if the caller is root.
7421 14. Ignore overlarge ACL variable number when reading spool file - insurance
7422 against a later release with more variables having written the file.
7424 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7425 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7426 this, as well as the form without the "IPv6" on the front.
7428 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7429 OS/Makefile-Darwin file.
7431 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7432 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7434 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7435 variable. However, some OS use ints for inodes. I've added cast to long int
7436 to get rid of the compiler warning.
7438 19. I had forgotten to lock out "/../" in configuration file names when
7439 ALT_CONFIG_PREFIX was set.
7441 20. Routers used for verification do not need to specify transports. However,
7442 if such a router generated a host list, and callout was configured, Exim
7443 crashed, because it could not find a port number from the (non-existent)
7444 transport. It now assumes port 25 in this circumstance.
7446 21. Added the -t option to exigrep.
7448 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7449 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7450 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7451 them is set, LOOKUP_LSEARCH is forced.
7453 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7455 24. Added sender and host information to the "rejected by local_scan()" log
7456 line; previously there was no indication of these.
7458 25. Added .include_if_exists.
7460 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7461 Linux. It turns out that not all file systems support this. Apparently some
7462 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7463 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7464 not supported on the file descriptor, is now ignored when Exim is trying to
7465 sync a directory. This applies only to Linux.
7467 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7469 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7470 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7473 29. Added support for maildirsize files from supplied patch (modified a bit).
7475 30. The use of :fail: followed by an empty string could lead Exim to respond to
7476 sender verification failures with (e.g.):
7478 550 Verification failed for <xxx>
7479 550 Sender verify failed
7481 where the first response line was missing the '-' that indicates it is not
7482 the final line of the response.
7484 31. The loop for finding the name of the user that called Exim had a hardwired
7485 limit of 10; it now uses the value of finduser_retries, which is used for
7486 all other user lookups.
7488 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7491 33. Exim was neglecting to zero errno before one call of strtol() when
7492 expanding a string and expecting an integer value. On some systems this
7493 resulted in spurious "integer overflow" errors. Also, it was casting the
7494 result into an int without checking.
7496 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7497 did not work. The code looks as if it has *never* worked, though it appears
7498 to have been documented since at least release 1.62. I have made it work.
7500 35. The "timeout_DNS" error in retry rules, also documented since at least
7501 1.62, also never worked. As it isn't clear exactly what this means, and
7502 clearly it isn't a major issue, I have abolished the feature by treating it
7503 as "timeout", and writing a warning to the main and panic logs.
7505 36. The display of retry rules for -brt wasn't always showing the error code
7508 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7509 timeout_connect_A, timeout_connect_MX.
7511 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7512 to the empty sender.
7514 39. The daemon was not analysing the content of -oX till after it had closed
7515 stderr and disconnected from the controlling terminal. This meant that any
7516 syntax errors were only noted on the panic log, and the return code from
7517 the command was 0. By re-arranging the code a little, I've made the
7518 decoding happen first, so such errors now appear on stderr, and the return
7519 code is 1. However, the actual setting up of the sockets still happens in
7520 the disconnected process, so errors there are still only recorded on the
7523 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7524 connections (as happens on some IP stacks) was logged at start up time as
7525 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7526 it from "IPv6 and IPv4", which means that two separate sockets are being
7529 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7530 says whether AF_INET or AF_INET6 was passed as an argument.
7532 42. Exiwhat output was messed up when time zones were included in log
7535 43. Exiwhat now gives more information about the daemon's listening ports,
7536 and whether -tls-on-connect was used.
7538 44. The "port" option of the smtp transport is now expanded.
7540 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7541 silently ignored. Now an error message is written to the main and panic
7544 46. There's a new ACL modifier called "logwrite" which writes to a log file
7545 as soon as it is encountered.
7547 47. Added $local_user_uid and $local_user_gid at routing time.
7549 48. Exim crashed when trying to verify a sender address that was being
7552 49. Exim was recognizing only a space character after ".include". It now also
7553 recognizes a tab character.
7555 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7556 extracting the relevant information from the specification. The man page no
7557 longer contains scrambled data for the -d option, and I've added a section
7558 at the front about calling Exim under different names.
7560 51. Added "extra_headers" argument to the "mail" command in filter files.
7562 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7565 53. Installed eximstats 1.29.
7567 54. Added transport_filter_timeout as a generic transport option.
7569 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7570 Cc: header lines. This was required by RFC 822, but it not required by RFC
7573 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7574 incoming messages that did not have them. Now it does so only if the
7575 message originates locally, that is, if there is no associated remote host
7576 address. When Resent- header lines are present, this applies to the Resent-
7577 lines rather than the non-Resent- lines.
7579 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7580 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7582 58. Messages for configuration errors now include the name of the main
7583 configuration file - useful now that there may be more than one file in a
7584 list (.included file names were always shown).
7586 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7587 for those rare installations that do not start the daemon as root or run it
7588 setuid root. I've cut out the call to initgroups() if the daemon is not
7591 60. The Exim user and group can now be bound into the binary as text strings
7592 that are looked up at the start of Exim's processing.
7594 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7596 62. Added $mailstore_basename variable.
7598 63. Installed patch to sieve.c from Michael Haardt.
7600 64. When Exim failed to open the panic log after failing to open the main log,
7601 the original message it was trying to log was written to stderr and debug
7602 output, but if they were not available (the usual case in production), it
7603 was lost. Now it is written to syslog before the two lines that record the
7604 failures to open the logs.
7606 65. Users' Exim filters run in subprocesses under the user's uid. It is
7607 possible for a "deliver" command or an alias in a "personal" command to
7608 provoke an address rewrite. If logging of address rewriting is configured,
7609 this fails because the process is not running as root or exim. There may be
7610 a better way of dealing with this, but for the moment (because 4.30 needs
7611 to be released), I have disabled address rewrite logging when running a
7612 filter in a non-root, non-exim process.
7618 1. The buildconfig auxiliary program wasn't quoting the value set for
7619 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7620 not defined. This bug was masked in 4.22 by the effect that was fixed in
7623 2. Some messages that were rejected after a message id was allocated were
7624 shown as "incomplete" by exigrep. It no longer does this for messages that
7625 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7627 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7628 have allow_domain_literals set, the ID did not get logged in the <= line.
7629 Domain literals are now always recognized in Message-ID: header lines.
7631 4. The first argument for a ${extract expansion item is the key name or field
7632 number. Leading and trailing spaces in this item were not being ignored,
7633 causing some misleading effects.
7635 5. When deliver_drop_privilege was set, single queue runner processes started
7636 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7637 same command in the process it spins off) were not dropping privilege.
7639 6. When the daemon running as "exim" started a queue runner, it always
7640 re-executed Exim in the spun-off process. This is a waste of effort when
7641 deliver_drop_privilege is set. The new process now just calls the
7642 queue-runner function directly.
7648 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7651 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7652 ignored. Though the use of -oP was forcing the writing of a pid file, it
7653 was always written to the default place.
7655 3. If the message "no IP address found for host xxxx" is generated during
7656 incoming verification, it is now followed by identification of the incoming
7657 connection (so you can more easily find what provoked it).
7659 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7661 5. Added some features to "harden" Exim a bit more against certain attacks:
7663 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7664 be put in Local/Makefile. This is like the never_users runtime option,
7665 but it cannot be overridden. The default setting is "root".
7667 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7668 prefix string with which any file named in a -C command line option
7671 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7672 is retained for -C and -D only if the caller of Exim is root. Without
7673 it, the exim user may also use -C and -D and retain privilege.
7675 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7676 command line option is disabled.
7678 6. Macro names set by the -D option must start with an upper case letter, just
7679 like macro names defined in the configuration file.
7681 7. Added "dereference=" facility to LDAP.
7683 8. Two instances of the typo "uknown" in the source files are fixed.
7685 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7686 the Configure-Makefile script screwed up while processing it.
7688 10. Incorporated PCRE 4.4.
7690 11. The SMTP synchronization check was not operating right at the start of an
7691 SMTP session. For example, it could not catch a HELO sent before the client
7692 waited for the greeting. There is now a check for outstanding input at the
7693 point when the greeting is written. Because of the duplex, asynchronous
7694 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7695 way, but not yet received, when the check is performed.
7697 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7698 on TCP/IP sockets, because this apparently causes some broken clients to
7701 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7702 unchanged) from the Cygwin maintainer.
7704 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7705 was supported instead of testing for mbx. Effectively a typo.
7707 15. The spa authenticator server code was not checking that the input it
7708 received was valid base64.
7710 16. The debug output line for the "set" modifier in ACLs was not showing the
7711 name of the variable that was being set.
7713 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7715 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7716 lookup on its own. Something else was needed to trigger the lookup. For
7717 example, a match in host_lookup or the need for a host name in a host list.
7718 Now, if $sender_host_name is referenced and the host name has not yet been
7719 looked up, a lookup is performed. If the lookup fails, the variable remains
7720 empty, and $host_lookup_failed is set to "1".
7722 19. Added "eqi" as a case-independent comparison operator.
7724 20. The saslauthd authentication condition could segfault if neither service
7725 nor realm was specified.
7727 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7728 error message that was logged was misleading, and incoming SMTP
7729 connections were dropped. The message is now more accurate, and temporary
7730 errors are given to SMTP connections.
7732 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7733 (except RSET) with a 421 error, until QUIT is received. However, it was
7734 failing to send a response to QUIT.
7736 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7737 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7738 ACL is now run after the helo_try_verify_hosts code.
7740 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7741 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7742 case-independent, so other case variants are also recognized). Apparently
7743 some systems use these upper case variants.
7745 25. If more than two messages were waiting for the same host, and a transport
7746 filter was specified for the transport, Exim sent two messages over the
7747 same TCP/IP connection, and then failed with "socket operation on non-
7748 socket" when it tried to send the third.
7750 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7752 27. The extern definition of crypt16() in expand.c was not being excluded when
7753 the OS had its own crypt16() function.
7755 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7756 as a preferred synonym for return_size_limit, both as an option and as an
7759 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7761 30. Changed the default configuration ACL to relax the local part checking rule
7762 for addresses that are not in any local domains. For these addresses,
7763 slashes and pipe symbols are allowed within local parts, but the sequence
7764 /../ is explicitly forbidden.
7766 31. SPA server authentication was not clearing the challenge buffer before
7769 32. log_message in a "warn" ACL statement was writing to the reject log as
7770 well as to the main log, which contradicts the documentation and doesn't
7771 seem right (because no rejection is happening). So I have stopped it.
7773 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7774 However, I am unable to do any testing of this.
7776 34. Fixed an infelicity in the appendfile transport. When checking directories
7777 for a mailbox, to see if any needed to be created, it was accidentally
7778 using path names with one or more superfluous leading slashes; tracing
7779 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7781 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7782 contents are added to the log line that is written for every discarded
7783 recipient. (Previously a log_message setting was ignored.)
7785 36. The ${quote: operator now quotes the string if it is empty.
7787 37. The install script runs exim in order to find its version number. If for
7788 some reason other than non-existence or emptiness, which it checks, it
7789 could not run './exim', it was installing it with an empty version number,
7790 i.e. as "exim-". This error state is now caught, and the installation is
7793 38. An argument was missing from the function that creates an error message
7794 when Exim fails to connect to the socket for saslauthd authentication.
7795 This could cause Exim to crash, or give a corrupted message.
7797 39. Added isip, isip4, and isip6 to ${if conditions.
7799 40. The ACL variables $acl_xx are now saved with the message, and can be
7800 accessed later in routers, transports, and filters.
7802 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7803 strings in the file are not string-expanded.
7805 42. If a MAIL command specified a SIZE value that was too large to fit into an
7806 int variable, the check against message_size_limit failed. Such values are
7807 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7808 day this will have to be increased, but I don't think I want to be around
7809 when emails are that large.
7816 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7817 iconv() is not standard in FreeBSD.
7819 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7820 IPv6 enabled. The observed symptom was a segmentation fault on return from
7821 the function os_common_find_running_interfaces() in src/os.c.
7823 3. In the check_special_case() function in daemon.c I had used "errno" as an
7824 argument name, which causes warnings on some systems. This was basically a
7825 typo, since it was named "eno" in the comments!
7827 4. The code that waits for the clock to tick (at a resolution of some fraction
7828 of a second) so as to ensure message-id uniqueness was always waiting for
7829 at least one whole tick, when it could have waited for less. [This is
7830 almost certainly not relevant at current processor speeds, where it is
7831 unlikely to ever wait at all. But we try to future-proof.]
7833 5. The function that sleeps for a time interval that includes fractions of a
7834 second contained a race. It did not block SIGALRM between setting the
7835 timer, and suspending (a couple of lines later). If the interval was short
7836 and the sigsuspend() was delayed until after it had expired, the suspension
7837 never ended. On busy systems this could lead to processes getting stuck for
7840 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7841 process, before it forks any delivery processes. The open lookup caching
7842 mechanism meant that the open file or database connection was passed into
7843 the delivery process. The problem was that delivery processes always tidy
7844 up cached lookup data. This could cause a problem for the next delivery
7845 process started by the queue runner, because the external queue runner
7846 process does not know about the closure. So the next delivery process
7847 still has data in the lookup cache. In the case of a file lookup, there was
7848 no problem because closing a file descriptor in a subprocess doesn't affect
7849 the parent. However, if the lookup was caching a connection to a database,
7850 the connection was closed, and the second delivery process was likely to
7851 see errors such as "PGSQL: query failed: server closed the connection
7852 unexpectedly". The problem has been fixed by closing all cached lookups
7853 in a queue runner before running a delivery process.
7855 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7856 seem to have the "const" qualifier which it has on other OS. I've
7859 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7860 *of the same type* (client or server) with the same public name that an
7861 error should be diagnosed.
7863 9. When Exim looked up a host name for an IP address, but failed to find the
7864 original IP address when looking up the host name (a safety check), it
7865 output the message "<ip address> does not match any IP for NULL", which was
7866 confusing, to say the least. The bug was that the host name should have
7867 appeared instead of "NULL".
7869 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7870 user that is built into the binary, and the -C or -D options is used, root
7871 privilege is dropped before the configuration file is read. In addition,
7872 logging is switched to stderr instead of the normal log files. If the
7873 configuration then re-defines the Exim user, the unprivileged environment
7874 is probably not what is expected, so Exim logs a panic warning message (but
7877 However, if deliver_drop_privilege is set, the unprivileged state may well
7878 be exactly what is intended, so the warning has been cut out in that case,
7879 and Exim is allowed to try to write to its normal log files.
7885 1. smtp_return_error_details was not giving details for temporary sender
7886 or receiver verification errors.
7888 2. Diagnose a configuration error if two authenticators have the same public
7891 3. Exim used not to create the message log file for a message until the first
7892 delivery attempt. This could be confusing when incoming messages were held
7893 for policy or load reasons. The message log file is now created at the time
7894 the message is received, and an initial "Received" line is written to it.
7896 4. The automatically generated man page for command line options had a minor
7897 bug that caused no ill effects; however, a more serious problem was that
7898 the procedure for building the man page automatically didn't always
7899 operate. Consequently, release 4.20 contains an out-of-date version. This
7900 shouldn't happen again.
7902 5. When building Exim with embedded Perl support, the script that builds the
7903 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7904 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7906 6. The freeze_tell option was not being used for messages that were frozen on
7907 arrival, either by an ACL or by local_scan().
7909 7. Added the smtp_incomplete_transaction log selector.
7911 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7912 was accepting AUTH without a new EHLO.
7914 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7915 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7918 10. Logging of TCP/IP connections (when configured) now happens in the main
7919 daemon process instead of the child process, so that the TCP/IP connection
7920 count is more accurate (but it can never be perfect).
7922 11. The use of "drop" in a nested ACL was not being handled correctly in the
7923 outer ACL. Now, if condition failure induced by the nested "drop" causes
7924 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7925 or "require"), the connection is dropped.
7927 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7928 that yield "discard" can now be used with an "accept" or a "discard" verb,
7929 but an error is generated for any others (because I can't see a useful way
7930 to define what should happen).
7932 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7933 are now processed in the same way as lines in the Exim configuration file.
7934 In particular, continuation lines are supported.
7936 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7938 15. Added -ti meaning -t -i.
7940 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7941 domains, and warn by logging if others are found.
7943 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7944 returned from GIFCONF when Exim is trying to find the list of interfaces on
7945 a host. The code in os.c has been modified to copy each ifreq to an aligned
7946 structure in all cases.
7948 Also, in some cases, the returned ifreq's were being copied to a 'struct
7949 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7950 means the last couple of bytes of an IPv6 address could be chopped if the
7951 ifreq contained only a normal sockaddr (14 bytes storage).
7953 18. Named domain lists were not supported in the hosts_treat_as_local option.
7954 An entry such as +xxxx was not recognized, and was treated as a literal
7957 19. Ensure that header lines added by a DATA ACL are included in the reject log
7958 if the ACL subsequently rejects the message.
7960 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7961 MD5 (which is deprecated).
7963 21. When testing a filter file using -bf, Exim was writing a message when it
7964 took the sender from a "From " line in the message, but it was not doing so
7965 when it took $return_path from a Return-Path: header line. It now does.
7967 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7968 with a valid header line field name (a series of printing characters
7969 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7971 23. Changed "disc" in the source to "disk" to conform to the documentation and
7972 the book and for uniformity.
7974 24. Ignore Sendmail's -Ooption=value command line item.
7976 25. When execve() failed while trying to run a command in a pipe transport,
7977 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7978 could be confused with a return value of 69 from the command itself. This
7979 has been changed to 127, the value the shell returns if it is asked to run
7980 a non-existent command. The wording for the related log line suggests a
7981 non-existent command as the problem.
7983 26. If received_header_text expands to an empty string, do not add a Received:
7984 header line to the message. (Well, it adds a token one on the spool, but
7985 marks it "old" so that it doesn't get used or transmitted.)
7987 27. Installed eximstats 1.28 (addition of -nt option).
7989 28. There was no check for failure on the call to getsockname() in the daemon
7990 code. This can fail if there is a shortage of resources on the system, with
7991 ENOMEM, for example. A temporary error is now given on failure.
7993 29. Contrary to the C standard, it seems that in some environments, the
7994 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7995 program. Exim now does this explicitly; it affects the formatting of
7996 timestamps using strftime().
7998 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7999 complaints. I've now initialized all the variables, to avoid this.
8001 32. Header lines added by a system filter were not being "seen" during
8002 transport-time rewrites.
8004 33. The info_callback() function passed to OpenSSL is set up with type void
8005 (*)(SSL *, int, int), as described somewhere. However, when calling the
8006 function (actually a macro) that sets it up, the type void(*)() is
8007 expected. I've put in a cast to prevent warnings from picky compilers.
8009 34. If a DNS black list lookup found a CNAME record, but there were no A
8010 records associated with the domain it pointed at, Exim crashed.
8012 35. If a DNS black list lookup returned more than one A record, Exim ignored
8013 all but the first. It now scans all returned addresses if a particular IP
8014 value is being sought. In this situation, the contents of the
8015 $dnslist_value variable are a list of all the addresses, separated by a
8018 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8019 to accept a host name and all its aliases if the forward lookup for any of
8020 them yielded the IP address of the incoming connection. Now it accepts only
8021 those names whose forward lookup yields the correct IP address. Any other
8022 names are discarded. This closes a loophole whereby a rogue DNS
8023 administrator could create reverse DNS records to break through a
8024 wildcarded host restriction in an ACL.
8026 37. If a user filter or a system filter that ran in a subprocess used any of
8027 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8028 the wrong values were passed to the pipe command ($thisaddress had the
8029 value of $0, $0 had the value of $1, etc). This bug was introduced by
8030 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8032 38. Improved the line breaking for long SMTP error messages from ACLs.
8033 Previously, if there was no break point between 40 and 75 characters, Exim
8034 left the rest of the message alone. Two changes have been made: (a) I've
8035 reduced the minimum length to 35 characters; (b) if it can't find a break
8036 point between 35 and 75 characters, it looks ahead and uses the first one
8037 that it finds. This may give the occasional overlong line, but at least the
8038 remaining text gets split now.
8040 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8041 file descriptors might be low, and that setting 1000 would always raise it.
8042 It turns out that in some environments, the limit is already over 1000 and
8043 that lowering it causes trouble. So now Exim takes care not to decrease it.
8045 40. When delivering a message, the value of $return_path is set to $sender_
8046 address at the start of routing (routers may change the value). By an
8047 oversight, this default was not being set up when an address was tested by
8048 -bt or -bv, which affected the outcome if any router or filter referred to
8051 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8052 log only when "message" or "log_message" are set. However, if one of the
8053 conditions was an address verification, or a call to a nested ACL, the
8054 messages generated by the underlying test were being passed through. This
8055 no longer happens. The underlying message is available in $acl_verify_
8056 message for both "message" and "log_message" expansions, so it can be
8057 passed through if needed.
8059 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8060 new expansion $bh_ to give the encoded byte string without charset
8061 translation. Translation happens only if iconv() is available; HAVE_ICONV
8062 indicates this at build time. HEADERS_CHARSET gives the charset to
8063 translate to; headers_charset can change it in the configuration, and
8064 "headers charset" can change it in an individual filter file.
8066 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8067 that creates RFC 2047 encoded "words" labels them as that charset instead
8068 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8069 expansion operator; (ii) when Exim creates a From: line for a local
8070 message; (iii) when a header line is rewritten to include a "phrase" part.
8072 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8073 buggy, causing it to skip the first lines of messages whose message ID
8074 ended in 'D'. This would not have bitten before Exim release 4.14, because
8075 message IDs were unlikely to end in 'D' before then. The effect was to have
8076 incorrect size information for certain domains.
8078 45. #include "config.h" was missing at the start of the crypt16.c module. This
8079 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8082 46. If there was a timeout during a "random" callout check, Exim treated it as
8083 a failure of the random address, and carried on sending RSET and the real
8084 address. If the delay was just some slowness somewhere, the response to the
8085 original RCPT would be taken as a response to RSET and so on, causing
8086 mayhem of various kinds.
8088 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8089 when I implemented it. It didn't allow for the fact that some option values
8090 may legitimately be negative (e.g. size_addition), and it didn't even do
8091 the right test for positive values.
8093 48. Domain names in DNS records are case-independent. Exim always looks them up
8094 in lower case. Some resolvers return domain names in exactly the case they
8095 appear in the zone file, that is, they may contain uppercase letters. Not
8096 all resolvers do this - some return always lower case. Exim was treating a
8097 change of case by a resolver as a change of domain, similar to a widening
8098 of a domain abbreviation. This triggered its re-routing code and so it was
8099 trying to route what was effectively the same domain again. This normally
8100 caused routing to fail (because the router wouldn't handle the domain
8101 twice). Now Exim checks for this case specially, and just changes the
8102 casing of the domain that it ultimately uses when it transmits the message
8105 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8108 50. If a filter generated a file delivery with a non-absolute name (possible if
8109 no home directory exists for the router), the forbid_file option was not
8112 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8113 the existing equality matching.
8115 52. Exim was using ints instead of ino_t variables in some places where it was
8116 dealing with inode numbers.
8118 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8119 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8120 called TMPDIR, and if it finds it is different, it changes its value.
8122 54. The smtp_printf() function is now made available to local_scan() so
8123 additional output lines can be written before returning. There is also an
8124 smtp_fflush() function to enable the detection of a dropped connection.
8125 The variables smtp_input and smtp_batched_input are exported to
8128 55. Changed the default runtime configuration: the message "Unknown user"
8129 has been removed from the ACL, and instead placed on the localuser router,
8130 using the cannot_route_message feature. This means that any verification
8131 failures that generate their own messages won't get overridden. Similarly,
8132 the "Unrouteable address" message that was in the ACL for unverifiable
8133 relay addresses has also been removed.
8135 56. Added hosts_avoid_esmtp to the smtp transport.
8137 57. The exicyclog script was not checking for the esoteric option
8138 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8139 will work only if exicyclog is run under the appropriate euid.
8141 58. Following a discussion on the list, the rules by which Exim recognises line
8142 endings on incoming messages have been changed. The -dropcr and drop_cr
8143 options are now no-ops, retained only for backwards compatibility. The
8144 following line terminators are recognized: LF CRLF CR. However, special
8145 processing applies to CR:
8147 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8148 nor a local message in the state where . is a terminator.
8150 (ii) If a bare CR is encountered in a header line, an extra space is added
8151 after the line terminator so as not to end the header. The reasoning
8152 behind this is that bare CRs in header lines are most likely either
8153 to be mistakes, or people trying to play silly games.
8155 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8156 was being incorrectly given as 18 bytes larger than it should have been.
8157 This is a VOB (very old bug).
8159 60. This may never have affected anything current, but just in case it has:
8160 When the local host is found other than at the start of a list of hosts,
8161 the local host, those with the same MX, and any that follow, are discarded.
8162 When the list in question was part of a longer list of hosts, the following
8163 hosts (not currently being processed) were also being discarded. This no
8164 longer happens. I'm not sure if this situation could ever has previously
8167 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8170 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8171 header. This is something that is recommended in a new Internet Draft, and
8172 is something that is documented as being done by Sendmail. There are two
8173 possible values. For messages generated by the autoreply transport, Exim
8176 Auto-Submitted: auto-replied
8178 whereas for all other generated messages (e.g. bounces) it adds
8180 Auto-Submitted: auto-generated
8182 63. The "personal" condition in filters now includes a test for the
8183 Auto-Submitted: header. If it contains the string "auto-" the message it
8184 not considered personal.
8186 64. Added rcpt_include_affixes as a generic transport option.
8188 65. Added queue_only_override (default true).
8190 66. Added the syslog_duplication option.
8192 67. If what should have been the first header line of a message consisted of
8193 a space followed by a colon, Exim was mis-interpreting it as a header line.
8194 It isn't of course - it is syntactically invalid and should therefore be
8195 treated as the start of the message body. The misbehaviour could have
8196 caused a number of strange effects, including loss of data in subsequent
8197 header lines, and spool format errors.
8199 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8200 client host had authenticated. This control can now be exercised by an ACL
8201 for more flexibility.
8203 69. By default, callouts do not happen when testing with -bh. There is now a
8204 variant, -bhc, which does actually run the callout code, including
8205 consulting and updating the callout cache.
8207 70. Added support for saslauthd authentication, courtesy of Alexander
8210 71. If statvfs() failed on the spool or log directories while checking their
8211 size for availability, Exim confusingly gave the error "space shortage".
8212 Furthermore, in debugging mode it crashed with a floating point exception.
8213 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8214 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8215 serious problem, Exim now writes to the main and panic logs when this
8216 happens, with details of the failure. It then refuses to accept the
8217 incoming message, giving the message "spool directory problem" or "log
8218 directory problem" with a 421 code for SMTP messages.
8220 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8221 0, 1, and 2 exist, because some OS complain for execs without them (see
8222 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8223 descriptors. However, the code omitted to check that the open succeeded,
8224 causing mysterious errors if for some reason the permissions on /dev/null
8225 got screwed. Now Exim writes a message to the main and panic logs, and
8226 bombs out if it can't open /dev/null.
8228 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8229 interact so that it is all more flexible. It is supposed to remain
8230 backwards compatible. Also added extra_local_interfaces.
8232 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8233 to bomb out with an assertion failure - to the client this appears as a
8234 connection drop. This problem occurs in the part of the code that was taken
8235 from the Samba project. Fortunately, the assertion is in a very simple
8236 function, so I have fixed this by reproducing the function inline in the
8237 one place where it is called, and arranging for authentication to fail
8238 instead of killing the process with assert().
8240 75. The SPA client code was not working when the server requested OEM rather
8241 than Unicode encoding.
8243 76. Added code to make require_files with a specific uid setting more usable in
8244 the case where statting the file as root fails - usually a non-root-mounted
8245 NFS file system. When this happens and the failure is EACCES, Exim now
8246 forks a subprocess and does the per-uid checking as the relevant uid.
8248 77. Added process_log_path.
8250 78. If log_file_path was not explicitly set, a setting of check_log_space or
8251 check_log_inodes was ignored.
8253 79. If a space check for the spool or log partitions fails, the incident is now
8254 logged. Of course, in the latter case the data may get lost...
8256 80. Added the %p formatting code to string_format() so that it can be used to
8257 print addresses in debug_print(). Adjusted all the address printing in the
8258 debugging in store.c to use %p rather than %d.
8260 81. There was a concern that a line of code in smtp_in.c could overflow a
8261 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8262 initially expressed, the concern was not well-founded, because trailing
8263 spaces are removed early. However, if the trailing spaces were followed by
8264 a NULL, they did not get removed, so the overflow was possible. Two fixes
8267 (a) I re-wrote the offending code in a cleaner fashion.
8268 (b) If an incoming SMTP command contains a NULL character, it is rejected
8271 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8272 runs initgroups(), so that if the Exim user is in any additional groups,
8273 they will be used during message reception.
8279 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob