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)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
20 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
21 after reception to before a subsequent reception. This should
22 mean slightly faster delivery, and also confirmation of reception
25 JH/02 Move from using the pcre library to pcre2. The former is no longer
26 being developed or supported (by the original developer).
28 JH/03 Constification work in the filters module required a major version
29 bump for the local-scan API. Specifically, the "headers_charset"
30 global which is visible via the API is now const and may therefore
31 not be modified by local-scan code.
33 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
34 sendfile() didi not account for the way the ClamAV driver code called it.
36 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
37 time check was being done for every character; replace that with one
40 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
41 sent was prefixed with a length byte.
43 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
44 RFC 5321. Previously Dovecot (at least) would log errors during
47 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
48 in 4.61 without a reason logged. Binaries will be bigger, which might
49 matter on diskspace-constrained systems, but debug is easier.
51 JH/09 Fix macro-definition during "-be" expansion testing. The move to
52 write-protected store for macros had not accounted for these runtime
53 additions; fix by removing this protection for "-be" mode.
55 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
56 handing out large-numbered file descriptors, violating the usual Unix
57 assumption (and required by Posix) that the lowest possible number will be
58 allocated by the kernel when a new one is needed. In the daemon, and any
59 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
60 are not useable for FD_SET() [and hence select()] and overwrite the stack.
61 Assorted crashes happen.
63 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
64 main-section options or in a connect ACL, the value from the first ever
65 connection was never replaced for subsequent connections. Found by
68 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
69 though only once PCRE2 was introduced: the memory accounting used under
70 debug offset allocations by an int, giving a hard trap in early startup.
71 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
73 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
74 with underbars is given. The write-protection of configuration introduced
75 in 4.95 trapped when normalisation was applied to an option not needing
78 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
80 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
81 failing to close the certificates directory, every hour or any time it
84 JH/16 Debugging initiated by an ACL control now continues through into routing
85 and transport processes. Previously debugging stopped any time Exim
86 re-execs, or for processing a queued message.
88 JH/17 The "expand" debug selector now gives more detail, specifically on the
89 result of expansion operators and items.
91 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
92 bad comparison between the option value and the name of the file to
93 be included was done, and a mismatch was wrongly identified.
94 4.88 to 4.95 are affected.
96 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
98 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
99 supplied as the db file-pair basename. Previously, if a directory
100 path was given, for example via the autoreply "once" option, the DB
101 file.pag and file.dir files would be created in that directory's
104 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
105 "taint" log_selector. These were previously deprecated.
107 JH/22 Fix static address-list lookups to properly return the matched item.
108 Previously only the domain part was returned.
110 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
111 the call into OpenSSL to send a TLS Close was being repeated; this
112 resulted in the library waiting for the peer's Close. If that was never
113 sent we waited forever. Fix by tracking send calls.
115 JH/24 The ${run} expansion item now expands its command string elements after
116 splitting. Previously it was before; the new ordering makes handling
117 zero-length arguments simpler. The old ordering can be obtained by
118 appending a new option "preexpand", after a comma, to the "run".
120 JH/25 Taint-check exec arguments for transport-initiated external processes.
121 Previously, tainted values could be used. This affects "pipe", "lmtp" and
122 "queryprogram" transport, transport-filter, and ETRN commands.
123 The ${run} expansion is also affected: in "preexpand" mode no part of
124 the command line may be tainted, in default mode the executable name
127 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
128 the the facility was not passed across execs, and only the first message
129 passed over a connection could use BDAT; any further ones using DATA.
131 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
132 uses $sending_ip_address and an interface is specified.
133 Previously any use of the local address in the EHLO name disabled
134 PIPECONNECT, the common case being to use the rDNS of it.
136 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
137 resumption. Previously verify failed because no certificate status is
138 passed on the wire for the restarted session. Fix by using the recorded
139 ocsp status of the stored session for the new connection.
141 JH/29 TLS resumption: the key for session lookup in the client now includes
142 more info that a server could potentially use in configuring a TLS
143 session, avoiding oferring mismatching sessions to such a server.
144 Previously only the server IP was used.
146 JH/30 Fix string_copyn() for limit greater than actual string length.
147 Previously the copied amount was the limit, which could result in a
148 overlapping memcpy for newly allocated destination soon after a
149 source string shorter than the limit. Found/investigated by KM.
151 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
152 close; it may be needed for a subsequent connection. This caused a
153 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
155 JH/32 Fix CHUNKING for a second message on a connection when the first was
156 rejected. Previously we did not reset the chunking-offered state, and
157 erroneously rejected the BDAT command. Investigation help from
164 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
165 related applications. Previously an "H" was used where available info
166 says that "M" should be, so change to match.
168 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
169 as arguments, so an implementation trying to copy these into a local
170 buffer was taking a taint-enforcement trap. Fix by using dynamically
171 created buffers. Similar fix for radius expansion condition.
173 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
174 reasonable, eg. to count headers. Fix by using dynamically created
175 buffers rather than a local. Do similar fixes for ACL actions "dcc",
176 "log_reject_target", "malware" and "spam"; the arguments are expanded
177 so could be handling tainted values.
179 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
180 broken the (no-op) support for this sendmail command. Restore it
181 to doing nothing, silently, and returning good status.
183 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
184 record path was given (or the default used) without a leading directory
185 path, an error occurred on trying to open it. Use the transport's working
188 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
189 transport. Previously it was the name on the DNS A-record; use instead
190 the head of the CNAME chain leading there (if there is one). This seems
191 to align better with RFC 6125.
193 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
194 smtp_accept_max_per_host allocated resources which were not released
195 when the limit was exceeded. This eventually crashed the daemon. Fix
196 by adding a release action in that path.
198 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
199 expanded; previously using tainted values was rejected. Fix by using
200 dynamically-created buffers.
202 JH/09 Relax restrictions on ACL verify condition needing access to message
203 headers. Previously they were only permitted in data and non-smtp ACLs;
204 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
205 not_blind, header_sender and header_names_ascii verification.
207 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
208 Previously a macro used one argument twice; when called with the
209 argument as an expression having side-effects, incorrect operation
210 resulted. Use an inlineable function.
212 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
213 held open for a verify callout. Previously this wan not accounted for
214 and a corrupt onward SMTP conversation resulted.
216 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
217 Previously they were embedded naively in the constructed address; when
218 needed, strip the quoting and quote the entire local_part.
219 Also make the inbound_srs expansion condition handle quoting.
221 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
222 excluded, not matching the documentation.
224 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
225 was given for the sqlite_dbfile a trap resulted.
227 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
228 "name" argument resulted in a trap. There is no reason to disallow such;
229 this was a coding error.
231 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
232 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
233 spent suspended, ignoring the POSIX definition. Previously we assumed
234 it did and a constant offset from real time could be used as a correction.
235 Change to using the same clock source for the start-of-message and the
236 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
237 exists, just to get a clock slightly more aligned to reality.
239 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
240 RFC says it is optional some validators care. The missing char was not
241 intended but triggered by a line-wrap alignment. Discovery and fix by
242 Guillaume Outters, hacked on by JH.
244 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
245 name being quoted was tainted a trap would be taken. Fix by using
246 dynamically created buffers. The routine could have been called by a
247 rewrite with the "h" flag, by using the "-F" command-line option, or
248 by using a "name=" option on a control=submission ACL modifier.
250 JH/19 SPF: change the Authentication-Results expansion component to give
251 smtp.helo when the sender domain is empty. Previously it gave
254 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
255 not in 127.0.0.0/8 to help in spotting list domains taken over by a
256 domain-parking registrar.
258 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
259 Previously when a whitespace character was specified it was not inserted
260 after removing the newline.
262 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
263 the domain part of the recipient address. This overrides any tls_sni
264 option set, which was previously used.
266 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
269 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
270 is_tainted() had an off-by-one error in the overenthusiastic direction.
271 Find and fix by Gavan. Although NetBSD is not a supported platform for
272 4.94 this bug could affect other platforms.
274 PP/01 Fix default prime selection to be consistent.
275 One path used ike23 still, instead of exim.dev.20160529.3; now both
276 execution flows will use the same DH primes (currently
277 exim.dev.20160529.3).
279 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
280 option in smtp transport, to match the documentation. Previously
281 verification was not being done in some cases where it should have been.
283 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
284 than one server was defined and depending on the platform memory layout
285 details, an internal consistency trap could be hit while walking the list
288 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
289 files. The value is used by the authresults expansion item. Previously
290 if this was used in a router or transport, a crash could result.
292 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
293 have a sufficient incidence of EINTR returns from select that an
294 interaction with other operations done by the main daemon loop exposed
295 a bug in the error-handling. This was benign apart from the log
298 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
299 for consistency with delivered "=>" and failed "**" lines. While we're
300 there, handle PRX and TFO.
302 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
303 applied. This resulted, if any header-line rewrite rules were configured,
304 in a panic-log triggerable by sending a message with a long address in
305 a header. Fix by increasing the arbitrary limit to larger than a single
306 (dewrapped) 5322 header line maximum size.
308 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
309 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
310 RFC 6648 which deprecates X- options in protocols as a general practice.
311 Changeover between the implementations is handled by the mechanisms
314 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
315 a remote which uses callout/hold. Previously the local was queued.
317 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
320 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
321 log line. The asterisk is now consistently appended to the remote IP
322 (and port, if given), and will also be provided on defer and fail log
323 lines. Previously it could be placed on the local IP if that was being
324 logged, and was only provided on delivery lines.
326 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
328 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
329 authenticator client_send option. Previously the next char, after a pair
330 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
331 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
332 documentation. There is still no way to get a leading ^ immediately
333 after a NUL (ie. for the password of a PLAIN method authenticator.
335 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
336 files. For bad sizes read, delete the record and whine to paniclog.
338 JH/38 When logging an AUTH failure, as server, do not include sensitive
339 information. Previously, the credentials would be included if given
340 as part of the AUTH command line and an ACL denied authentication.
342 JH/39 Bug 2691: fix $local_part_data. When the matching list element
343 referred to a file, bad data was returned. This likely also affected
346 JH/40 The gsasl authenticator now supports caching of the salted password
347 generated by the client-side implementation. This required the addition
348 of a new variable: $auth4.
350 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
351 left undeleted; the attempt to re-create it then failed - resulting in
352 the usual "SIGHUP tp have daemon reload configuration" to not work.
353 This affected any platform not supporting "abstract" Unix-domain
354 sockets (i.e. not Linux).
356 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
357 recipients" response to RCPT in a later response, with a 250. The
358 previous coding assumed this would not happen, and under PIPELINING
359 would result in both lost and duplicate recipients for a message.
361 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
362 Previously the weighting was incorrectly applied. Similar fix for socks
363 proxies. Found and fixed by Heiko Schlichting.
365 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
366 not handle sub-lists included using the +namedlist syntax. While
367 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
368 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
371 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
372 Previously the permanent pool was used, so the sore could not be freed.
373 This meant a connection with many messages would use continually-growing
376 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
377 per-pool so as not to waste too much space. Previously a constant size
378 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
379 DOS attacks harder. The cost is wasted memory use in the larger blocks.
381 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
382 same space repeatedly, and a smaller process footprint.
384 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
385 for GnuTLS. Previously the name started "%s" which, while not a bug,
386 looked as if if might be one.
388 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
389 "continued connection") make the $proxy_* variables available. Previously
390 the information was not passed across the exec() call for subsequent
391 transport executions. This also mean that the log lines for the
392 messages can show the proxy information.
394 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
395 receive time. With modern systems the difference is significant.
396 The historical behaviour can be restored by disabling (a new) log_selector
397 "queue_time_exclusive".
399 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
400 content) were specifically tested for. Now, also cover expansions
401 resulting in ACL names and inline ACL content.
403 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
404 making it unusable in complex expressions.
406 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
407 fail with a taint-check on the spoolfile name, and leave the message
410 HS/01 Enforce absolute PID file path name.
412 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
414 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
416 PP/02 Bug 2643: Correct TLS DH constants.
417 A missing NUL termination in our code-generation tool had led to some
418 incorrect Diffie-Hellman constants in the Exim source.
419 Reported by kylon94, code-gen tool fix by Simon Arlott.
421 PP/03 Impose security length checks on various command-line options.
422 Fixes CVE-2020-SPRSS reported by Qualys.
424 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
425 better. Reported by Qualys.
427 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
428 providing a particularly obnoxious sender full name.
431 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
433 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
436 PP/08 Change default for recipients_max from unlimited to 50,000.
438 PP/09 Fix security issue with too many recipients on a message (to remove a
439 known security problem if someone does set recipients_max to unlimited,
440 or if local additions add to the recipient list).
441 Fixes CVE-2020-RCPTL reported by Qualys.
443 PP/10 Fix security issue in SMTP verb option parsing
444 Fixes CVE-2020-EXOPT reported by Qualys.
446 PP/11 Fix security issue in BDAT state confusion.
447 Ensure we reset known-good where we know we need to not be reading BDAT
448 data, as a general case fix, and move the places where we switch to BDAT
449 mode until after various protocol state checks.
450 Fixes CVE-2020-BDATA reported by Qualys.
452 HS/03 Die on "/../" in msglog file names
454 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
455 the Exim runtime user are allowed to create files.
457 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
460 QS/03 When reading the output from interpreted forward files we do not
461 pass the pipe between the parent and the interpreting process to
462 executed child processes (if any).
464 QS/04 Always die if requested from internal logging, even is logging is
467 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
468 the API; compilation noo longer completes with DMARC support included.
469 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
470 been reported on other platforms.
472 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
474 JH/56 Make the majority of info read from config files readonly, for defence-in-
475 depth against exploits. Suggestion by Qualys.
476 Not supported on Solaris 10.
478 JH/57 Fix control=fakreject for a custom message containing tainted data.
479 Previously this resulted in a log complaint, due to a re-expansion present
480 since fakereject was originally introduced.
482 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
483 modifier was given, a loop resulted.
485 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
486 pipelined SMTP command followed the BDAT LAST then it would be
487 incorrectly treated as part of the message body, causing a verification
490 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
491 require looser validation rules than those for 821-level addresses,
492 which only permit IP addresses.
498 JH/01 Avoid costly startup code when not strictly needed. This reduces time
499 for some exim process initialisations. It does mean that the logging
500 of TLS configuration problems is only done for the daemon startup.
502 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
504 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
505 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
508 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
509 DKIM signing is being done. Previously a transport_filter would always
510 disable CHUNKING, falling back to traditional DATA.
512 JH/05 Regard command-line recipients as tainted.
514 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
516 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
517 PAM library frees one of the arguments given to it, despite the
518 documentation. Therefore a plain malloc must be used.
520 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
521 on-stack buffers were used, resulting in a taint trap when DSN information
522 copied from a received message was written into the buffer.
524 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
525 the ordering of its ARC headers. This caused a crash.
527 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
528 a new record was being constructed with information from the peer, a trap
531 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
532 installation would get error messages from DMARC verify, when it hit the
533 nonexistent file indicated by the default. Distros wanting DMARC enabled
534 should both provide the file and set the option.
535 Also enforce no DMARC verification for command-line sourced messages.
537 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
538 could, depending on the platform, hang at the STARTTLS response.
540 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
541 message on a connection. Previously if one message had ARC headers and
542 the following one did not, a crash could result when adding an
543 Authentication-Results: header.
545 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
546 Exim main code and Exim-related utities. The introduction of taint
547 tracking also did many adjustments to string handling. Since then, eximon
548 frequently terminated with an assert failure.
550 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
551 check for 452 responses. This slightly helps the inefficieny of doing
552 a large alias-expansion into a recipient-limited target. The max_rcpt
553 transport option still applies (and at the current default, will override
554 the new feature). The check is done for either cause of synch, and forces
555 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
556 connection. The new facility is not tunable at this time.
558 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
559 library live data was being used, so the results became garbage. Make
560 copies while it is still usable.
562 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
563 on delivery deferred (==) and failed (**) lines (if a delivery was
564 attemtped). Previously it was only on completion (=>) lines.
566 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
567 for the expansion of the server_scram_iter and server_scram_salt options.
569 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
570 are now specifically given a NO_DATA response without hitting the system
571 resolver. The library goes on to do the now-standard TXT lookup.
572 Use of dnsdb lookups is not affected.
574 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
575 only retrieve the errormessage once. Previously two calls to dlerror()
576 were used, and the second one (for mainlog/paniclog) retrieved null
579 JH/20 Taint checking: disallow use of tainted data for
580 - the appendfile transport file and directory options
581 - the pipe transport command
582 - the autoreply transport file, log and once options
583 - file names used by the redirect router (including filter files)
585 - paths used by single-key lookups
586 Previously this was permitted.
588 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
589 adjusted the size of a major service buffer; this failed because the
590 buffer was in use at the time. Change to a compile-time increase in the
591 buffer size, when this authenticator is compiled into exim.
593 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
594 previous fast-mode was untenable in the face of glibs using mmap to
595 support larger malloc requests.
597 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
598 New values supported, if defined on system where compiled:
599 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
600 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
602 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
603 running a limited number of proceses in parallel, a benefit is gained. The
604 amount varies with the platform hardware and load. The use of the option
605 queue_run_in_order means we cannot do this, as ordering becomes
608 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
609 had introduced a string-copy (for ensuring NUL-termination) which was not
610 appropriate for that case, which can include embedded NUL bytes in the
611 block of data. Investigation showed the copy to actually be needless, the
612 data being length-specified.
614 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
615 done during a receiving connection, and both used TLS, global info was
616 used rather than per-connection info for tracking the state of data
617 queued for transmission. This could result in a connection hang.
619 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
620 Previously, when delivering serveral messages down a single connection
621 only the first would provide a SIZE. This was due to the size information
622 not being properly tracked.
624 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
625 TAI (at 37 seconds currently), pretend to be in UTC for time-related
626 expansion and logging. Previously, spurious values such as a future
627 minute could be seen.
629 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
630 it could crash from a null-deref. This could also affect the
631 ${addresses: } operator and ${readsock } item.
633 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
634 message following a mime one, the variable was not reset.
636 JH/30 When an pipelined-connect fails at the first response, assume incorrect
637 cached capability (perhaps the peer reneged?) and immediately retry in
640 JH/31 Fix spurious detection of timeout while writing to transport filter.
642 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
643 an attempt to copy the string was made before checking it.
645 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
646 taint of the lookup key was maintained; we now regard the presence in the
647 filesystem as sufficient validation.
649 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
650 argument is supplied.
652 JH/35 The dsearch lookup now requires that the directory is an absolute path.
653 Previously this was not checked, and nonempty relative paths made an
654 access under Exim's current working directory.
656 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
657 Previously no event was raised.
659 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
660 parameter supplied by the sender MAIL FROM command. Previously it was
661 ignored, and only the check_spool_space option value for the required
664 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
665 the size of the signing public-key. Previously it was instead giving
666 the size of the signature hash.
668 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
669 the default. See the (new) dkim_verify_min_keysizes option.
671 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
672 and an ACL use a lookup for checking either the local_part or domain,
673 stale data could be accessed. Ensure that variable references are
674 dropped between messages.
676 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
677 by the client was not checked as pointing within response data before
678 being used. A malicious client could thus cause an out-of-bounds read and
679 possibly gain authentication. Fix by adding the check.
681 JH/42 Internationalisation: change the default for downconversion in the smtp
682 transport to be "if needed". Previously it was "as previously set" for
683 the message, which usually meant "if needed" for message-submission but
684 "no" for everything else. However, MTAs have been seen using SMTPUTF8
685 even when the envelope addresses did not need it, resulting in forwarding
686 failures to non-supporting MTAs. A downconvert in such cases will be
687 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
688 transport. The change does mean that addresses needing conversion will
689 be converted when previously a delivery failure would occur.
691 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
692 response was received it would be used unchecked in a fail-DSN, violating
693 standards on line-length limits. Truncate if needed.
695 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
696 doesn't work. (Additionally add a new main config option to configure the
703 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
704 side, to decode a TLS 1.3 packet capture.
706 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
707 Previously the default library behaviour applied, sending two, each in
710 JH/03 Debug output for ACL now gives the config file name and line number for
713 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
715 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
717 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
718 buffer overrun for (non-chunking) other transports.
720 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
721 TLS1.3, means that a server rejecting a client certificate is not visible
722 to the client until the first read of encrypted data (typically the
723 response to EHLO). Add detection for that case and treat it as a failed
724 TLS connection attempt, so that the normal retry-in-clear can work (if
725 suitably configured).
727 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
728 and/or domain. Found and fixed by Jason Betts.
730 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
731 configuration). If a CNAME target was not a wellformed name pattern, a
734 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
735 the OS reports them interleaved with other addresses.
737 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
738 used both for input and for a verify callout, both encrypted, SMTP
739 responses being sent by the server could be lost. This resulted in
740 dropped connections and sometimes bounces generated by a peer sending
743 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
744 string. Previously it was possible to cause undefined behaviour in a
745 library routine (usually a crash). Found by "zerons".
747 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
750 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
751 API was removed, so update to use the newer ones.
753 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
754 any timeout set, is taking a long time. Previously we would hang on to a
755 rotated logfile "forever" if the input was arriving with long gaps
756 (a previous attempt to fix addressed lack, for a long time, of initial
759 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
760 shared (NFS) environment. The length of the tempfile name is now
761 4 + 16 ("hdr.$message_exim_id") which might break on file
762 systems which restrict the file name length to lower values.
765 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
766 shared (NFS) environment.
768 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
769 did for all versions <4.90). Notably -M, -m, --invert, -I may be
772 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
773 on some platforms for bit 31.
775 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
776 to changes apparently associated with TLS1.3 handling some of the APIs
777 previously used were either nonfunctional or inappropriate. Strings
778 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
779 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
780 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
781 This affects log line X= elements, the $tls_{in,out}_cipher variables,
782 and the use of specific cipher names in the encrypted= ACL condition.
784 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
786 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
787 verification result was not updated unless hosts_require_ocsp applied.
789 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
790 queue_list_requires_admin set to false, non-admin users were denied the
793 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
794 directory-of-certs mode. Previously they were advertised despite the
797 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
798 A single TCP connection by a client will now hold a TLS connection open
799 for multiple message deliveries, by default. Previously the default was to
802 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
803 default. If built with the facility, DANE will be used. The facility
804 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
806 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
807 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
808 must be defined and you must still, unless you define DISABLE_TLS, manage
809 the the include-dir and library-file requirements that go with that
810 choice. Non-TLS builds are still supported.
812 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
815 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
816 default. If the platform supports and has the facility enabled, it will
817 be requested on all coneections.
819 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
820 controlled by the build-time option SUPPORT_PIPE_CONNECT.
822 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
824 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
825 success-DSN messages. Previously the From: header was always the default
826 one for these; the option was ignored.
828 JH/28 Fix the timeout on smtp response to apply to the whole response.
829 Previously it was reset for every read, so a teergrubing peer sending
830 single bytes within the time limit could extend the connection for a
831 long time. Credit to Qualsys Security Advisory Team for the discovery.
833 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
834 delivery address, which leaked information of the results of local
835 forwarding. Change to the original envelope recipient address, per
838 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
839 requested. Previously not bounce was generated and a log entry of
840 error ignored was made.
842 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
844 JH/32 Introduce a general tainting mechanism for values read from the input
845 channel, and values derived from them. Refuse to expand any tainted
846 values, to catch one form of exploit.
848 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
849 was unused and the unexpanded text used for the test. Found and
850 fixed by Ruben Jenster.
852 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
853 an attempt to use a TLS library read routine dereffed a nul pointer,
856 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
857 them in our smtp response.
859 JH/36 Have the general router option retry_use_local_part default to true when
860 any of the restrictive preconditions are set (to anything). Previously it
861 was only for check_local user. The change removes one item of manual
862 configuration which is required for proper retries when a remote router
863 handles a subset of addresses for a domain.
865 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
866 link count into consideration.
868 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
869 caused the extension of big_buffer, the following lines were ignored.
871 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
872 accordance with RFC 2308. Previously there was no expiry, so a longlived
873 receive process (eg. due to ACL delays) versus a short SOA value could
876 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
878 JH/39 Promote DMARC support to mainline.
880 JH/40 Bug 2452: Add a References: header to DSNs.
882 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
883 parameters. The relevant library call is documented as "Deprecated: This
884 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
885 3.6.0, DH parameters are negotiated following RFC7919."
887 HS/06 Change the default of dnssec_request_domains to "*"
889 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
890 carried on and emitted a BDAT command, even when PIPELINING was not
893 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
894 buffer was used for the filename, resulting in a trap when tainted
895 arguments (eg. $domain) were used.
897 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
898 recommended to avoid a possible server-load attack. The feature can be
899 re-enabled via the openssl_options main cofiguration option.
901 JH/45 local_scan API: documented the current smtp_printf() call. This changed
902 for version 4.90 - adding a "more data" boolean to the arguments.
903 Bumped the ABI version number also, this having been missed previously;
904 release versions 4.90 to 4.92.3 inclusive were effectively broken in
905 respect of usage of smtp_printf() by either local_scan code or libraries
906 accessed via the ${dlfunc } expansion item. Both will need coding
907 adjustment for any calls to smtp_printf() to match the new function
908 signature; a FALSE value for the new argument is always safe.
910 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
911 the file-offset (which the Linux syscall does, and exim expects); this
912 resulted in an indefinite loop.
914 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
915 to do ARC verification. The Authentication-Results: header line added
916 by the configuration then had no ARC item.
918 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
919 an incorrect mode was used for reading the result, resulting in it being
926 JH/01 Remove code calling the customisable local_scan function, unless a new
927 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
929 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
930 non-signal-safe functions being used.
932 JH/03 Bug 2269: When presented with a received message having a stupidly large
933 number of DKIM-Signature headers, disable DKIM verification to avoid
934 a resource-consumption attack. The limit is set at twenty.
936 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
937 report of oldest_pass in ${authres } in consequence, and separate out
938 some descriptions of reasons for verification fail.
940 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
941 files in the spool were present and unlocked. A queue-runner could spot
942 them, resulting in a duplicate delivery. Fix that by doing the unlock
943 after the unlink. Investigation by Tim Stewart. Take the opportunity to
944 add more error-checking on spoolfile handling while that code is being
947 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
948 No known attacks, no CVE, this is defensive hardening.
950 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
951 a queue-runner could start a delivery while other operations were ongoing.
952 Cutthrough delivery was a common victim, resulting in duplicate delivery.
953 Found and investigated by Tim Stewart. Fix by using the open message data
954 file handle rather than opening another, and not locally closing it (which
955 releases a lock) for that case, while creating the temporary .eml format
956 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
958 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
959 $sender_verify_failure/$recipient_verify_failure to "random".
961 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
964 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
965 Previously this would segfault.
967 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
970 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
971 like zero, since the resolver should be doing this for us, But we need one
972 as a CNAME but no MX presence gets the CNAME returned; we need to check
973 that doesn't point to an MX to declare it "no MX returned" rather than
974 "error, loop". A new main option is added so the older capability of
975 following some limited number of chain links is maintained.
977 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
979 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
980 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
981 now-deprecated earlier definitions used only specified the range up to TLS
982 1.2 (in the older-version library docs).
984 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
986 JH/15 Rework TLS client-side context management. Stop using a global, and
987 explicitly pass a context around. This enables future use of TLS for
988 connections to service-daemons (eg. malware scanning) while a client smtp
989 connection is using TLS; with cutthrough connections this is quite likely.
991 JH/16 Fix ARC verification to do AS checks in reverse order.
993 JH/17 Support a "tls" option on the ${readsocket } expansion item.
995 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
996 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
997 Previously the "utf8" would be re-prepended for every additional message.
999 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1000 Previously thery were accepted, resulting in issues when attempting to
1001 forward messages to a non-supporting MTA.
1003 PP/02 Let -n work with printing macros too, not just options.
1005 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1006 one parent address was copied, and bogus data was used at delivery-logging
1007 time. Either a crash (after delivery) or bogus log data could result.
1008 Discovery and analysis by Tim Stewart.
1010 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1011 Previously if the string ended mid-character, we did not insert the
1012 promised '?' replacement.
1014 PP/04 Documentation: current string operators work on bytes, not codepoints.
1016 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1017 should pack well giving a smaller memory footprint so better caching and
1018 therefore performance. Group the declarations where this can't be done so
1019 that the byte-sized flag variables are not interspersed among pointer
1020 variables, giving a better chance of good packing by the compiler.
1022 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1023 non-null, to avoid issues with sites running BATV. Previously reports were
1024 sent with an empty envelope sender so looked like bounces.
1026 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1027 The ignore_error flag wasn't being returned from the filter subprocess so
1028 was not set for later routers. Investigation and fix by Matthias Kurz.
1030 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1031 and a msg:complete for the whole, when a message is manually removed using
1032 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1034 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1035 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1036 hope that is portable enough.
1038 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1039 requiring a known-CA anchor certificate; make it now rely entirely on the
1040 TLSA as an anchor. Checking the name on the leaf cert against the name
1041 on the A-record for the host is still done for TA (but not for EE mode).
1043 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1044 included in delivery lines for non-proxied connections, when compiled with
1045 SUPPORT_SOCKS and running with proxy logging enabled.
1047 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1048 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1049 move the existing event to fire before the normal logging of message
1050 failure so that custom logging is bracketed by normal logging.
1052 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1053 msg:fail:internal event. Developement by Matthias Kurz.
1055 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1056 far too small for todays use of crypto signatures stored there. Go all
1057 the way to the max DNS message size of 64kB, even though this might be
1058 overmuch for IOT constrained device use.
1060 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1061 copy a string over itself. The library routine is documented as not
1062 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1064 JH/32 For main options check_spool_space and check_inode_space, where the
1065 platform supports 64b integers, support more than the previous 2^31 kB
1066 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1067 the previous G, M, k.
1069 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1070 $authenticated_fail_id variable on authentication failure. Previously
1073 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1074 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1075 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1076 GNUTLS_SEC_PARAM_MEDIUM.
1078 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1079 side. Previously we would continue as if no SNI had been received.
1081 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1082 "<" character, which should be a list-separator specification, we walked
1083 off past the nul-terimation.
1085 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1086 causes) even when the retry time is not yet met. Previously they were
1087 not, meaning that when (say) an account was over-quota and temp-rejecting,
1088 and multiple senders' messages were queued, only one sender would get
1089 notified on each configured delay_warning cycle.
1091 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1093 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1094 cipher-suites, an error can be left on the stack even for a succeeding
1095 accept; this results in impossible error messages when a later operation
1098 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1099 return error codes indicating retry. Under TLS1.3 this becomes required.
1101 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1102 it only wrote the new authenticators, resulting in a lack of tracking of
1103 peer changes of ESMTP extensions until the next cache flush.
1105 JH/41 Fix the loop reading a message header line to check for integer overflow,
1106 and more-often against header_maxsize. Previously a crafted message could
1107 induce a crash of the recive process; now the message is cleanly rejected.
1109 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1110 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1116 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1117 When redis_servers is set to a list of > 1 element, and the Redis servers
1118 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1119 case of MOVED into a DEFER case instead, thus moving the query onto the
1120 next server in the list. For a cluster of N elements, all N servers must
1121 be defined in redis_servers.
1123 GF/02 Catch and remove uninitialized value warning in exiqsumm
1124 Check for existence of @ARGV before looking at $ARGV[0]
1126 JH/01 Replace the store_release() internal interface with store_newblock(),
1127 which internalises the check required to safely use the old one, plus
1128 the allocate and data copy operations duplicated in both (!) of the
1129 extant use locations.
1131 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1132 modifier. This matches the restriction on the commandline.
1134 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1135 Previously only the last row was returned.
1137 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1138 we assumed that tags in the header were well-formed, and parsed the
1139 element content after inspecting only the first char of the tag.
1140 Assumptions at that stage could crash the receive process on malformed
1143 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1144 While running the DKIM ACL we operate on the Permanent memory pool so that
1145 variables created with "set" persist to the DATA ACL. Also (at any time)
1146 DNS lookups that fail create cache records using the Permanent pool. But
1147 expansions release any allocations made on the current pool - so a dnsdb
1148 lookup expansion done in the DKIM ACL releases the memory used for the
1149 DNS negative-cache, and bad things result. Solution is to switch to the
1150 Main pool for expansions.
1151 While we're in that code, add checks on the DNS cache during store_reset,
1152 active in the testsuite.
1153 Problem spotted, and debugging aided, by Wolfgang Breyha.
1155 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1156 When none of the hosts presented to a transport match an already-open
1157 connection, close it and proceed with the list. Previously we would
1158 queue the message. Spotted by Lena with Yahoo, probably involving
1161 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1162 Previously a spurious "250 OK id=" response was appended to the proper
1165 JH/08 The "support for" informational output now, which built with Content
1166 Scanning support, has a line for the malware scanner interfaces compiled
1167 in. Interface can be individually included or not at build time.
1169 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1170 by the template makefile "src/EDITME". The "STREAM" support for an older
1171 ClamAV interface method is removed.
1173 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1174 rows affected is given instead).
1176 JH/11 The runtime Berkeley DB library version is now additionally output by
1177 "exim -d -bV". Previously only the compile-time version was shown.
1179 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1180 SMTP connection. Previously, when one had more recipients than the
1181 first, an abortive onward connection was made. Move to full support for
1182 multiple onward connections in sequence, handling cutthrough connection
1183 for all multi-message initiating connections.
1185 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1186 routers. Previously, a multi-recipient message would fail to match the
1187 onward-connection opened for the first recipient, and cause its closure.
1189 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1190 a timeout on read on a GnuTLS initiating connection, resulting in the
1191 initiating connection being dropped. This mattered most when the callout
1192 was marked defer_ok. Fix to keep the two timeout-detection methods
1195 JH/15 Relax results from ACL control request to enable cutthrough, in
1196 unsupported situations, from error to silently (except under debug)
1197 ignoring. This covers use with PRDR, frozen messages, queue-only and
1200 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1202 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1203 metadata, resulting in a crash in free().
1205 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1206 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1207 Broken also in d185889f4, with init system revamp.
1209 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1210 Previously we abruptly closed the connection after reading a malware-
1211 found indication; now we go on to read the "scan ok" response line,
1214 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1215 ACL. Previously, a crash would result.
1217 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1218 macro text after a replacement (previously it was only once per line) and
1219 by skipping builtin macros when searching for an uppercase lead character.
1221 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1222 for the build is renamed.
1224 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1225 was allocated for every new TLS startup, meaning one per message. Fix
1226 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1228 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1229 reported the original. Fix to report (as far as possible) the ACL
1230 result replacing the original.
1232 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1233 OpenSSL. Certificate information is loaded for every new TLS startup,
1234 and the resources needed to be freed.
1236 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1238 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1241 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1242 DATA response info to the (existing) per-recipient response info for
1243 the "C=" log element. It can have useful tracking info from the
1244 destination system. Patch from Simon Arlott.
1246 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1247 length value. Previously this would segfault.
1249 HS/02 Support Avast multiline protoocol, this allows passing flags to
1250 newer versions of the scanner.
1252 JH/28 Ensure that variables possibly set during message acceptance are marked
1253 dead before release of memory in the daemon loop. This stops complaints
1254 about them when the debug_store option is enabled. Discovered specifically
1255 for sender_rate_period, but applies to a whole set of variables.
1256 Do the same for the queue-runner and queue-list loops, for variables set
1257 from spool message files. Do the same for the SMTP per-message loop, for
1258 certain variables indirectly set in ACL operations.
1260 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1261 as a multi-recipient message from a mailinglist manager). The coding had
1262 an arbitrary cutoff number of characters while checking for more input;
1263 enforced by writing a NUL into the buffer. This corrupted long / fast
1264 input. The problem was exposed more widely when more pipelineing of SMTP
1265 responses was introduced, and one Exim system was feeding another.
1266 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1267 receiving system, and refused recipients seen by the sending system
1268 (propating to people being dropped from mailing lists).
1269 Discovered and pinpointed by David Carter.
1271 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1272 replaced by the ${authresults } expansion.
1274 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1276 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1277 allows proper process termination in container environments.
1279 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1280 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1282 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1283 and "err_perm", deprecated since 4.83 when the RFC-defined words
1284 "temperror" and "permerror" were introduced.
1286 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1287 transport-filters or DKIM-signing. The restriction was lost in the
1288 consolidation of verify-callout and delivery SMTP handling.
1289 Extend the restriction to also cover ARC-signing.
1291 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1292 in defer=pass mode supply a 450 to the initiator. Previously the message
1295 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1296 tls_require_ciphers is used as before.
1298 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1299 "pass_unscanned". Only tmpfails from the scanner are written to
1300 the paniclog, as they may require admin intervention (permission
1301 denied, license issues). Other scanner errors (like decompression
1302 bombs) do not cause a paniclog entry.
1304 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1305 Previously it was possible to log spurious information in receive log
1308 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1309 triggered odd behaviour from Outlook Express clients.
1311 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1314 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1315 since the IETF WG has not yet settled on that versus the original
1316 "bare" representation.
1318 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1319 Previously the millisecond value corrupted the output.
1320 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1321 corrupted the output.
1327 JH/01 Rework error string handling in TLS interface so that the caller in
1328 more cases is responsible for logging. This permits library-sourced
1329 string to be attached to addresses during delivery, and collapses
1330 pairs of long lines into single ones.
1332 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1333 during configuration. Wildcards are allowed and expanded.
1335 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1336 This permits better logging.
1338 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1339 do not close it down and have the child transport start it up again on
1340 the passed-on TCP connection. Instead, proxy the child (and any
1341 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1342 affected: the continued delivery log lines do not have any DNSSEC, TLS
1343 Certificate or OCSP information. TLS cipher information is still logged.
1345 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1346 identical IP addresses on different listening ports. Will also affect
1349 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1350 add noisy ifdef guards to special-case this sillyness.
1351 Patch from Bernd Kuhls.
1353 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1354 than 255 are no longer allowed.
1356 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1357 Disable the session-cache too, which might reduce our load. Since we
1358 currrectly use a new context for every connection, both as server and
1359 client, there is no benefit for these.
1360 GnuTLS appears to not support tickets server-side by default (we don't
1361 call gnutls_session_ticket_enable_server()) but client side is enabled
1362 by default on recent versions (3.1.3 +) unless the PFS priority string
1365 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1366 <https://reproducible-builds.org/specs/source-date-epoch/>.
1368 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1369 the check for any unsuccessful recipients did not notice the limit, and
1370 erroneously found still-pending ones.
1372 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1373 MSG_MORE. Only in-clear (not on TLS connections).
1375 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1376 transport-filter is involved we can buffer the headers in memory for
1377 creating the signature, and read the spool data file once for the
1378 signature and again for transmission.
1380 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1381 4.77 as the kernel support then wasn't solid, having issues in 64bit
1382 mode. Now, it's been long enough. Add support for FreeBSD also.
1384 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1385 case where the routing stage had gathered several addresses to send to
1386 a host before calling the transport for the first, we previously failed
1387 to close down TLS in the old transport process before passing the TCP
1388 connection to the new process. The new one sent a STARTTLS command
1389 which naturally failed, giving a failed delivery and bloating the retry
1390 database. Investigation and fix prototype from Wolfgang Breyha.
1392 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1393 false-negatives in the check that the sender had not preempted a response
1394 or prompt from Exim (running as a server), due to that code's lack of
1395 awareness of the SMTP input buffering.
1397 PP/04 Add commandline_checks_require_admin option.
1398 Exim drops privileges sanely, various checks such as -be aren't a
1399 security problem, as long as you trust local users with access to their
1400 own account. When invoked by services which pass untrusted data to
1401 Exim, this might be an issue. Set this option in main configuration
1402 AND make fixes to the calling application, such as using `--` to stop
1405 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1406 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1407 responses to those, into a single TLS record each way (this usually means
1408 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1411 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1412 affects you only if you're dancing at the edge of the param size limits.
1413 If you are, and this message makes sense to you, then: raise the
1414 configured limit or use OpenSSL 1.1. Nothing we can do for older
1417 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1418 cmdline element to get the documented default one. Previously it was
1421 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1422 are unsuable for non-message-specific reasons - eg. connection timeout,
1423 banner-time rejection.
1425 JH/16 Fix logging of delivery remote port, when specified by router, under
1428 PP/06 Repair manualroute's ability to take options in any order, even if one
1429 is the name of a transport.
1432 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1434 JH/17 Change the list-building routines interface to use the expanding-string
1435 triplet model, for better allocation and copying behaviour.
1437 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1438 Previously it was constructed the first time a possibly-matching string
1439 was met in the configuration file input during startup; now it is done
1442 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1443 compatible one, to avoid the (poorly documented) possibility of a config
1444 file in the working directory redirecting the DB files, possibly correpting
1445 some existing file. CVE-2017-10140 assigned for BDB.
1447 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1448 cache-hot. Previously, although the result was properly cached, the
1449 initial verify call returned a defer.
1451 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1452 the main verify for receipient in uncached-mode.
1454 JH/22 Retire historical build files to an "unsupported" subdir. These are
1455 defined as "ones for which we have no current evidence of testing".
1457 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1458 if present. Previously it was ignored.
1460 JH/24 Start using specified-initialisers in C structure init coding. This is
1461 a C99 feature (it's 2017, so now considered safe).
1463 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1464 if was a fixed-sized field and bitmask ops via macros; it is now more
1467 PP/07 GitHub PR 56: Apply MariaDB build fix.
1468 Patch provided by Jaroslav Škarvada.
1470 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1471 during Coverity cleanups [4.87 JH/47]
1472 Diagnosis and fix provided by Michael Fischer v. Mollard.
1474 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1475 the right size to place the terminating semicolon on its own folded
1476 line, the header hash was calculated to an incorrect value thanks to
1477 the (relaxed) space the fold became.
1479 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1480 and confused the parent.
1482 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1483 which could crash as a result. This could lead to undeliverable messages.
1485 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1488 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1489 matters on fast-turnover and PID-randomising systems, which were getting
1490 out-of-order delivery.
1492 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1493 a possibly-overlapping copy. The symptom was that "Remote host closed
1494 connection in response to HELO" was logged instead of the actual 4xx
1497 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1498 Previously only that bufferd was discarded, resulting in SYMTP command
1501 JH/32 DKIM: when a message has multiple signatures matching an identity given
1502 in dkim_verify_signers, run the dkim acl once for each. Previously only
1503 one run was done. Bug 2189.
1505 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1506 "panic the current process" to "deliberately defer". The panic log is
1507 still written with the problem list name; the mail and reject logs now
1508 get a temp-reject line for the message that was being handled, saying
1509 something like "domains check lookup or other defer". The SMTP 451
1510 message is still "Temporary local problem".
1512 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1513 A crafted sequence of BDAT commands could result in in-use memory beeing
1514 freed. CVE-2017-16943.
1516 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1517 from SMTP input. Previously it was always done; now only done for DATA
1518 and not BDAT commands. CVE-2017-16944.
1520 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1521 to the message (such as an overlong header line). Previously this was
1522 not done and we did not exit BDAT mode. Followon from the previous item
1523 though a different problem.
1529 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1530 than -2003 did; needs libidn2 in addition to libidn.
1532 JH/02 The path option on a pipe transport is now expanded before use.
1534 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1535 Patch provided by "Björn", documentation fix added too.
1537 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1538 missing a wire-to-host endian conversion.
1540 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1541 close after a BDAT command line could be taken as a following command,
1542 giving a synch failure. Fix by only checking for synch immediately
1543 before acknowledging the chunk.
1545 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1546 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1547 macro. Patches provided by Josh Soref.
1549 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1550 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1551 by IANA as a service extension. However, John Klensin suggests that we
1554 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1555 the dkim code may be unix-mode line endings rather than smtp wire-format
1556 CRLF, so prepend a CR to any bare LF.
1558 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1559 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1561 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1562 the first were themselves being wrongly included in the feed into dkim
1563 processing; with most chunk sizes in use this resulted in an incorrect
1564 body hash calculated value.
1566 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1567 DKIM signature block, for verification. Although advised against by
1568 standards it is specifically not ruled illegal.
1570 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1572 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1573 missing a body hash (the bh= tag).
1575 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1576 It seems that HAProxy sends the Proxy Protocol information in clear and
1577 only then does a TLS startup, so do the same.
1579 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1580 TCP connections (such as for Spamd) unless the daemon successfully set
1581 Fast Open mode on its listening sockets. This fixes breakage seen on
1582 too-old kernels or those not configured for Fast Open, at the cost of
1583 requiring both directions being enabled for TFO, and TFO never being used
1584 by non-daemon-related Exim processes.
1586 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1587 endings, at least on the first header line. Try to canonify any that get
1588 past that check, despite the cost.
1590 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1591 now limited to an arbitrary five deep, while parsing addresses with the
1592 strip_excess_angle_brackets option enabled.
1594 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1595 instead leave the unprompted TLS handshake in socket buffer for the
1596 TLS library to consume.
1598 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1600 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1602 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1603 region in one operation, for speed, and this leaves assigned pointers
1604 dangling. Add checks run only under the testsuite which checks all
1605 variables at a store-reset and panics on a dangling pointer; add code
1606 explicitly nulling out all the variables discovered. Fixes one known
1607 bug: a transport crash, where a dangling pointer for $sending_ip_address
1608 originally assigned in a verify callout, is re-used.
1610 PP/06 Drop '.' from @INC in various Perl scripts.
1612 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1614 PP/08 Reduce a number of compilation warnings under clang; building with
1615 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1616 should be warning-free.
1618 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1620 HS/01 Fix portability problems introduced by PP/08 for platforms where
1621 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1623 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1624 chunk. This allows us to accept broken chunked messages. We need a more
1625 general solution here.
1627 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1628 already-broken messages in the queue.
1630 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1632 JH/19 Fix reference counting bug in routing-generated-address tracking.
1638 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1639 supports it and a size is available (ie. the sending peer gave us one).
1641 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1642 years of being deprecated). The replacements are the ACLs
1643 acl_smtp_mime and acl_not_smtp_mime.
1645 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1646 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1647 or even an in-clear connection were permitted. Now, if the host lookup
1648 was dnssec and dane was requested then the host is only used if the
1649 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1650 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1651 if one fails this test.
1652 This means that a poorly-configured remote DNS will make it incommunicado;
1653 but it protects against a DNS-interception attack on it.
1655 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1656 non-noisy when a race steals the message being considered.
1658 JH/05 If main configuration option tls_certificate is unset, generate a
1659 self-signed certificate for inbound TLS connections.
1661 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1662 in rewrites and routers.
1664 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1665 and logged a warning sing 4.83; now they are a configuration file error.
1667 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1668 (lacking @domain). Apply the same qualification processing as RCPT.
1670 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1672 JH/10 Support ${sha256:} applied to a string (as well as the previous
1675 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1676 a cutthrough deliver is pending, as we always want to make a connection.
1677 This also avoids re-routing the message when later placing the cutthrough
1678 connection after a verify cache hit.
1679 Do not update it with the verify result either.
1681 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1682 when routing results in more than one destination address.
1684 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1685 signing (which inhibits the cutthrough capability). Previously only
1686 the presence of an option was tested; now an expansion evaluating as
1687 empty is permissible (obviously it should depend only on data available
1688 when the cutthrough connection is made).
1690 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1691 the relevant preceding SMTP command did not note the pipelining mode.
1693 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1694 Previously they were not counted.
1696 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1697 as one having no matching records. Previously we deferred the message
1698 that needed the lookup.
1700 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1701 distinguished as "(=".
1703 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1704 for missing MX records. Previously it only worked for missing A records.
1706 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1708 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1709 after the data-go-ahead and data-ack. Patch from Jason Betts.
1711 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1712 even for a "none" policy. Patch from Tony Meyer.
1714 JH/22 Fix continued use of a connection for further deliveries. If a port was
1715 specified by a router, it must also match for the delivery to be
1718 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1719 When one of the recipients of a message was unsuitable for the connection
1720 (has no matching addresses), we lost track of needing to mark it
1721 deferred. As a result mail would be lost.
1723 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1725 JH/25 Decoding ACL controls is now done using a binary search; the source code
1726 takes up less space and should be simpler to maintain. Merge the ACL
1727 condition decode tables also, with similar effect.
1729 JH/26 Fix problem with one_time used on a redirect router which returned the
1730 parent address unchanged. A retry would see the parent address marked as
1731 delivered, so not attempt the (identical) child. As a result mail would
1734 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1735 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1736 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1739 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1740 Main config options check_{log,spool}_{inodes,space} are now
1741 100 inodes, 10MB unless set otherwise in the configuration.
1743 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1744 Previously it only applied to the main-section connection policy
1747 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1749 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1750 by me. Added RFC7919 DH primes as an alternative.
1752 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1753 are not in the system include path.
1755 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1756 GnuTLS, when a session startup failed (eg because the client disconnected)
1757 Exim did stdio operations after fclose. This was exposed by a recent
1758 change which nulled out the file handle after the fclose.
1760 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1761 signed directly by the cert-signing cert, rather than an intermediate
1762 OCSP-signing cert. This is the model used by LetsEncrypt.
1764 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1766 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1767 an incoming connection.
1769 HS/02 Bug 1802: Do not half-close the connection after sending a request
1772 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1773 fallback to "prime256v1".
1775 JH/34 SECURITY: Use proper copy of DATA command in error message.
1776 Could leak key material. Remotely exploitable. CVE-2016-9963.
1782 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1783 and 3.4.4 - once the server is enabled to respond to an OCSP request
1784 it does even when not requested, resulting in a stapling non-aware
1785 client dropping the TLS connection.
1787 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1788 support variable-length bit vectors. No functional change.
1790 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1791 The I= interface field on outgoing lines is now after the H= remote
1792 host field, same as incoming lines. There is a separate
1793 outgoing_interface log selector which allows you to disable the
1796 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1797 If not running log_selector +smtp_connection the mainlog would be held
1798 open indefinitely after a "too many connections" event, including to a
1799 deleted file after a log rotate. Leave the per net connection logging
1800 leaving it open for efficiency as that will be quickly detected by the
1801 check on the next write.
1803 HS/01 Bug 1671: Fix post transport crash.
1804 Processing the wait-<transport> messages could crash the delivery
1805 process if the message IDs didn't exist for some reason. When
1806 using 'split_spool_directory=yes' the construction of the spool
1807 file name failed already, exposing the same netto behaviour.
1809 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1810 mime_regex ACL conditions.
1812 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1813 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1814 to HELO, local diagnostic string.
1816 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1817 log line to debug. Even when configured with a tls authenticator many
1818 client connections are expected to not authenticate in this way, so
1819 an authenticate fail is not an error.
1821 HS/02 Add the Exim version string to the process info. This way exiwhat
1822 gives some more detail about the running daemon.
1824 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1825 matter for fast-change records such as DNSBLs.
1827 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1828 retry record, even if constant. There may be multiple transports with
1829 different interface settings and the retry behaviour needs to be kept
1832 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1834 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1836 JH/10 Bug 840: fix log_defer_output option of pipe transport
1838 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1839 hosts_randomize. This matters a lot when combined with mua_wrapper.
1841 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1842 ${quote_pgsql:<string>} operator.
1844 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1846 JH/14 Change header file name for freeradius-client. Relevant if compiling
1847 with Radius support; from the Gentoo tree and checked under Fedora.
1849 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1851 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1852 option transport_filter, meaning no filtering.
1854 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1856 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1857 defaults to "*" (all hosts). The variable is now available when not built
1858 with TLS, default unset, mainly to enable keeping the testsuite sane.
1859 If a server certificate is not supplied (via tls_certificate) an error is
1860 logged, and clients will find TLS connections fail on startup. Presumably
1861 they will retry in-clear.
1862 Packagers of Exim are strongly encouraged to create a server certificate
1863 at installation time.
1865 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1866 with the $config_file variable.
1868 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1869 in transport context, after the attempt, and per-recipient. The latter type
1870 is per host attempted. The event data is the error message, and the errno
1871 information encodes the lookup type (A vs. MX) used for the (first) host,
1872 and the trailing two digits of the smtp 4xx response.
1874 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1875 to write to mainlog (or rejectlog, paniclog) in the window between file
1876 creation and permissions/ownership being changed. Particularly affects
1877 installations where exicyclog is run as root, rather than exim user;
1878 result is that the running daemon panics and dies.
1880 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1882 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1883 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1884 "pri" and "weight". Note that the previous implicit priority given by the
1885 list order is no longer honoured.
1887 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1888 for DKIM processing.
1890 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1891 by defining SUPPORT_SOCKS.
1893 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1894 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1895 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1896 variables are renamed to proxy_{local,external}_{address,port}.
1898 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1899 for a build by defining SUPPORT_I18N
1901 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1902 of the query string, and make ${quote_redis:} do that quoting.
1904 JH/29 Move Events support from Experimental to mainline, enabled by default
1905 and removable for a build by defining DISABLE_EVENT.
1907 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1909 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1910 cached by the daemon.
1912 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1913 by defining LOOKUP_REDIS. The libhiredis library is required.
1915 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1916 keys are given for lookup.
1918 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1919 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1920 only supported when built with TLS support. The PolarSSL SHA routines
1921 are still used when the TLS library is too old for convenient support.
1923 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1924 openssl_options), for security. OpenSSL forces this from version 1.1.0
1925 server-side so match that on older versions.
1927 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1928 allocation for $value could be released as the expansion processing
1929 concluded, but leaving the global pointer active for it.
1931 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1932 and to use the domains and local_parts ACL conditions.
1934 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1935 incorrectly not doubled on cutthrough transmission, hence seen as a
1936 body-termination at the receiving system - resulting in truncated mails.
1937 Commonly the sender saw a TCP-level error, and retransmitted the message
1938 via the normal store-and-forward channel. This could result in duplicates
1939 received - but deduplicating mailstores were liable to retain only the
1940 initial truncated version.
1942 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1944 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1946 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1947 we're in there, support oversigning also; bug 1309.
1949 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1951 HS/04 Add support for keep_environment and add_environment options.
1953 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1954 either intentional arithmetic overflow during PRNG, or testing config-
1957 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1958 delivery resulted in actual delivery. Cancel cutthrough before DATA
1961 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1962 rejects a recipient: pass the reject to the originator.
1964 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1965 Many were false-positives and ignorable, but it's worth fixing the
1968 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1969 for the new environment-manipulation done at startup. Move the routines
1970 from being local to tls.c to being global via the os.c file.
1972 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1973 an extract embedded as result-arg for a map, the first arg for extract
1974 is unavailable so we cannot tell if this is a numbered or keyed
1975 extraction. Accept either.
1981 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1984 JH/02 The smtp transport option "multi_domain" is now expanded.
1986 JH/03 The smtp transport now requests PRDR by default, if the server offers
1989 JH/04 Certificate name checking on server certificates, when exim is a client,
1990 is now done by default. The transport option tls_verify_cert_hostnames
1991 can be used to disable this per-host. The build option
1992 EXPERIMENTAL_CERTNAMES is withdrawn.
1994 JH/05 The value of the tls_verify_certificates smtp transport and main options
1995 default to the word "system" to access the system default CA bundle.
1996 For GnuTLS, only version 3.0.20 or later.
1998 JH/06 Verification of the server certificate for a TLS connection is now tried
1999 (but not required) by default. The verification status is now logged by
2000 default, for both outbound TLS and client-certificate supplying inbound
2003 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2006 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2007 Status Notification (bounce) messages are now MIME format per RFC 3464.
2008 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2009 under the control of the dsn_advertise_hosts option, and routers may
2010 have a dsn_lasthop option.
2012 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2013 default, modifiable by a malware= option. The list separator for
2014 the options can now be changed in the usual way. Bug 68.
2016 JH/10 The smtp_receive_timeout main option is now expanded before use.
2018 JH/11 The incoming_interface log option now also enables logging of the
2019 local interface on delivery outgoing connections.
2021 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2022 if the interface and destination host and port all match.
2024 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2027 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2028 Patch from Andrew Lewis.
2030 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2031 now supports optional time-restrictions, weighting, and priority
2032 modifiers per server. Patch originally by <rommer@active.by>.
2034 JH/16 The spamd_address main option now supports a mixed list of local
2035 and remote servers. Remote servers can be IPv6 addresses, and
2036 specify a port-range.
2038 JH/17 Bug 68: The spamd_address main option now supports an optional
2039 timeout value per server.
2041 JH/18 Bug 1581: Router and transport options headers_add/remove can
2042 now have the list separator specified.
2044 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2047 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2050 JH/21 Support for the A6 type of dns record is withdrawn.
2052 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2053 rather than the verbs used.
2055 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2056 from 255 to 1024 chars.
2058 JH/24 Verification callouts now attempt to use TLS by default.
2060 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2061 are generic router options now. The defaults didn't change.
2063 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2064 Original patch from Alexander Shikoff, worked over by JH.
2066 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2067 files if ZCAT_COMMAND is not executable.
2069 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2071 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2073 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2074 Normally benign, it bites when the pair was led to by a CNAME;
2075 modern usage is to not canonicalize the domain to a CNAME target
2076 (and we were inconsistent anyway for A-only vs AAAA+A).
2078 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2080 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2081 when evaluating $sender_host_dnssec.
2083 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2084 $sender_helo_dnssec variable.
2086 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2088 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2090 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2092 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2093 documented as working, but never had. Support all but $spam_report.
2095 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2096 added for tls authenticator.
2098 HS/03 Add perl_taintmode main config option
2104 TL/01 When running the test suite, the README says that variables such as
2105 no_msglog_check are global and can be placed anywhere in a specific
2106 test's script, however it was observed that placement needed to be near
2107 the beginning for it to behave that way. Changed the runtest perl
2108 script to read through the entire script once to detect and set these
2109 variables, reset to the beginning of the script, and then run through
2110 the script parsing/test process like normal.
2112 TL/02 The BSD's have an arc4random API. One of the functions to induce
2113 adding randomness was arc4random_stir(), but it has been removed in
2114 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2115 function when detected.
2117 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2118 cause callback expansion.
2120 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2121 syntax errors in an expansion can be treated as a string instead of
2122 logging or causing an error, due to the internal use of bool_lax
2123 instead of bool when processing it.
2125 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2126 server certificates when making smtp deliveries.
2128 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2130 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2132 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2134 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2135 Merged patch from Sebastian Wiedenroth.
2137 JH/05 Fix results-pipe from transport process. Several recipients, combined
2138 with certificate use, exposed issues where response data items split
2139 over buffer boundaries were not parsed properly. This eventually
2140 resulted in duplicates being sent. This issue only became common enough
2141 to notice due to the introduction of connection certificate information,
2142 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2144 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2145 size buffer was used, resulting in syntax errors when an expansion
2148 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2149 version 3.3.6 or later.
2151 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2152 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2153 both become "event_action", the variables become $event_name, $event_data
2154 and $event_defer_errno. There is a new variable $verify_mode, usable in
2155 routers, transports and related events. The tls:cert event is now also
2156 raised for inbound connections, if the main configuration event_action
2159 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2160 early OCSP support, but no stapling (appears to be less than 1.0.0).
2162 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2163 server certificate names available under the smtp transport option
2164 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2167 JH/10 Time-related extraction expansions from certificates now use the main
2168 option "timezone" setting for output formatting, and are consistent
2169 between OpenSSL and GnuTLS compilations. Bug 1541.
2171 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2172 encoded parameter in the incoming message. Bug 1558.
2174 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2175 include certificate info, eximon was claiming there were spoolfile
2178 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2180 JH/14 Log delivery-related information more consistently, using the sequence
2181 "H=<name> [<ip>]" wherever possible.
2183 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2184 are problematic for Debian distribution, omit them from the release
2187 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2189 JH/16 Fix string representation of time values on 64bit time_t architectures.
2192 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2193 output list separator was used.
2198 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2199 checkers that were complaining about end of non-void function with no
2202 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2203 This was a regression introduced in 4.83 by another bugfix.
2205 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2207 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2208 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2214 TF/01 Correctly close the server side of TLS when forking for delivery.
2216 When a message was received over SMTP with TLS, Exim failed to clear up
2217 the incoming connection properly after forking off the child process to
2218 deliver the message. In some situations the subsequent outgoing
2219 delivery connection happened to have the same fd number as the incoming
2220 connection previously had. Exim would try to use TLS and fail, logging
2221 a "Bad file descriptor" error.
2223 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2224 utilities have not been installed.
2226 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2227 temporary space as the ACL may create new global variables.
2229 TL/01 LDAP support uses per connection or global context settings, depending
2230 upon the detected version of the libraries at build time.
2232 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2233 to extract and use the src ip:port in logging and expansions as if it
2234 were a direct connection from the outside internet. PPv2 support was
2235 updated based on HAProxy spec change in May 2014.
2237 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2239 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2240 Properly escape header and check for NULL return.
2242 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2245 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2247 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2248 characters in header names, implemented as a verify condition.
2249 Contributed by Michael Fischer v. Mollard.
2251 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2252 results permerror and temperror. Previous values are deprecated but
2253 still accepted. In a future release, err_perm and err_temp will be
2254 completely removed, which will be a backward incompatibility if the
2255 ACL tests for either of these two old results. Patch contributed by
2256 user bes-internal on the mailing list.
2258 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2260 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2261 selectors, in both main and reject logs.
2263 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2266 JH/07 Add malware type "sock" for talking to simple daemon.
2268 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2270 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2271 routers/transports under cutthrough routing.
2273 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2274 numbers. Touch up "bool" conditional to keep the same definition.
2276 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2278 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2280 JH/12 Expand items in router/transport headers_add or headers_remove lists
2281 individually rather than the list as a whole. Bug 1452.
2283 Required for reasonable handling of multiple headers_ options when
2284 they may be empty; requires that headers_remove items with embedded
2285 colons must have them doubled (or the list-separator changed).
2287 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2288 view the policy declared in the DMARC record. Currently, $dmarc_status
2289 is a combined value of both the record presence and the result of the
2292 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2294 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2295 dnslookup router and the smtp transport (applying to the forward
2298 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2299 of ldap servers used for a specific lookup. Patch provided by Heiko
2302 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2303 New variable $lookup_dnssec_authenticated for observability.
2305 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2306 Patch submitted by Lars Timman.
2308 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2310 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2311 Requires trusted mode and valid format message id, aborts otherwise.
2312 Patch contributed by Heiko Schlichting.
2314 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2315 certextract with support for various fields. Bug 1358.
2317 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2318 is requested by default, modifiable by smtp transport option
2321 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2322 operate on certificate variables to give certificate fingerprints
2323 Also new ${sha256:cert_variable}.
2325 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2327 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2328 Christian Aistleitner.
2330 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2332 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2333 file. Patch from Wolfgang Breyha.
2335 JH/25 Expand the coverage of the delivery $host and $host_address to
2336 client authenticators run in verify callout. Bug 1476.
2338 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2339 align with daemon_smtp_ports. Bug 72.
2341 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2342 support and error reporting did not work properly.
2344 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2345 and is readable. Patch from Andrew Colin Kissa.
2347 TL/14 Enhance documentation of ${run expansion and how it parses the
2348 commandline after expansion, particularly in the case when an
2349 unquoted variable expansion results in an empty value.
2351 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2353 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2354 by renaming away from it. Addresses GH issue 15, reported by
2357 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2359 TL/15 SECURITY: prevent double expansion in math comparison functions
2360 (can expand unsanitized data). Not remotely exploitable.
2367 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2369 PP/02 Make -n do something, by making it not do something.
2370 When combined with -bP, the name of an option is not output.
2372 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2375 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2376 $sender_host_name and config options to manage this, and basic check
2379 PP/05 DSCP support for outbound connections and control modifier for inbound.
2381 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2382 (Only plugin which currently uses this is kerberos4, which nobody should
2383 be using, but we should make it available and other future plugins might
2384 conceivably use it, even though it would break NAT; stuff *should* be
2385 using channel bindings instead).
2387 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2388 name; added for Sendmail compatibility; requires admin caller.
2389 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2390 just ignore it); requires trusted caller.
2391 Also parse but ignore: -Ac -Am -X<logfile>
2394 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2396 TL/02 Add +smtp_confirmation as a default logging option.
2398 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2399 Patch by Magnus Holmgren from 2007-02-20.
2401 TL/04 Bugzilla 1281 - Spec typo.
2402 Bugzilla 1283 - Spec typo.
2403 Bugzilla 1290 - Spec grammar fixes.
2405 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2407 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2409 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2410 mailing list by by Dmitry Isaikin.
2412 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2414 JH/02 Support "G" suffix to numbers in ${if comparisons.
2416 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2418 NM/01 Bugzilla 1197 - Spec typo
2419 Bugzilla 1196 - Spec examples corrections
2421 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2423 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2424 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2427 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2428 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2430 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2431 "acl {{name}{arg}...}", and optional args on acl condition
2434 JH/05 Permit multiple router/transport headers_add/remove lines.
2436 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2438 JH/07 Avoid using a waiting database for a single-message-only transport.
2439 Performance patch from Paul Fisher. Bugzilla 1262.
2441 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2444 JH/09 Add $headers_added variable, with content from use of ACL modifier
2445 add_header (but not yet added to the message). Bugzilla 199.
2447 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2448 Pulled from Bugzilla 817 by Wolfgang Breyha.
2450 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2452 (nb: this is the same fix as in Exim 4.80.1)
2454 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2457 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2460 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2462 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2463 advertise SMTP AUTH mechanism to us, instead of a generic
2464 protocol violation error. Also, make Exim more robust to bad
2465 data from the Dovecot auth socket.
2467 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2469 When a queue runner is handling a message, Exim first routes the
2470 recipient addresses, during which it prunes them based on the retry
2471 hints database. After that it attempts to deliver the message to
2472 any remaining recipients. It then updates the hints database using
2475 So if a recipient address works intermittently, it can get repeatedly
2476 deferred at routing time. The retry hints record remains fresh so the
2477 address never reaches the final cutoff time.
2479 This is a fairly common occurrence when a user is bumping up against
2480 their storage quota. Exim had some logic in its local delivery code
2481 to deal with this. However it did not apply to per-recipient defers
2482 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2484 This change adds a proper retry rule check during routing so that the
2485 final cutoff time is checked against the message's age. We only do
2486 this check if there is an address retry record and there is not a
2487 domain retry record; this implies that previous attempts to handle
2488 the address had the retry_use_local_parts option turned on. We use
2489 this as an approximation for the destination being like a local
2490 delivery, as in LMTP.
2492 I suspect this new check makes the old local delivery cutoff check
2493 redundant, but I have not verified this so I left the code in place.
2495 TF/02 Correct gecos expansion when From: is a prefix of the username.
2497 Test 0254 submits a message to Exim with the header
2501 When I ran the test suite under the user fanf2, Exim expanded
2502 the header to contain my full name, whereas it should have added
2503 a Resent-Sender: header. It erroneously treats any prefix of the
2504 username as equal to the username.
2506 This change corrects that bug.
2508 GF/01 DCC debug and logging tidyup
2509 Error conditions log to paniclog rather than rejectlog.
2510 Debug lines prefixed by "DCC: " to remove any ambiguity.
2512 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2514 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2515 Bug spotted by Jeremy Harris; was flawed since initial commit.
2516 Would have resulted in OCSP responses post-SNI triggering an Exim
2517 NULL dereference and crash.
2519 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2521 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2522 Bug detection, analysis and fix by Samuel Thibault.
2523 Bugzilla 1331, Debian bug #698092.
2525 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2527 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2528 Server implementation by Todd Lyons, client by JH.
2529 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2530 config variable "prdr_enable" controls whether the server
2531 advertises the facility. If the client requests PRDR a new
2532 acl_data_smtp_prdr ACL is called once for each recipient, after
2533 the body content is received and before the acl_smtp_data ACL.
2534 The client is controlled by both of: a hosts_try_prdr option
2535 on the smtp transport, and the server advertisement.
2536 Default client logging of deliveries and rejections involving
2537 PRDR are flagged with the string "PRDR".
2539 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2540 fclose(). Diagnosis by Todd Lyons.
2542 PP/17 Update configure.default to handle IPv6 localhost better.
2543 Patch by Alain Williams (plus minor tweaks).
2546 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2547 This is now consistent with GnuTLS, and is now documented: the
2548 previous undocumented portable approach to treating the option as
2549 unset was to force an expansion failure. That still works, and
2550 an empty string is now equivalent.
2552 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2553 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2554 not performing validation itself.
2556 PP/20 Added force_command boolean option to pipe transport.
2557 Patch from Nick Koston, of cPanel Inc.
2559 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2562 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2564 PP/21 Fix eximon continuous updating with timestamped log-files.
2565 Broken in a format-string cleanup in 4.80, missed when I repaired the
2566 other false fix of the same issue.
2567 Report and fix from Heiko Schlichting.
2570 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2571 Report from Prashanth Katuri.
2573 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2574 It's SecureTransport, so affects any MacOS clients which use the
2575 system-integrated TLS libraries, including email clients.
2577 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2578 using a MIME ACL for non-SMTP local injection.
2579 Report and assistance in diagnosis by Warren Baker.
2581 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2583 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2585 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2586 last id that failed so it may be referenced in subsequent ACL's.
2588 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2591 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2592 ldap library initialization, allowing self-signed CA's to be
2593 used. Also properly sets require_cert option later in code by
2594 using NULL (global ldap config) instead of ldap handle (per
2595 session). Bug diagnosis and testing by alxgomz.
2597 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2598 the src/util/ subdirectory.
2600 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2601 renamed to Transport Post Delivery Action by Jeremy Harris, as
2604 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2605 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2606 redis_servers = needs to be configured which will be used by the redis
2607 lookup. Patch from Warren Baker, of The Packet Hub.
2609 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2611 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2612 hostname or reverse DNS when processing a host list. Used suggestions
2613 from multiple comments on this bug.
2615 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2617 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2618 Missed a few lines, added it to make the runtest require no keyboard
2621 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2622 contains upper case chars. Make router use caseful_local_part.
2624 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2625 support when GnuTLS has been built with p11-kit.
2631 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2633 This, or similar/improved, will also be change PP/11 of 4.82.
2639 PP/01 Handle short writes when writing local log-files.
2640 In practice, only affects FreeBSD (8 onwards).
2641 Bugzilla 1053, with thanks to Dmitry Isaikin.
2643 NM/01 Bugzilla 949 - Documentation tweak
2645 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2648 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2650 PP/02 Implemented gsasl authenticator.
2652 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2654 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2655 `pkg-config foo` for cflags/libs.
2657 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2658 with rest of GSASL and with heimdal_gssapi.
2660 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2661 `pkg-config foo` for cflags/libs for the TLS implementation.
2663 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2664 properties get this fed in as external SSF. A number of robustness
2665 and debugging improvements to the cyrus_sasl authenticator.
2667 PP/08 cyrus_sasl server now expands the server_realm option.
2669 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2670 Patch by Jeremy Harris.
2672 PP/10 Added dbmjz lookup type.
2674 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2676 PP/12 MAIL args handles TAB as well as SP, for better interop with
2677 non-compliant senders.
2678 Analysis and variant patch by Todd Lyons.
2680 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2681 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2682 Patch from Dirk Mueller <dmueller@suse.com>
2684 PP/13 tls_peerdn now print-escaped for spool files.
2685 Observed some $tls_peerdn in wild which contained \n, which resulted
2686 in spool file corruption.
2688 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2689 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2690 or write after TLS renegotiation, which otherwise led to messages
2693 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2694 as a tracking header (ie: a signed header comes before the signature).
2695 Patch from Wolfgang Breyha.
2697 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2698 comma-sep list; embedded commas doubled.
2700 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2702 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2704 Report and patch from Dmitry Banschikov.
2706 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2707 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2708 failures. We appear to now support TLS1.1+ with Exim.
2710 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2711 lets Exim select keys and certificates based upon TLS SNI from client.
2712 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2713 before an outbound SMTP session. New log_selector, +tls_sni.
2715 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2716 NULL dereference. Report and patch from Alun Jones.
2718 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2719 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2720 Not seeing resolver debug output on NetBSD, but suspect this is a
2721 resolver implementation change.
2723 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2724 Left warnings. Added "eximon gdb" invocation mode.
2726 PP/21 Defaulting "accept_8bitmime" to true, not false.
2728 PP/22 Added -bw for inetd wait mode support.
2730 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2731 locate the relevant includes and libraries. Made this the default.
2733 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2734 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2736 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2737 This may cause build issues on older platforms.
2739 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2740 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2741 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2742 Added SNI support via GnuTLS too.
2743 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2745 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2747 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2748 Applied second patch from Janne, implementing suggestion to default
2749 multiple-strings-in-record handling to match SPF spec.
2751 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2753 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2754 read-only, out of scope).
2755 Patch from Wolfgang Breyha, report from Stuart Northfield.
2757 PP/29 Fix three issues highlighted by clang analyser static analysis.
2758 Only crash-plausible issue would require the Cambridge-specific
2759 iplookup router and a misconfiguration.
2760 Report from Marcin Mirosław.
2762 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2764 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2765 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2766 As part of this, removing so much warning spew let me fix some minor
2767 real issues in debug logging.
2769 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2770 assignment on my part. Fixed.
2772 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2773 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2774 Janne Snabb (who went above and beyond: thank you).
2776 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2777 string otherwise requires a connection and a bunch more work and it's
2778 relatively easy to get wrong. Should also expose TLS library linkage
2781 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2782 64-bit ${eval} (JH/03).
2784 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2785 GNU libc to support some of the 64-bit stuff, should not lead to
2786 conflicts. Defined before os.h is pulled in, so if a given platform
2787 needs to override this, it can.
2789 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2790 protection layer was required, which is not implemented.
2791 Bugzilla 1254, patch from Wolfgang Breyha.
2793 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2794 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2795 tls_dhparam take prime identifiers. Also unbreak combination of
2796 OpenSSL+DH_params+TLSSNI.
2798 PP/39 Disable SSLv2 by default in OpenSSL support.
2804 PP/01 Solaris build fix for Oracle's LDAP libraries.
2805 Bugzilla 1109, patch from Stephen Usher.
2807 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2809 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2812 TF/02 Fix a couple more cases where we did not log the error message
2813 when unlink() failed. See also change 4.74-TF/03.
2815 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2816 lock up or crash if it happened to be inside a call to libc when it
2817 got a SIGUSR1 from exiwhat.
2819 The SIGUSR1 handler appends the current process status to the process
2820 log which is later printed by exiwhat. It used to use the general
2821 purpose logging code to do this, but several functions it calls are
2822 not safe for signals.
2824 The new output code in the SIGUSR1 handler is specific to the process
2825 log, and simple enough that it's easy to inspect for signal safety.
2826 Removing some special cases also simplifies the general logging code.
2827 Removing the spurious timestamps from the process log simplifies
2830 TF/04 Improved ratelimit ACL condition.
2832 The /noupdate option has been deprecated in favour of /readonly which
2833 has clearer semantics. The /leaky, /strict, and /readonly update modes
2834 are mutually exclusive. The update mode is no longer included in the
2835 database key; it just determines when the database is updated. (This
2836 means that when you upgrade Exim will forget old rate measurements.)
2838 Exim now checks that the per_* options are used with an update mode that
2839 makes sense for the current ACL. For example, when Exim is processing a
2840 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2841 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2842 must specify per_mail/readonly. If you omit the update mode it defaults to
2843 /leaky where that makes sense (as before) or /readonly where required.
2845 The /noupdate option is now undocumented but still supported for
2846 backwards compatibility. It is equivalent to /readonly except that in
2847 ACLs where /readonly is required you may specify /leaky/noupdate or
2848 /strict/noupdate which are treated the same as /readonly.
2850 A useful new feature is the /count= option. This is a generalization
2851 of the per_byte option, so that you can measure the throughput of other
2852 aggregate values. For example, the per_byte option is now equivalent
2853 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2855 The per_rcpt option has been generalized using the /count= mechanism
2856 (though it's more complicated than the per_byte equivalence). When it is
2857 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2858 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2859 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2860 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2861 Note that using per_rcpt with a non-readonly update mode in more than
2862 one ACL will cause the recipients to be double-counted. (The per_mail
2863 and per_byte options don't have this problem.)
2865 The handling of very low rates has changed slightly. If the computed rate
2866 is less than the event's count (usually one) then this event is the first
2867 after a long gap. In this case the rate is set to the same as this event's
2868 count, so that the first message of a spam run is counted properly.
2870 The major new feature is a mechanism for counting the rate of unique
2871 events. The new per_addr option counts the number of different
2872 recipients that someone has sent messages to in the last time period. It
2873 behaves like per_rcpt if all the recipient addresses are different, but
2874 duplicate recipient addresses do not increase the measured rate. Like
2875 the /count= option this is a general mechanism, so the per_addr option
2876 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2877 example, measure the rate that a client uses different sender addresses
2878 with the options per_mail/unique=$sender_address. There are further
2879 details in the main documentation.
2881 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2883 TF/06 Removed a few PCRE remnants.
2885 TF/07 Automatically extract Exim's version number from tags in the git
2886 repository when doing development or release builds.
2888 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2889 Bugzilla 879. Patch from Paul Fisher.
2891 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2892 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2895 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2897 PP/05 Variable $av_failed, true if the AV scanner deferred.
2898 Bugzilla 1078. Patch from John Horne.
2900 PP/06 Stop make process more reliably on build failure.
2901 Bugzilla 1087. Patch from Heiko Schlittermann.
2903 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2904 Bugzilla 1089. Patch from Heiko Schlittermann.
2906 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2907 Bugzilla 1131. Patch from Holger Weiß.
2909 PP/09 Handle IPv6 addresses with SPF.
2910 Bugzilla 860. Patch from Wolfgang Breyha.
2912 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2914 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2917 PP/11 match_* no longer expand right-hand-side by default.
2918 New compile-time build option, EXPAND_LISTMATCH_RHS.
2919 New expansion conditions, "inlist", "inlisti".
2921 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2923 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2925 PP/14 fix log_write() format string regression from TF/03.
2926 Bugzilla 1152. Patch from Dmitry Isaikin.
2932 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2934 PP/02 Harmonised TLS library version reporting; only show if debugging.
2935 Layout now matches that introduced for other libraries in 4.74 PP/03.
2937 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2939 PP/04 New "dns_use_edns0" global option.
2941 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2944 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2945 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2947 TK/01 Updated PolarSSL code to 0.14.2.
2948 Bugzilla 1097. Patch from Andreas Metzler.
2950 PP/07 Catch divide-by-zero in ${eval:...}.
2951 Fixes bugzilla 1102.
2953 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2956 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2957 format-string attack -- SECURITY: remote arbitrary code execution.
2959 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2960 time unintentionally subject to list matching rules, letting the header
2961 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2962 arbitrary string expansion). This allowed for information disclosure.
2964 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2965 INT_MIN/-1 -- value coerced to INT_MAX.
2971 NM/01 Workaround for PCRE version dependency in version reporting
2974 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2975 This fixes portability to compilers other than gcc, notably
2976 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2978 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2979 makefiles for portability to HP-UX and POSIX correctness.
2981 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2982 Also via indented variable definition in the Makefile.
2983 (Debugging by Oliver Heesakkers).
2985 PP/02 Restore caching of spamd results with expanded spamd_address.
2986 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2988 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2989 Improves build reliability. Fix from: Frank Elsner
2991 NM/02 Fix wide character breakage in the rfc2047 coding
2992 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2994 NM/03 Allow underscore in dnslist lookups
2995 Fixes bug 1026. Patch from Graeme Fowler
2997 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2998 Code patches from Adam Ciarcinski of NetBSD.
3000 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3003 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3004 is logged, to avoid truncation. Patch from John Horne.
3006 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3007 Patch from Jakob Hirsch.
3009 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3010 SQL string expansion failure details.
3011 Patch from Andrey Oktyabrski.
3013 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3014 Patch from Simon Arlott.
3016 PP/09 New lookups functionality failed to compile on old gcc which rejects
3017 extern declarations in function scope.
3018 Patch from Oliver Fleischmann
3020 PP/10 Use sig_atomic_t for flags set from signal handlers.
3021 Check getgroups() return and improve debugging.
3022 Fixed developed for diagnosis in bug 927 (which turned out to be
3025 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3026 Patch from Mark Zealey.
3028 PP/12 Bugzilla 1056: Improved spamd server selection.
3029 Patch from Mark Zealey.
3031 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3032 Based on patch from Heiko Schlittermann.
3034 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3035 Patch from Uwe Doering, sign-off by Michael Haardt.
3037 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3038 variable declaration deep within a block. Bug and patch from
3041 PP/15 lookups-Makefile IRIX compatibility coercion.
3043 PP/16 Make DISABLE_DKIM build knob functional.
3045 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3046 Patch by Simon Arlott
3048 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3049 variable argument macros. Our copy now differs from upstream.
3055 TF/01 Failure to get a lock on a hints database can have serious
3056 consequences so log it to the panic log.
3058 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3059 controlled using the smtp_confirmation log selector.
3061 TF/03 Include the error message when we fail to unlink a spool file.
3063 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3064 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3065 for maintaining out-of-tree patches for some time.
3067 PP/01 Bugzilla 139: Documentation and portability issues.
3068 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3069 Handle per-OS dynamic-module compilation flags.
3071 PP/02 Let /dev/null have normal permissions.
3072 The 4.73 fixes were a little too stringent and complained about the
3073 permissions on /dev/null. Exempt it from some checks.
3074 Reported by Andreas M. Kirchwitz.
3076 PP/03 Report version information for many libraries, including
3077 Exim version information for dynamically loaded libraries. Created
3078 version.h, now support a version extension string for distributors
3079 who patch heavily. Dynamic module ABI change.
3081 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3082 privilege escalation vulnerability whereby the Exim run-time user
3083 can cause root to append content of the attacker's choosing to
3086 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3089 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3090 If dropping privileges for untrusted macros, we disabled normal logging
3091 on the basis that it would fail; for the Exim run-time user, this is not
3092 the case, and it resulted in successful deliveries going unlogged.
3093 Fixed. Reported by Andreas Metzler.
3099 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3100 only prepend for the Resent-* case. Fixes regression introduced in
3101 Exim 4.70 by NM/22 for Bugzilla 607.
3103 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3104 increasing numbers of administrators be bitten by this.
3106 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3108 PP/03 Bugzilla 994: added openssl_options main configuration option.
3110 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3112 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3114 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3116 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3117 without a peer certificate, leading to a segfault because of an
3118 assumption that peers always have certificates. Be a little more
3119 paranoid. Problem reported by Martin Tscholak.
3121 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3122 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3123 NB: ClamAV planning to remove STREAM in "middle of 2010".
3124 CL also introduces -bmalware, various -d+acl logging additions and
3125 more caution in buffer sizes.
3127 PP/09 Implemented reverse_ip expansion operator.
3129 PP/10 Bugzilla 937: provide a "debug" ACL control.
3131 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3133 PP/12 Bugzilla 973: Implement --version.
3135 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3137 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3139 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3141 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3142 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3143 ignore trailing whitespace.
3145 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3147 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3150 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3151 Notification from Dr Andrew Aitchison.
3153 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3154 ExtendedDetectionInfo response format.
3155 Notification from John Horne.
3157 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3160 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3161 XSL and documented dependency on system catalogs, with examples of how
3164 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3167 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3168 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3169 configuration file which is writeable by the Exim user or group.
3171 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3172 of configuration files to cover files specified with the -C option if
3173 they are going to be used with root privileges, not just the default
3176 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3177 option (effectively making it always true).
3179 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3180 files to be used while preserving root privileges.
3182 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3183 that rogue child processes cannot use them.
3185 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3186 run-time user, instead of root.
3188 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3189 Exim run-time user without dropping privileges.
3191 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3192 result string, instead of calling string_vformat() twice with the same
3195 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3196 for other users. Others should always drop root privileges if they use
3197 -C on the command line, even for a whitelisted configure file.
3199 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3201 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3207 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3208 $data_path, and $header_path variables; fixed documentation bugs and
3211 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3212 exipick to access non-standard spools, including the "frozen" queue
3215 NM/01 Bugzilla 965: Support mysql stored procedures.
3216 Patch from Alain Williams
3218 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3220 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3221 Patch from Andreas Metzler
3223 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3224 Patch from Kirill Miazine
3226 NM/05 Bugzilla 671: Added umask to procmail example.
3228 JJ/03 installed exipick 20100323.0, fixing doc bug
3230 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3231 directory. Notification and patch from Dan Rosenberg.
3233 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3235 TK/02 Improve log output when DKIM signing operation fails.
3237 MH/01 Treat the transport option dkim_domain as a colon separated
3238 list, not as a single string, and sign the message with each element,
3239 omitting multiple occurences of the same signer.
3241 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3242 Bugzilla 985, 986. Patch by Simon Arlott
3244 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3245 Patch by Simon Arlott
3247 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3248 MBX locking. Notification from Dan Rosenberg.
3254 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3256 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3258 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3260 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3262 NM/04 Bugzilla 915: Flag broken perl installation during build.
3268 TK/01 Added patch by Johannes Berg that expands the main option
3269 "spamd_address" if it starts with a dollar sign.
3271 TK/02 Write list of recipients to X-Envelope-Sender header when building
3272 the mbox-format spool file for content scanning (suggested by Jakob
3275 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3276 (http://www.dcc-servers.net/) support via dccifd. Activated by
3277 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3279 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3280 by Mark Daniel Reidel <mr@df.eu>.
3282 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3283 When building exim an external PCRE library is now needed -
3284 PCRE is a system library on the majority of modern systems.
3285 See entry on PCRE_LIBS in EDITME file.
3287 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3288 conversation. Added nologin parameter to request.
3289 Patch contributed by Kirill Miazine.
3291 TF/01 Do not log submission mode rewrites if they do not change the address.
3293 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3295 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3296 log files in place. Contributed by Roberto Lima.
3298 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3300 TF/03 Bugzilla 615: When checking the local_parts router precondition
3301 after a local_part_suffix or local_part_prefix option, Exim now
3302 does not use the address's named list lookup cache, since this
3303 contains cached lookups for the whole local part.
3305 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3306 Robert Millan. Documentation is in experimental-spec.txt.
3308 TF/04 Bugzilla 668: Fix parallel build (make -j).
3310 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3312 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3313 Patch provided by Jan Srzednicki.
3315 TF/05 Leading white space used to be stripped from $spam_report which
3316 wrecked the formatting. Now it is preserved.
3318 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3319 that they are available at delivery time.
3321 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3323 TF/08 TLS error reporting now respects the incoming_interface and
3324 incoming_port log selectors.
3326 TF/09 Produce a more useful error message if an SMTP transport's hosts
3327 setting expands to an empty string.
3329 NM/06 Bugzilla 744: EXPN did not work under TLS.
3330 Patch provided by Phil Pennock.
3332 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3333 Patch provided by Richard Godbee.
3335 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3336 acl_smtp_notquit, added index entry.
3338 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3339 Patch provided by Eugene Bujak.
3341 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3342 accept(). Patch provided by Maxim Dounin.
3344 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3345 Patch provided by Phil Pennock.
3347 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3349 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3350 Patch provided by Brad "anomie" Jorsch.
3352 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3353 Patch provided by Dean Brooks.
3355 TK/05 Add native DKIM support (does not depend on external libraries).
3357 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3358 Patch provided by Graeme Fowler.
3360 NM/16 Bugzilla 851: Documentation example syntax fix.
3362 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3364 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3367 NM/19 Bugzilla 745: TLS version reporting.
3368 Patch provided by Phil Pennock.
3370 NM/20 Bugzilla 167: bool: condition support.
3371 Patch provided by Phil Pennock.
3373 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3374 clients. Patch provided by Phil Pennock.
3376 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3377 Patch provided by Brad "anomie" Jorsch.
3379 NM/23 Bugzilla 687: Fix misparses in eximstats.
3380 Patch provided by Heiko Schlittermann.
3382 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3383 Patch provided by Heiko Schlittermann.
3385 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3386 plus update to original patch.
3388 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3390 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3391 Patch provided by David Brownlee.
3393 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3395 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3397 NM/30 Bugzilla 888: TLS documentation bugfixes.
3399 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3401 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3402 Unlike the original bugzilla I have changed all shell scripts in src tree.
3404 NM/33 Bugzilla 898: Transport filter timeout fix.
3405 Patch by Todd Rinaldo.
3407 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3408 Patch by Serge Demonchaux.
3410 NM/35 Bugzilla 39: Base64 decode bug fixes.
3411 Patch by Jakob Hirsch.
3413 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3415 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3417 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3419 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3425 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3426 ALT-N's libdkim that I have put here:
3427 http://duncanthrax.net/exim-experimental/
3429 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3430 were called 'true' and it seems that C99 defines that as a reserved
3431 keyword to be used with 'bool' variable types. That means you could
3432 not include C99-style headers which use bools without triggering
3433 build errors in sieve.c.
3435 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3436 as mailq or other aliases. Changed the --help handling significantly
3437 to do whats expected. exim_usage() emits usage/help information.
3439 SC/01 Added the -bylocaldomain option to eximstats.
3441 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3443 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3445 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3451 PH/01 Another patch from the Sieve maintainer.
3453 PH/02 When an IPv6 address is converted to a string for single-key lookup
3454 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3455 dots are used instead of colons so that keys in lsearch files need not
3456 contain colons. This was done some time before quoting was made available
3457 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3458 (notated using the quote facility) so as to distinguish them from IPv4
3459 keys. This meant that lookups for IP addresses in host lists did not work
3460 for iplsearch lookups.
3462 This has been fixed by arranging for IPv6 addresses to be expressed with
3463 colons if the lookup type is iplsearch. This is not incompatible, because
3464 previously such lookups could never work.
3466 The situation is now rather anomalous, since one *can* have colons in
3467 ordinary lsearch keys. However, making the change in all cases is
3468 incompatible and would probably break a number of configurations.
3470 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3473 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3474 conversion specification without a maximum field width, thereby enabling
3475 a rogue spamd server to cause a buffer overflow. While nobody in their
3476 right mind would setup Exim to query an untrusted spamd server, an
3477 attacker that gains access to a server running spamd could potentially
3478 exploit this vulnerability to run arbitrary code as the Exim user.
3480 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3481 $primary_hostname instead of what libspf2 thinks the hosts name is.
3483 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3484 a directory entry by the name of the lookup key. Previously, if a
3485 symlink pointed to a non-existing file or a file in a directory that
3486 Exim lacked permissions to read, a lookup for a key matching that
3487 symlink would fail. Now it is enough that a matching directory entry
3488 exists, symlink or not. (Bugzilla 503.)
3490 PH/03 The body_linecount and body_zerocount variables are now exported in the
3493 PH/04 Added the $dnslist_matched variable.
3495 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3496 This means they are set thereafter only if the connection becomes
3499 PH/06 Added the client_condition to authenticators so that some can be skipped
3500 by clients under certain conditions.
3502 PH/07 The error message for a badly-placed control=no_multiline_responses left
3503 "_responses" off the end of the name.
3505 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3507 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3508 (without spaces) instead of just copying the configuration text.
3510 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3512 PH/11 Added $max_received_linelength.
3514 PH/12 Added +ignore_defer and +include_defer to host lists.
3516 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3517 way in which PCRE > 7.0 is built.
3519 PH/14 Implemented queue_only_load_latch.
3521 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3522 MAIL command. The effect was to mangle the value on 64-bit systems.
3524 PH/16 Another patch from the Sieve maintainer.
3526 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3528 PH/18 If a system quota error occurred while trying to create the file for
3529 a maildir delivery, the message "Mailbox is full" was not appended to the
3530 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3531 only to a quota excession during the actual writing of the file.
3533 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3534 characters?) which causes problems in log lines. The DN values are now
3535 passed through string_printing() before being added to log lines.
3537 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3538 and InterBase are left for another time.)
3540 PH/21 Added message_body_newlines option.
3542 PH/22 Guard against possible overflow in moan_check_errorcopy().
3544 PH/23 POSIX allows open() to be a macro; guard against that.
3546 PH/24 If the recipient of an error message contained an @ in the local part
3547 (suitably quoted, of course), incorrect values were put in $domain and
3548 $local_part during the evaluation of errors_copy.
3554 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3555 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3558 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3559 issue a MAIL command.
3561 PH/02 In an ACL statement such as
3563 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3565 if a client was not listed at all, or was listed with a value other than
3566 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3567 the condition was not true (as it should be), so access was not denied.
3568 The bug was that the ! inversion was incorrectly passed on to the second
3569 item. This has been fixed.
3571 PH/03 Added additional dnslists conditions == and =& which are different from
3572 = and & when the dns lookup returns more than one IP address.
3574 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3575 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3577 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3578 FSYNC, which compiles an option called disable_fsync that allows for
3579 bypassing fsync(). The documentation is heavily laced with warnings.
3581 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3583 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3584 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3585 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3586 including adding "make clean"; (3) Added -fPIC when compiling the test
3587 dynamically loaded module, to get rid of a warning.
3589 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3590 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3591 = 0s. The bug is otherwise harmless.
3593 PH/07 There was a bug in the dovecot authenticator such that the value of
3594 $auth1 could be overwritten, and so not correctly preserved, after a
3595 successful authentication. This usually meant that the value preserved by
3596 the server_setid option was incorrect.
3598 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3600 PH/09 Installed PCRE release 7.0.
3602 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3603 run for batched SMTP input. It is now run at the start of every message
3604 in the batch. While fixing this I discovered that the process information
3605 (output by running exiwhat) was not always getting set for -bs and -bS
3606 input. This is fixed, and it now also says "batched" for BSMTP.
3608 PH/11 Added control=no_pipelining.
3610 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3611 patch, slightly modified), and move the expansion of helo_data till after
3612 the connection is made in the smtp transport (so it can use these
3615 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3617 PH/14 Added log_selector = +pid.
3619 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3621 PH/16 Add ${if forany and ${if forall.
3623 PH/17 Added dsn_from option to vary the From: line in DSNs.
3625 PH/18 Flush SMTP output before performing a callout, unless control =
3626 no_callout_flush is set.
3628 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3629 was true (the default) a successful delivery failed to delete the retry
3630 item, thus causing premature timeout of the address. The bug is now
3633 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3635 PH/21 Long custom messages for fakedefer and fakereject are now split up
3636 into multiline responses in the same way that messages for "deny" and
3637 other ACL rejections are.
3639 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3640 with slight modification.
3642 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3643 draft, changing the syntax and factoring some duplicate code".
3645 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3646 for deliveries of the second and subsequent messages over the same SMTP
3649 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3650 ${reduce, with only minor "tidies".
3652 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3654 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3655 expansion side effects.
3657 PH/27 When a message times out after an over-quota error from an Exim-imposed
3658 quota, the bounce message says "mailbox is full". This message was not
3659 being given when it was a system quota that was exceeded. It now should
3662 MH/03 Made $recipients available in local_scan(). local_scan() already has
3663 better access to the recipient list through recipients_list[], but
3664 $recipients can be useful in postmaster-provided expansion strings.
3666 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3667 in the case of a MAIL command with additional options following the
3668 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3669 were accidentally chopped off.
3671 PH/29 SMTP synchronization checks are implemented when a command is read -
3672 there is a check that no more input is waiting when there shouldn't be
3673 any. However, for some commands, a delay in an ACL can mean that it is
3674 some time before the response is written. In this time, more input might
3675 arrive, invalidly. So now there are extra checks after an ACL has run for
3676 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3677 pipelining has not been advertised.
3679 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3681 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3682 returned in the SMTP response when smtp_return_error_details was false.
3683 This has been fixed.
3685 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3686 instead of the C I/O that was originally supplied, because problems were
3687 reported on Solaris.
3689 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3690 Exim which did not show up earlier: it was assuming that a call to
3691 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3692 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3693 was a macro that became an assignment, so it seemed to work. This has
3694 changed to a proper function call with a void return, hence the compile
3695 error. Exim's code has been fixed.
3697 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3700 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3702 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3704 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3705 because it turns out to be too big or there is a timeout) there is no
3706 "Completed" line in the log. When some messages of this type were
3707 selected by exigrep, they were listed as "not completed". Others were
3708 picked up by some special patterns. I have improved the selection
3709 criteria to be more general.
3711 PH/38 The host_find_failed option in the manualroute router can now be set
3712 to "ignore", to completely ignore a host whose IP address cannot be
3713 found. If all hosts are ignored, the behaviour is controlled by the new
3714 host_all_ignored option.
3716 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3717 homing or because of multiple MX records with /mx) generated more than
3718 one IP address, and the following item turned out to be the local host,
3719 all the secondary addresses of the first item were incorrectly removed
3720 from the list, along with the local host and any following hosts (which
3721 is what is supposed to happen).
3723 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3724 whoever called Exim into the -H file. In the case of the daemon it was
3725 behaving confusingly. When first started, it used values for whoever
3726 started the daemon, but after a SIGHUP it used the Exim user (because it
3727 calls itself on a restart). I have changed the code so that it now always
3730 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3731 message are rejected with the same error (e.g. no authentication or bad
3732 sender address), and a DATA command is nevertheless sent (as can happen
3733 with PIPELINING or a stupid MUA), the error message that was given to the
3734 RCPT commands is included in the rejection of the DATA command. This is
3735 intended to be helpful for MUAs that show only the final error to their
3738 PH/42 Another patch from the Sieve maintainer.
3740 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3741 Eximstats - Fixed some broken HTML links and added missing column headers
3743 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3744 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3746 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3748 PH/43 Yet another patch from the Sieve maintainer.
3750 PH/44 I found a way to check for a TCP/IP connection going away before sending
3751 the response to the final '.' that terminates a message, but only in the
3752 case where the client has not sent further data following the '.'
3753 (unfortunately, this is allowed). However, in many cases there won't be
3754 any further data because there won't be any more messages to send. A call
3755 to select() can be used: if it shows that the input is "ready", there is
3756 either input waiting, or the socket has been closed. An attempt to read
3757 the next input character can distinguish the two cases. Previously, Exim
3758 would have sent an OK response which the client would never have see.
3759 This could lead to message repetition. This fix should cure that, at
3760 least in a lot of common cases.
3762 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3763 advertised in response to EHLO.
3769 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3770 fixed by 4.65/MH/01 (is this a record?) are fixed:
3772 (i) An empty string was always treated as zero by the numeric comparison
3773 operators. This behaviour has been restored.
3775 (ii) It is documented that the numeric comparison operators always treat
3776 their arguments as decimal numbers. This was broken in that numbers
3777 starting with 0 were being interpreted as octal.
3779 While fixing these problems I realized that there was another issue that
3780 hadn't been noticed. Values of message_size_limit (both the global option
3781 and the transport option) were treated as octal if they started with 0.
3782 The documentation was vague. These values are now always treated as
3783 decimal, and I will make that clear in the documentation.
3789 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3790 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3793 MH/01 Don't check that the operands of numeric comparison operators are
3794 integers when their expansion is in "skipping" mode (fixes bug
3795 introduced by 4.64-PH/07).
3797 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3798 child addresses, Exim now panics and dies. Previously, because the count
3799 is held in a short int, deliveries were likely to be lost. As such a
3800 large number of recipients for a single message is ridiculous
3801 (performance will be very, very poor), I have chosen to impose a limit
3802 rather than extend the field.
3808 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3809 leftover -K file (the existence of which was triggered by #402).
3810 While we were at it, introduced process PID as part of the -K
3811 filename. This should rule out race conditions when creating
3814 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3815 processing considerably. Previous code took too long for large mails,
3816 triggering a timeout which in turn triggers #401.
3818 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3819 in the DK code in transports.c. sendfile() is not really portable,
3820 hence the _LINUX specificness.
3822 TF/01 In the add_headers option to the mail command in an Exim filter,
3823 there was a bug that Exim would claim a syntax error in any
3824 header after the first one which had an odd number of characters
3827 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3828 callout verification, Exim cached a "reject" for the entire domain. This
3829 is correct for most verifications, but it is not correct for a recipient
3830 verification with use_sender or use_postmaster set, because in that case
3831 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3832 case of MAIL FROM:<> rejection from other early rejections (e.g.
3833 rejection of HELO). When verifying a recipient using a non-null MAIL
3834 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3835 Whatever the result of the callout, the value of the domain cache is
3836 left unchanged (for any other kind of callout, getting as far as trying
3837 RCPT means that the domain itself is ok).
3839 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3842 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3843 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3844 session, the daemon ignores ECONNECT errors and logs others; it now
3845 ignores EPIPE as well.
3847 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3848 (quoted-printable decoding).
3850 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3851 later the small subsequent patch to fix an introduced bug.
3853 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3855 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3857 PH/08 An error is now given if message_size_limit is specified negative.
3859 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3860 to be given (somewhat) arbitrary names.
3862 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3865 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3866 miscellaneous code fixes
3868 PH/10 Added the log_reject_target ACL modifier to specify where to log
3871 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3872 hostname. This is wrong, because it relates to the incoming message (and
3873 probably the interface on which it is arriving) and not to the outgoing
3874 callout (which could be using a different interface). This has been
3875 changed to use the value of the helo_data option from the smtp transport
3876 instead - this is what is used when a message is actually being sent. If
3877 there is no remote transport (possible with a router that sets up host
3878 addresses), $smtp_active_hostname is used.
3880 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3881 tweaks were necessary in order to get it to work (see also 21 below):
3882 (a) The code assumed that strncpy() returns a negative number on buffer
3883 overflow, which isn't the case. Replaced with Exim's string_format()
3885 (b) There were several signed/unsigned issues. I just did the minimum
3886 hacking in of casts. There is scope for a larger refactoring.
3887 (c) The code used strcasecmp() which is not a standard C function.
3888 Replaced with Exim's strcmpic() function.
3889 (d) The code set only $1; it now sets $auth1 as well.
3890 (e) A simple test gave the error "authentication client didn't specify
3891 service in request". It would seem that Dovecot has changed its
3892 interface. Fortunately there's a specification; I followed it and
3893 changed what the client sends and it appears to be working now.
3895 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3898 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3899 address A is aliased to B and C, where B exists and C does not. Without
3900 -v the output is "A verified" because verification stops after a
3901 successful redirection if more than one address is generated. However,
3902 with -v the child addresses are also verified. Exim was outputting "A
3903 failed to verify" and then showing the successful verification for C,
3904 with its parentage. It now outputs "B failed to verify", showing B's
3905 parentage before showing the successful verification of C.
3907 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3908 look up a TXT record in a specific list after matching in a combined
3911 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3912 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3913 they consult the DNS. I had assumed they would set it the way they
3914 wanted; and indeed my experiments on Linux seem to show that in some
3915 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3916 To be on the safe side, however, I have now made the interface to
3917 host_find_byname() similar to host_find_bydns(), with an argument
3918 containing the DNS resolver options. The host_find_byname() function now
3919 sets these options at its start, just as host_find_bydns() does. The smtp
3920 transport options dns_qualify_single and dns_search_parents are passed to
3921 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3922 of host_find_byname() use the default settings of RES_DEFNAMES
3923 (qualify_single) but not RES_DNSRCH (search_parents).
3925 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3926 spool_read_header() do less string testing, by means of a preliminary
3927 switch on the second character of optional "-foo" lines. (This is
3928 overdue, caused by the large number of possibilities that now exist.
3929 Originally there were few.) While I was there, I also converted the
3930 str(n)cmp tests so they don't re-test the leading "-" and the first
3931 character, in the hope this might squeeze out yet more improvement.
3933 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3934 flag allowing group syntax was set by the header_syntax check but not
3935 turned off, possible causing trouble later; (2) The flag was not being
3936 set at all for the header_verify test, causing "group"-style headers to
3937 be rejected. I have now set it in this case, and also caused header_
3938 verify to ignore an empty address taken from a group. While doing this, I
3939 came across some other cases where the code for allowing group syntax
3940 while scanning a header line wasn't quite right (mostly, not resetting
3941 the flag correctly in the right place). These bugs could have caused
3942 trouble for malformed header lines. I hope it is now all correct.
3944 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3945 with the "reply" argument non-NULL. The code, however (which originally
3946 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3947 but it didn't always do it. This confused somebody who was copying the
3948 code for some other use. I have removed all the tests.
3950 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3951 feature that was used to support insecure browsers during the U.S. crypto
3952 embargo. It requires special client support, and Exim is probably the
3953 only MTA that supported it -- and would never use it because real RSA is
3954 always available. This code has been removed, because it had the bad
3955 effect of slowing Exim down by computing (never used) parameters for the
3956 RSA_EXPORT functionality.
3958 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3959 authenticator to fail if there's a tab character in the incoming data
3960 (there should never be unless someone is messing about, as it's supposed
3961 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3962 if the connection is using TLS or if the remote IP is the same as the
3963 local IP, and the "valid-client-cert option" if a client certificate has
3966 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3967 authenticators. This can be used for authorization after authentication
3968 succeeds. (In the case of plaintext, it servers for both authentication
3971 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3972 if any retry times were supplied.
3974 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3975 connection, where there is no client IP address to check. In this
3976 situation, the verify now always succeeds.
3978 PH/25 Applied John Jetmore's -Mset patch.
3980 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3982 PH/27 In a string expansion for a processed (not raw) header when multiple
3983 headers of the same name were present, leading whitespace was being
3984 removed from all of them, but trailing whitespace was being removed only
3985 from the last one. Now trailing whitespace is removed from each header
3986 before concatenation. Completely empty headers in a concatenation (as
3987 before) are ignored.
3989 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3990 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3992 PH/29 [Removed. This was a change that I later backed out, and forgot to
3993 correct the ChangeLog entry (that I had efficiently created) before
3994 committing the later change.]
3996 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3997 address errors (4xx response to RCPT) over the same connection as other
3998 messages routed to the same hosts. Such deliveries are always "forced",
3999 so retry times are not inspected. This resulted in far too many retries
4000 for the affected addresses. The effect occurred only when there were more
4001 hosts than the hosts_max_try setting in the smtp transport when it had
4002 the 4xx errors. Those hosts that it had tried were not added to the list
4003 of hosts for which the message was waiting, so if all were tried, there
4004 was no problem. Two fixes have been applied:
4006 (i) If there are any address or message errors in an SMTP delivery, none
4007 of the hosts (tried or untried) are now added to the list of hosts
4008 for which the message is waiting, so the message should not be a
4009 candidate for sending over the same connection that was used for a
4010 successful delivery of some other message. This seems entirely
4011 reasonable: after all the message is NOT "waiting for some host".
4012 This is so "obvious" that I'm not sure why it wasn't done
4013 previously. Hope I haven't missed anything, but it can't do any
4014 harm, as the worst effect is to miss an optimization.
4016 (ii) If, despite (i), such a delivery is accidentally attempted, the
4017 routing retry time is respected, so at least it doesn't keep
4018 hammering the server.
4020 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4021 in ${readsocket because some servers need this prod.
4023 PH/32 Added some extra debug output when updating a wait-xxx database.
4025 PH/33 The hint "could be header name not terminated by colon", which has been
4026 given for certain expansion errors for a long time, was not being given
4027 for the ${if def:h_colon_omitted{... case.
4029 PH/34 The spec says: "With one important exception, whenever a domain list is
4030 being scanned, $domain contains the subject domain." There was at least
4031 one case where this was not true.
4033 PH/35 The error "getsockname() failed: connection reset by peer" was being
4034 written to the panic log as well as the main log, but it isn't really
4035 panic-worthy as it just means the connection died rather early on. I have
4036 removed the panic log writing for the ECONNRESET error when getsockname()
4039 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4040 runs only) independently of the message's sender address. This meant
4041 that, if the 4xx error was in fact related to the sender, a different
4042 message to the same recipient with a different sender could confuse
4043 things. In particular, this can happen when sending to a greylisting
4044 server, but other circumstances could also provoke similar problems.
4045 I have changed the default so that the retry time for these errors is now
4046 based a combination of the sender and recipient addresses. This change
4047 can be overridden by setting address_retry_include_sender=false in the
4050 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4051 remote server are returned as part of bounce messages. This was not
4052 happening for LMTP over a pipe (the lmtp transport), but now it is the
4053 same for both kinds of LMTP.
4055 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4056 in header lines that were in fact CNAMEs. This is no longer the case.
4058 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4059 and queue runs started by the daemon processed all messages. This has
4060 been fixed so that -R and -S can now usefully be given with -q<time>.
4062 PH/40 Import PCRE release 6.7 (fixes some bugs).
4064 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4066 PH/42 Give an error if -q is specified more than once.
4068 PH/43 Renamed the variables $interface_address and $interface_port as
4069 $received_ip_address and $received_port, to make it clear that these
4070 values apply to message reception, and not to the outgoing interface when
4071 a message is delivered. (The old names remain recognized, of course.)
4073 PH/44 There was no timeout on the connect() call when using a Unix domain
4074 socket in the ${readsocket expansion. There now is.
4076 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4077 be meaningful with "accept".
4079 SC/01 Eximstats V1.43
4080 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4082 SC/02 Eximstats V1.44
4083 Use a glob alias rather than an array ref in the generated
4084 parser. This improves both readability and performance.
4086 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4087 Collect SpamAssassin and rejection statistics.
4088 Don't display local sender or destination tables unless
4089 there is data to show.
4090 Added average volumes into the top table text output.
4092 SC/04 Eximstats V1.46
4093 Collect data on the number of addresses (recipients)
4094 as well as the number of messages.
4096 SC/05 Eximstats V1.47
4097 Added 'Message too big' to the list of mail rejection
4098 reasons (thanks to Marco Gaiarin).
4100 SC/06 Eximstats V1.48
4101 Mainlog lines which have GMT offsets and are too short to
4102 have a flag are now skipped.
4104 SC/07 Eximstats V1.49 (Alain Williams)
4105 Added the -emptyok flag.
4107 SC/08 Eximstats V1.50
4108 Fixes for obtaining the IP address from reject messages.
4110 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4111 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4112 whitespace changes from 4.64-PH/27
4114 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4117 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4118 are found, allow negative numbers in numeric criteria)
4120 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4122 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4123 to match changes made in 4.64-PH/43
4125 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4127 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4128 "domain" is actually an IP address, and force a failure. This locks out
4129 those revolvers/nameservers that support "A-for-A" lookups, in
4130 contravention of the specifications.
4132 PH/48 When a host name was looked up from an IP address, and the subsequent
4133 forward lookup of the name timed out, the host name was left in
4134 $sender_host_name, contrary to the specification.
4136 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4137 restricted to single-key lookups, Exim was not diagnosing an error if
4138 * or *@ was used with a query-style lookup.
4140 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4142 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4143 long ago, but noone interested enough thought of it. Let's just say that
4144 the "1.1" means that there are some new functions that weren't there at
4145 some point in the past.
4147 PH/51 Error processing for expansion failure of helo_data from an smtp
4148 transport during callout processing was broken.
4150 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4151 tested/used via the -bh/-bhc/-bs options.
4153 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4154 bug, fixed in subsequent PCRE releases).
4156 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4157 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4159 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4165 SC/01 Use a glob alias rather than an array ref in eximstats generated
4166 parser. This improves both readability and performance.
4168 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4169 Don't display local sender or destination tables in eximstats unless
4170 there is data to show.
4171 Added average volumes into the eximstats top table text output.
4173 SC/03 Collect data on the number of addresses (recipients) as well
4174 as the number of messages in eximstats.
4176 TF/01 Correct an error in the documentation for the redirect router. Exim
4177 does (usually) call initgroups() when daemonizing.
4179 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4180 with consistent privilege compared to when running as a daemon.
4182 TF/03 Note in the spec that $authenticated_id is not set for local
4183 submissions from trusted users.
4185 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4186 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4188 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4189 by adding some example configuration directives to the default
4190 configuration file. A little bit of work is required to uncomment the
4191 directives and define how usernames and passwords are checked, but
4192 there is now a framework to start from.
4194 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4195 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4196 without this. I don't know how relevant this is to other LDAP libraries.
4198 PH/02 Add the verb name to the "unknown ACL verb" error.
4200 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4202 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4204 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4205 directory not expanded when it should be if an expanded home directory
4206 was set for the address (which is overridden by the transport).
4208 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4211 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4212 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4213 because it is too late at that time, and has no effect.
4215 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4216 security issue with \' (bugzilla #107). I could not use the
4217 PQescapeStringConn() function, because it needs a PGconn value as one of
4220 PH/08 When testing addresses using -bt, indicate those final addresses that
4221 are duplicates that would not cause an additional delivery. At least one
4222 person was confused, thinking that -bt output corresponded to deliveries.
4223 (Suppressing duplicates isn't a good idea as you lose the information
4224 about possibly different redirections that led to the duplicates.)
4226 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4227 systems where poll() doesn't work, in particular OS X.
4229 PH/10 Added more information to debugging output for retry time not reached.
4231 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4232 operations in malware.c.
4234 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4237 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4238 syslog_duplication set false, log lines that would normally be written
4239 both the the main log and to the reject log were not written to syslog at
4242 PH/14 In the default configuration, change the use of "message" in ACL warn
4243 statements to "add_header".
4245 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4246 not followed by a command (e.g. "seen endif").
4248 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4249 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4252 PH/17 Added extra conditions to the default value of delay_warning_condition
4256 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4257 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4258 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4261 The Auto-Submitted: and various List- headers are standardised, whereas I
4262 don't think Precedence: ever was.
4264 PH/18 Refactored debugging code in route_finduser() to show more information,
4265 in particular, the error code if getpwnam() issues one.
4267 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4268 This is apparently needed in addition to the PH/07 change above to avoid
4269 any possible encoding problems.
4271 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4272 but not after initializing Perl.
4274 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4275 output them only if debugging. By default they are written stderr,
4276 apparently, which is not desirable.
4278 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4281 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4284 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4286 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4287 authenticated or an ident call has been made. Suppress the default
4288 values for $authenticated_id and $authenticated_sender (but permit -oMai
4289 and -oMas) when testing with -bh.
4291 PH/24 Re-jigged the order of the tests in the default configuration so that the
4292 tests for valid domains and recipients precede the DNS black list and CSA
4293 tests, on the grounds that those ones are more expensive.
4295 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4296 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4297 This bug exists in every version of Exim that I still have, right back to
4300 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4301 However, an attempt to turn on case-sensitivity in a regex key by
4302 including (?-i) didn't work because the subject string was already
4303 lowercased, and the effects were non-intuitive. It turns out that a
4304 one-line patch can be used to allow (?-i) to work as expected.
4310 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4311 other effects) broke the use of negated acl sub-conditions.
4313 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4316 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4317 "Deny" causes Exim to reject the incoming connection with a 554 error.
4318 Unfortunately, if there is a major crisis, such as a disk failure,
4319 tcp-wrappers gives "deny", whereas what one would like would be some
4320 kind of temporary error. A kludge has been added to help with this.
4321 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4322 554 error is used if errno is still zero or contains ENOENT (which occurs
4323 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4326 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4328 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4329 errors. Otherwise a message that provokes a temporary error (when other
4330 messages do not) can cause a whole host to time out.
4332 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4333 addresses were routed directly to files or pipes from a redirect router.
4334 File deliveries just didn't batch; pipe deliveries might have suffered
4337 PH/06 A failure to get a lock for a hints database would erroneously always say
4338 "Failed to get write lock", even when it was really a read lock.
4340 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4341 of 0600. This has been changed to use the value of the lockfile_mode
4342 option (which defaults to 0600).
4344 PH/08 Applied small patch from the Sieve maintainer.
4346 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4347 folder from quota calculations, a direct delivery into this folder messed
4348 up the contents of the maildirsize file. This was because the regex was
4349 used only to exclude .Trash (or whatever) when the size of the mailbox
4350 was calculated. There was no check that a delivery was happening into an
4351 excluded directory. This bug has been fixed by ignoring all quota
4352 processing for deliveries into excluded directories.
4354 PH/10 Added the maildirfolder_create_regex option to appendfile.
4360 PH/01 The code for finding all the local interface addresses on a FreeBSD
4361 system running IPv6 was broken. This may well have applied to all BSD
4362 systems, as well as to others that have similar system calls. The broken
4363 code found IPv4 interfaces correctly, but gave incorrect values for the
4364 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4365 that it would not match correctly against @[] and not recognize the IPv6
4368 PH/02 The ipliteral router was not recognizing addresses of the form user@
4369 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4371 PH/03 Added disable_ipv6.
4373 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4374 decoded form, because it is most often used to construct To: headers
4375 lines in autoreplies, and the decoded form may well be syntactically
4376 invalid. However, $reply_address has leading white space removed, and all
4377 newlines turned into spaces so that the autoreply transport does not
4380 PH/05 If group was specified without a user on a router, and no group or user
4381 was specified on a transport, the group from the router was ignored.
4383 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4384 for visible compile-time settings that can be used to change these
4385 numbers, for those that want even more. Backwards compatibility with old
4386 spool files has been maintained. However, going back to a previous Exim
4387 release will lost any variables that are in spool files.
4389 PH/07 Two small changes when running in the test harness: increase delay when
4390 passing a TCP/IP connection to a new process, in case the original
4391 process has to generate a bounce, and remove special handling of
4392 127.0.0.2 (sic), which is no longer necessary.
4394 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4395 be the same on different OS.
4397 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4400 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4401 whether --show-vars was specified or not
4403 JJ/02 exipick: Added support for new ACL variable spool format introduced
4406 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4407 syntactically invalid From: or Reply-to: line, and a filter used this to
4408 generate an autoreply, and therefore failed to obtain an address for the
4409 autoreply, Exim could try to deliver to a non-existent relative file
4410 name, causing unrelated and misleading errors. What now happens is that
4411 it logs this as a hard delivery error, but does not attempt to create a
4414 PH/11 The exinext utility has a -C option for testing purposes, but although
4415 the given file was scanned by exinext itself; it wasn't being passed on
4416 when Exim was called.
4418 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4419 an end-of-file indication when reading a command response.
4421 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4422 compiled. In many other places in Exim, IPv6 addresses are always
4423 recognized, so I have changed this. It also means that IPv4 domain
4424 literals of the form [IPV4:n.n.n.n] are now always recognized.
4426 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4427 used if the router is not running as root, for example, when verifying at
4428 ACL time, or when using -bh. The debugging output from this situation was
4429 non-existent - all you got was a failure to exec. I have made two
4432 (a) Failures to set uid/gid, the current directory, or a process leader
4433 in a subprocess such as that created by queryprogram now generate
4434 suitable debugging output when -d is set.
4436 (b) The queryprogram router detects when it is not running as root,
4437 outputs suitable debugging information if -d is set, and then runs
4438 the subprocess without attempting to change uid/gid.
4440 PH/15 Minor change to Makefile for building test_host (undocumented testing
4443 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4444 additional section of a DNS packet that returns MX or SRV records.
4445 Instead, it always explicitly searches for A/AAAA records. This avoids
4446 major problems that occur when a DNS server includes only records of one
4447 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4448 fixed another bug: if SRV records were looked up and the corresponding
4449 address records were *not* found in the additional section, the port
4450 values from the SRV records were lost.
4452 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4453 using the correct key (the original address) when searching the retry
4454 rules in order to find which one to use for generating the retry hint.
4456 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4457 adding the default one. Similarly, if it contains a Reply-To: header, the
4458 errors_reply_to option, if set, is not used.
4460 PH/19 When calculating a retry time, Exim used to measure the "time since
4461 failure" by looking at the "first failed" field in the retry record. Now
4462 it does not use this if it is later than than the arrival time of the
4463 message. Instead it uses the arrival time. This makes for better
4464 behaviour in cases where some deliveries succeed, thus re-setting the
4465 "first failed" field. An example is a quota failure for a huge message
4466 when small messages continue to be delivered. Without this change, the
4467 "time since failure" will always be short, possible causing more frequent
4468 delivery attempts for the huge message than are intended.
4469 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4471 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4472 $1, $2, $3) because the numerical variables can be reset during some
4473 expansion items (e.g. "match"), thereby losing the authentication data.
4475 PH/21 Make -bV show the size of off_t variables so that the test suite can
4476 decide whether to run tests for quotas > 2G.
4478 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4479 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4480 filecount value is greater than 2G or if a quota value is greater than 2G
4481 on a system where the size of off_t is not greater than 4, a panic error
4484 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4485 never match. The debug and -bh output now contains an explicit error
4486 message indicating a malformed IPv4 address or mask.
4488 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4489 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4490 PH/23 above applies.
4492 PH/25 Do not write to syslog when running in the test harness. The only
4493 occasion when this arises is a failure to open the main or panic logs
4494 (for which there is an explicit test).
4496 PH/26 Added the /no_tell option to "control=freeze".
4498 PH/27 If a host name lookup failed very early in a connection, for example, if
4499 the IP address matched host_lookup and the reverse lookup yielded a name
4500 that did not have a forward lookup, an error message of the form "no IP
4501 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4502 could be logged. Now it outputs the IP address instead of "NULL".
4504 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4505 allows the sender and the authenticated sender to be set when
4506 submitting a message from within Exim. Since child_open_exim() is
4507 documented for local_scan(), the new function should be too.
4509 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4510 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4511 results in an empty string is now treated as unset.
4513 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4515 PH/31 Added sender_verify_fail logging option.
4517 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4518 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4519 tidied the source and removed it altogether.
4521 PH/33 When a queue run was abandoned because the load average was too high, a
4522 log line was always written; now it is written only if the queue_run log
4523 selector is set. In addition, the log line for abandonment now contains
4524 information about the queue run such as the pid. This is always present
4525 in "start" and "stop" lines but was omitted from the "abandon" line.
4527 PH/34 Omit spaces between a header name and the colon in the error message that
4528 is given when verify = headers_syntax fails (if there are lots of them,
4529 the message gets confusing).
4531 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4532 names, as there are now some PTR records that contain slashes. This check
4533 is only to protect against broken name servers that fall over on strange
4534 characters, so the fact that it applies to all lookups doesn't matter.
4536 PH/36 Now that the new test suite is complete, we can remove some of the
4537 special code in Exim that was needed for the old test suite. For example,
4538 sorting DNS records because real resolvers return them in an arbitrary
4539 order. The new test suite's fake resolver always returns records in the
4542 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4543 bounces) except when queue_only is set, to avoid logging races between
4544 the different processes.
4546 PH/38 Panic-die if .include specifies a non-absolute path.
4548 PH/39 A tweak to the "H" retry rule from its user.
4550 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4551 a label. They prevented compilation on older perls.
4553 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4554 a warning to be raised on newish perls.
4556 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4557 on queue. Changes to match documented behaviour of showing count of
4558 messages matching specified criteria.
4560 PH/40 Changed the default ident timeout from 30s to 5s.
4562 PH/41 Added support for the use of login_cap features, on those BSD systems
4563 that have them, for controlling the resources used by pipe deliveries.
4565 PH/42 The content-scanning code uses fopen() to create files in which to put
4566 message data. Previously it was not paying any attention to the mode of
4567 the files. Exim runs with umask(0) because the rest of the code creates
4568 files with open(), and sets the required mode explicitly. Thus, these
4569 files were ending up world-writeable. This was not a big issue, because,
4570 being within the spool directory, they were not world-accessible. I have
4571 created a function called modefopen, which takes an additional mode
4572 argument. It sets umask(777), creates the file, chmods it to the required
4573 mode, then resets the umask. All the relevant calls to fopen() in the
4574 content scanning code have been changed to use this function.
4576 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4577 to 24 hours. This avoids potential overflow problems when processing G
4578 and H retry rules. I suspect nobody ever tinkers with this value.
4580 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4582 PH/45 When the plaintext authenticator is running as a client, the server's
4583 challenges are checked to ensure they are valid base64 strings. By
4584 default, the authentication attempt is cancelled if an invalid string is
4585 received. Setting client_ignore_invalid_base64 true ignores these errors.
4586 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4587 they are received. Thus, the responses can be made to depend on the
4588 challenges. If an invalid string is ignored, an empty string is placed in
4591 PH/46 Messages that are created by the autoreply transport now contains a
4592 References: header, in accordance with RFCs 2822 and 3834.
4594 PH/47 Added authenticated_sender_force to the smtp transport.
4596 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4598 PH/49 Installed latest patch from the Sieve maintainer.
4600 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4601 was also set, the appendfile transport was unnecessarily scanning a
4602 directory of message files (e.g. for maildir delivery) to find the count
4603 of files (along with the size), even though it did not need this
4604 information. It now does the scan only if it needs to find either the
4605 size of the count of files.
4607 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4609 PH/52 Two bugs concerned with error handling when the smtp transport is
4612 (i) Exim was not creating retry information for temporary errors given
4613 for individual recipients after the DATA command when the smtp transport
4614 was used in LMTP mode. This meant that they could be retried too
4615 frequently, and not timed out correctly.
4617 (ii) Exim was setting the flag that allows error details to be returned
4618 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4619 recipients that were returned after the DATA command.
4621 PH/53 This is related to PH/52, but is more general: for any failing address,
4622 when detailed error information was permitted to be returned to the
4623 sender, but the error was temporary, then after the final timeout, only
4624 "retry timeout exceeded" was returned. Now it returns the full error as
4625 well as "retry timeout exceeded".
4627 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4628 do this, and (what is worse) MTAs that accept it.
4630 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4631 will now be deprecated.
4633 PH/56 New os.c-cygwin from the Cygwin maintainer.
4635 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4636 formats (previously only available in exim formats via -bpr, -bpru,
4637 and -bpra. Now also available in native and exiqgrep formats)
4639 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4640 with very large, slow to parse queues
4642 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4644 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4646 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4647 responses to authentication challenges, though it was showing the
4648 challenges; (ii) I've removed the CR characters from the debug output for
4651 PH/58 Allow for the insertion of a newline as well as a space when a string
4652 is turned into more than one encoded-word during RFC 2047 encoding. The
4653 Sieve code now uses this.
4655 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4656 data_4xx, lost_connection, tls_required.
4658 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4659 message was being sent as an SMTP response.
4661 PH/61 Add -l and -k options to exicyclog.
4663 PH/62 When verifying, if an address was redirected to one new address, so that
4664 verification continued, and the new address failed or deferred after
4665 having set something in $address_data, the value of $address_data was not
4666 passed back to the ACL. This was different to the case when no
4667 redirection occurred. The value is now passed back in both cases.
4669 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4670 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4671 use login_cap.h, so on its own it isn't the distinguishing feature. The
4672 new name refers directly to the setclassresources() function.
4674 PH/65 Added configuration files for NetBSD3.
4676 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4678 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4679 is preferred over IPv4.
4681 PH/68 The bounce_return_message and bounce_return_body options were not being
4682 honoured for bounces generated during the reception of non-SMTP messages.
4683 In particular, this applied to messages rejected by the ACL. This bug has
4684 been fixed. However, if bounce_return_message is true and bounce_return_
4685 body is false, the headers that are returned for a non-SMTP message
4686 include only those that have been read before the error was detected.
4687 (In the case of an ACL rejection, they have all been read.)
4689 PH/69 The HTML version of the specification is now built in a directory called
4690 spec_html instead of spec.html, because the latter looks like a path with
4691 a MIME-type, and this confuses some software.
4693 PH/70 Catch two compiler warnings in sieve.c.
4695 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4696 function verify_get_ident() calls ip_connect() to connect a socket, but
4697 if the "connect()" function timed out, ip_connect() used to close the
4698 socket. However, verify_get_ident() also closes the socket later, and in
4699 between Exim writes to the log, which may get opened at this point. When
4700 the socket was closed in ip_connect(), the log could get the same file
4701 descriptor number as the socket. This naturally causes chaos. The fix is
4702 not to close the socket in ip_connect(); the socket should be closed by
4703 the function that creates it. There was only one place in the code where
4704 this was missing, in the iplookup router, which I don't think anybody now
4705 uses, but I've fixed it anyway.
4707 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4708 well as to direct DNS lookups. Otherwise the handling of names in host
4709 lists is inconsistent and therefore confusing.
4715 PH/01 Two changes to the default runtime configuration:
4717 (1) Move the checks for relay_from_hosts and authenticated clients from
4718 after to before the (commented out) DNS black list checks.
4720 (2) Add control=submission to the relay_from_hosts and authenticated
4721 clients checks, on the grounds that messages accepted by these
4722 statements are most likely to be submissions.
4724 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4726 (1) Generate an error if the third argument for the ${prvs expansion is
4729 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4732 (3) Reset the variables that are obtained from the first argument of
4733 ${prvscheck and used in the second argument before leaving the code,
4734 because their memory is reclaimed, so using them afterwards may do
4737 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4738 one (it's much easier than Tom thought :-).
4740 (5) Because of (4), we can now allow for the use of $prvscheck_result
4741 inside the third argument.
4743 PH/03 For some reason, the default setting of PATH when running a command from
4744 a pipe transport was just "/usr/bin". I have changed it to
4747 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4748 anything to be listed in the output from -bV.
4750 PH/05 When a filter generated an autoreply, the entire To: header line was
4751 quoted in the delivery log line, like this:
4753 => >A.N.Other <ano@some.domain> <original@ddress> ...
4755 This has been changed so that it extracts the operative address. There
4756 may be more than one such address. If so, they are comma-separated, like
4759 => >ano@some.domain,ona@other.domain <original@ddress> ...
4761 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4762 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4763 not being looked up in the rDNS to get a host name, Exim was showing the
4764 IP address twice in Received: lines, even though the IP addresses were
4765 identical. For example:
4767 Received: from [1.2.3.4] (helo=[1.2.3.4])
4769 However, if the real host name was known, it was omitting the HELO data
4770 if it matched the actual IP address. This has been tidied up so that it
4771 doesn't show the same IP address twice.
4773 PH/07 When both +timestamp and +memory debugging was on, the value given by
4774 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4775 called by the debug printing, thereby overwriting the timestamp buffer.
4776 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4778 PH/08 When the original message was included in an autoreply transport, it
4779 always said "this is a copy of the message, including all the headers",
4780 even if body_only or headers_only was set. It now gives an appropriate
4783 PH/09 Applied a patch from the Sieve maintainer which:
4785 o fixes some comments
4786 o adds the (disabled) notify extension core
4787 o adds some debug output for the result of if/elsif tests
4788 o points to the current vacation draft in the documentation
4789 and documents the missing references header update
4793 o fixes a bug in processing the envelope test (when testing
4794 multiple envelope elements, the last element determined the
4797 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4798 Electronic Mail") by including:
4800 Auto-submitted: auto-generated
4802 in the messages that it generates (bounce messages and others, such as
4803 warnings). In the case of bounce messages for non-SMTP messages, there was
4804 also a typo: it was using "Auto_submitted" (underscore instead of
4805 hyphen). Since every message generated by Exim is necessarily in response
4806 to another message, thes have all been changed to:
4808 Auto-Submitted: auto-replied
4810 in accordance with these statements in the RFC:
4812 The auto-replied keyword:
4814 - SHOULD be used on messages sent in direct response to another
4815 message by an automatic process,
4817 - MUST NOT be used on manually-generated messages,
4819 - MAY be used on Delivery Status Notifications (DSNs) and Message
4820 Disposition Notifications (MDNs),
4822 - MUST NOT be used on messages generated by automatic or periodic
4823 processes, except for messages which are automatic responses to
4826 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4827 to the default Received: header definition.
4829 PH/12 Added log selector acl_warn_skipped (default on).
4831 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4832 variables because (a) they are in the wrong storage pool and (b) even if
4833 they were copied, it wouldn't work properly because of the caching.
4835 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4836 checking when decoding. Apparently there are clients that generate
4837 overlong encoded strings. Why am I not surprised?
4839 PH/15 If the first argument of "${if match_address" was not empty, but did not
4840 contain an "@" character, Exim crashed. Now it writes a panic log message
4841 and treats the condition as false.
4843 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4845 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4846 extension "envelope-auth". The code is finished and in agreement with
4847 other implementations, but there is no documentation so far and in fact,
4848 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4849 not changing the active code.
4851 Print executed "if" and "elsif" statements when debugging is used. This
4852 helps a great deal to understand what a filter does.
4854 Document more things not specified clearly in RFC3028. I had all this
4855 sorted out, when out of a sudden new issues came to my mind. Oops."
4857 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4860 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4861 canonical form (as documented). However, after a host name lookup from
4862 the IP address, check_host() was doing a simple string comparison with
4863 addresses acquired from the DNS when checking that the found name did
4864 have the original IP as one of its addresses. Since any found IPv6
4865 addresses are likely to be in abbreviated form, the comparison could
4866 fail. Luckily, there already exists a function for doing the comparison
4867 by converting both addresses to binary, so now that is used instead of
4868 the text comparison.
4870 PH/20 There was another similar case to PH/19, when a complete host name was
4871 given in a host list; looking up its IP address could give an abbreviated
4872 form, whereas the current host's name might or might not be abbreviated.
4873 The same fix has been applied.
4879 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4880 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4883 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4884 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4886 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4888 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4889 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4890 Also, for "Auto-Submitted", treat anything other than "no" as
4891 non-personal, in accordance with RFC 3834. (Previously it treated
4892 anything starting "auto-" as non-personal.)
4894 TF/01 The control=submission/name=... option had a problem with syntax
4895 errors if the name included a slash character. The /name= option
4896 now slurps the rest of the string, so it can include any characters
4897 but it must come last in the list of options (after /sender_retain
4900 PH/05 Some modifications to the interface to the fake nameserver for the new
4908 TK/01 Added the "success_on_redirect" address verification option. See
4909 NewStuff for rationale and an example.
4911 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4913 PH/02 Patch to exigrep to allow it to work on syslog lines.
4915 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4916 fread() to skip over the body file's header line, because in Cygwin the
4917 header line is locked and is inaccessible.
4919 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4920 co-exist for some time) to make it clear that it is the Exim ID that is
4921 referenced, not the Message-ID: header line.
4923 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4924 string_format() function, because snprintf() does not exist on all
4927 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4928 ${stat: expansion item.
4930 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4931 protocol synchronization error", to keep the pedants happy.
4933 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4934 well as for IRIX systems, when gcc is being used. See the host.c source
4937 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4939 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4942 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4943 transport and to the smtp transport in LMTP mode.
4945 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4947 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4948 run a filter in a subprocess. This could lead to confusion in subsequent
4949 lookups in the parent process. There should also be a search_tidyup() at
4950 the end of the subprocess.
4952 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4953 only if the host matched helo_try_verify_hosts, which caused the
4954 verification to occur when the EHLO/HELO command was issued. The ACL just
4955 tested the remembered result. Now, if a previous verification attempt has
4956 not happened, "verify = helo" does it there and then.
4958 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4960 TK/03 Fix log output including CR from clamd.
4962 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4963 exist provoked a memory error which could cause a segfault.
4965 PH/15 Installed PCRE 6.2
4967 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4969 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4970 of the problem. Specifically, suggested +O2 rather than +O1 for the
4973 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4975 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4976 record, the port was not correctly logged when the outgoing_port log
4977 selector was set (it logged the transort's default port).
4979 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4980 fallback_hosts, and "hosts" in the smtp transport.
4982 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4983 host errors such as "Connection refused".
4985 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4986 authentication with radiusclient 0.4.9:
4988 - Error returned from rc_read_config was caught wrongly
4989 - Username/password not passed on to radius server due to wrong length.
4991 The presumption is that some radiusclient API changes for 4.51/PH/17
4992 were not taken care of correctly. The code is still untested by me (my
4993 Linux distribution still has 0.3.2 of radiusclient), but it was
4994 contributed by a Radius user.
4996 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4997 expanding the "port" option of the smtp transport.
4999 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5000 while reading a MIME header. Thanks to Tom Hughes for a patch.
5002 PH/24 Include config.h inside local_scan.h so that configuration settings are
5005 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5006 that in an ACL for RCPT (for example), you can examine exactly what was
5009 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5010 commands, but it was not correctly comparing the address with the actual
5011 client host address. Thus, it would show the EHLO address in Received:
5012 header lines when this was not necessary.
5014 PH/27 Added the % operator to ${eval:}.
5016 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5017 it should be ignoring failures (because with -C, for example, it has lost
5018 privilege). It wasn't ignoring creation failures other than "already
5021 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5024 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5025 because an input line was too long, either on its own, or by virtue of
5026 too many continuations, the temporary file was not being removed, and the
5027 return code was incorrect.
5029 PH/31 Missing "BOOL" in function definition in filtertest.c.
5031 PH/32 Applied Sieve patches from the maintainer.
5033 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5035 PH/33 Added "verify = not_blind".
5037 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5038 Local/Makefile (with some defaults set). These are used in built scripts
5039 such as exicyclog, but they have never been used in the exim_install
5040 script (though there are many overriding facilities there). I have
5041 arranged that the exim_install script now takes note of these two
5044 PH/35 Installed configuration files for Dragonfly.
5046 PH/36 When a locally submitted message by a trusted user did not contain a
5047 From: header, and the sender address was obtained from -f or from an SMTP
5048 MAIL command, and the trusted user did not use -F to supply a sender
5049 name, $originator_name was incorrectly used when constructing a From:
5050 header. Furthermore, $originator_name was used for submission mode
5051 messages from external hosts without From: headers in a similar way,
5052 which is clearly wrong.
5054 PH/37 Added control=suppress_local_fixups.
5056 PH/38 When log_selector = +received_sender was set, and the addition of the
5057 sender made the log line's construction buffer exactly full, or one byte
5058 less than full, an overflow happened when the terminating "\n" was
5061 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5062 when the result of a list match is failure because a DNS lookup failed.
5064 PH/40 RM_COMMAND is now used in the building process.
5066 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5067 the "build-* directories that it finds.
5069 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5070 address in a domain literal was a prefix of an interface address.
5072 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5073 when verifying a sender address, unless rewrite_headers is false.
5075 PH/44 Wrote a long comment about why errors_to addresses are verified as
5076 recipients, not senders.
5078 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5079 the ratelimit ACL was added.
5081 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5083 PH/46 Added extra information about PostgreSQL errors to the error string.
5085 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5086 avoiding the need to install special zones in a real server. This is
5087 backwards compatible; if it can't find the fake resolver, it drops back.
5088 Thus, both old and new test suites can be run.
5090 TF/02 Added util/ratelimit.pl
5092 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5093 clock is set back in time.
5095 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5096 Candler <B.Candler@pobox.com>.
5098 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5099 OK for addresses that are the result of redirections.
5101 PH/48 A number of further additions for the benefit of the new test suite,
5102 including a fake gethostbyname() that interfaces to the fake DNS resolver
5105 TF/06 The fix for widen_domains has also been applied to qualify_single and
5106 search_parents which are the other dnslookup options that can cause
5109 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5112 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5114 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5115 service" if it encountered an MX record with an empty target hostname.
5116 The message is now "an MX or SRV record indicated no SMTP service".
5118 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5119 if the DNS of the sending site is misconfigured. This is quite a
5120 common situation. This change restores the behaviour of treating a
5121 helo verification defer as a failure.
5123 PH/51 If self=fail was set on a router, the bounce message did not include the
5124 actual error message.
5130 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5132 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5133 command itself ended in error, the underlying message about the transport
5134 filter timeout was being overwritten with the pipe command error. Now the
5135 underlying error message should be appended to the second error message.
5137 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5139 PH/02 Reduce the amount of output that "make" produces by default. Full output
5140 can still be requested.
5142 PH/03 The warning log line about a condition test deferring for a "warn" verb
5143 was being output only once per connection, rather than after each
5144 occurrence (because it was using the same function as for successful
5145 "warn" verbs). This seems wrong, so I have changed it.
5147 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5148 it should not have, which might have caused a crash in the right
5149 circumstances, but probably never did.
5151 PH/04 Installed a modified version of Tony Finch's patch to make submission
5152 mode fix the return path as well as the Sender: header line, and to
5153 add a /name= option so that you can make the user's friendly name appear
5156 TF/03 Added the control = fakedefer ACL modifier.
5158 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5159 Mark Lowes for thorough testing.
5161 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5163 TK/03 Merged latest SRS patch from Miles Wilton.
5165 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5166 with the definition in sysexits.h (which is #included earlier).
5167 Fortunately, Exim does not actually use EX_OK. The code used to try to
5168 preserve the sysexits.h value, by assuming that macro definitions were
5169 scanned for macro replacements. I have been disabused of this notion,
5170 so now the code just undefines EX_OK before #including unistd.h.
5172 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5173 in the smtp transport. When a block could not be written in a single
5174 write() function, the timeout was being re-applied to each part-write.
5175 This seems wrong - if the receiver was accepting one byte at a time it
5176 would take for ever. The timeout is now adjusted when this happens. It
5177 doesn't have to be particularly precise.
5179 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5180 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5182 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5183 without a domain if the check to <postmaster@domain> fails.
5185 SC/01 Eximstats: added -xls and the ability to specify output files
5186 (patch written by Frank Heydlauf).
5188 SC/02 Eximstats: use FileHandles for outputting results.
5190 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5192 SC/04 Eximstats: fixed display of large numbers with -nvr option
5194 SC/05 Eximstats: fixed merging of reports with empty tables.
5196 SC/06 Eximstats: added the -include_original_destination flag
5198 SC/07 Eximstats: removed tabs and trailing whitespace.
5200 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5202 TK/06 MBOX spool code: Add real "From " MBOX separator line
5203 so the .eml file is really in mbox format (even though
5204 most programs do not really care). Patch from Alex Miller.
5206 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5207 The latter is generated from $received_to and is only set if the
5208 message has one envelope recipient. SA can use these headers,
5209 obviously out-of-the-box. Patch from Alex Miller.
5211 PH/08 The ${def test on a variable was returning false if the variable's
5212 value was "0", contrary to what the specification has always said!
5213 The result should be true unless the variable is empty.
5215 PH/09 The syntax error of a character other than { following "${if
5216 def:variable_name" (after optional whitespace) was not being diagnosed.
5217 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5218 accidental colon was present, for example, could give incorrect results.
5220 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5221 result is used (not including appendfile, where other changes are about
5224 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5225 This involved changing a lot of size variables from int to off_t. It
5226 should work with maildirs and everything.
5228 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5229 spamd dying while we are connected to it.
5231 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5234 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5235 function for BDB 4.3.
5237 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5239 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5240 the sources for PCRE have been re-arranged and more files are now
5243 PH/15 The code I had for printing potentially long long variables in PH/11
5244 above was not the best (it lost precision). The length of off_t variables
5245 is now inspected at build time, and an appropriate printing format (%ld
5246 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5247 to be "long long int" or "long int". This is needed for the internal
5248 formatting function string_vformat().
5250 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5251 the configuration file to be ":syslog", then the script "guesses" where
5252 the logs files are, rather than using the compiled in default. In our
5253 case the guess is not the same as the compiled default, so the script
5254 suddenly stopped working when I started to use syslog. The patch checks
5255 to see if log_file_path is "". If so, it attempts to read it from exim
5256 with no configuration file to get the compiled in version, before it
5257 falls back to the previous guessing code."
5259 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5260 implementing BATV in an Exim configuration. See NewStuff for the gory
5263 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5264 Makefile that are specific to HP-UX.
5266 PH/18 If the "use_postmaster" option was set for a recipient callout together
5267 with the "random" option, the postmaster address was used as the MAIL
5268 FROM address for the random test, but not for the subsequent recipient
5269 test. It is now used for both.
5271 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5272 patch removes a few documentation additions to RFC 3028, because the
5273 latest draft now contains them. It adds the new en;ascii-case comparator
5274 and a new error check for 8bit text in MIME parts. Comparator and
5275 require names are now matched exactly. I enabled the subaddress
5276 extension, but it is not well tested yet (read: it works for me)."
5278 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5279 rework some of the code of TK/09 above to avoid the hardwired use of
5280 "%lld" and "long long". Replaced the call to snprintf() with a call to
5283 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5284 records point to non-existent hosts", "retry timeout exceeded", and
5285 "retry time not reached for any host after a long failure period".
5287 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5288 experimental DomainKeys support:
5290 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5291 (2) On an error such as an illegally used "control", the wrong name for
5292 the control was given.
5294 These problems did NOT occur unless DomainKeys support was compiled.
5296 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5298 PH/24 Added ${if match_ip condition.
5300 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5301 databases so that it will be absolutely obvious if a crash occurs in the
5302 DB library. This is a regular occurrence (often caused by mis-matched
5305 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5306 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5307 picked up on a user's system that detects such things. There doesn't seem
5308 to be a gcc warning option for this - only an attribute that has to be
5309 put on the function's prototype. It seems that in Fedora Core 4 they have
5310 set this on a number of new functions. No doubt there will be more in due
5313 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5314 specify a transport. However, if an address that was verified by such a
5315 router was the subject of a callout, Exim crashed because it tried to
5316 read the rcpt_include_affixes from the non-existent transport. Now it
5317 just assumes that the setting of that option is false. This bug was
5318 introduced by 4.51/PH/31.
5320 PH/28 Changed -d+all to exclude +memory, because that information is very
5321 rarely of interest, but it makes the output a lot bigger. People tend to
5322 do -d+all out of habit.
5324 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5325 code in os-type was giving problems when libc.so lives in lib64, like on
5328 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5329 aren't the modern standard, and it seems that some systems' include files
5330 don't always have them. Exim was already checking for some of the newer
5331 ones like T_AAAA, and defining it itself. I've added checks for all the
5332 record types that Exim uses.
5334 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5335 not automatically generating a new one, as it is supposed to. This
5336 prevented TLS from working. If the file did exist, but contained invalid
5337 data, a new version was generated, as expected. It was only the case of a
5338 non-existent file that was broken.
5340 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5341 with a change in libdomainkeys > 0.64.
5343 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5344 from DNS. If the selector record carries the flag, it now has
5345 precedence over the domain-wide flag.
5347 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5349 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5350 the use of a port name that isn't defined in /etc/services) occurred, the
5351 message was deferred as in a normal delivery, and thus remained on the
5352 spool, instead of being failed because of the mua_wrapper setting. This
5353 is now fixed, and I tidied up some of the mua_wrapper messages at the
5356 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5357 the messages in a hash of arrays rather than using individual hashes.
5358 This is a bit cleaner and results in dramatic memory savings, albeit
5359 at a slight CPU cost.
5361 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5362 as requested by Marc Sherman.
5364 SC/10 Eximstats: added histograms for user specified patterns as requested
5367 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5369 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5370 fopen() in the content-scanning modules that did not already have it.
5376 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5377 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5379 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5381 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5383 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5384 to test Sieve filters that use "vacation".
5386 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5387 that changes the way the GnuTLS parameters are stored in the cache file.
5388 The new format can be generated externally. For backward compatibility,
5389 if the data in the cache doesn't make sense, Exim assumes it has read an
5390 old-format file, and it generates new data and writes a new file. This
5391 means that you can't go back to an older release without removing the
5394 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5395 work if there are any delivery delays because "one_time" forces the
5396 parent to be marked "delivered", so its unseen clone is never tried
5397 again. For this reason, Exim now forbids the simultaneous setting of
5400 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5401 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5402 entry if you want to know the details. The fix, however, neglected to
5403 consider the case when local delivery batching is involved. The test for
5404 "previously delivered" was not happening when checking to see if an
5405 address could be batched with a previous (undelivered) one; under
5406 certain circumstances this could lead to multiple deliveries to the same
5409 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5410 in its include files, and this causes problems building Exim.
5412 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5413 header_syntax) but Exim was just ignoring anything given after a slash.
5414 In particular, this caused confusion with an attempt to use "verify =
5415 reverse_host_lookup/defer_ok". An error is now given when options are
5416 supplied for verify items that do not have them. (Maybe reverse_host_
5417 lookup should have a defer_ok option, but that's a different point.)
5419 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5420 defined by RFC 821) to 2048, because there were problems with some AUTH
5421 commands, and RFC 1869 says the size should be increased for extended
5422 SMTP commands that take arguments.
5424 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5427 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5428 "unknown" error; now it says that the functionality isn't in the binary.
5430 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5431 an address' error message when a string expansion fails (syntax or
5432 whatever). Otherwise the password may appear in the log. Following change
5433 PH/42 below, there is no longer a chance of it appearing in a bounce
5436 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5438 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5439 of its addresses was ever tried. (Bugzilla bug #2.)
5441 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5442 the result incorrectly in the debug output. (It correctly added a newline
5443 to what was transported.)
5445 TF/01 Added $received_time.
5447 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5448 commented out examples of how to interface to a virus scanner and to
5449 SpamAssassin. Also added commented examples of av_scanner and
5450 spamd_address settings.
5452 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5453 and controls are allowed in which ACLs. There were a couple of minor
5454 errors. Some of the entries in the conditions table (which is a table of
5455 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5456 negation of where the condition IS allowed.
5458 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5460 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5461 header file does not have a version number, so I've had to invent a new
5462 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5463 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5464 radiusclient), but it was contributed by a Radius user.
5466 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5467 files or directories, for OpenSSL.
5469 PH/19 When an Exim process that is running as root has to create an Exim log
5470 file, it does so in a subprocess that runs as exim:exim so as to get the
5471 ownership right at creation (otherwise, other Exim processes might see
5472 the file with the wrong ownership). There was no test for failure of this
5473 fork() call, which would lead to the process getting stuck as it waited
5474 for a non-existent subprocess. Forks do occasionally fail when resources
5475 run out. I reviewed all the other calls to fork(); they all seem to check
5478 PH/20 When checking for unexpected SMTP input at connect time (before writing
5479 the banner), Exim was not dealing correctly with a non-positive return
5480 from the read() function. If the client had disconnected by this time,
5481 the result was a log entry for a synchronization error with an empty
5482 string after "input=" when read() returned zero. If read() returned -1
5483 (an event I could not check), uninitialized data bytes were printed.
5484 There were reports of junk text (parts of files, etc) appearing after
5487 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5489 PH/22 Added support for macro redefinition, and (re)definition in between
5490 driver and ACL definitions.
5492 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5493 forgetting to use the resulting value; it was using the unexpanded value.
5495 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5496 hadn't been configured. The fix is from Juergen Kreileder, who
5497 understands it better than I do:
5499 "Here's what I see happening with three configured cyrus_sasl
5500 authenticators configured (plain, login, cram-md5):
5502 On startup auth_cyrus_sasl_init() gets called for each of these.
5503 This means three calls to sasl_listmech() without a specified mech_list.
5504 => SASL tests which mechs of all available mechs actually work
5505 => three warnings about OTP not working
5506 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5508 With the patch, sasl_listmech() also gets called three times. But now
5509 SASL's mech_list option is set to the server_mech specified in the the
5510 authenticator. Or in other words, the answer from sasl_listmech()
5511 gets limited to just the mech you're testing for (which is different
5513 => the return list contains just 'plain' or 'login', 'cram-md5' or
5514 nothing depending on the value of ob->server_mech.
5516 I've just tested the patch: Authentication still works fine,
5517 unavailable mechs specified in the exim configuration are still
5518 caught, and the auth.log warnings about OTP are gone."
5520 PH/25 When debugging is enabled, the contents of the command line are added
5521 to the debugging output, even when log_selector=+arguments is not
5524 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5525 answer is "GNU", and only if the return is "GNU/something" is the answer
5528 PH/27 $acl_verify_message is now set immediately after the failure of a
5529 verification in an ACL, and so is available in subsequent modifiers. In
5530 particular, the message can be preserved by coding like this:
5532 warn !verify = sender
5533 set acl_m0 = $acl_verify_message
5535 Previously, $acl_verify_message was set only while expanding "message"
5536 and "log_message" when a very denied access.
5538 PH/28 Modified OS/os.c-Linux with
5540 -#ifndef OS_LOAD_AVERAGE
5541 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5543 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5544 nomenclature these days.)
5546 PH/29 Installed patch from the Sieve maintainer that adds the options
5547 sieve_useraddress and sieve_subaddress to the redirect router.
5549 PH/30 In these circumstances:
5550 . Two addresses routed to the same list of hosts;
5551 . First host does not offer TLS;
5552 . First host accepts first address;
5553 . First host gives temporary error to second address;
5554 . Second host offers TLS and a TLS session is established;
5555 . Second host accepts second address.
5556 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5557 and peerdn, if requested) that were in fact used only for the second
5560 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5561 attention to any local part prefix or suffix that was matched by the
5562 router that accepted the address. It now behaves in the same way as it
5563 does for delivery: the affixes are removed from the local part unless
5564 rcpt_include_affixes is set on the transport.
5566 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5567 timeout during the DATA phase of an incoming message.
5569 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5570 applied a patch sanctioned by the Sieve maintainer.
5572 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5573 the uid or gid is negative. A case of a negative gid caused this to be
5574 noticed. The fix allows for either to be negative.
5576 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5577 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5580 PH/36 The change PH/12 above was broken. Fixed it.
5582 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5583 the grounds that routing the same address twice would always produce the
5584 same answer. This might have been true once, but it is certainly no
5585 longer true now. Routing a child address may depend on the previous
5586 routing that produced that child. Some complicated redirection strategies
5587 went wrong when messages had multiple recipients, and made Exim's
5588 behaviour dependent on the order in which the addresses were given.
5590 I have moved the duplicate checking until after the routing is complete.
5591 Exim scans the addresses that are assigned to local and remote
5592 transports, and removes any duplicates. This means that more work will be
5593 done, as duplicates will always all be routed, but duplicates are
5594 presumably rare, so I don't expect this is of any significance.
5596 For deliveries to pipes, files, and autoreplies, the duplicate checking
5597 still happens during the routing process, since they are not going to be
5600 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5601 It corrects a timeout issue with spamd. This is Ian's comment: "The
5602 background is that sometimes spamd either never reads data from a
5603 connection it has accepted, or it never writes response data. The exiscan
5604 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5605 blindly assumes that writes won't block so it may never time out."
5607 PH/39 Allow G after quota size as well as K and M.
5609 PH/40 The value set for $authenticated_id in an authenticator may not contain
5610 binary zeroes or newlines because the value is written to log lines and
5611 to spool files. There was no check on this. Now the value is run through
5612 the string_printing() function so that such characters are converted to
5613 printable escape sequences.
5615 PH/41 $message_linecount is a new variable that contains the total number of
5616 lines in the message. Compare $body_linecount, which is the count for the
5619 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5620 bounce and delay warning messages, except in certain special cases, which
5623 (a) An SMTP error message from a remote host;
5624 (b) A message specified in a :fail: redirection;
5625 (c) A message specified in a "fail" command in a system filter;
5626 (d) A message specified in a FAIL return from the queryprogram router;
5627 (e) A message specified by the cannot_route_message router option.
5629 In these cases only, Exim does include the error details in bounce and
5630 warning messages. There are also a few cases where bland messages such
5631 as "unrouteable address" or "local delivery error" are given.
5633 PH/43 $value is now also set for the "else" part of a ${run expansion.
5635 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5636 being worked on, but at least Exim now implements the latest version to
5639 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5640 process to complete was treated as a delivery failure, a timeout while
5641 writing the message to the pipe was logged, but erroneously treated as a
5642 successful delivery. Such timeouts include transport filter timeouts. For
5643 consistency with the overall process timeout, these timeouts are now
5644 treated as errors, giving rise to delivery failures by default. However,
5645 there is now a new Boolean option for the pipe transport called
5646 timeout_defer, which, if set TRUE, converts the failures into defers for
5647 both kinds of timeout. A transport filter timeout is now identified in
5650 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5651 systems where "make" and "gmake" are different, calling "gmake" at top
5652 level broke things. I've arranged for the value of $(MAKE) to be passed
5653 from the Makefile to this script so that it can call the same version of
5657 A note about Exim versions 4.44 and 4.50
5658 ----------------------------------------
5660 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5661 changes of various kinds. As a consequence, a big documentation update was
5662 needed. This delayed the release for rather longer than seemed good, especially
5663 in the light of a couple of (minor) security issues. Therefore, the changes
5664 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5665 release. So 4.44 and 4.50 are in effect two different branches that both start
5668 I have left the 4.50 change log unchanged; it contains all the changes since
5669 4.43. The change log for 4.44 is below; many of its items are identical to
5670 those for 4.50. This seems to be the most sensible way to preserve the
5671 historical information.
5677 1. Minor wording change to the doc/README.SIEVE file.
5679 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5680 computation of the current number of files was incorrect.
5682 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5683 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5686 4. Give more explanation in the error message when the command for a transport
5687 filter fails to execute.
5689 5. There are several places where Exim runs a non-Exim command in a
5690 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5691 was being done only for the command run by the queryprogram router. It is
5692 now done for all such subprocesses. The other cases are: ${run, transport
5693 filters, and the commands run by the lmtp and pipe transports.
5695 6. Added CONFIGURE_GROUP build-time option.
5697 7. Some older OS have a limit of 256 on the maximum number of file
5698 descriptors. Exim was using setrlimit() to set 1000 as a large value
5699 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5700 systems. I've change it so that if it can't get 1000, it tries for 256.
5702 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5703 was an oversight, and furthermore, ever since the addition of extra
5704 controls (e.g. 4.43/32), the checks on when to allow different forms of
5705 "control" were broken. There should now be diagnostics for all cases when a
5706 control that does not make sense is encountered.
5708 9. Added the /retain_sender option to "control=submission".
5710 10. $recipients is now available in the predata ACL (oversight).
5712 11. Tidy the search cache before the fork to do a delivery from a message
5713 received from the command line. Otherwise the child will trigger a lookup
5714 failure and thereby defer the delivery if it tries to use (for example) a
5715 cached ldap connection that the parent has called unbind on.
5717 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5718 of $address_data from the recipient verification was clobbered by the
5719 sender verification.
5721 13. The value of address_data from a sender verification is now available in
5722 $sender_address_data in subsequent conditions in the ACL statement.
5724 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5726 15. Added a new option "connect=<time>" to callout options, to set a different
5729 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5730 was its contents. (It was OK if the option was not defined at all.)
5732 17. A "Completed" log line is now written for messages that are removed from
5733 the spool by the -Mrm option.
5735 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5736 information about exactly what failed.
5738 19. Added -dd to debug only the daemon process.
5740 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5741 handles timeouts, both on the server side and network timeouts. Renamed the
5742 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5744 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5745 It is now set to "smtps".
5747 22. $host_address is now set to the target address during the checking of
5748 ignore_target_hosts.
5750 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5751 being passed; this would have caused $sender_host_name to have been used if
5752 matching the list had actually called for a host name (not very likely,
5753 since this list is usually IP addresses). A host name is now passed as
5756 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5757 code that specifies a non-restarting handler (typically sigaction() in
5758 modern systems) in an attempt to fix a rare and obscure crash bug.
5760 25. Narrowed the window for a race in the daemon that could cause it to ignore
5761 SIGCHLD signals. This is not a major problem, because they are used only to
5762 wake it up if nothing else does.
5764 26. A malformed maildirsize file could cause Exim to calculate negative values
5765 for the mailbox size or file count. Odd effects could occur as a result.
5766 The maildirsize information is now recalculated if the size or filecount
5769 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5770 support for a long time. Removed HAVE_SYS_VFS_H.
5772 28. Installed the latest version of exipick from John Jetmore.
5774 29. In an address list, if the pattern was not a regular expression, an empty
5775 subject address (from a bounce message) matched only if the pattern was an
5776 empty string. Non-empty patterns were not even tested. This was the wrong
5777 because it is perfectly reasonable to use an empty address as part of a
5778 database query. An empty address is now tested by patterns that are
5779 lookups. However, all the other forms of pattern expect the subject to
5780 contain a local part and a domain, and therefore, for them, an empty
5781 address still always fails if the pattern is not itself empty.
5783 30. Exim went into a mad DNS loop when attempting to do a callout where the
5784 host was specified on an smtp transport, and looking it up yielded more
5785 than one IP address.
5787 31. Re-factored the code for checking spool and log partition space into a
5788 function that finds that data and another that does the check. The former
5789 is then used to implement four new variables: $spool_space, $log_space,
5790 $spool_inodes, and $log_inodes.
5792 32. The RFC2047 encoding function was originally intended for short strings
5793 such as real names; it was not keeping to the 75-character limit for
5794 encoded words that the RFC imposes. It now respects the limit, and
5795 generates multiple encoded words if necessary. To be on the safe side, I
5796 have increased the buffer size for the ${rfc2047: expansion operator from
5799 33. It is now permitted to omit both strings after an "if" condition; if the
5800 condition is true, the result is "true". As before, when the second string
5801 is omitted, a false condition yields an empty string. This makes it less
5802 cumbersome to write custom ACL and router conditions.
5804 34. Failure to deliver a bounce message always caused it to be frozen, even if
5805 there was an errors_to setting on the router. The errors_to setting is now
5808 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5809 canonical form (fully expanded) before being placed in
5810 $sender_host_address.
5812 36. The table in the code that translates DNS record types into text (T_A to
5813 "A" for instance) was missing entries for NS and CNAME. It is just possible
5814 that this could have caused confusion if both these types were looked up
5815 for the same domain, because the text type is used as part of Exim's
5816 per-process caching. But the chance of anyone hitting this buglet seems
5819 37. The dnsdb lookup has been extended in a number of ways.
5821 (1) There is a new type, "zns", which walks up the domain tree until it
5822 finds some nameserver records. It should be used with care.
5824 (2) There is a new type, "mxh", which is like "mx" except that it returns
5825 just the host names, not the priorities.
5827 (3) It is now possible to give a list of domains (or IP addresses) to be
5828 looked up. The behaviour when one of the lookups defers can be
5829 controlled by a keyword.
5831 (4) It is now possible to specify the separator character for use when
5832 multiple records are returned.
5834 38. The dnslists ACL condition has been extended: it is now possible to supply
5835 a list of IP addresses and/or domains to be looked up in a particular DNS
5838 39. Added log_selector=+queue_time_overall.
5840 40. When running the queue in the test harness, wait just a tad after forking a
5841 delivery process, to get repeatability of debugging output.
5843 41. Include certificate and key file names in error message when GnuTLS fails
5844 to set them up, because the GnuTLS error message doesn't include the name
5845 of the failing file when there is a problem reading it.
5847 42. Allow both -bf and -bF in the same test run.
5849 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5851 44. The "Exiscan patch" is now merged into the mainline Exim source.
5853 45. Sometimes the final signoff response after QUIT could fail to get
5854 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5855 < 0 before doing a fflush(). This bug looks as though it goes back to the
5856 introduction of TLS in release 3.20, but "sometimes" must have been rare
5857 because the tests only now provoked it.
5859 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5860 (this can affect the format of dates).
5862 47. exim_tidydb, when checking for the continued existence of a message for
5863 which it has found a message-specific retry record, was not finding
5864 messages that were in split spool directories. Consequently, it was
5865 deleting retry records that should have stayed in existence.
5867 48. Steve fixed some bugs in eximstats.
5869 49. The SPA authentication driver was not abandoning authentication and moving
5870 on to the next authenticator when an expansion was forced to fail,
5871 contradicting the general specification for all authenticators. Instead it
5872 was generating a temporary error. It now behaves as specified.
5874 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5875 (the order specifies the preference for clients). The order is now AES256,
5876 AES128, 3DES, ARCFOUR128.
5878 51. Small patch to Sieve code - explicitly set From: when generating an
5881 52. Exim crashed if a remote delivery caused a very long error message to be
5882 recorded - for instance if somebody sent an entire SpamAssassin report back
5883 as a large number of 550 error lines. This bug was coincidentally fixed by
5884 increasing the size of one of Exim's internal buffers (big_buffer) that
5885 happened as part of the Exiscan merge. However, to be on the safe side, I
5886 have made the code more robust (and fixed the comments that describe what
5889 53. Now that there can be additional text after "Completed" in log lines (if
5890 the queue_time_overall log selector is set), a one-byte patch to exigrep
5891 was needed to allow it to recognize "Completed" as not the last thing in
5894 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5895 patch that reportedly fixes this has been added. I am not expert enough to
5896 create a test for it. This is what the patch creator wrote:
5898 "I found a little strange behaviour of ldap code when working with
5899 Windows 2003 AD Domain, where users was placed in more than one
5900 Organization Units. When I tried to give exim partial DN, the exit code
5901 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5902 But simultaneously result of request was absolutely normal ldap result,
5903 so I produce this patch..."
5905 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5906 REFERENCE, so I have modified the code to exclude the patch when that macro
5909 55. Some experimental protocols are using DNS PTR records for new purposes. The
5910 keys for these records are domain names, not reversed IP addresses. The
5911 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5912 leaves it alone. Component reversal etc. now happens only for IP addresses.
5915 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5917 57. Double the size of the debug message buffer (to 2048) so that more of very
5918 long debug lines gets shown.
5920 58. The exicyclog utility now does better if the number of log files to keep
5921 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5923 59. Two changes related to the smtp_active_hostname option:
5925 (1) $smtp_active_hostname is now available as a variable.
5926 (2) The default for smtp_banner uses $smtp_active_hostname instead
5927 of $primary_hostname.
5929 60. The host_aton() function is supposed to be passed a string that is known
5930 to be a valid IP address. However, in the case of IPv6 addresses, it was
5931 not checking this. This is a hostage to fortune. Exim now panics and dies
5932 if the condition is not met. A case was found where this could be provoked
5933 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5934 components; fortuitously, this particular loophole had already been fixed
5935 by change 4.50/55 above.
5937 If there are any other similar loopholes, the new check in host_aton()
5938 itself should stop them being exploited. The report I received stated that
5939 data on the command line could provoke the exploit when Exim was running as
5940 exim, but did not say which command line option was involved. All I could
5941 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5942 running as the user.
5945 61. There was a buffer overflow vulnerability in the SPA authentication code
5946 (which came originally from the Samba project). I have added a test to the
5947 spa_base64_to_bits() function which I hope fixes it.
5950 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5951 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5953 63. The daemon start-up calls getloadavg() while still root for those OS that
5954 need the first call to be done as root, but it missed one case: when
5955 deliver_queue_load_max is set with deliver_drop_privilege. This is
5956 necessary for the benefit of the queue runner, because there is no re-exec
5957 when deliver_drop_privilege is set.
5959 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5960 This has been fixed.
5962 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5963 was in use, was not putting the data itself into the right store pool;
5964 consequently, it could be overwritten for a subsequent message in the same
5965 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5968 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5970 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5971 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5972 the function were treating the return as a boolean value, which happened to
5973 work because 0=false and not-0=true, but is not correct code.
5975 68. The host_aton() function was not handling scoped IPv6 addresses (those
5976 with, for example, "%eth0" on the end) correctly.
5978 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5979 negated items (that is, ~something) in unsigned ints. Some compilers
5980 apparently mutter when there is no cast.
5982 70. If an address verification called from an ACL failed, and did not produce a
5983 user-specific message (i.e. there was only a "system" message), nothing was
5984 put in $acl_verify_message. In this situation, it now puts the system
5987 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5988 session; change 4.31/43 added the unwanted input to the log line - except
5989 that it did not do this in the start of session case. It now does.
5991 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5992 This is wrong and can cause the other end to generate a synchronization
5993 error if it is another Exim or anything else that does the synchronization
5994 check. A QUIT command is no longer sent after a timeout.
5996 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5997 during host lookups.
5999 74. The defer_ok option of callout verification was not working if it was used
6000 when verifying addresses in header lines, that is, for this case:
6002 verify = header_sender/callout=defer_ok
6004 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6005 those file descriptors could be used for SMTP connections. If anything
6006 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6007 could be sent to the SMTP client, causing chaos. The daemon now opens
6008 stdin, stdout, and stderr to /dev/null when it puts itself into the
6011 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6012 log by default. The user can override this with suitable Perl magic.
6014 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6015 the log message when discard triggers, was not working for the DATA ACL or
6016 for the non-SMTP ACL.
6018 78. Error message wording change in sieve.c.
6020 79. If smtp_accept_max_per_host was set, the number of connections could be
6021 restricted to fewer than expected, because the daemon was trying to set up
6022 a new connection before checking whether the processes handling previous
6023 connections had finished. The check for completed processes is now done
6024 earlier. On busy systems, this bug wouldn't be noticed because something
6025 else would have woken the daemon, and it would have reaped the completed
6028 80. If a message was submitted locally by a user whose login name contained one
6029 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6030 It caused a spool format error. I have fixed the spool reading code. A
6031 related problem was that the "from" clause in the Received: line became
6032 illegal because of the space(s). It is now covered by ${quote_local_part.
6034 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6037 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6039 83. Patch from Sieve maintainer to fix unterminated string problem in
6040 "vacation" handling.
6042 84. Some minor changes to the Linux configuration files to help with other
6043 OS variants using glibc.
6045 85. One more patch for Sieve to update vacation handling to latest spec.
6048 ----------------------------------------------------
6049 See the note above about the 4.44 and 4.50 releases.
6050 ----------------------------------------------------
6056 1. Change 4.43/35 introduced a bug that caused file counts to be
6057 incorrectly computed when quota_filecount was set in an appendfile
6060 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6061 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6064 3. Give more explanation in the error message when the command for a transport
6065 filter fails to execute.
6067 4. There are several places where Exim runs a non-Exim command in a
6068 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6069 was being done only for the command run by the queryprogram router. It is
6070 now done for all such subprocesses. The other cases are: ${run, transport
6071 filters, and the commands run by the lmtp and pipe transports.
6073 5. Some older OS have a limit of 256 on the maximum number of file
6074 descriptors. Exim was using setrlimit() to set 1000 as a large value
6075 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6076 systems. I've change it so that if it can't get 1000, it tries for 256.
6078 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6079 was an oversight, and furthermore, ever since the addition of extra
6080 controls (e.g. 4.43/32), the checks on when to allow different forms of
6081 "control" were broken. There should now be diagnostics for all cases when a
6082 control that does not make sense is encountered.
6084 7. $recipients is now available in the predata ACL (oversight).
6086 8. Tidy the search cache before the fork to do a delivery from a message
6087 received from the command line. Otherwise the child will trigger a lookup
6088 failure and thereby defer the delivery if it tries to use (for example) a
6089 cached ldap connection that the parent has called unbind on.
6091 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6092 of $address_data from the recipient verification was clobbered by the
6093 sender verification.
6095 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6096 was its contents. (It was OK if the option was not defined at all.)
6098 11. A "Completed" log line is now written for messages that are removed from
6099 the spool by the -Mrm option.
6101 12. $host_address is now set to the target address during the checking of
6102 ignore_target_hosts.
6104 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6105 being passed; this would have caused $sender_host_name to have been used if
6106 matching the list had actually called for a host name (not very likely,
6107 since this list is usually IP addresses). A host name is now passed as
6110 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6111 code that specifies a non-restarting handler (typically sigaction() in
6112 modern systems) in an attempt to fix a rare and obscure crash bug.
6114 15. Narrowed the window for a race in the daemon that could cause it to ignore
6115 SIGCHLD signals. This is not a major problem, because they are used only to
6116 wake it up if nothing else does.
6118 16. A malformed maildirsize file could cause Exim to calculate negative values
6119 for the mailbox size or file count. Odd effects could occur as a result.
6120 The maildirsize information is now recalculated if the size or filecount
6123 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6124 support for a long time. Removed HAVE_SYS_VFS_H.
6126 18. Updated exipick to current release from John Jetmore.
6128 19. Allow an empty sender to be matched against a lookup in an address list.
6129 Previously the only cases considered were a regular expression, or an
6132 20. Exim went into a mad DNS lookup loop when doing a callout where the
6133 host was specified on the transport, if the DNS lookup yielded more than
6136 21. The RFC2047 encoding function was originally intended for short strings
6137 such as real names; it was not keeping to the 75-character limit for
6138 encoded words that the RFC imposes. It now respects the limit, and
6139 generates multiple encoded words if necessary. To be on the safe side, I
6140 have increased the buffer size for the ${rfc2047: expansion operator from
6143 22. Failure to deliver a bounce message always caused it to be frozen, even if
6144 there was an errors_to setting on the router. The errors_to setting is now
6147 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6148 canonical form (fully expanded) before being placed in
6149 $sender_host_address.
6151 24. Updated eximstats to version 1.33
6153 25. Include certificate and key file names in error message when GnuTLS fails
6154 to set them up, because the GnuTLS error message doesn't include the name
6155 of the failing file when there is a problem reading it.
6157 26. Expand error message when OpenSSL has problems setting up cert/key files.
6160 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6161 (this can affect the format of dates).
6163 28. exim_tidydb, when checking for the continued existence of a message for
6164 which it has found a message-specific retry record, was not finding
6165 messages that were in split spool directories. Consequently, it was
6166 deleting retry records that should have stayed in existence.
6168 29. eximstats updated to version 1.35
6169 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6170 1.35 - bugfix such that pie charts by volume are generated correctly
6172 30. The SPA authentication driver was not abandoning authentication and moving
6173 on to the next authenticator when an expansion was forced to fail,
6174 contradicting the general specification for all authenticators. Instead it
6175 was generating a temporary error. It now behaves as specified.
6177 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6178 (the order specifies the preference for clients). The order is now AES256,
6179 AES128, 3DES, ARCFOUR128.
6181 31. Small patch to Sieve code - explicitly set From: when generating an
6184 32. Exim crashed if a remote delivery caused a very long error message to be
6185 recorded - for instance if somebody sent an entire SpamAssassin report back
6186 as a large number of 550 error lines. This bug was coincidentally fixed by
6187 increasing the size of one of Exim's internal buffers (big_buffer) that
6188 happened as part of the Exiscan merge. However, to be on the safe side, I
6189 have made the code more robust (and fixed the comments that describe what
6192 33. Some experimental protocols are using DNS PTR records for new purposes. The
6193 keys for these records are domain names, not reversed IP addresses. The
6194 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6195 leaves it alone. Component reversal etc. now happens only for IP addresses.
6198 34. The host_aton() function is supposed to be passed a string that is known
6199 to be a valid IP address. However, in the case of IPv6 addresses, it was
6200 not checking this. This is a hostage to fortune. Exim now panics and dies
6201 if the condition is not met. A case was found where this could be provoked
6202 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6203 components; fortuitously, this particular loophole had already been fixed
6204 by change 4.50/55 or 4.44/33 above.
6206 If there are any other similar loopholes, the new check in host_aton()
6207 itself should stop them being exploited. The report I received stated that
6208 data on the command line could provoke the exploit when Exim was running as
6209 exim, but did not say which command line option was involved. All I could
6210 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6211 running as the user.
6214 35. There was a buffer overflow vulnerability in the SPA authentication code
6215 (which came originally from the Samba project). I have added a test to the
6216 spa_base64_to_bits() function which I hope fixes it.
6219 36. The daemon start-up calls getloadavg() while still root for those OS that
6220 need the first call to be done as root, but it missed one case: when
6221 deliver_queue_load_max is set with deliver_drop_privilege. This is
6222 necessary for the benefit of the queue runner, because there is no re-exec
6223 when deliver_drop_privilege is set.
6225 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6226 was in use, was not putting the data itself into the right store pool;
6227 consequently, it could be overwritten for a subsequent message in the same
6228 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6231 38. Sometimes the final signoff response after QUIT could fail to get
6232 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6233 < 0 before doing a fflush(). This bug looks as though it goes back to the
6234 introduction of TLS in release 3.20, but "sometimes" must have been rare
6235 because the tests only now provoked it.
6241 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6242 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6243 Now it returns an integer. A number of calls were still expecting a T/F
6244 return. Fortuitously, in all cases, the tests worked in OK situations,
6245 which is the norm. However, things would have gone wrong on any write
6246 failures on the smtp file descriptor. This function is used when sending
6247 messages over SMTP and also when doing verify callouts.
6249 2. When Exim is called to do synchronous delivery of a locally submitted
6250 message (the -odf or -odi options), it no longer closes stderr before doing
6253 3. Implemented the mua_wrapper option.
6255 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6257 5. Implemented the functions header_remove(), header_testname(),
6258 header_add_at_position(), and receive_remove_recipient(), and exported them
6261 6. If an ACL "warn" statement specified the addition of headers, Exim already
6262 inserted X-ACL-Warn: at the start if there was no header name. However, it
6263 was not making this test for the second and subsequent header lines if
6264 there were newlines in the string. This meant that an invalid header could
6265 be inserted if Exim was badly configured.
6267 7. Allow an ACL "warn" statement to add header lines at the start or after all
6268 the Received: headers, as well as at the end.
6270 8. Added the rcpt_4xx retry error code.
6272 9. Added postmaster_mailfrom=xxx to callout verification option.
6274 10. Added mailfrom=xxxx to the callout verification option, for verify=
6277 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6278 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6280 12. Inserted some casts to stop certain compilers warning when using pointer
6281 differences as field lengths or precisions in printf-type calls (mostly
6282 affecting debugging statements).
6284 13. Added optional readline() support for -be (dynamically loaded).
6286 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6287 same clock tick as a message's arrival, so that its received time was the
6288 same as the "first fail" time on the retry record, and that message
6289 remained on the queue past the ultimate address timeout, every queue runner
6290 would try a delivery (because it was past the ultimate address timeout) but
6291 after another failure, the ultimate address timeout, which should have then
6292 bounced the address, did not kick in. This was a "< instead of <=" error;
6293 in most cases the first failure would have been in the next clock tick
6294 after the received time, and all would be well.
6296 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6297 being recognized when the domain list was tested by the match_domain
6298 condition in an expansion string.
6300 16. Added the ${str2b64: operator.
6302 17. Exim was always calling setrlimit() to set a large limit for the number of
6303 processes, without checking whether the existing limit was already
6304 adequate. (It did check for the limit on file descriptors.) Furthermore,
6305 errors from getrlimit() and setrlimit() were being ignored. Now they are
6306 logged to the main and panic logs, but Exim does carry on, to try to do its
6307 job under whatever limits there are.
6309 18. Imported PCRE 5.0.
6311 19. Trivial typo in log message " temporarily refused connection" (the leading
6314 20. If the log selector return_path_on_delivery was set and an address was
6315 redirected to /dev/null, the delivery process crashed because it assumed
6316 that a return path would always be set for a "successful" delivery. In this
6317 case, the whole delivery is bypassed as an optimization, and therefore no
6320 21. Internal re-arrangement: the function for sending a challenge and reading
6321 a response while authentication was assuming a zero-terminated challenge
6322 string. It's now changed to take a pointer and a length, to allow for
6323 binary data in such strings.
6325 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6327 23. Exim was not respecting finduser_retries when seeking the login of the
6328 uid under which it was called; it was always trying 10 times. (The default
6329 setting of finduser_retries is zero.) Also, it was sleeping after the final
6330 failure, which is pointless.
6332 24. Implemented tls_on_connect_ports.
6334 25. Implemented acl_smtp_predata.
6336 26. If the domain in control=submission is set empty, Exim assumes that the
6337 authenticated id is a complete email address when it generates From: or
6338 Sender: header lines.
6340 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6341 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6342 chown and chgrp in /bin and hostname in /usr/bin.
6344 28. Exim was keeping the "process log" file open after each use, just as it
6345 does for the main log. This opens the possibility of it remaining open for
6346 long periods when the USR1 signal hits a daemon. Occasional processlog
6347 errors were reported, that could have been caused by this. Anyway, it seems
6348 much more sensible not to leave this file open at all, so that is what now
6351 29. The long-running daemon process does not normally write to the log once it
6352 has entered its main loop, and it closes the log before doing so. This is
6353 so that log files can straightforwardly be renamed and moved. However,
6354 there are a couple of unusual error situations where the daemon does write
6355 log entries, and I had neglected to close the log afterwards.
6357 30. The text of an SMTP error response that was received during a remote
6358 delivery was being truncated at 512 bytes. This is too short for some of
6359 the long messages that one sometimes sees. I've increased the limit to
6362 31. It is now possible to make retry rules that apply only when a message has a
6363 specific sender, in particular, an empty sender.
6365 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6366 it possible to be selective about when SMTP synchronization is enforced.
6368 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6370 32. Implemented hosts_connection_nolog.
6372 33. Added an ACL for QUIT.
6374 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6377 35. Added mailbox_size and mailbox_filecount to appendfile.
6379 36. Added control = no_multiline_responses to ACLs.
6381 37. There was a bug in the logic of the code that waits for the clock to tick
6382 in the case where the clock went backwards by a substantial amount such
6383 that the microsecond fraction of "now" was more than the microsecond
6384 fraction of "then" (but the whole seconds number was less).
6386 38. Added support for the libradius Radius client library this is found on
6387 FreeBSD (previously only the radiusclient library was supported).
6393 1. When certain lookups returned multiple values in the form name=value, the
6394 quoting of the values was not always being done properly. Specifically:
6395 (a) If the value started with a double quote, but contained no whitespace,
6397 (b) If the value contained whitespace other than a space character (i.e.
6398 tabs or newlines or carriage returns) it was not quoted.
6399 This fix has been applied to the mysql and pgsql lookups by writing a
6400 separate quoting function and calling it from the lookup code. The fix
6401 should probably also be applied to nisplus, ibase and oracle lookups, but
6402 since I cannot test any of those, I have not disturbed their existing code.
6404 2. A hit in the callout cache for a specific address caused a log line with no
6405 reason for rejecting RCPT. Now it says "Previous (cached) callout
6406 verification failure".
6408 3. There was an off-by-one bug in the queryprogram router. An over-long
6409 return line was truncated at 256 instead of 255 characters, thereby
6410 overflowing its buffer with the terminating zero. As well as fixing this, I
6411 have increased the buffer size to 1024 (and made a note to document this).
6413 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6414 when Exim is waiting for an SMTP response from a remote server, Exim
6415 restarts its select() call on the socket, thereby resetting its timeout.
6416 This is not a problem when such interrupts are rare. Somebody set up a cron
6417 job to run exiwhat every 2 minutes, which is less than the normal select()
6418 timeout (5 or 10 minutes). This meant that the select() timeout never
6419 kicked in because it was always reset. I have fixed this by comparing the
6420 time when an interrupt arrives with the time at the start of the first call
6421 to select(). If more time than the timeout has elapsed, the interrupt is
6422 treated as a timeout.
6424 5. Some internal re-factoring in preparation for the addition of Sieve
6425 extensions (by MH). In particular, the "personal" test is moved to a
6426 separate function, and given an option for scanning Cc: and Bcc: (which is
6427 not set for Exim filters).
6429 6. When Exim created an email address using the login of the caller as the
6430 local part (e.g. when creating a From: or Sender: header line), it was not
6431 quoting the local part when it contained special characters such as @.
6433 7. Installed new OpenBSD configuration files.
6435 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6436 try to make them clearer.
6438 9. Callout options, other than the timeout value, were being ignored when
6439 verifying sender addresses in header lines. For example, when using
6441 verify = header_sender/callout=no_cache
6443 the cache was (incorrectly) being used.
6445 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6446 only the Cygwin environment.
6448 11. When return_path_on_delivery was set as a log selector, if different remote
6449 addresses in the same message used different return paths and parallel
6450 remote delivery occurred, the wrong values would sometimes be logged.
6451 (Whenever a remote delivery process finished, the return path value from
6452 the most recently started remote delivery process was logged.)
6454 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6455 lines when AUTH and/or TLS are in use. This is the "received protocol"
6456 field. Exim used to use "asmtp" for authenticated SMTP, without any
6457 indication (in the protocol name) for TLS use. Now it follows the RFC and
6458 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6459 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6460 names appear in log lines as well as in Received: header lines.
6462 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6463 extensions, and comparison tests, and to fix some bugs.
6465 14. Changes to the "personal" filter test:
6467 (1) The test was buggy in that it was just doing the equivalent of
6468 "contains" tests on header lines. For example, if a user's address was
6469 anne@some.where, the "personal" test would incorrectly be true for
6471 To: susanne@some.where
6473 This test is now done by extracting each address from the header in turn,
6474 and checking the entire address. Other tests that are part of "personal"
6475 are now done using regular expressions (for example, to check local parts
6476 of addresses in From: header lines).
6478 (2) The list of non-personal local parts in From: addresses has been
6479 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6480 taken from the Sieve specification recommendations.
6482 (3) If the message contains any header line starting with "List-" it is
6483 treated as non-personal.
6485 (4) The test for "circular" in the Subject: header line has been removed
6486 because it now seems ill-conceived.
6488 15. Minor typos in src/EDITME comments corrected.
6490 16. Installed latest exipick from John Jetmore.
6492 17. If headers_add on a router specified a text string that was too long for
6493 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6494 of string_sprintf() is now avoided.
6496 18. $message_body_size was not set (it was always zero) when running the DATA
6497 ACL and the local_scan() function.
6499 19. For the "mail" command in an Exim filter, no default was being set for
6500 the once_repeat time, causing a random time value to be used if "once" was
6501 specified. (If the value happened to be <= 0, no repeat happened.) The
6502 default is now 0s, meaning "never repeat". The "vacation" command was OK
6503 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6504 (I found it when inspecting the code).
6506 20. There is now an overall timeout for performing a callout verification. It
6507 defaults to 4 times the callout timeout, which applies to individual SMTP
6508 commands during the callout. The overall timeout applies when there is more
6509 than one host that can be tried. The timeout is checked before trying the
6510 next host. This prevents very long delays if there are a large number of
6511 hosts and all are timing out (e.g. when the network connections are timing
6512 out). The value of the overall timeout can be changed by specifying an
6513 additional sub-option for "callout", called "maxwait". For example:
6515 verify = sender/callout=5s,maxwait=20s
6517 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6518 to the end before writing, but this should make it even safer).
6520 22. Exim was forgetting that it had advertised PIPELINING for the second and
6521 subsequent messages on an SMTP connection. It was also not resetting its
6522 memory on STARTTLS and an internal HELO.
6524 23. When Exim logs an SMTP synchronization error within a session, it now
6525 records whether PIPELINING has been advertised or not.
6527 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6528 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6529 rather than long int.
6531 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6533 26. Added the never_mail option to autoreply.
6539 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6540 crash if the getsockname() call failed; this can happen if a connection is
6541 closed very soon after it is established. The problem was simply in the
6542 order in which certain operations were done, causing Exim to try to write
6543 to the SMTP stream before it had set up the file descriptor. The bug has
6544 been fixed by making things happen in the correct order.
6550 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6551 before the connection was closed, thus losing the rejection response.
6553 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6554 some early Solaris releases, but causes trouble in current releases where
6555 socklen_t is defined.
6557 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6560 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6563 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6564 that releasing the top of it at the end releases what was used for sub-
6565 expansions (unless the block got too big). However, discard this block if
6566 the first thing is a variable or header, so that we can use its block when
6567 it is dynamic (useful for very large $message_headers, for example).
6569 6. Lookups now cache *every* query, not just the most recent. A new, separate
6570 store pool is used for this. It can be recovered when all lookup caches are
6571 flushed. Lookups now release memory at the end of their result strings.
6572 This has involved some general refactoring of the lookup sources.
6574 7. Some code has been added to the store_xxx() functions to reduce the amount
6575 of flapping under certain conditions.
6577 8. log_incoming_interface used to affect only the <= reception log lines. Now
6578 it causes the local interface and port to be added to several more SMTP log
6579 lines, for example "SMTP connection from", and rejection lines.
6581 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6583 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6585 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6586 could be overwritten at the end of the current message (or the start of a
6587 new message if it was set in a HELO ACL). The value is now preserved for
6588 the duration of the SMTP connection.
6590 12. If a transport had a headers_rewrite setting, and a matching header line
6591 contained an unqualified address, that address was qualified, even if it
6592 did not match any rewriting rules. The underlying bug was that the values
6593 of the flags that permit the existence of unqualified sender and recipient
6594 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6595 non-local messages, and by -bnq for local messages) were not being
6596 preserved with the message after it was received.
6598 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6599 "next input=" as part of the text comprising the host identity instead of
6600 the correct text. The code was using the same buffer for two different
6601 strings. However, depending on which order the printing function evaluated
6602 its arguments, the bug did not always show up. Under Linux, for example, my
6603 test suite worked just fine.
6605 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6606 doesn't work with some older versions of Perl. It has been changed to "my",
6607 which in any case is probably the better facility to use.
6609 15. A really picky compiler found some instances of statements for creating
6610 error messages that either had too many or two few arguments for the format
6613 16. The size of the buffer for calls to the DNS resolver has been increased
6614 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6615 for addresses that have a lot of PTR records. This alleviates a problem; it
6616 does not fully solve it.
6618 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6619 buffer now truncates the list and logs the incident, which is the same
6620 action as happens when Exim is looking up a host name and its aliases.
6621 Previously in this situation something unpredictable would happen;
6622 sometimes it was "internal error: store_reset failed".
6624 18. If a server dropped the connection unexpectedly when an Exim client was
6625 using GnuTLS and trying to read a response, the client delivery process
6626 crashed while trying to generate an error log message.
6628 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6629 string, for example:
6631 warn message = H1: something\nH2: something
6633 the text was added as a single header line from Exim's point of view
6634 though it ended up OK in the delivered message. However, searching for the
6635 second and subsequent header lines using $h_h2: did not work. This has been
6636 fixed. Similarly, if a system filter added multiple headers in this way,
6637 the routers could not see them.
6639 20. Expanded the error message when iplsearch is called with an invalid key to
6640 suggest using net-iplsearch in a host list.
6642 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6643 ACLs are no longer actually imposed (and a message to that effect is
6646 22. If a "gecos" field in a passwd entry contained escaped characters, in
6647 particular, if it contained a \" sequence, Exim got it wrong when building
6648 a From: or a Sender: header from that name. A second bug also caused
6649 incorrect handling when an unquoted " was present following a character
6650 that needed quoting.
6652 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6653 was not being matched caselessly.
6655 24. Arranged for all hyphens in the exim.8 source to be escaped with
6658 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6659 properly. Recipient callout cache records were still being keyed to include
6660 the sender, even when use_sender was set false. This led to far more
6661 callouts that were necessary. The sender is no longer included in the key
6662 when use_sender is false.
6664 26. Added "control = submission" modifier to ACLs.
6666 27. Added the ${base62d: operator to decode base 62 numbers.
6668 28. dnsdb lookups can now access SRV records.
6670 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6671 the configuration file.
6673 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6674 (-v) mode. This makes the output for a verbose queue run more intelligible.
6676 31. Added a use_postmaster feature to recipient callouts.
6678 32. Added the $body_zerocount variable, containing the number of binary zero
6679 bytes in the message body.
6681 33. The time of last modification of the "new" subdirectory is now used as the
6682 "mailbox time last read" when there is a quota error for a maildir
6685 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6687 35. Added +ignore_unknown as a special item in host lists.
6689 36. Code for decoding IPv6 addresses in host lists is now included, even if
6690 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6691 address was recognized as an IP address, but was then not correctly decoded
6692 into binary, causing unexpected and incorrect effects when compared with
6699 1. Very minor rewording of debugging text in manualroute to say "list of
6700 hosts" instead of "hostlist".
6702 2. If verify=header_syntax was set, and a header line with an unqualified
6703 address (no domain) and a large number of spaces between the end of the
6704 name and the colon was received, the reception process suffered a buffer
6705 overflow, and (when I tested it) crashed. This was caused by some obsolete
6706 code that should have been removed. The fix is to remove it!
6708 3. When running in the test harness, delay a bit after writing a bounce
6709 message to get a bit more predictability in the log output.
6711 4. Added a call to search_tidyup() just before forking a reception process. In
6712 theory, someone could use a lookup in the expansion of smtp_accept_max_
6713 per_host which, without the tidyup, could leave open a database connection.
6715 5. Added the variables $recipient_data and $sender_data which get set from a
6716 lookup success in an ACL "recipients" or "senders" condition, or a router
6717 "senders" option, similar to $domain_data and $local_part_data.
6719 6. Moved the writing of debug_print from before to after the "senders" test
6722 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6723 problems for message scanning, either using a data ACL, or using
6724 local_scan() because the Received: header was not generated till after they
6725 were called (in order to set the time as the time of reception completion).
6726 I have revised the way this works. The header is now generated after the
6727 body is received, but before the ACL or local_scan() are called. After they
6728 are run, the timestamp in the header is updated.
6734 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6735 before starting a queue runner without re-exec. This happened only when
6736 deliver_drop_privilege was set or when the Exim user was set to root. The
6737 effect of the bug was that timeouts during subsequent deliveries caused
6738 crashes instead of being properly handled. The handler is now left at its
6739 default (and expected) setting.
6741 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6742 message, again when deliver_drop_privilege is set or Exim is run as root.
6743 The bug described in (1) was not present in this case, but the tidying up
6744 of the other signals was missing. I have made the two cases consistent.
6746 3. The ignore_target_hosts setting on a manualroute router was being ignored
6747 for hosts that were looked up using the /MX notation.
6749 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6752 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6753 operated on the sender address. After changing the $sender_address to <>
6754 for the sender address verify, Exim was re-instated it as the original
6755 (before rewriting) address, but remembering that it had rewritten it, so it
6756 wasn't rewriting it again. This bug also had the effect of breaking the
6757 sender address verification caching when the sender address was rewritten.
6759 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6760 This has been changed so that if the ip literal address matches
6761 ignore_target_hosts, the router declines.
6763 7. Added expansion conditions match_domain, match_address, and match_local_
6764 part (NOT match_host).
6766 8. The placeholder for the Received: header didn't have a length field set.
6768 9. Added code to Exim itself and to exim_lock to test for a specific race
6769 condition that could lead to file corruption when using MBX delivery. The
6770 issue is with the lockfile that is created in /tmp. If this file is removed
6771 after a process has opened it but before that process has acquired a lock,
6772 there is the potential for a second process to recreate the file and also
6773 acquire a lock. This could lead to two Exim processes writing to the file
6774 at the same time. The added code performs the same test as UW imapd; it
6775 checks after acquiring the lock that its file descriptor still refers to
6776 the same named file.
6778 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6779 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6782 11. Added the smtp_active_hostname option. If used, this will typically be made
6783 to depend on the incoming interface address. Because $interface_address is
6784 not set up until the daemon has forked a reception process, error responses
6785 that can happen earlier (such as "too many connections") no longer contain
6788 12. If an expansion in a condition on a "warn" statement fails because a lookup
6789 defers, the "warn" statement is abandoned, and the next ACL statement is
6790 processed. Previously this caused the whole ACL to be aborted.
6792 13. Added the iplsearch lookup type.
6794 14. Added ident_timeout as a log selector.
6796 15. Added tls_certificate_verified as a log selector.
6798 16. Added a global option tls_require_ciphers (compare the smtp transport
6799 option of the same name). This controls incoming TLS connections.
6801 17. I finally figured out how to make tls_require_ciphers do a similar thing
6802 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6803 before starting the TLS session.
6805 18. Tabs are now shown as \t in -bP output.
6807 19. If the log selector return_path_on_delivery was set, Exim crashed when
6808 bouncing a message because it had too many Received: header lines.
6810 20. If two routers both had headers_remove settings, and the first one included
6811 a superfluous trailing colon, the final name in the first list and the
6812 first name in the second list were incorrectly joined into one item (with a
6813 colon in the middle).
6819 1. Added -C and -D options to the exinext utility, mainly to make it easier
6820 to include in the automated testing, but these could be helpful when
6821 multiple configurations are in use.
6823 2. The exinext utility was not formatting the output nicely when there was
6824 an alternate port involved in the retry record key, nor when there was a
6825 message id as well (for retries that were specific to a specific message
6826 and a specific host). It was also confused by IPv6 addresses, because of
6827 the additional colons they contain. I have fixed the IPv4 problem, and
6828 patched it up to do a reasonable job for IPv6.
6830 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6831 delivery, the log line now contains "pipelined" if PIPELINING was used.
6833 4. An SMTP transport process used to panic and die if the bind() call to set
6834 an explicit outgoing interface failed. This has been changed; it is now
6835 treated in the same way as a connect() failure.
6837 5. A reference to $sender_host_name in the part of a conditional expansion
6838 that was being skipped was still causing a DNS lookup. This no longer
6841 6. The def: expansion condition was not recognizing references to header lines
6842 that used bh_ and bheader_.
6844 7. Added the _cache feature to named lists.
6846 8. The code for checking quota_filecount in the appendfile transport was
6847 allowing one more file than it should have been.
6849 9. For compatibility with Sendmail, the command line option
6857 and sets the incoming protocol and host name (for trusted callers). The
6858 host name and its colon can be omitted when only the protocol is to be set.
6859 Note the Exim already has two private options, -pd and -ps, that refer to
6860 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6861 "s", but I don't think that's a major issue.
6863 10. A number of refactoring changes to the code, none of which should affect
6866 (a) The number of logging options was getting close to filling up the
6867 32-bit word that was used as a bit map. I have split them into two classes:
6868 those that are passed in the argument to log_write(), and those that are
6869 only ever tested independently outside of that function. These are now in
6870 separate 32-bit words, so there is plenty of room for expansion again.
6871 There is no change in the user interface or the logging behaviour.
6873 (b) When building, for example, log lines, the code previously used a
6874 macro that called string_cat() twice, in order to add two strings. This is
6875 not really sufficiently general. Furthermore, there was one instance where
6876 it was actually wrong because one of the argument was used twice, and in
6877 one call a function was used. (As it happened, calling the function twice
6878 did not affect the overall behaviour.) The macro has been replaced by a
6879 function that can join an arbitrary number of extra strings onto a growing
6882 (c) The code for expansion conditions now uses a table and a binary chop
6883 instead of a serial search (which was left over from when there were very
6884 few conditions). Also, it now recognizes conditions like "pam" even when
6885 the relevant support is not compiled in: a suitably worded error message is
6886 given if an attempt is made to use such a condition.
6888 11. Added ${time_interval:xxxxx}.
6890 12. A bug was causing one of the ddress fields not to be passed back correctly
6891 from remote delivery subprocesses. The field in question was not being
6892 subsequently used, so this caused to problems in practice.
6894 13. Added new log selectors queue_time and deliver_time.
6896 14. Might have fixed a bug in maildirsizefile handling that threw up
6897 "unexpected character" debug warnings, and recalculated the data
6898 unnecessarily. In any case, I expanded the warning message to give more
6901 15. Added the message "Restricted characters in address" to the statements in
6902 the default ACL that block characters like @ and % in local parts.
6904 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6905 Three changes have been made:
6907 (a) There was a serious bug; a negative response to MAIL caused the whole
6908 recipient domain to be cached as invalid, thereby blocking all messages
6909 to all local parts at the same domain, from all senders. This bug has
6910 been fixed. The domain is no longer cached after a negative response to
6911 MAIL if the sender used is not empty.
6913 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6916 (c) A new callout option, "use_sender" has been added for people who want
6917 the modified behaviour.
6923 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6926 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6927 indeed breaks things for older releases.
6929 3. Added additional logging to the case where there is a problem reading data
6930 from a filter that is running in a subprocess using a pipe, in order to
6931 try to track down a specific problem.
6933 4. Testing facility fudge: when running in the test harness and attempting
6934 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6935 getting "No route to host". Convert this to a timeout.
6937 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6940 6. Some OS don't have socklen_t but use size_t instead. This affects the
6941 fifth argument of getsockopt() amongst other things. This is now
6942 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6943 can be set for individual OS. I have set it for SunOS5, OSF1, and
6944 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6945 some earlier ones do not.
6947 7. Change 4.30/15 was not doing the test caselessly.
6949 8. The standard form for an IPv6 address literal was being rejected by address
6950 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6951 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6952 this, as well as the form without the "IPv6" on the front (but only when
6953 address literals are enabled, of course).
6955 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6957 10. Exim crashed if a message with an empty sender address specified by -f
6958 encountered a router with an errors_to setting. This could be provoked only
6959 by a command such as
6963 where an empty string was supplied; "<>" did not hit this bug.
6965 11. Installed PCRE release 4.5.
6967 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6968 remained set. It is now erased.
6970 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6971 times from message ids (which are base 36 rather than the normal 62).
6973 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6974 were being counted as actual protocol errors, and logged if the log
6975 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6976 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6977 and DATA following a set of rejected RCPTs do not count as protocol errors.
6978 In other words, Exim assumes they were pipelined, though this may not
6979 actually be the case. Of course, in all cases the client gets an
6980 appropriate error code.
6982 15. If a lookup fails in an ACL condition, a message about the failure may
6983 be available; it is used if testing the ACL cannot continue, because most
6984 such messages specify what the cause of the deferral is. However, some
6985 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6986 that caused an old message to be retained and used if a later statement
6987 caused a defer, replacing the real cause of the deferral.
6989 16. If an IP address had so many PTR records that the DNS lookup buffer
6990 was not large enough to hold them, Exim could crash while trying to process
6991 the truncated data. It now detects and logs this case.
6993 17. Further to 4.21/58, another change has been made: if (and only if) the
6994 first line of a message (the first header line) ends with CRLF, a bare LF
6995 in a subsequent header line has a space inserted after it, so as not to
6996 terminate the header.
6998 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6999 unnecessarily, used atoi() instead of strtol(), and didn't check the
7000 termination when getting file sizes from file names by regex.
7002 19. Completely re-implemented the support for maildirsize files, in the light
7003 of a number of problems with the previous contributed implementation
7004 (4.30/29). In particular:
7006 . If the quota is zero, the maildirsize file is maintained, but no quota is
7009 . If the maildir directory does not exist, it is created before any attempt
7010 to write a maildirsize file.
7012 . The quota value in the file is just a cache; if the quota is changed in
7013 the transport, the new value overrides.
7015 . A regular expression is available for excluding directories from the
7018 20. The autoreply transport checks the characters in options that define the
7019 message's headers; it allows continued headers, but it was checking with
7020 isspace() after an embedded newline instead of explicitly looking for a
7023 21. If all the "regular" hosts to which an address was routed had passed their
7024 expiry times, and had not reached their retry times, the address was
7025 bounced, even if fallback hosts were defined. Now Exim should go on to try
7028 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7029 equivalent code in the SMTP transport. Some hosts send humungous responses
7030 to HELO/EHLO, more than 1024 it seems.
7032 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7033 gives compiler warnings in some environments. I've now done it "properly",
7036 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7037 (because of problems with the built-in one) was declared to return uschar *
7038 instead of char *, causing compiler failure.
7040 25. Fixed a file descriptor leak when processing alias/forward files.
7042 26. Fixed a minor format string issue in dbfn.c.
7044 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7046 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7047 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7050 29. When a sender address is verified, it is cached, to save repeating the test
7051 when there is more than one recipient in a message. However, when the
7052 verification involves a callout, it is possible for different callout
7053 options to be set for different recipients. It is too complicated to keep
7054 track of this in the cache, so now Exim always runs a verification when a
7055 callout is required, relying on the callout cache for the optimization.
7056 The overhead is duplication of the address routing, but this should not be
7059 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7060 to be verified with callout=postmaster, and the main callout worked but the
7061 postmaster check failed, the verification correctly failed. However, if a
7062 subsequent RCPT command asked for sender verification *without* the
7063 postmaster check, incorrect caching caused this verification also to fail,
7066 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7067 it was not caching the DNS options (qualify_single, search_parents) that
7068 were used when the lookup failed. A subsequent lookup with different
7069 options therefore always gave the same answer, though there were cases
7070 where it should not have. (Example: a "domains = !$mx_any" option on a
7071 dnslookup router: the "domains" option is always processed without any
7072 widening, but the router might have qualify_single set.) Now Exim uses the
7073 cached value only when the same options are set.
7075 32. Added John Jetmore's "exipick" utility to the distribution.
7077 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7078 than a timeout (e.g. a certificate is required, and is not provided), an
7079 Exim server now closes the connection immediately. Previously it waited for
7080 the client to close - but if the client is SSL, it seems that they each
7081 wait for each other, leading to a delay before one of them times out.
7083 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7084 maintained 0.8.x compatibility because I don't think many are using it, and
7085 it is clearly obsolete.
7087 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7090 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7091 client certificate was expired. A simple patch fixes this, though I don't
7092 understand the full logic of why the verify callback is called multiple
7095 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7096 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7097 which causes problems with some clients (such as the Certicom SSL Plus
7098 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7099 disables the coutermeasure allowing Eudora to connect."
7101 38. Exim was not checking that a write() to a log file succeeded. This could
7102 lead to Bad Things if a log got too big, in particular if it hit a file
7103 size limit. Exim now panics and dies if it cannot write to a log file, just
7104 as it does if it cannot open a log file.
7106 39. Modified OS/Makefile-Linux so that it now contains
7108 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7110 The two -D definitions ensure that Exim is compiled with large file
7111 support, which makes it possible to handle log files that are bigger than
7114 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7115 instance) a domain was checked against a named list that involved a lookup,
7116 causing $domain_data to be set, then another domain was checked against the
7117 same list, then the first domain was re-checked, the value of $domain_data
7118 after the final check could be wrong. In particular, if the second check
7119 failed, it could be set empty. This bug probably also applied to
7122 41. The strip_trailing_dot option was not being applied to the address given
7123 with the -f command-line option.
7125 42. The code for reading a message's header from the spool was incrementing
7126 $received_count, but never initializing it. This meant that the value was
7127 incorrect (doubled) while delivering a message in the same process in which
7128 it was received. In the most common configuration of Exim, this never
7129 happens - a fresh exec is done - but it can happen when
7130 deliver_drop_privilege is set.
7132 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7133 it now includes up to 150 characters of the unexpected data in the log
7136 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7137 and building data strings. The size of both of these buffers was 10 000
7138 bytes - far larger than anybody would *ever* want, thought I. Needless to
7139 say, somebody hit the limit. I have increased the maximum line length to
7140 20 000 and the maximum data length of concatenated lines to 100 000. I have
7141 also fixed two bugs, because there was no checking on these buffers. Tsk,
7142 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7143 buffer is too small.
7145 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7146 lsearch lookups. Now it does.
7148 46. When parsing a route_list item in a manualroute router, a fixed-length
7149 buffer was used for the list of hosts. I made this 1024 bytes long,
7150 thinking that nobody would ever have a list of hosts that long. Wrong.
7151 Somebody had a whole pile of complicated expansion conditions, and the
7152 string was silently truncated, leading to an expansion error. It turns out
7153 that it is easier to change to an unlimited length (owing to other changes
7154 that have happened since this code was originally written) than to build
7155 structure for giving a limitation error. The length of the item that
7156 expands into the list of hosts is now unlimited.
7158 47. The lsearch lookup could not handle data where the length of text line was
7159 more than 4095 characters. Such lines were truncated, leading to shortened
7160 data being returned. It should now handle lines of any length.
7162 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7163 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7166 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7167 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7168 no longer happens in comment lines. A list of replacements is now placed
7169 at the head of all of the source files, except those whose only change is
7170 to replace PERL_COMMAND in the very first #! line.
7172 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7173 messages on the queue, with a bottom-up merge sort, using code contributed
7174 by Michael Haardt. This should make operations like -bp somewhat faster on
7175 large queues. It won't affect queue runners, except when queue_run_in_order
7178 51. Installed eximstats 1.31 in the distribution.
7180 52. Added support for SRV lookups to the dnslookup router.
7182 53. If an ACL referred to $message_body or $message_body_end, the value was not
7183 reset for any messages that followed in the same SMTP session.
7185 54. The store-handling optimization for building very long strings was not
7186 differentiating between the different store pools. I don't think this
7187 actually made any difference in practice, but I've tidied it.
7189 55. While running the routers to verify a sender address, $sender_address
7190 was still set to the sender address. This is wrong, because when routing to
7191 send a bounce to the sender, it would be empty. Therefore, I have changed
7192 it so that, while verifying a sender address, $sender_address is set to <>.
7193 (There is no change to what happens when verifying a recipient address.)
7195 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7196 target A or AAAA records (if not already returned) without resetting the
7197 qualify_single or search_parents options of the DNS resolver. These are
7198 inappropriate in this case because the targets of MX and SRV records must
7199 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7200 target that, when qualified, matched something in the local domain. These
7201 two options are now turned off when doing these lookups.
7203 57. It seems that at least some releases of Reiserfs (which does not have the
7204 concept of a fixed number of inodes) returns zero and not -1 for the
7205 number of available inodes. This interacted badly with check_spool_inodes,
7206 which assumed that -1 was the "no such thing" setting. What I have done is
7207 to check that the total number of inodes is greater than zero before doing
7208 the test of how many are available.
7210 58. When a "warn" ACL statement has a log_message modifier, the message is
7211 remembered, and not repeated. This is to avoid a lot of repetition when a
7212 message has many recipients that cause the same warning to be written.
7213 However, Exim was preserving the list of already written lines for an
7214 entire SMTP session, which doesn't seem right. The memory is now reset if a
7215 new message is started.
7217 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7218 debugging output unless log_rewrite was also set.
7220 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7221 of (int)(handle) when we know that handle contains (void *)(-1).
7223 61. The Exim daemon panic-logs an error return when it closes the incoming
7224 connection. However "connection reset by peer" seems to be common, and
7225 isn't really an error worthy of noting specially, so that particular error
7228 62. When Exim is trying to find all the local interfaces, it used to panic and
7229 die if the ioctl to get the interface flags failed. However, it seems that
7230 on at least one OS (Solaris 9) it is possible to have an interface that is
7231 included in the list of interfaces, but for which you get a failure error
7232 for this call. This happens when the interface is not "plumbed" into a
7233 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7234 failure of the "get flags" call assumes that the interface is down.
7236 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7237 makes life easier for people who are doing arithmetic on fields extracted
7238 from dates, where you often get leading zeros that should not be
7239 interpreted as octal.
7241 64. Added qualify_domain to the redirect router, to override the global
7244 65. If a pathologically long header line contained very many addresses (the
7245 report of this problem mentioned 10 000) and each of them was rewritten,
7246 Exim could use up a very large amount of memory. (It kept on making new
7247 copies of the header line as it rewrote, and never released the old ones.)
7248 At the expense of a bit more processing, the header rewriting function has
7249 been changed so that it no longer eats memory in this way.
7251 66. The generation of the Received: header has been moved from the time that a
7252 message starts to be received, to the time that it finishes. The timestamp
7253 in the Received: header should now be very close to that of the <= log
7254 line. There are two side-effects of this change:
7256 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7257 logged header lines no longer include the local Received: line, because
7258 it has not yet been created. The same applies to a copy of the message
7259 that is returned to a non-SMTP sender when a message is rejected.
7261 (b) When a filter file is tested using -bf, no additional Received: header
7262 is added to the test message. After some thought, I decided that this
7265 This change does not affect the value of $received_for. It is still set
7266 after address rewriting, but before local_scan() is called.
7268 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7270 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7271 gave an unhelpful panic error message, and a defer error. I have managed to
7272 change this behaviour so that it now rejects any supplied certificate,
7273 which seems right, as the list of acceptable certificates is empty.
7275 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7276 gave an unhelpful defer error. I have not managed to make this reject any
7277 supplied certificates, but the error message it gives is "no certificate
7278 supplied", which is not helpful.
7280 70. exigrep's output now also includes lines that are not associated with any
7281 message, but which match the given pattern. Implemented by a patch from
7282 Martin Sluka, which also tidied up the Perl a bit.
7284 71. Recipient callout verification, like sender verification, was using <> in
7285 the MAIL FROM command. This isn't really the right thing, since the actual
7286 sender may affect whether the remote host accepts the recipient or not. I
7287 have changed it to use the actual sender in the callout; this means that
7288 the cache record is now keyed on a recipient/sender pair, not just the
7289 recipient address. There doesn't seem to be a real danger of callout loops,
7290 since a callout by the remote host to check the sender would use <>.
7291 [SEE ABOVE: changed after hitting problems.]
7293 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7294 temporary errors. However, in the case of such a code being given after
7295 the end of a data transmission (i.e. after ".") Exim was failing to write
7296 a retry record for the message. (Yes, there was some broken host that was
7297 actually sending 8xx at this point.)
7299 73. An unknown lookup type in a host list could cause Exim to panic-die when
7300 the list was checked. (An example that provoked this was putting <; in the
7301 middle of a list instead of at the start.) If this happened during a DATA
7302 ACL check, a -D file could be left lying around. This kind of configuration
7303 error no longer causes Exim to die; instead it causes a defer error. The
7304 incident is still logged to the main and panic logs.
7306 74. Buglet left over from Exim 3 conversion. The message "too many messages
7307 in one connection" was written to the rejectlog but not the mainlog, except
7308 when address rewriting (yes!) was being logged.
7310 75. Added write_rejectlog option.
7312 76. When a system filter was run not as root (that is, when system_filter_user
7313 was set), the values of the $n variables were not being returned to the
7314 main process; thus, they were not subsequently available in the $sn
7317 77. Added +return_path_on_delivery log selector.
7319 78. A connection timeout was being treated differently from recipients deferred
7320 when testing hosts_max_try with a message that was older than the host's
7321 retry timeout. (The host should not be counted, thus allowing all hosts to
7322 be tried at least once before bouncing.) This may have been the cause of an
7323 occasionally reported bug whereby a message would remain on the queue
7324 longer than the retry timeout, but would be bounced if a delivery was
7325 forced. I say "may" because I never totally pinned down the problem;
7326 setting up timeout/retry tests is difficult. See also the next item.
7328 79. The ultimate address timeout was not being applied to errors that involved
7329 a combination of host plus message (for example, a timeout on a MAIL
7330 command). When an address resolved to a number of possible hosts, and they
7331 were not all tried for each delivery (e.g. because of hosts_max_try), a
7332 message could remain on the queue longer than the retry timeout.
7334 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7337 81. Fixed an obscure SMTP outgoing bug which required at least the following
7338 conditions: (a) there was another message waiting for the same server;
7339 (b) the server returned 5xx to all RCPT commands in the first message so
7340 that the message was not completed; (c) the server dropped the connection
7341 or gave a negative response to the RSET that Exim sends to abort the
7342 transaction. The observed case was a dropped connection after DATA that had
7343 been sent in pipelining mode. That is, the server had advertised PIPELINING
7344 but was not implementing it correctly. The effect of the bug was incorrect
7345 behaviour, such as trying another host, and this could lead to a crash.
7351 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7352 and daemon.c were passed as pointers to ints; they should have been
7353 pointers to socklen_t variables (which are typically unsigned ints).
7355 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7358 3. Fixed a really odd bug that affected only the testing scheme; patching a
7359 certain fixed string in the binary changed the value of another string that
7360 happened to be identical to the end of the original first string.
7362 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7363 name", it returns that address as the IP address. On some operating
7364 systems (e.g. Solaris), it also passes back the IP address string as the
7365 "host name". However, on others (e.g. Linux), it passes back an empty
7366 string. Exim wasn't checking for this, and was changing the host name to an
7367 empty string, assuming it had been canonicalized.
7369 5. Although rare, it is permitted to have more than one PTR record for a given
7370 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7371 all the names associated with an address, because they do in Solaris.
7372 However, it seems that they do not in Linux for data that comes from the
7373 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7374 I found this out when I moved to a new Linux workstation and tried to run
7375 the Exim test suite.
7377 To get round this problem I have changed the code so that it now does its
7378 own call to the DNS to look up PTR records when searching for a host name.
7379 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7380 addresses that are only in /etc/hosts are still found.
7382 This behaviour is, however, controlled by an option called host_lookup_
7383 order, which defaults to "bydns:byaddr". If people want to use the other
7384 order, or indeed, just use one or the other means of lookup, they can
7385 specify it in this variable.
7387 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7388 some operating systems, this comes back from gethostbyaddr() as an empty
7389 string, and this is what Exim used to test for. However, it seems that in
7390 other systems, "." is yielded. Exim now tests for this case too.
7392 7. The values of check_spool_space and check_log_space are now held internally
7393 as a number of kilobytes instead of an absolute number of bytes. If a
7394 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7395 kilobyte. This means that much larger values can be stored.
7397 8. Exim monitor: an attempt to get the action menu when not actually pointing
7398 at a message produces an empty menu entitled "No message selected". This
7399 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7400 no entries in it ("Shell widget menu has zero width and/or height"). So I
7401 have added a single, blank menu entry in this case.
7403 9. Added ${quote_local_part.
7405 10. MIME decoding is now applied to the contents of Subject: header lines when
7408 11. Now that a reference to $sender_host_address automatically causes a reverse
7409 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7410 host lookup before query-style lookups in lists that might use this
7411 variable. This has therefore been abolished, and the "net-" prefix is no
7412 longer necessary for query-style lookups.
7414 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7415 have been a typo for LFLAGS, so it has been changed.
7417 13. The install script calls Exim with "-C /dev/null" in order to find the
7418 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7419 to be output. However, since Exim outputs its version number before the
7420 error, it didn't break the script. It just looked ugly. I fixed this by
7421 always allowing "-C /dev/null" if the caller is root.
7423 14. Ignore overlarge ACL variable number when reading spool file - insurance
7424 against a later release with more variables having written the file.
7426 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7427 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7428 this, as well as the form without the "IPv6" on the front.
7430 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7431 OS/Makefile-Darwin file.
7433 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7434 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7436 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7437 variable. However, some OS use ints for inodes. I've added cast to long int
7438 to get rid of the compiler warning.
7440 19. I had forgotten to lock out "/../" in configuration file names when
7441 ALT_CONFIG_PREFIX was set.
7443 20. Routers used for verification do not need to specify transports. However,
7444 if such a router generated a host list, and callout was configured, Exim
7445 crashed, because it could not find a port number from the (non-existent)
7446 transport. It now assumes port 25 in this circumstance.
7448 21. Added the -t option to exigrep.
7450 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7451 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7452 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7453 them is set, LOOKUP_LSEARCH is forced.
7455 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7457 24. Added sender and host information to the "rejected by local_scan()" log
7458 line; previously there was no indication of these.
7460 25. Added .include_if_exists.
7462 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7463 Linux. It turns out that not all file systems support this. Apparently some
7464 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7465 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7466 not supported on the file descriptor, is now ignored when Exim is trying to
7467 sync a directory. This applies only to Linux.
7469 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7471 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7472 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7475 29. Added support for maildirsize files from supplied patch (modified a bit).
7477 30. The use of :fail: followed by an empty string could lead Exim to respond to
7478 sender verification failures with (e.g.):
7480 550 Verification failed for <xxx>
7481 550 Sender verify failed
7483 where the first response line was missing the '-' that indicates it is not
7484 the final line of the response.
7486 31. The loop for finding the name of the user that called Exim had a hardwired
7487 limit of 10; it now uses the value of finduser_retries, which is used for
7488 all other user lookups.
7490 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7493 33. Exim was neglecting to zero errno before one call of strtol() when
7494 expanding a string and expecting an integer value. On some systems this
7495 resulted in spurious "integer overflow" errors. Also, it was casting the
7496 result into an int without checking.
7498 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7499 did not work. The code looks as if it has *never* worked, though it appears
7500 to have been documented since at least release 1.62. I have made it work.
7502 35. The "timeout_DNS" error in retry rules, also documented since at least
7503 1.62, also never worked. As it isn't clear exactly what this means, and
7504 clearly it isn't a major issue, I have abolished the feature by treating it
7505 as "timeout", and writing a warning to the main and panic logs.
7507 36. The display of retry rules for -brt wasn't always showing the error code
7510 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7511 timeout_connect_A, timeout_connect_MX.
7513 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7514 to the empty sender.
7516 39. The daemon was not analysing the content of -oX till after it had closed
7517 stderr and disconnected from the controlling terminal. This meant that any
7518 syntax errors were only noted on the panic log, and the return code from
7519 the command was 0. By re-arranging the code a little, I've made the
7520 decoding happen first, so such errors now appear on stderr, and the return
7521 code is 1. However, the actual setting up of the sockets still happens in
7522 the disconnected process, so errors there are still only recorded on the
7525 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7526 connections (as happens on some IP stacks) was logged at start up time as
7527 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7528 it from "IPv6 and IPv4", which means that two separate sockets are being
7531 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7532 says whether AF_INET or AF_INET6 was passed as an argument.
7534 42. Exiwhat output was messed up when time zones were included in log
7537 43. Exiwhat now gives more information about the daemon's listening ports,
7538 and whether -tls-on-connect was used.
7540 44. The "port" option of the smtp transport is now expanded.
7542 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7543 silently ignored. Now an error message is written to the main and panic
7546 46. There's a new ACL modifier called "logwrite" which writes to a log file
7547 as soon as it is encountered.
7549 47. Added $local_user_uid and $local_user_gid at routing time.
7551 48. Exim crashed when trying to verify a sender address that was being
7554 49. Exim was recognizing only a space character after ".include". It now also
7555 recognizes a tab character.
7557 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7558 extracting the relevant information from the specification. The man page no
7559 longer contains scrambled data for the -d option, and I've added a section
7560 at the front about calling Exim under different names.
7562 51. Added "extra_headers" argument to the "mail" command in filter files.
7564 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7567 53. Installed eximstats 1.29.
7569 54. Added transport_filter_timeout as a generic transport option.
7571 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7572 Cc: header lines. This was required by RFC 822, but it not required by RFC
7575 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7576 incoming messages that did not have them. Now it does so only if the
7577 message originates locally, that is, if there is no associated remote host
7578 address. When Resent- header lines are present, this applies to the Resent-
7579 lines rather than the non-Resent- lines.
7581 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7582 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7584 58. Messages for configuration errors now include the name of the main
7585 configuration file - useful now that there may be more than one file in a
7586 list (.included file names were always shown).
7588 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7589 for those rare installations that do not start the daemon as root or run it
7590 setuid root. I've cut out the call to initgroups() if the daemon is not
7593 60. The Exim user and group can now be bound into the binary as text strings
7594 that are looked up at the start of Exim's processing.
7596 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7598 62. Added $mailstore_basename variable.
7600 63. Installed patch to sieve.c from Michael Haardt.
7602 64. When Exim failed to open the panic log after failing to open the main log,
7603 the original message it was trying to log was written to stderr and debug
7604 output, but if they were not available (the usual case in production), it
7605 was lost. Now it is written to syslog before the two lines that record the
7606 failures to open the logs.
7608 65. Users' Exim filters run in subprocesses under the user's uid. It is
7609 possible for a "deliver" command or an alias in a "personal" command to
7610 provoke an address rewrite. If logging of address rewriting is configured,
7611 this fails because the process is not running as root or exim. There may be
7612 a better way of dealing with this, but for the moment (because 4.30 needs
7613 to be released), I have disabled address rewrite logging when running a
7614 filter in a non-root, non-exim process.
7620 1. The buildconfig auxiliary program wasn't quoting the value set for
7621 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7622 not defined. This bug was masked in 4.22 by the effect that was fixed in
7625 2. Some messages that were rejected after a message id was allocated were
7626 shown as "incomplete" by exigrep. It no longer does this for messages that
7627 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7629 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7630 have allow_domain_literals set, the ID did not get logged in the <= line.
7631 Domain literals are now always recognized in Message-ID: header lines.
7633 4. The first argument for a ${extract expansion item is the key name or field
7634 number. Leading and trailing spaces in this item were not being ignored,
7635 causing some misleading effects.
7637 5. When deliver_drop_privilege was set, single queue runner processes started
7638 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7639 same command in the process it spins off) were not dropping privilege.
7641 6. When the daemon running as "exim" started a queue runner, it always
7642 re-executed Exim in the spun-off process. This is a waste of effort when
7643 deliver_drop_privilege is set. The new process now just calls the
7644 queue-runner function directly.
7650 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7653 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7654 ignored. Though the use of -oP was forcing the writing of a pid file, it
7655 was always written to the default place.
7657 3. If the message "no IP address found for host xxxx" is generated during
7658 incoming verification, it is now followed by identification of the incoming
7659 connection (so you can more easily find what provoked it).
7661 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7663 5. Added some features to "harden" Exim a bit more against certain attacks:
7665 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7666 be put in Local/Makefile. This is like the never_users runtime option,
7667 but it cannot be overridden. The default setting is "root".
7669 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7670 prefix string with which any file named in a -C command line option
7673 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7674 is retained for -C and -D only if the caller of Exim is root. Without
7675 it, the exim user may also use -C and -D and retain privilege.
7677 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7678 command line option is disabled.
7680 6. Macro names set by the -D option must start with an upper case letter, just
7681 like macro names defined in the configuration file.
7683 7. Added "dereference=" facility to LDAP.
7685 8. Two instances of the typo "uknown" in the source files are fixed.
7687 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7688 the Configure-Makefile script screwed up while processing it.
7690 10. Incorporated PCRE 4.4.
7692 11. The SMTP synchronization check was not operating right at the start of an
7693 SMTP session. For example, it could not catch a HELO sent before the client
7694 waited for the greeting. There is now a check for outstanding input at the
7695 point when the greeting is written. Because of the duplex, asynchronous
7696 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7697 way, but not yet received, when the check is performed.
7699 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7700 on TCP/IP sockets, because this apparently causes some broken clients to
7703 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7704 unchanged) from the Cygwin maintainer.
7706 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7707 was supported instead of testing for mbx. Effectively a typo.
7709 15. The spa authenticator server code was not checking that the input it
7710 received was valid base64.
7712 16. The debug output line for the "set" modifier in ACLs was not showing the
7713 name of the variable that was being set.
7715 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7717 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7718 lookup on its own. Something else was needed to trigger the lookup. For
7719 example, a match in host_lookup or the need for a host name in a host list.
7720 Now, if $sender_host_name is referenced and the host name has not yet been
7721 looked up, a lookup is performed. If the lookup fails, the variable remains
7722 empty, and $host_lookup_failed is set to "1".
7724 19. Added "eqi" as a case-independent comparison operator.
7726 20. The saslauthd authentication condition could segfault if neither service
7727 nor realm was specified.
7729 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7730 error message that was logged was misleading, and incoming SMTP
7731 connections were dropped. The message is now more accurate, and temporary
7732 errors are given to SMTP connections.
7734 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7735 (except RSET) with a 421 error, until QUIT is received. However, it was
7736 failing to send a response to QUIT.
7738 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7739 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7740 ACL is now run after the helo_try_verify_hosts code.
7742 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7743 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7744 case-independent, so other case variants are also recognized). Apparently
7745 some systems use these upper case variants.
7747 25. If more than two messages were waiting for the same host, and a transport
7748 filter was specified for the transport, Exim sent two messages over the
7749 same TCP/IP connection, and then failed with "socket operation on non-
7750 socket" when it tried to send the third.
7752 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7754 27. The extern definition of crypt16() in expand.c was not being excluded when
7755 the OS had its own crypt16() function.
7757 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7758 as a preferred synonym for return_size_limit, both as an option and as an
7761 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7763 30. Changed the default configuration ACL to relax the local part checking rule
7764 for addresses that are not in any local domains. For these addresses,
7765 slashes and pipe symbols are allowed within local parts, but the sequence
7766 /../ is explicitly forbidden.
7768 31. SPA server authentication was not clearing the challenge buffer before
7771 32. log_message in a "warn" ACL statement was writing to the reject log as
7772 well as to the main log, which contradicts the documentation and doesn't
7773 seem right (because no rejection is happening). So I have stopped it.
7775 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7776 However, I am unable to do any testing of this.
7778 34. Fixed an infelicity in the appendfile transport. When checking directories
7779 for a mailbox, to see if any needed to be created, it was accidentally
7780 using path names with one or more superfluous leading slashes; tracing
7781 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7783 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7784 contents are added to the log line that is written for every discarded
7785 recipient. (Previously a log_message setting was ignored.)
7787 36. The ${quote: operator now quotes the string if it is empty.
7789 37. The install script runs exim in order to find its version number. If for
7790 some reason other than non-existence or emptiness, which it checks, it
7791 could not run './exim', it was installing it with an empty version number,
7792 i.e. as "exim-". This error state is now caught, and the installation is
7795 38. An argument was missing from the function that creates an error message
7796 when Exim fails to connect to the socket for saslauthd authentication.
7797 This could cause Exim to crash, or give a corrupted message.
7799 39. Added isip, isip4, and isip6 to ${if conditions.
7801 40. The ACL variables $acl_xx are now saved with the message, and can be
7802 accessed later in routers, transports, and filters.
7804 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7805 strings in the file are not string-expanded.
7807 42. If a MAIL command specified a SIZE value that was too large to fit into an
7808 int variable, the check against message_size_limit failed. Such values are
7809 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7810 day this will have to be increased, but I don't think I want to be around
7811 when emails are that large.
7818 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7819 iconv() is not standard in FreeBSD.
7821 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7822 IPv6 enabled. The observed symptom was a segmentation fault on return from
7823 the function os_common_find_running_interfaces() in src/os.c.
7825 3. In the check_special_case() function in daemon.c I had used "errno" as an
7826 argument name, which causes warnings on some systems. This was basically a
7827 typo, since it was named "eno" in the comments!
7829 4. The code that waits for the clock to tick (at a resolution of some fraction
7830 of a second) so as to ensure message-id uniqueness was always waiting for
7831 at least one whole tick, when it could have waited for less. [This is
7832 almost certainly not relevant at current processor speeds, where it is
7833 unlikely to ever wait at all. But we try to future-proof.]
7835 5. The function that sleeps for a time interval that includes fractions of a
7836 second contained a race. It did not block SIGALRM between setting the
7837 timer, and suspending (a couple of lines later). If the interval was short
7838 and the sigsuspend() was delayed until after it had expired, the suspension
7839 never ended. On busy systems this could lead to processes getting stuck for
7842 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7843 process, before it forks any delivery processes. The open lookup caching
7844 mechanism meant that the open file or database connection was passed into
7845 the delivery process. The problem was that delivery processes always tidy
7846 up cached lookup data. This could cause a problem for the next delivery
7847 process started by the queue runner, because the external queue runner
7848 process does not know about the closure. So the next delivery process
7849 still has data in the lookup cache. In the case of a file lookup, there was
7850 no problem because closing a file descriptor in a subprocess doesn't affect
7851 the parent. However, if the lookup was caching a connection to a database,
7852 the connection was closed, and the second delivery process was likely to
7853 see errors such as "PGSQL: query failed: server closed the connection
7854 unexpectedly". The problem has been fixed by closing all cached lookups
7855 in a queue runner before running a delivery process.
7857 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7858 seem to have the "const" qualifier which it has on other OS. I've
7861 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7862 *of the same type* (client or server) with the same public name that an
7863 error should be diagnosed.
7865 9. When Exim looked up a host name for an IP address, but failed to find the
7866 original IP address when looking up the host name (a safety check), it
7867 output the message "<ip address> does not match any IP for NULL", which was
7868 confusing, to say the least. The bug was that the host name should have
7869 appeared instead of "NULL".
7871 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7872 user that is built into the binary, and the -C or -D options is used, root
7873 privilege is dropped before the configuration file is read. In addition,
7874 logging is switched to stderr instead of the normal log files. If the
7875 configuration then re-defines the Exim user, the unprivileged environment
7876 is probably not what is expected, so Exim logs a panic warning message (but
7879 However, if deliver_drop_privilege is set, the unprivileged state may well
7880 be exactly what is intended, so the warning has been cut out in that case,
7881 and Exim is allowed to try to write to its normal log files.
7887 1. smtp_return_error_details was not giving details for temporary sender
7888 or receiver verification errors.
7890 2. Diagnose a configuration error if two authenticators have the same public
7893 3. Exim used not to create the message log file for a message until the first
7894 delivery attempt. This could be confusing when incoming messages were held
7895 for policy or load reasons. The message log file is now created at the time
7896 the message is received, and an initial "Received" line is written to it.
7898 4. The automatically generated man page for command line options had a minor
7899 bug that caused no ill effects; however, a more serious problem was that
7900 the procedure for building the man page automatically didn't always
7901 operate. Consequently, release 4.20 contains an out-of-date version. This
7902 shouldn't happen again.
7904 5. When building Exim with embedded Perl support, the script that builds the
7905 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7906 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7908 6. The freeze_tell option was not being used for messages that were frozen on
7909 arrival, either by an ACL or by local_scan().
7911 7. Added the smtp_incomplete_transaction log selector.
7913 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7914 was accepting AUTH without a new EHLO.
7916 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7917 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7920 10. Logging of TCP/IP connections (when configured) now happens in the main
7921 daemon process instead of the child process, so that the TCP/IP connection
7922 count is more accurate (but it can never be perfect).
7924 11. The use of "drop" in a nested ACL was not being handled correctly in the
7925 outer ACL. Now, if condition failure induced by the nested "drop" causes
7926 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7927 or "require"), the connection is dropped.
7929 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7930 that yield "discard" can now be used with an "accept" or a "discard" verb,
7931 but an error is generated for any others (because I can't see a useful way
7932 to define what should happen).
7934 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7935 are now processed in the same way as lines in the Exim configuration file.
7936 In particular, continuation lines are supported.
7938 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7940 15. Added -ti meaning -t -i.
7942 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7943 domains, and warn by logging if others are found.
7945 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7946 returned from GIFCONF when Exim is trying to find the list of interfaces on
7947 a host. The code in os.c has been modified to copy each ifreq to an aligned
7948 structure in all cases.
7950 Also, in some cases, the returned ifreq's were being copied to a 'struct
7951 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7952 means the last couple of bytes of an IPv6 address could be chopped if the
7953 ifreq contained only a normal sockaddr (14 bytes storage).
7955 18. Named domain lists were not supported in the hosts_treat_as_local option.
7956 An entry such as +xxxx was not recognized, and was treated as a literal
7959 19. Ensure that header lines added by a DATA ACL are included in the reject log
7960 if the ACL subsequently rejects the message.
7962 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7963 MD5 (which is deprecated).
7965 21. When testing a filter file using -bf, Exim was writing a message when it
7966 took the sender from a "From " line in the message, but it was not doing so
7967 when it took $return_path from a Return-Path: header line. It now does.
7969 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7970 with a valid header line field name (a series of printing characters
7971 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7973 23. Changed "disc" in the source to "disk" to conform to the documentation and
7974 the book and for uniformity.
7976 24. Ignore Sendmail's -Ooption=value command line item.
7978 25. When execve() failed while trying to run a command in a pipe transport,
7979 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7980 could be confused with a return value of 69 from the command itself. This
7981 has been changed to 127, the value the shell returns if it is asked to run
7982 a non-existent command. The wording for the related log line suggests a
7983 non-existent command as the problem.
7985 26. If received_header_text expands to an empty string, do not add a Received:
7986 header line to the message. (Well, it adds a token one on the spool, but
7987 marks it "old" so that it doesn't get used or transmitted.)
7989 27. Installed eximstats 1.28 (addition of -nt option).
7991 28. There was no check for failure on the call to getsockname() in the daemon
7992 code. This can fail if there is a shortage of resources on the system, with
7993 ENOMEM, for example. A temporary error is now given on failure.
7995 29. Contrary to the C standard, it seems that in some environments, the
7996 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7997 program. Exim now does this explicitly; it affects the formatting of
7998 timestamps using strftime().
8000 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8001 complaints. I've now initialized all the variables, to avoid this.
8003 32. Header lines added by a system filter were not being "seen" during
8004 transport-time rewrites.
8006 33. The info_callback() function passed to OpenSSL is set up with type void
8007 (*)(SSL *, int, int), as described somewhere. However, when calling the
8008 function (actually a macro) that sets it up, the type void(*)() is
8009 expected. I've put in a cast to prevent warnings from picky compilers.
8011 34. If a DNS black list lookup found a CNAME record, but there were no A
8012 records associated with the domain it pointed at, Exim crashed.
8014 35. If a DNS black list lookup returned more than one A record, Exim ignored
8015 all but the first. It now scans all returned addresses if a particular IP
8016 value is being sought. In this situation, the contents of the
8017 $dnslist_value variable are a list of all the addresses, separated by a
8020 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8021 to accept a host name and all its aliases if the forward lookup for any of
8022 them yielded the IP address of the incoming connection. Now it accepts only
8023 those names whose forward lookup yields the correct IP address. Any other
8024 names are discarded. This closes a loophole whereby a rogue DNS
8025 administrator could create reverse DNS records to break through a
8026 wildcarded host restriction in an ACL.
8028 37. If a user filter or a system filter that ran in a subprocess used any of
8029 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8030 the wrong values were passed to the pipe command ($thisaddress had the
8031 value of $0, $0 had the value of $1, etc). This bug was introduced by
8032 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8034 38. Improved the line breaking for long SMTP error messages from ACLs.
8035 Previously, if there was no break point between 40 and 75 characters, Exim
8036 left the rest of the message alone. Two changes have been made: (a) I've
8037 reduced the minimum length to 35 characters; (b) if it can't find a break
8038 point between 35 and 75 characters, it looks ahead and uses the first one
8039 that it finds. This may give the occasional overlong line, but at least the
8040 remaining text gets split now.
8042 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8043 file descriptors might be low, and that setting 1000 would always raise it.
8044 It turns out that in some environments, the limit is already over 1000 and
8045 that lowering it causes trouble. So now Exim takes care not to decrease it.
8047 40. When delivering a message, the value of $return_path is set to $sender_
8048 address at the start of routing (routers may change the value). By an
8049 oversight, this default was not being set up when an address was tested by
8050 -bt or -bv, which affected the outcome if any router or filter referred to
8053 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8054 log only when "message" or "log_message" are set. However, if one of the
8055 conditions was an address verification, or a call to a nested ACL, the
8056 messages generated by the underlying test were being passed through. This
8057 no longer happens. The underlying message is available in $acl_verify_
8058 message for both "message" and "log_message" expansions, so it can be
8059 passed through if needed.
8061 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8062 new expansion $bh_ to give the encoded byte string without charset
8063 translation. Translation happens only if iconv() is available; HAVE_ICONV
8064 indicates this at build time. HEADERS_CHARSET gives the charset to
8065 translate to; headers_charset can change it in the configuration, and
8066 "headers charset" can change it in an individual filter file.
8068 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8069 that creates RFC 2047 encoded "words" labels them as that charset instead
8070 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8071 expansion operator; (ii) when Exim creates a From: line for a local
8072 message; (iii) when a header line is rewritten to include a "phrase" part.
8074 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8075 buggy, causing it to skip the first lines of messages whose message ID
8076 ended in 'D'. This would not have bitten before Exim release 4.14, because
8077 message IDs were unlikely to end in 'D' before then. The effect was to have
8078 incorrect size information for certain domains.
8080 45. #include "config.h" was missing at the start of the crypt16.c module. This
8081 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8084 46. If there was a timeout during a "random" callout check, Exim treated it as
8085 a failure of the random address, and carried on sending RSET and the real
8086 address. If the delay was just some slowness somewhere, the response to the
8087 original RCPT would be taken as a response to RSET and so on, causing
8088 mayhem of various kinds.
8090 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8091 when I implemented it. It didn't allow for the fact that some option values
8092 may legitimately be negative (e.g. size_addition), and it didn't even do
8093 the right test for positive values.
8095 48. Domain names in DNS records are case-independent. Exim always looks them up
8096 in lower case. Some resolvers return domain names in exactly the case they
8097 appear in the zone file, that is, they may contain uppercase letters. Not
8098 all resolvers do this - some return always lower case. Exim was treating a
8099 change of case by a resolver as a change of domain, similar to a widening
8100 of a domain abbreviation. This triggered its re-routing code and so it was
8101 trying to route what was effectively the same domain again. This normally
8102 caused routing to fail (because the router wouldn't handle the domain
8103 twice). Now Exim checks for this case specially, and just changes the
8104 casing of the domain that it ultimately uses when it transmits the message
8107 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8110 50. If a filter generated a file delivery with a non-absolute name (possible if
8111 no home directory exists for the router), the forbid_file option was not
8114 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8115 the existing equality matching.
8117 52. Exim was using ints instead of ino_t variables in some places where it was
8118 dealing with inode numbers.
8120 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8121 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8122 called TMPDIR, and if it finds it is different, it changes its value.
8124 54. The smtp_printf() function is now made available to local_scan() so
8125 additional output lines can be written before returning. There is also an
8126 smtp_fflush() function to enable the detection of a dropped connection.
8127 The variables smtp_input and smtp_batched_input are exported to
8130 55. Changed the default runtime configuration: the message "Unknown user"
8131 has been removed from the ACL, and instead placed on the localuser router,
8132 using the cannot_route_message feature. This means that any verification
8133 failures that generate their own messages won't get overridden. Similarly,
8134 the "Unrouteable address" message that was in the ACL for unverifiable
8135 relay addresses has also been removed.
8137 56. Added hosts_avoid_esmtp to the smtp transport.
8139 57. The exicyclog script was not checking for the esoteric option
8140 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8141 will work only if exicyclog is run under the appropriate euid.
8143 58. Following a discussion on the list, the rules by which Exim recognises line
8144 endings on incoming messages have been changed. The -dropcr and drop_cr
8145 options are now no-ops, retained only for backwards compatibility. The
8146 following line terminators are recognized: LF CRLF CR. However, special
8147 processing applies to CR:
8149 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8150 nor a local message in the state where . is a terminator.
8152 (ii) If a bare CR is encountered in a header line, an extra space is added
8153 after the line terminator so as not to end the header. The reasoning
8154 behind this is that bare CRs in header lines are most likely either
8155 to be mistakes, or people trying to play silly games.
8157 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8158 was being incorrectly given as 18 bytes larger than it should have been.
8159 This is a VOB (very old bug).
8161 60. This may never have affected anything current, but just in case it has:
8162 When the local host is found other than at the start of a list of hosts,
8163 the local host, those with the same MX, and any that follow, are discarded.
8164 When the list in question was part of a longer list of hosts, the following
8165 hosts (not currently being processed) were also being discarded. This no
8166 longer happens. I'm not sure if this situation could ever has previously
8169 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8172 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8173 header. This is something that is recommended in a new Internet Draft, and
8174 is something that is documented as being done by Sendmail. There are two
8175 possible values. For messages generated by the autoreply transport, Exim
8178 Auto-Submitted: auto-replied
8180 whereas for all other generated messages (e.g. bounces) it adds
8182 Auto-Submitted: auto-generated
8184 63. The "personal" condition in filters now includes a test for the
8185 Auto-Submitted: header. If it contains the string "auto-" the message it
8186 not considered personal.
8188 64. Added rcpt_include_affixes as a generic transport option.
8190 65. Added queue_only_override (default true).
8192 66. Added the syslog_duplication option.
8194 67. If what should have been the first header line of a message consisted of
8195 a space followed by a colon, Exim was mis-interpreting it as a header line.
8196 It isn't of course - it is syntactically invalid and should therefore be
8197 treated as the start of the message body. The misbehaviour could have
8198 caused a number of strange effects, including loss of data in subsequent
8199 header lines, and spool format errors.
8201 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8202 client host had authenticated. This control can now be exercised by an ACL
8203 for more flexibility.
8205 69. By default, callouts do not happen when testing with -bh. There is now a
8206 variant, -bhc, which does actually run the callout code, including
8207 consulting and updating the callout cache.
8209 70. Added support for saslauthd authentication, courtesy of Alexander
8212 71. If statvfs() failed on the spool or log directories while checking their
8213 size for availability, Exim confusingly gave the error "space shortage".
8214 Furthermore, in debugging mode it crashed with a floating point exception.
8215 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8216 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8217 serious problem, Exim now writes to the main and panic logs when this
8218 happens, with details of the failure. It then refuses to accept the
8219 incoming message, giving the message "spool directory problem" or "log
8220 directory problem" with a 421 code for SMTP messages.
8222 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8223 0, 1, and 2 exist, because some OS complain for execs without them (see
8224 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8225 descriptors. However, the code omitted to check that the open succeeded,
8226 causing mysterious errors if for some reason the permissions on /dev/null
8227 got screwed. Now Exim writes a message to the main and panic logs, and
8228 bombs out if it can't open /dev/null.
8230 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8231 interact so that it is all more flexible. It is supposed to remain
8232 backwards compatible. Also added extra_local_interfaces.
8234 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8235 to bomb out with an assertion failure - to the client this appears as a
8236 connection drop. This problem occurs in the part of the code that was taken
8237 from the Samba project. Fortunately, the assertion is in a very simple
8238 function, so I have fixed this by reproducing the function inline in the
8239 one place where it is called, and arranging for authentication to fail
8240 instead of killing the process with assert().
8242 75. The SPA client code was not working when the server requested OEM rather
8243 than Unicode encoding.
8245 76. Added code to make require_files with a specific uid setting more usable in
8246 the case where statting the file as root fails - usually a non-root-mounted
8247 NFS file system. When this happens and the failure is EACCES, Exim now
8248 forks a subprocess and does the per-uid checking as the relevant uid.
8250 77. Added process_log_path.
8252 78. If log_file_path was not explicitly set, a setting of check_log_space or
8253 check_log_inodes was ignored.
8255 79. If a space check for the spool or log partitions fails, the incident is now
8256 logged. Of course, in the latter case the data may get lost...
8258 80. Added the %p formatting code to string_format() so that it can be used to
8259 print addresses in debug_print(). Adjusted all the address printing in the
8260 debugging in store.c to use %p rather than %d.
8262 81. There was a concern that a line of code in smtp_in.c could overflow a
8263 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8264 initially expressed, the concern was not well-founded, because trailing
8265 spaces are removed early. However, if the trailing spaces were followed by
8266 a NULL, they did not get removed, so the overflow was possible. Two fixes
8269 (a) I re-wrote the offending code in a cleaner fashion.
8270 (b) If an incoming SMTP command contains a NULL character, it is rejected
8273 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8274 runs initgroups(), so that if the Exim user is in any additional groups,
8275 they will be used during message reception.
8281 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob