1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes. Previously debugging stopped any time Exim
74 re-execs, or for processing a queued message.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
84 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
86 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
87 supplied as the db file-pair basename. Previously, if a directory
88 path was given, for example via the autoreply "once" option, the DB
89 file.pag and file.dir files would be created in that directory's
92 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
93 "taint" log_selector. These were previously deprecated.
95 JH/22 Fix static address-list lookups to properly return the matched item.
96 Previously only the domain part was returned.
98 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
99 the call into OpenSSL to send a TLS Close was being repeated; this
100 resulted in the library waiting for the peer's Close. If that was never
101 sent we waited forever. Fix by tracking send calls.
103 JH/24 The ${run} expansion item now expands its command string elements after
104 splitting. Previously it was before; the new ordering makes handling
105 zero-length arguments simpler. The old ordering can be obtained by
106 appending a new option "preexpand", after a comma, to the "run".
108 JH/25 Taint-check exec arguments for transport-initiated external processes.
109 Previously, tainted values could be used. This affects "pipe", "lmtp" and
110 "queryprogram" transport, transport-filter, and ETRN commands.
111 The ${run} expansion is also affected: in "preexpand" mode no part of
112 the command line may be tainted, in default mode the executable name
115 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
116 the the facility was not passed across execs, and only the first message
117 passed over a connection could use BDAT; any further ones using DATA.
119 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
120 uses $sending_ip_address and an interface is specified.
121 Previously any use of the local address in the EHLO name disabled
122 PIPECONNECT, the common case being to use the rDNS of it.
124 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
125 resumption. Previously verify failed because no certificate status is
126 passed on the wire for the restarted session. Fix by using the recorded
127 ocsp status of the stored session for the new connection.
129 JH/29 TLS resumption: the key for session lookup in the client now includes
130 more info that a server could potentially use in configuring a TLS
131 session, avoiding oferring mismatching sessions to such a server.
132 Previously only the server IP was used.
134 JH/30 Fix string_copyn() for limit greater than actual string length.
135 Previously the copied amount was the limit, which could result in a
136 overlapping memcpy for newly allocated destination soon after a
137 source string shorter than the limit. Found/investigated by KM.
139 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
140 close; it may be needed for a subsequent connection. This caused a
141 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
143 JH/32 Fix CHUNKING for a second message on a connection when the first was
144 rejected. Previously we did not reset the chunking-offered state, and
145 erroneously rejected the BDAT command. Investigation help from
152 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
153 related applications. Previously an "H" was used where available info
154 says that "M" should be, so change to match.
156 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
157 as arguments, so an implementation trying to copy these into a local
158 buffer was taking a taint-enforcement trap. Fix by using dynamically
159 created buffers. Similar fix for radius expansion condition.
161 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
162 reasonable, eg. to count headers. Fix by using dynamically created
163 buffers rather than a local. Do similar fixes for ACL actions "dcc",
164 "log_reject_target", "malware" and "spam"; the arguments are expanded
165 so could be handling tainted values.
167 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
168 broken the (no-op) support for this sendmail command. Restore it
169 to doing nothing, silently, and returning good status.
171 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
172 record path was given (or the default used) without a leading directory
173 path, an error occurred on trying to open it. Use the transport's working
176 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
177 transport. Previously it was the name on the DNS A-record; use instead
178 the head of the CNAME chain leading there (if there is one). This seems
179 to align better with RFC 6125.
181 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
182 smtp_accept_max_per_host allocated resources which were not released
183 when the limit was exceeded. This eventually crashed the daemon. Fix
184 by adding a release action in that path.
186 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
187 expanded; previously using tainted values was rejected. Fix by using
188 dynamically-created buffers.
190 JH/09 Relax restrictions on ACL verify condition needing access to message
191 headers. Previously they were only permitted in data and non-smtp ACLs;
192 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
193 not_blind, header_sender and header_names_ascii verification.
195 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
196 Previously a macro used one argument twice; when called with the
197 argument as an expression having side-effects, incorrect operation
198 resulted. Use an inlineable function.
200 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
201 held open for a verify callout. Previously this wan not accounted for
202 and a corrupt onward SMTP conversation resulted.
204 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
205 Previously they were embedded naively in the constructed address; when
206 needed, strip the quoting and quote the entire local_part.
207 Also make the inbound_srs expansion condition handle quoting.
209 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
210 excluded, not matching the documentation.
212 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
213 was given for the sqlite_dbfile a trap resulted.
215 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
216 "name" argument resulted in a trap. There is no reason to disallow such;
217 this was a coding error.
219 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
220 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
221 spent suspended, ignoring the POSIX definition. Previously we assumed
222 it did and a constant offset from real time could be used as a correction.
223 Change to using the same clock source for the start-of-message and the
224 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
225 exists, just to get a clock slightly more aligned to reality.
227 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
228 RFC says it is optional some validators care. The missing char was not
229 intended but triggered by a line-wrap alignment. Discovery and fix by
230 Guillaume Outters, hacked on by JH.
232 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
233 name being quoted was tainted a trap would be taken. Fix by using
234 dynamically created buffers. The routine could have been called by a
235 rewrite with the "h" flag, by using the "-F" command-line option, or
236 by using a "name=" option on a control=submission ACL modifier.
238 JH/19 SPF: change the Authentication-Results expansion component to give
239 smtp.helo when the sender domain is empty. Previously it gave
242 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
243 not in 127.0.0.0/8 to help in spotting list domains taken over by a
244 domain-parking registrar.
246 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
247 Previously when a whitespace character was specified it was not inserted
248 after removing the newline.
250 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
251 the domain part of the recipient address. This overrides any tls_sni
252 option set, which was previously used.
254 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
257 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
258 is_tainted() had an off-by-one error in the overenthusiastic direction.
259 Find and fix by Gavan. Although NetBSD is not a supported platform for
260 4.94 this bug could affect other platforms.
262 PP/01 Fix default prime selection to be consistent.
263 One path used ike23 still, instead of exim.dev.20160529.3; now both
264 execution flows will use the same DH primes (currently
265 exim.dev.20160529.3).
267 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
268 option in smtp transport, to match the documentation. Previously
269 verification was not being done in some cases where it should have been.
271 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
272 than one server was defined and depending on the platform memory layout
273 details, an internal consistency trap could be hit while walking the list
276 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
277 files. The value is used by the authresults expansion item. Previously
278 if this was used in a router or transport, a crash could result.
280 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
281 have a sufficient incidence of EINTR returns from select that an
282 interaction with other operations done by the main daemon loop exposed
283 a bug in the error-handling. This was benign apart from the log
286 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
287 for consistency with delivered "=>" and failed "**" lines. While we're
288 there, handle PRX and TFO.
290 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
291 applied. This resulted, if any header-line rewrite rules were configured,
292 in a panic-log triggerable by sending a message with a long address in
293 a header. Fix by increasing the arbitrary limit to larger than a single
294 (dewrapped) 5322 header line maximum size.
296 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
297 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
298 RFC 6648 which deprecates X- options in protocols as a general practice.
299 Changeover between the implementations is handled by the mechanisms
302 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
303 a remote which uses callout/hold. Previously the local was queued.
305 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
308 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
309 log line. The asterisk is now consistently appended to the remote IP
310 (and port, if given), and will also be provided on defer and fail log
311 lines. Previously it could be placed on the local IP if that was being
312 logged, and was only provided on delivery lines.
314 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
316 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
317 authenticator client_send option. Previously the next char, after a pair
318 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
319 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
320 documentation. There is still no way to get a leading ^ immediately
321 after a NUL (ie. for the password of a PLAIN method authenticator.
323 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
324 files. For bad sizes read, delete the record and whine to paniclog.
326 JH/38 When logging an AUTH failure, as server, do not include sensitive
327 information. Previously, the credentials would be included if given
328 as part of the AUTH command line and an ACL denied authentication.
330 JH/39 Bug 2691: fix $local_part_data. When the matching list element
331 referred to a file, bad data was returned. This likely also affected
334 JH/40 The gsasl authenticator now supports caching of the salted password
335 generated by the client-side implementation. This required the addition
336 of a new variable: $auth4.
338 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
339 left undeleted; the attempt to re-create it then failed - resulting in
340 the usual "SIGHUP tp have daemon reload configuration" to not work.
341 This affected any platform not supporting "abstract" Unix-domain
342 sockets (i.e. not Linux).
344 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
345 recipients" response to RCPT in a later response, with a 250. The
346 previous coding assumed this would not happen, and under PIPELINING
347 would result in both lost and duplicate recipients for a message.
349 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
350 Previously the weighting was incorrectly applied. Similar fix for socks
351 proxies. Found and fixed by Heiko Schlichting.
353 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
354 not handle sub-lists included using the +namedlist syntax. While
355 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
356 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
359 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
360 Previously the permanent pool was used, so the sore could not be freed.
361 This meant a connection with many messages would use continually-growing
364 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
365 per-pool so as not to waste too much space. Previously a constant size
366 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
367 DOS attacks harder. The cost is wasted memory use in the larger blocks.
369 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
370 same space repeatedly, and a smaller process footprint.
372 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
373 for GnuTLS. Previously the name started "%s" which, while not a bug,
374 looked as if if might be one.
376 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
377 "continued connection") make the $proxy_* variables available. Previously
378 the information was not passed across the exec() call for subsequent
379 transport executions. This also mean that the log lines for the
380 messages can show the proxy information.
382 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
383 receive time. With modern systems the difference is significant.
384 The historical behaviour can be restored by disabling (a new) log_selector
385 "queue_time_exclusive".
387 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
388 content) were specifically tested for. Now, also cover expansions
389 resulting in ACL names and inline ACL content.
391 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
392 making it unusable in complex expressions.
394 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
395 fail with a taint-check on the spoolfile name, and leave the message
398 HS/01 Enforce absolute PID file path name.
400 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
402 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
404 PP/02 Bug 2643: Correct TLS DH constants.
405 A missing NUL termination in our code-generation tool had led to some
406 incorrect Diffie-Hellman constants in the Exim source.
407 Reported by kylon94, code-gen tool fix by Simon Arlott.
409 PP/03 Impose security length checks on various command-line options.
410 Fixes CVE-2020-SPRSS reported by Qualys.
412 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
413 better. Reported by Qualys.
415 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
416 providing a particularly obnoxious sender full name.
419 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
421 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
424 PP/08 Change default for recipients_max from unlimited to 50,000.
426 PP/09 Fix security issue with too many recipients on a message (to remove a
427 known security problem if someone does set recipients_max to unlimited,
428 or if local additions add to the recipient list).
429 Fixes CVE-2020-RCPTL reported by Qualys.
431 PP/10 Fix security issue in SMTP verb option parsing
432 Fixes CVE-2020-EXOPT reported by Qualys.
434 PP/11 Fix security issue in BDAT state confusion.
435 Ensure we reset known-good where we know we need to not be reading BDAT
436 data, as a general case fix, and move the places where we switch to BDAT
437 mode until after various protocol state checks.
438 Fixes CVE-2020-BDATA reported by Qualys.
440 HS/03 Die on "/../" in msglog file names
442 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
443 the Exim runtime user are allowed to create files.
445 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
448 QS/03 When reading the output from interpreted forward files we do not
449 pass the pipe between the parent and the interpreting process to
450 executed child processes (if any).
452 QS/04 Always die if requested from internal logging, even is logging is
455 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
456 the API; compilation noo longer completes with DMARC support included.
457 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
458 been reported on other platforms.
460 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
462 JH/56 Make the majority of info read from config files readonly, for defence-in-
463 depth against exploits. Suggestion by Qualys.
464 Not supported on Solaris 10.
466 JH/57 Fix control=fakreject for a custom message containing tainted data.
467 Previously this resulted in a log complaint, due to a re-expansion present
468 since fakereject was originally introduced.
470 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
471 modifier was given, a loop resulted.
473 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
474 pipelined SMTP command followed the BDAT LAST then it would be
475 incorrectly treated as part of the message body, causing a verification
478 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
479 require looser validation rules than those for 821-level addresses,
480 which only permit IP addresses.
486 JH/01 Avoid costly startup code when not strictly needed. This reduces time
487 for some exim process initialisations. It does mean that the logging
488 of TLS configuration problems is only done for the daemon startup.
490 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
492 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
493 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
496 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
497 DKIM signing is being done. Previously a transport_filter would always
498 disable CHUNKING, falling back to traditional DATA.
500 JH/05 Regard command-line recipients as tainted.
502 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
504 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
505 PAM library frees one of the arguments given to it, despite the
506 documentation. Therefore a plain malloc must be used.
508 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
509 on-stack buffers were used, resulting in a taint trap when DSN information
510 copied from a received message was written into the buffer.
512 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
513 the ordering of its ARC headers. This caused a crash.
515 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
516 a new record was being constructed with information from the peer, a trap
519 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
520 installation would get error messages from DMARC verify, when it hit the
521 nonexistent file indicated by the default. Distros wanting DMARC enabled
522 should both provide the file and set the option.
523 Also enforce no DMARC verification for command-line sourced messages.
525 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
526 could, depending on the platform, hang at the STARTTLS response.
528 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
529 message on a connection. Previously if one message had ARC headers and
530 the following one did not, a crash could result when adding an
531 Authentication-Results: header.
533 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
534 Exim main code and Exim-related utities. The introduction of taint
535 tracking also did many adjustments to string handling. Since then, eximon
536 frequently terminated with an assert failure.
538 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
539 check for 452 responses. This slightly helps the inefficieny of doing
540 a large alias-expansion into a recipient-limited target. The max_rcpt
541 transport option still applies (and at the current default, will override
542 the new feature). The check is done for either cause of synch, and forces
543 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
544 connection. The new facility is not tunable at this time.
546 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
547 library live data was being used, so the results became garbage. Make
548 copies while it is still usable.
550 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
551 on delivery deferred (==) and failed (**) lines (if a delivery was
552 attemtped). Previously it was only on completion (=>) lines.
554 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
555 for the expansion of the server_scram_iter and server_scram_salt options.
557 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
558 are now specifically given a NO_DATA response without hitting the system
559 resolver. The library goes on to do the now-standard TXT lookup.
560 Use of dnsdb lookups is not affected.
562 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
563 only retrieve the errormessage once. Previously two calls to dlerror()
564 were used, and the second one (for mainlog/paniclog) retrieved null
567 JH/20 Taint checking: disallow use of tainted data for
568 - the appendfile transport file and directory options
569 - the pipe transport command
570 - the autoreply transport file, log and once options
571 - file names used by the redirect router (including filter files)
573 - paths used by single-key lookups
574 Previously this was permitted.
576 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
577 adjusted the size of a major service buffer; this failed because the
578 buffer was in use at the time. Change to a compile-time increase in the
579 buffer size, when this authenticator is compiled into exim.
581 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
582 previous fast-mode was untenable in the face of glibs using mmap to
583 support larger malloc requests.
585 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
586 New values supported, if defined on system where compiled:
587 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
588 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
590 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
591 running a limited number of proceses in parallel, a benefit is gained. The
592 amount varies with the platform hardware and load. The use of the option
593 queue_run_in_order means we cannot do this, as ordering becomes
596 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
597 had introduced a string-copy (for ensuring NUL-termination) which was not
598 appropriate for that case, which can include embedded NUL bytes in the
599 block of data. Investigation showed the copy to actually be needless, the
600 data being length-specified.
602 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
603 done during a receiving connection, and both used TLS, global info was
604 used rather than per-connection info for tracking the state of data
605 queued for transmission. This could result in a connection hang.
607 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
608 Previously, when delivering serveral messages down a single connection
609 only the first would provide a SIZE. This was due to the size information
610 not being properly tracked.
612 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
613 TAI (at 37 seconds currently), pretend to be in UTC for time-related
614 expansion and logging. Previously, spurious values such as a future
615 minute could be seen.
617 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
618 it could crash from a null-deref. This could also affect the
619 ${addresses: } operator and ${readsock } item.
621 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
622 message following a mime one, the variable was not reset.
624 JH/30 When an pipelined-connect fails at the first response, assume incorrect
625 cached capability (perhaps the peer reneged?) and immediately retry in
628 JH/31 Fix spurious detection of timeout while writing to transport filter.
630 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
631 an attempt to copy the string was made before checking it.
633 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
634 taint of the lookup key was maintained; we now regard the presence in the
635 filesystem as sufficient validation.
637 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
638 argument is supplied.
640 JH/35 The dsearch lookup now requires that the directory is an absolute path.
641 Previously this was not checked, and nonempty relative paths made an
642 access under Exim's current working directory.
644 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
645 Previously no event was raised.
647 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
648 parameter supplied by the sender MAIL FROM command. Previously it was
649 ignored, and only the check_spool_space option value for the required
652 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
653 the size of the signing public-key. Previously it was instead giving
654 the size of the signature hash.
656 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
657 the default. See the (new) dkim_verify_min_keysizes option.
659 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
660 and an ACL use a lookup for checking either the local_part or domain,
661 stale data could be accessed. Ensure that variable references are
662 dropped between messages.
664 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
665 by the client was not checked as pointing within response data before
666 being used. A malicious client could thus cause an out-of-bounds read and
667 possibly gain authentication. Fix by adding the check.
669 JH/42 Internationalisation: change the default for downconversion in the smtp
670 transport to be "if needed". Previously it was "as previously set" for
671 the message, which usually meant "if needed" for message-submission but
672 "no" for everything else. However, MTAs have been seen using SMTPUTF8
673 even when the envelope addresses did not need it, resulting in forwarding
674 failures to non-supporting MTAs. A downconvert in such cases will be
675 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
676 transport. The change does mean that addresses needing conversion will
677 be converted when previously a delivery failure would occur.
679 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
680 response was received it would be used unchecked in a fail-DSN, violating
681 standards on line-length limits. Truncate if needed.
683 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
684 doesn't work. (Additionally add a new main config option to configure the
691 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
692 side, to decode a TLS 1.3 packet capture.
694 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
695 Previously the default library behaviour applied, sending two, each in
698 JH/03 Debug output for ACL now gives the config file name and line number for
701 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
703 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
705 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
706 buffer overrun for (non-chunking) other transports.
708 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
709 TLS1.3, means that a server rejecting a client certificate is not visible
710 to the client until the first read of encrypted data (typically the
711 response to EHLO). Add detection for that case and treat it as a failed
712 TLS connection attempt, so that the normal retry-in-clear can work (if
713 suitably configured).
715 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
716 and/or domain. Found and fixed by Jason Betts.
718 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
719 configuration). If a CNAME target was not a wellformed name pattern, a
722 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
723 the OS reports them interleaved with other addresses.
725 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
726 used both for input and for a verify callout, both encrypted, SMTP
727 responses being sent by the server could be lost. This resulted in
728 dropped connections and sometimes bounces generated by a peer sending
731 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
732 string. Previously it was possible to cause undefined behaviour in a
733 library routine (usually a crash). Found by "zerons".
735 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
738 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
739 API was removed, so update to use the newer ones.
741 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
742 any timeout set, is taking a long time. Previously we would hang on to a
743 rotated logfile "forever" if the input was arriving with long gaps
744 (a previous attempt to fix addressed lack, for a long time, of initial
747 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
748 shared (NFS) environment. The length of the tempfile name is now
749 4 + 16 ("hdr.$message_exim_id") which might break on file
750 systems which restrict the file name length to lower values.
753 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
754 shared (NFS) environment.
756 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
757 did for all versions <4.90). Notably -M, -m, --invert, -I may be
760 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
761 on some platforms for bit 31.
763 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
764 to changes apparently associated with TLS1.3 handling some of the APIs
765 previously used were either nonfunctional or inappropriate. Strings
766 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
767 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
768 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
769 This affects log line X= elements, the $tls_{in,out}_cipher variables,
770 and the use of specific cipher names in the encrypted= ACL condition.
772 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
774 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
775 verification result was not updated unless hosts_require_ocsp applied.
777 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
778 queue_list_requires_admin set to false, non-admin users were denied the
781 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
782 directory-of-certs mode. Previously they were advertised despite the
785 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
786 A single TCP connection by a client will now hold a TLS connection open
787 for multiple message deliveries, by default. Previously the default was to
790 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
791 default. If built with the facility, DANE will be used. The facility
792 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
794 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
795 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
796 must be defined and you must still, unless you define DISABLE_TLS, manage
797 the the include-dir and library-file requirements that go with that
798 choice. Non-TLS builds are still supported.
800 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
803 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
804 default. If the platform supports and has the facility enabled, it will
805 be requested on all coneections.
807 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
808 controlled by the build-time option SUPPORT_PIPE_CONNECT.
810 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
812 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
813 success-DSN messages. Previously the From: header was always the default
814 one for these; the option was ignored.
816 JH/28 Fix the timeout on smtp response to apply to the whole response.
817 Previously it was reset for every read, so a teergrubing peer sending
818 single bytes within the time limit could extend the connection for a
819 long time. Credit to Qualsys Security Advisory Team for the discovery.
821 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
822 delivery address, which leaked information of the results of local
823 forwarding. Change to the original envelope recipient address, per
826 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
827 requested. Previously not bounce was generated and a log entry of
828 error ignored was made.
830 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
832 JH/32 Introduce a general tainting mechanism for values read from the input
833 channel, and values derived from them. Refuse to expand any tainted
834 values, to catch one form of exploit.
836 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
837 was unused and the unexpanded text used for the test. Found and
838 fixed by Ruben Jenster.
840 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
841 an attempt to use a TLS library read routine dereffed a nul pointer,
844 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
845 them in our smtp response.
847 JH/36 Have the general router option retry_use_local_part default to true when
848 any of the restrictive preconditions are set (to anything). Previously it
849 was only for check_local user. The change removes one item of manual
850 configuration which is required for proper retries when a remote router
851 handles a subset of addresses for a domain.
853 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
854 link count into consideration.
856 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
857 caused the extension of big_buffer, the following lines were ignored.
859 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
860 accordance with RFC 2308. Previously there was no expiry, so a longlived
861 receive process (eg. due to ACL delays) versus a short SOA value could
864 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
866 JH/39 Promote DMARC support to mainline.
868 JH/40 Bug 2452: Add a References: header to DSNs.
870 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
871 parameters. The relevant library call is documented as "Deprecated: This
872 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
873 3.6.0, DH parameters are negotiated following RFC7919."
875 HS/06 Change the default of dnssec_request_domains to "*"
877 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
878 carried on and emitted a BDAT command, even when PIPELINING was not
881 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
882 buffer was used for the filename, resulting in a trap when tainted
883 arguments (eg. $domain) were used.
885 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
886 recommended to avoid a possible server-load attack. The feature can be
887 re-enabled via the openssl_options main cofiguration option.
889 JH/45 local_scan API: documented the current smtp_printf() call. This changed
890 for version 4.90 - adding a "more data" boolean to the arguments.
891 Bumped the ABI version number also, this having been missed previously;
892 release versions 4.90 to 4.92.3 inclusive were effectively broken in
893 respect of usage of smtp_printf() by either local_scan code or libraries
894 accessed via the ${dlfunc } expansion item. Both will need coding
895 adjustment for any calls to smtp_printf() to match the new function
896 signature; a FALSE value for the new argument is always safe.
898 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
899 the file-offset (which the Linux syscall does, and exim expects); this
900 resulted in an indefinite loop.
902 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
903 to do ARC verification. The Authentication-Results: header line added
904 by the configuration then had no ARC item.
906 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
907 an incorrect mode was used for reading the result, resulting in it being
914 JH/01 Remove code calling the customisable local_scan function, unless a new
915 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
917 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
918 non-signal-safe functions being used.
920 JH/03 Bug 2269: When presented with a received message having a stupidly large
921 number of DKIM-Signature headers, disable DKIM verification to avoid
922 a resource-consumption attack. The limit is set at twenty.
924 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
925 report of oldest_pass in ${authres } in consequence, and separate out
926 some descriptions of reasons for verification fail.
928 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
929 files in the spool were present and unlocked. A queue-runner could spot
930 them, resulting in a duplicate delivery. Fix that by doing the unlock
931 after the unlink. Investigation by Tim Stewart. Take the opportunity to
932 add more error-checking on spoolfile handling while that code is being
935 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
936 No known attacks, no CVE, this is defensive hardening.
938 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
939 a queue-runner could start a delivery while other operations were ongoing.
940 Cutthrough delivery was a common victim, resulting in duplicate delivery.
941 Found and investigated by Tim Stewart. Fix by using the open message data
942 file handle rather than opening another, and not locally closing it (which
943 releases a lock) for that case, while creating the temporary .eml format
944 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
946 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
947 $sender_verify_failure/$recipient_verify_failure to "random".
949 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
952 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
953 Previously this would segfault.
955 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
958 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
959 like zero, since the resolver should be doing this for us, But we need one
960 as a CNAME but no MX presence gets the CNAME returned; we need to check
961 that doesn't point to an MX to declare it "no MX returned" rather than
962 "error, loop". A new main option is added so the older capability of
963 following some limited number of chain links is maintained.
965 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
967 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
968 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
969 now-deprecated earlier definitions used only specified the range up to TLS
970 1.2 (in the older-version library docs).
972 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
974 JH/15 Rework TLS client-side context management. Stop using a global, and
975 explicitly pass a context around. This enables future use of TLS for
976 connections to service-daemons (eg. malware scanning) while a client smtp
977 connection is using TLS; with cutthrough connections this is quite likely.
979 JH/16 Fix ARC verification to do AS checks in reverse order.
981 JH/17 Support a "tls" option on the ${readsocket } expansion item.
983 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
984 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
985 Previously the "utf8" would be re-prepended for every additional message.
987 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
988 Previously thery were accepted, resulting in issues when attempting to
989 forward messages to a non-supporting MTA.
991 PP/02 Let -n work with printing macros too, not just options.
993 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
994 one parent address was copied, and bogus data was used at delivery-logging
995 time. Either a crash (after delivery) or bogus log data could result.
996 Discovery and analysis by Tim Stewart.
998 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
999 Previously if the string ended mid-character, we did not insert the
1000 promised '?' replacement.
1002 PP/04 Documentation: current string operators work on bytes, not codepoints.
1004 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1005 should pack well giving a smaller memory footprint so better caching and
1006 therefore performance. Group the declarations where this can't be done so
1007 that the byte-sized flag variables are not interspersed among pointer
1008 variables, giving a better chance of good packing by the compiler.
1010 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1011 non-null, to avoid issues with sites running BATV. Previously reports were
1012 sent with an empty envelope sender so looked like bounces.
1014 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1015 The ignore_error flag wasn't being returned from the filter subprocess so
1016 was not set for later routers. Investigation and fix by Matthias Kurz.
1018 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1019 and a msg:complete for the whole, when a message is manually removed using
1020 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1022 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1023 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1024 hope that is portable enough.
1026 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1027 requiring a known-CA anchor certificate; make it now rely entirely on the
1028 TLSA as an anchor. Checking the name on the leaf cert against the name
1029 on the A-record for the host is still done for TA (but not for EE mode).
1031 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1032 included in delivery lines for non-proxied connections, when compiled with
1033 SUPPORT_SOCKS and running with proxy logging enabled.
1035 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1036 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1037 move the existing event to fire before the normal logging of message
1038 failure so that custom logging is bracketed by normal logging.
1040 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1041 msg:fail:internal event. Developement by Matthias Kurz.
1043 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1044 far too small for todays use of crypto signatures stored there. Go all
1045 the way to the max DNS message size of 64kB, even though this might be
1046 overmuch for IOT constrained device use.
1048 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1049 copy a string over itself. The library routine is documented as not
1050 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1052 JH/32 For main options check_spool_space and check_inode_space, where the
1053 platform supports 64b integers, support more than the previous 2^31 kB
1054 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1055 the previous G, M, k.
1057 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1058 $authenticated_fail_id variable on authentication failure. Previously
1061 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1062 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1063 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1064 GNUTLS_SEC_PARAM_MEDIUM.
1066 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1067 side. Previously we would continue as if no SNI had been received.
1069 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1070 "<" character, which should be a list-separator specification, we walked
1071 off past the nul-terimation.
1073 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1074 causes) even when the retry time is not yet met. Previously they were
1075 not, meaning that when (say) an account was over-quota and temp-rejecting,
1076 and multiple senders' messages were queued, only one sender would get
1077 notified on each configured delay_warning cycle.
1079 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1081 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1082 cipher-suites, an error can be left on the stack even for a succeeding
1083 accept; this results in impossible error messages when a later operation
1086 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1087 return error codes indicating retry. Under TLS1.3 this becomes required.
1089 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1090 it only wrote the new authenticators, resulting in a lack of tracking of
1091 peer changes of ESMTP extensions until the next cache flush.
1093 JH/41 Fix the loop reading a message header line to check for integer overflow,
1094 and more-often against header_maxsize. Previously a crafted message could
1095 induce a crash of the recive process; now the message is cleanly rejected.
1097 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1098 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1104 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1105 When redis_servers is set to a list of > 1 element, and the Redis servers
1106 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1107 case of MOVED into a DEFER case instead, thus moving the query onto the
1108 next server in the list. For a cluster of N elements, all N servers must
1109 be defined in redis_servers.
1111 GF/02 Catch and remove uninitialized value warning in exiqsumm
1112 Check for existence of @ARGV before looking at $ARGV[0]
1114 JH/01 Replace the store_release() internal interface with store_newblock(),
1115 which internalises the check required to safely use the old one, plus
1116 the allocate and data copy operations duplicated in both (!) of the
1117 extant use locations.
1119 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1120 modifier. This matches the restriction on the commandline.
1122 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1123 Previously only the last row was returned.
1125 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1126 we assumed that tags in the header were well-formed, and parsed the
1127 element content after inspecting only the first char of the tag.
1128 Assumptions at that stage could crash the receive process on malformed
1131 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1132 While running the DKIM ACL we operate on the Permanent memory pool so that
1133 variables created with "set" persist to the DATA ACL. Also (at any time)
1134 DNS lookups that fail create cache records using the Permanent pool. But
1135 expansions release any allocations made on the current pool - so a dnsdb
1136 lookup expansion done in the DKIM ACL releases the memory used for the
1137 DNS negative-cache, and bad things result. Solution is to switch to the
1138 Main pool for expansions.
1139 While we're in that code, add checks on the DNS cache during store_reset,
1140 active in the testsuite.
1141 Problem spotted, and debugging aided, by Wolfgang Breyha.
1143 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1144 When none of the hosts presented to a transport match an already-open
1145 connection, close it and proceed with the list. Previously we would
1146 queue the message. Spotted by Lena with Yahoo, probably involving
1149 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1150 Previously a spurious "250 OK id=" response was appended to the proper
1153 JH/08 The "support for" informational output now, which built with Content
1154 Scanning support, has a line for the malware scanner interfaces compiled
1155 in. Interface can be individually included or not at build time.
1157 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1158 by the template makefile "src/EDITME". The "STREAM" support for an older
1159 ClamAV interface method is removed.
1161 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1162 rows affected is given instead).
1164 JH/11 The runtime Berkeley DB library version is now additionally output by
1165 "exim -d -bV". Previously only the compile-time version was shown.
1167 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1168 SMTP connection. Previously, when one had more recipients than the
1169 first, an abortive onward connection was made. Move to full support for
1170 multiple onward connections in sequence, handling cutthrough connection
1171 for all multi-message initiating connections.
1173 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1174 routers. Previously, a multi-recipient message would fail to match the
1175 onward-connection opened for the first recipient, and cause its closure.
1177 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1178 a timeout on read on a GnuTLS initiating connection, resulting in the
1179 initiating connection being dropped. This mattered most when the callout
1180 was marked defer_ok. Fix to keep the two timeout-detection methods
1183 JH/15 Relax results from ACL control request to enable cutthrough, in
1184 unsupported situations, from error to silently (except under debug)
1185 ignoring. This covers use with PRDR, frozen messages, queue-only and
1188 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1190 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1191 metadata, resulting in a crash in free().
1193 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1194 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1195 Broken also in d185889f4, with init system revamp.
1197 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1198 Previously we abruptly closed the connection after reading a malware-
1199 found indication; now we go on to read the "scan ok" response line,
1202 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1203 ACL. Previously, a crash would result.
1205 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1206 macro text after a replacement (previously it was only once per line) and
1207 by skipping builtin macros when searching for an uppercase lead character.
1209 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1210 for the build is renamed.
1212 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1213 was allocated for every new TLS startup, meaning one per message. Fix
1214 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1216 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1217 reported the original. Fix to report (as far as possible) the ACL
1218 result replacing the original.
1220 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1221 OpenSSL. Certificate information is loaded for every new TLS startup,
1222 and the resources needed to be freed.
1224 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1226 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1229 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1230 DATA response info to the (existing) per-recipient response info for
1231 the "C=" log element. It can have useful tracking info from the
1232 destination system. Patch from Simon Arlott.
1234 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1235 length value. Previously this would segfault.
1237 HS/02 Support Avast multiline protoocol, this allows passing flags to
1238 newer versions of the scanner.
1240 JH/28 Ensure that variables possibly set during message acceptance are marked
1241 dead before release of memory in the daemon loop. This stops complaints
1242 about them when the debug_store option is enabled. Discovered specifically
1243 for sender_rate_period, but applies to a whole set of variables.
1244 Do the same for the queue-runner and queue-list loops, for variables set
1245 from spool message files. Do the same for the SMTP per-message loop, for
1246 certain variables indirectly set in ACL operations.
1248 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1249 as a multi-recipient message from a mailinglist manager). The coding had
1250 an arbitrary cutoff number of characters while checking for more input;
1251 enforced by writing a NUL into the buffer. This corrupted long / fast
1252 input. The problem was exposed more widely when more pipelineing of SMTP
1253 responses was introduced, and one Exim system was feeding another.
1254 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1255 receiving system, and refused recipients seen by the sending system
1256 (propating to people being dropped from mailing lists).
1257 Discovered and pinpointed by David Carter.
1259 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1260 replaced by the ${authresults } expansion.
1262 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1264 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1265 allows proper process termination in container environments.
1267 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1268 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1270 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1271 and "err_perm", deprecated since 4.83 when the RFC-defined words
1272 "temperror" and "permerror" were introduced.
1274 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1275 transport-filters or DKIM-signing. The restriction was lost in the
1276 consolidation of verify-callout and delivery SMTP handling.
1277 Extend the restriction to also cover ARC-signing.
1279 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1280 in defer=pass mode supply a 450 to the initiator. Previously the message
1283 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1284 tls_require_ciphers is used as before.
1286 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1287 "pass_unscanned". Only tmpfails from the scanner are written to
1288 the paniclog, as they may require admin intervention (permission
1289 denied, license issues). Other scanner errors (like decompression
1290 bombs) do not cause a paniclog entry.
1292 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1293 Previously it was possible to log spurious information in receive log
1296 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1297 triggered odd behaviour from Outlook Express clients.
1299 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1302 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1303 since the IETF WG has not yet settled on that versus the original
1304 "bare" representation.
1306 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1307 Previously the millisecond value corrupted the output.
1308 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1309 corrupted the output.
1315 JH/01 Rework error string handling in TLS interface so that the caller in
1316 more cases is responsible for logging. This permits library-sourced
1317 string to be attached to addresses during delivery, and collapses
1318 pairs of long lines into single ones.
1320 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1321 during configuration. Wildcards are allowed and expanded.
1323 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1324 This permits better logging.
1326 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1327 do not close it down and have the child transport start it up again on
1328 the passed-on TCP connection. Instead, proxy the child (and any
1329 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1330 affected: the continued delivery log lines do not have any DNSSEC, TLS
1331 Certificate or OCSP information. TLS cipher information is still logged.
1333 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1334 identical IP addresses on different listening ports. Will also affect
1337 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1338 add noisy ifdef guards to special-case this sillyness.
1339 Patch from Bernd Kuhls.
1341 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1342 than 255 are no longer allowed.
1344 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1345 Disable the session-cache too, which might reduce our load. Since we
1346 currrectly use a new context for every connection, both as server and
1347 client, there is no benefit for these.
1348 GnuTLS appears to not support tickets server-side by default (we don't
1349 call gnutls_session_ticket_enable_server()) but client side is enabled
1350 by default on recent versions (3.1.3 +) unless the PFS priority string
1353 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1354 <https://reproducible-builds.org/specs/source-date-epoch/>.
1356 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1357 the check for any unsuccessful recipients did not notice the limit, and
1358 erroneously found still-pending ones.
1360 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1361 MSG_MORE. Only in-clear (not on TLS connections).
1363 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1364 transport-filter is involved we can buffer the headers in memory for
1365 creating the signature, and read the spool data file once for the
1366 signature and again for transmission.
1368 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1369 4.77 as the kernel support then wasn't solid, having issues in 64bit
1370 mode. Now, it's been long enough. Add support for FreeBSD also.
1372 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1373 case where the routing stage had gathered several addresses to send to
1374 a host before calling the transport for the first, we previously failed
1375 to close down TLS in the old transport process before passing the TCP
1376 connection to the new process. The new one sent a STARTTLS command
1377 which naturally failed, giving a failed delivery and bloating the retry
1378 database. Investigation and fix prototype from Wolfgang Breyha.
1380 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1381 false-negatives in the check that the sender had not preempted a response
1382 or prompt from Exim (running as a server), due to that code's lack of
1383 awareness of the SMTP input buffering.
1385 PP/04 Add commandline_checks_require_admin option.
1386 Exim drops privileges sanely, various checks such as -be aren't a
1387 security problem, as long as you trust local users with access to their
1388 own account. When invoked by services which pass untrusted data to
1389 Exim, this might be an issue. Set this option in main configuration
1390 AND make fixes to the calling application, such as using `--` to stop
1393 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1394 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1395 responses to those, into a single TLS record each way (this usually means
1396 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1399 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1400 affects you only if you're dancing at the edge of the param size limits.
1401 If you are, and this message makes sense to you, then: raise the
1402 configured limit or use OpenSSL 1.1. Nothing we can do for older
1405 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1406 cmdline element to get the documented default one. Previously it was
1409 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1410 are unsuable for non-message-specific reasons - eg. connection timeout,
1411 banner-time rejection.
1413 JH/16 Fix logging of delivery remote port, when specified by router, under
1416 PP/06 Repair manualroute's ability to take options in any order, even if one
1417 is the name of a transport.
1420 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1422 JH/17 Change the list-building routines interface to use the expanding-string
1423 triplet model, for better allocation and copying behaviour.
1425 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1426 Previously it was constructed the first time a possibly-matching string
1427 was met in the configuration file input during startup; now it is done
1430 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1431 compatible one, to avoid the (poorly documented) possibility of a config
1432 file in the working directory redirecting the DB files, possibly correpting
1433 some existing file. CVE-2017-10140 assigned for BDB.
1435 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1436 cache-hot. Previously, although the result was properly cached, the
1437 initial verify call returned a defer.
1439 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1440 the main verify for receipient in uncached-mode.
1442 JH/22 Retire historical build files to an "unsupported" subdir. These are
1443 defined as "ones for which we have no current evidence of testing".
1445 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1446 if present. Previously it was ignored.
1448 JH/24 Start using specified-initialisers in C structure init coding. This is
1449 a C99 feature (it's 2017, so now considered safe).
1451 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1452 if was a fixed-sized field and bitmask ops via macros; it is now more
1455 PP/07 GitHub PR 56: Apply MariaDB build fix.
1456 Patch provided by Jaroslav Škarvada.
1458 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1459 during Coverity cleanups [4.87 JH/47]
1460 Diagnosis and fix provided by Michael Fischer v. Mollard.
1462 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1463 the right size to place the terminating semicolon on its own folded
1464 line, the header hash was calculated to an incorrect value thanks to
1465 the (relaxed) space the fold became.
1467 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1468 and confused the parent.
1470 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1471 which could crash as a result. This could lead to undeliverable messages.
1473 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1476 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1477 matters on fast-turnover and PID-randomising systems, which were getting
1478 out-of-order delivery.
1480 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1481 a possibly-overlapping copy. The symptom was that "Remote host closed
1482 connection in response to HELO" was logged instead of the actual 4xx
1485 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1486 Previously only that bufferd was discarded, resulting in SYMTP command
1489 JH/32 DKIM: when a message has multiple signatures matching an identity given
1490 in dkim_verify_signers, run the dkim acl once for each. Previously only
1491 one run was done. Bug 2189.
1493 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1494 "panic the current process" to "deliberately defer". The panic log is
1495 still written with the problem list name; the mail and reject logs now
1496 get a temp-reject line for the message that was being handled, saying
1497 something like "domains check lookup or other defer". The SMTP 451
1498 message is still "Temporary local problem".
1500 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1501 A crafted sequence of BDAT commands could result in in-use memory beeing
1502 freed. CVE-2017-16943.
1504 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1505 from SMTP input. Previously it was always done; now only done for DATA
1506 and not BDAT commands. CVE-2017-16944.
1508 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1509 to the message (such as an overlong header line). Previously this was
1510 not done and we did not exit BDAT mode. Followon from the previous item
1511 though a different problem.
1517 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1518 than -2003 did; needs libidn2 in addition to libidn.
1520 JH/02 The path option on a pipe transport is now expanded before use.
1522 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1523 Patch provided by "Björn", documentation fix added too.
1525 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1526 missing a wire-to-host endian conversion.
1528 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1529 close after a BDAT command line could be taken as a following command,
1530 giving a synch failure. Fix by only checking for synch immediately
1531 before acknowledging the chunk.
1533 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1534 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1535 macro. Patches provided by Josh Soref.
1537 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1538 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1539 by IANA as a service extension. However, John Klensin suggests that we
1542 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1543 the dkim code may be unix-mode line endings rather than smtp wire-format
1544 CRLF, so prepend a CR to any bare LF.
1546 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1547 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1549 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1550 the first were themselves being wrongly included in the feed into dkim
1551 processing; with most chunk sizes in use this resulted in an incorrect
1552 body hash calculated value.
1554 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1555 DKIM signature block, for verification. Although advised against by
1556 standards it is specifically not ruled illegal.
1558 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1560 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1561 missing a body hash (the bh= tag).
1563 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1564 It seems that HAProxy sends the Proxy Protocol information in clear and
1565 only then does a TLS startup, so do the same.
1567 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1568 TCP connections (such as for Spamd) unless the daemon successfully set
1569 Fast Open mode on its listening sockets. This fixes breakage seen on
1570 too-old kernels or those not configured for Fast Open, at the cost of
1571 requiring both directions being enabled for TFO, and TFO never being used
1572 by non-daemon-related Exim processes.
1574 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1575 endings, at least on the first header line. Try to canonify any that get
1576 past that check, despite the cost.
1578 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1579 now limited to an arbitrary five deep, while parsing addresses with the
1580 strip_excess_angle_brackets option enabled.
1582 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1583 instead leave the unprompted TLS handshake in socket buffer for the
1584 TLS library to consume.
1586 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1588 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1590 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1591 region in one operation, for speed, and this leaves assigned pointers
1592 dangling. Add checks run only under the testsuite which checks all
1593 variables at a store-reset and panics on a dangling pointer; add code
1594 explicitly nulling out all the variables discovered. Fixes one known
1595 bug: a transport crash, where a dangling pointer for $sending_ip_address
1596 originally assigned in a verify callout, is re-used.
1598 PP/06 Drop '.' from @INC in various Perl scripts.
1600 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1602 PP/08 Reduce a number of compilation warnings under clang; building with
1603 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1604 should be warning-free.
1606 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1608 HS/01 Fix portability problems introduced by PP/08 for platforms where
1609 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1611 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1612 chunk. This allows us to accept broken chunked messages. We need a more
1613 general solution here.
1615 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1616 already-broken messages in the queue.
1618 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1620 JH/19 Fix reference counting bug in routing-generated-address tracking.
1626 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1627 supports it and a size is available (ie. the sending peer gave us one).
1629 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1630 years of being deprecated). The replacements are the ACLs
1631 acl_smtp_mime and acl_not_smtp_mime.
1633 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1634 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1635 or even an in-clear connection were permitted. Now, if the host lookup
1636 was dnssec and dane was requested then the host is only used if the
1637 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1638 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1639 if one fails this test.
1640 This means that a poorly-configured remote DNS will make it incommunicado;
1641 but it protects against a DNS-interception attack on it.
1643 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1644 non-noisy when a race steals the message being considered.
1646 JH/05 If main configuration option tls_certificate is unset, generate a
1647 self-signed certificate for inbound TLS connections.
1649 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1650 in rewrites and routers.
1652 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1653 and logged a warning sing 4.83; now they are a configuration file error.
1655 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1656 (lacking @domain). Apply the same qualification processing as RCPT.
1658 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1660 JH/10 Support ${sha256:} applied to a string (as well as the previous
1663 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1664 a cutthrough deliver is pending, as we always want to make a connection.
1665 This also avoids re-routing the message when later placing the cutthrough
1666 connection after a verify cache hit.
1667 Do not update it with the verify result either.
1669 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1670 when routing results in more than one destination address.
1672 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1673 signing (which inhibits the cutthrough capability). Previously only
1674 the presence of an option was tested; now an expansion evaluating as
1675 empty is permissible (obviously it should depend only on data available
1676 when the cutthrough connection is made).
1678 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1679 the relevant preceding SMTP command did not note the pipelining mode.
1681 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1682 Previously they were not counted.
1684 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1685 as one having no matching records. Previously we deferred the message
1686 that needed the lookup.
1688 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1689 distinguished as "(=".
1691 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1692 for missing MX records. Previously it only worked for missing A records.
1694 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1696 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1697 after the data-go-ahead and data-ack. Patch from Jason Betts.
1699 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1700 even for a "none" policy. Patch from Tony Meyer.
1702 JH/22 Fix continued use of a connection for further deliveries. If a port was
1703 specified by a router, it must also match for the delivery to be
1706 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1707 When one of the recipients of a message was unsuitable for the connection
1708 (has no matching addresses), we lost track of needing to mark it
1709 deferred. As a result mail would be lost.
1711 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1713 JH/25 Decoding ACL controls is now done using a binary search; the source code
1714 takes up less space and should be simpler to maintain. Merge the ACL
1715 condition decode tables also, with similar effect.
1717 JH/26 Fix problem with one_time used on a redirect router which returned the
1718 parent address unchanged. A retry would see the parent address marked as
1719 delivered, so not attempt the (identical) child. As a result mail would
1722 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1723 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1724 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1727 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1728 Main config options check_{log,spool}_{inodes,space} are now
1729 100 inodes, 10MB unless set otherwise in the configuration.
1731 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1732 Previously it only applied to the main-section connection policy
1735 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1737 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1738 by me. Added RFC7919 DH primes as an alternative.
1740 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1741 are not in the system include path.
1743 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1744 GnuTLS, when a session startup failed (eg because the client disconnected)
1745 Exim did stdio operations after fclose. This was exposed by a recent
1746 change which nulled out the file handle after the fclose.
1748 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1749 signed directly by the cert-signing cert, rather than an intermediate
1750 OCSP-signing cert. This is the model used by LetsEncrypt.
1752 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1754 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1755 an incoming connection.
1757 HS/02 Bug 1802: Do not half-close the connection after sending a request
1760 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1761 fallback to "prime256v1".
1763 JH/34 SECURITY: Use proper copy of DATA command in error message.
1764 Could leak key material. Remotely exploitable. CVE-2016-9963.
1770 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1771 and 3.4.4 - once the server is enabled to respond to an OCSP request
1772 it does even when not requested, resulting in a stapling non-aware
1773 client dropping the TLS connection.
1775 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1776 support variable-length bit vectors. No functional change.
1778 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1779 The I= interface field on outgoing lines is now after the H= remote
1780 host field, same as incoming lines. There is a separate
1781 outgoing_interface log selector which allows you to disable the
1784 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1785 If not running log_selector +smtp_connection the mainlog would be held
1786 open indefinitely after a "too many connections" event, including to a
1787 deleted file after a log rotate. Leave the per net connection logging
1788 leaving it open for efficiency as that will be quickly detected by the
1789 check on the next write.
1791 HS/01 Bug 1671: Fix post transport crash.
1792 Processing the wait-<transport> messages could crash the delivery
1793 process if the message IDs didn't exist for some reason. When
1794 using 'split_spool_directory=yes' the construction of the spool
1795 file name failed already, exposing the same netto behaviour.
1797 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1798 mime_regex ACL conditions.
1800 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1801 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1802 to HELO, local diagnostic string.
1804 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1805 log line to debug. Even when configured with a tls authenticator many
1806 client connections are expected to not authenticate in this way, so
1807 an authenticate fail is not an error.
1809 HS/02 Add the Exim version string to the process info. This way exiwhat
1810 gives some more detail about the running daemon.
1812 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1813 matter for fast-change records such as DNSBLs.
1815 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1816 retry record, even if constant. There may be multiple transports with
1817 different interface settings and the retry behaviour needs to be kept
1820 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1822 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1824 JH/10 Bug 840: fix log_defer_output option of pipe transport
1826 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1827 hosts_randomize. This matters a lot when combined with mua_wrapper.
1829 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1830 ${quote_pgsql:<string>} operator.
1832 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1834 JH/14 Change header file name for freeradius-client. Relevant if compiling
1835 with Radius support; from the Gentoo tree and checked under Fedora.
1837 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1839 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1840 option transport_filter, meaning no filtering.
1842 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1844 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1845 defaults to "*" (all hosts). The variable is now available when not built
1846 with TLS, default unset, mainly to enable keeping the testsuite sane.
1847 If a server certificate is not supplied (via tls_certificate) an error is
1848 logged, and clients will find TLS connections fail on startup. Presumably
1849 they will retry in-clear.
1850 Packagers of Exim are strongly encouraged to create a server certificate
1851 at installation time.
1853 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1854 with the $config_file variable.
1856 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1857 in transport context, after the attempt, and per-recipient. The latter type
1858 is per host attempted. The event data is the error message, and the errno
1859 information encodes the lookup type (A vs. MX) used for the (first) host,
1860 and the trailing two digits of the smtp 4xx response.
1862 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1863 to write to mainlog (or rejectlog, paniclog) in the window between file
1864 creation and permissions/ownership being changed. Particularly affects
1865 installations where exicyclog is run as root, rather than exim user;
1866 result is that the running daemon panics and dies.
1868 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1870 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1871 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1872 "pri" and "weight". Note that the previous implicit priority given by the
1873 list order is no longer honoured.
1875 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1876 for DKIM processing.
1878 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1879 by defining SUPPORT_SOCKS.
1881 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1882 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1883 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1884 variables are renamed to proxy_{local,external}_{address,port}.
1886 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1887 for a build by defining SUPPORT_I18N
1889 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1890 of the query string, and make ${quote_redis:} do that quoting.
1892 JH/29 Move Events support from Experimental to mainline, enabled by default
1893 and removable for a build by defining DISABLE_EVENT.
1895 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1897 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1898 cached by the daemon.
1900 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1901 by defining LOOKUP_REDIS. The libhiredis library is required.
1903 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1904 keys are given for lookup.
1906 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1907 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1908 only supported when built with TLS support. The PolarSSL SHA routines
1909 are still used when the TLS library is too old for convenient support.
1911 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1912 openssl_options), for security. OpenSSL forces this from version 1.1.0
1913 server-side so match that on older versions.
1915 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1916 allocation for $value could be released as the expansion processing
1917 concluded, but leaving the global pointer active for it.
1919 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1920 and to use the domains and local_parts ACL conditions.
1922 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1923 incorrectly not doubled on cutthrough transmission, hence seen as a
1924 body-termination at the receiving system - resulting in truncated mails.
1925 Commonly the sender saw a TCP-level error, and retransmitted the message
1926 via the normal store-and-forward channel. This could result in duplicates
1927 received - but deduplicating mailstores were liable to retain only the
1928 initial truncated version.
1930 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1932 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1934 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1935 we're in there, support oversigning also; bug 1309.
1937 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1939 HS/04 Add support for keep_environment and add_environment options.
1941 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1942 either intentional arithmetic overflow during PRNG, or testing config-
1945 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1946 delivery resulted in actual delivery. Cancel cutthrough before DATA
1949 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1950 rejects a recipient: pass the reject to the originator.
1952 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1953 Many were false-positives and ignorable, but it's worth fixing the
1956 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1957 for the new environment-manipulation done at startup. Move the routines
1958 from being local to tls.c to being global via the os.c file.
1960 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1961 an extract embedded as result-arg for a map, the first arg for extract
1962 is unavailable so we cannot tell if this is a numbered or keyed
1963 extraction. Accept either.
1969 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1972 JH/02 The smtp transport option "multi_domain" is now expanded.
1974 JH/03 The smtp transport now requests PRDR by default, if the server offers
1977 JH/04 Certificate name checking on server certificates, when exim is a client,
1978 is now done by default. The transport option tls_verify_cert_hostnames
1979 can be used to disable this per-host. The build option
1980 EXPERIMENTAL_CERTNAMES is withdrawn.
1982 JH/05 The value of the tls_verify_certificates smtp transport and main options
1983 default to the word "system" to access the system default CA bundle.
1984 For GnuTLS, only version 3.0.20 or later.
1986 JH/06 Verification of the server certificate for a TLS connection is now tried
1987 (but not required) by default. The verification status is now logged by
1988 default, for both outbound TLS and client-certificate supplying inbound
1991 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1994 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1995 Status Notification (bounce) messages are now MIME format per RFC 3464.
1996 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1997 under the control of the dsn_advertise_hosts option, and routers may
1998 have a dsn_lasthop option.
2000 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2001 default, modifiable by a malware= option. The list separator for
2002 the options can now be changed in the usual way. Bug 68.
2004 JH/10 The smtp_receive_timeout main option is now expanded before use.
2006 JH/11 The incoming_interface log option now also enables logging of the
2007 local interface on delivery outgoing connections.
2009 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2010 if the interface and destination host and port all match.
2012 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2015 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2016 Patch from Andrew Lewis.
2018 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2019 now supports optional time-restrictions, weighting, and priority
2020 modifiers per server. Patch originally by <rommer@active.by>.
2022 JH/16 The spamd_address main option now supports a mixed list of local
2023 and remote servers. Remote servers can be IPv6 addresses, and
2024 specify a port-range.
2026 JH/17 Bug 68: The spamd_address main option now supports an optional
2027 timeout value per server.
2029 JH/18 Bug 1581: Router and transport options headers_add/remove can
2030 now have the list separator specified.
2032 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2035 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2038 JH/21 Support for the A6 type of dns record is withdrawn.
2040 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2041 rather than the verbs used.
2043 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2044 from 255 to 1024 chars.
2046 JH/24 Verification callouts now attempt to use TLS by default.
2048 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2049 are generic router options now. The defaults didn't change.
2051 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2052 Original patch from Alexander Shikoff, worked over by JH.
2054 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2055 files if ZCAT_COMMAND is not executable.
2057 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2059 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2061 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2062 Normally benign, it bites when the pair was led to by a CNAME;
2063 modern usage is to not canonicalize the domain to a CNAME target
2064 (and we were inconsistent anyway for A-only vs AAAA+A).
2066 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2068 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2069 when evaluating $sender_host_dnssec.
2071 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2072 $sender_helo_dnssec variable.
2074 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2076 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2078 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2080 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2081 documented as working, but never had. Support all but $spam_report.
2083 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2084 added for tls authenticator.
2086 HS/03 Add perl_taintmode main config option
2092 TL/01 When running the test suite, the README says that variables such as
2093 no_msglog_check are global and can be placed anywhere in a specific
2094 test's script, however it was observed that placement needed to be near
2095 the beginning for it to behave that way. Changed the runtest perl
2096 script to read through the entire script once to detect and set these
2097 variables, reset to the beginning of the script, and then run through
2098 the script parsing/test process like normal.
2100 TL/02 The BSD's have an arc4random API. One of the functions to induce
2101 adding randomness was arc4random_stir(), but it has been removed in
2102 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2103 function when detected.
2105 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2106 cause callback expansion.
2108 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2109 syntax errors in an expansion can be treated as a string instead of
2110 logging or causing an error, due to the internal use of bool_lax
2111 instead of bool when processing it.
2113 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2114 server certificates when making smtp deliveries.
2116 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2118 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2120 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2122 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2123 Merged patch from Sebastian Wiedenroth.
2125 JH/05 Fix results-pipe from transport process. Several recipients, combined
2126 with certificate use, exposed issues where response data items split
2127 over buffer boundaries were not parsed properly. This eventually
2128 resulted in duplicates being sent. This issue only became common enough
2129 to notice due to the introduction of connection certificate information,
2130 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2132 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2133 size buffer was used, resulting in syntax errors when an expansion
2136 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2137 version 3.3.6 or later.
2139 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2140 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2141 both become "event_action", the variables become $event_name, $event_data
2142 and $event_defer_errno. There is a new variable $verify_mode, usable in
2143 routers, transports and related events. The tls:cert event is now also
2144 raised for inbound connections, if the main configuration event_action
2147 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2148 early OCSP support, but no stapling (appears to be less than 1.0.0).
2150 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2151 server certificate names available under the smtp transport option
2152 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2155 JH/10 Time-related extraction expansions from certificates now use the main
2156 option "timezone" setting for output formatting, and are consistent
2157 between OpenSSL and GnuTLS compilations. Bug 1541.
2159 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2160 encoded parameter in the incoming message. Bug 1558.
2162 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2163 include certificate info, eximon was claiming there were spoolfile
2166 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2168 JH/14 Log delivery-related information more consistently, using the sequence
2169 "H=<name> [<ip>]" wherever possible.
2171 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2172 are problematic for Debian distribution, omit them from the release
2175 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2177 JH/16 Fix string representation of time values on 64bit time_t architectures.
2180 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2181 output list separator was used.
2186 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2187 checkers that were complaining about end of non-void function with no
2190 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2191 This was a regression introduced in 4.83 by another bugfix.
2193 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2195 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2196 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2202 TF/01 Correctly close the server side of TLS when forking for delivery.
2204 When a message was received over SMTP with TLS, Exim failed to clear up
2205 the incoming connection properly after forking off the child process to
2206 deliver the message. In some situations the subsequent outgoing
2207 delivery connection happened to have the same fd number as the incoming
2208 connection previously had. Exim would try to use TLS and fail, logging
2209 a "Bad file descriptor" error.
2211 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2212 utilities have not been installed.
2214 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2215 temporary space as the ACL may create new global variables.
2217 TL/01 LDAP support uses per connection or global context settings, depending
2218 upon the detected version of the libraries at build time.
2220 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2221 to extract and use the src ip:port in logging and expansions as if it
2222 were a direct connection from the outside internet. PPv2 support was
2223 updated based on HAProxy spec change in May 2014.
2225 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2227 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2228 Properly escape header and check for NULL return.
2230 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2233 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2235 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2236 characters in header names, implemented as a verify condition.
2237 Contributed by Michael Fischer v. Mollard.
2239 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2240 results permerror and temperror. Previous values are deprecated but
2241 still accepted. In a future release, err_perm and err_temp will be
2242 completely removed, which will be a backward incompatibility if the
2243 ACL tests for either of these two old results. Patch contributed by
2244 user bes-internal on the mailing list.
2246 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2248 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2249 selectors, in both main and reject logs.
2251 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2254 JH/07 Add malware type "sock" for talking to simple daemon.
2256 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2258 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2259 routers/transports under cutthrough routing.
2261 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2262 numbers. Touch up "bool" conditional to keep the same definition.
2264 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2266 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2268 JH/12 Expand items in router/transport headers_add or headers_remove lists
2269 individually rather than the list as a whole. Bug 1452.
2271 Required for reasonable handling of multiple headers_ options when
2272 they may be empty; requires that headers_remove items with embedded
2273 colons must have them doubled (or the list-separator changed).
2275 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2276 view the policy declared in the DMARC record. Currently, $dmarc_status
2277 is a combined value of both the record presence and the result of the
2280 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2282 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2283 dnslookup router and the smtp transport (applying to the forward
2286 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2287 of ldap servers used for a specific lookup. Patch provided by Heiko
2290 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2291 New variable $lookup_dnssec_authenticated for observability.
2293 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2294 Patch submitted by Lars Timman.
2296 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2298 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2299 Requires trusted mode and valid format message id, aborts otherwise.
2300 Patch contributed by Heiko Schlichting.
2302 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2303 certextract with support for various fields. Bug 1358.
2305 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2306 is requested by default, modifiable by smtp transport option
2309 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2310 operate on certificate variables to give certificate fingerprints
2311 Also new ${sha256:cert_variable}.
2313 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2315 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2316 Christian Aistleitner.
2318 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2320 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2321 file. Patch from Wolfgang Breyha.
2323 JH/25 Expand the coverage of the delivery $host and $host_address to
2324 client authenticators run in verify callout. Bug 1476.
2326 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2327 align with daemon_smtp_ports. Bug 72.
2329 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2330 support and error reporting did not work properly.
2332 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2333 and is readable. Patch from Andrew Colin Kissa.
2335 TL/14 Enhance documentation of ${run expansion and how it parses the
2336 commandline after expansion, particularly in the case when an
2337 unquoted variable expansion results in an empty value.
2339 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2341 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2342 by renaming away from it. Addresses GH issue 15, reported by
2345 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2347 TL/15 SECURITY: prevent double expansion in math comparison functions
2348 (can expand unsanitized data). Not remotely exploitable.
2355 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2357 PP/02 Make -n do something, by making it not do something.
2358 When combined with -bP, the name of an option is not output.
2360 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2363 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2364 $sender_host_name and config options to manage this, and basic check
2367 PP/05 DSCP support for outbound connections and control modifier for inbound.
2369 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2370 (Only plugin which currently uses this is kerberos4, which nobody should
2371 be using, but we should make it available and other future plugins might
2372 conceivably use it, even though it would break NAT; stuff *should* be
2373 using channel bindings instead).
2375 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2376 name; added for Sendmail compatibility; requires admin caller.
2377 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2378 just ignore it); requires trusted caller.
2379 Also parse but ignore: -Ac -Am -X<logfile>
2382 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2384 TL/02 Add +smtp_confirmation as a default logging option.
2386 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2387 Patch by Magnus Holmgren from 2007-02-20.
2389 TL/04 Bugzilla 1281 - Spec typo.
2390 Bugzilla 1283 - Spec typo.
2391 Bugzilla 1290 - Spec grammar fixes.
2393 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2395 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2397 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2398 mailing list by by Dmitry Isaikin.
2400 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2402 JH/02 Support "G" suffix to numbers in ${if comparisons.
2404 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2406 NM/01 Bugzilla 1197 - Spec typo
2407 Bugzilla 1196 - Spec examples corrections
2409 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2411 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2412 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2415 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2416 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2418 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2419 "acl {{name}{arg}...}", and optional args on acl condition
2422 JH/05 Permit multiple router/transport headers_add/remove lines.
2424 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2426 JH/07 Avoid using a waiting database for a single-message-only transport.
2427 Performance patch from Paul Fisher. Bugzilla 1262.
2429 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2432 JH/09 Add $headers_added variable, with content from use of ACL modifier
2433 add_header (but not yet added to the message). Bugzilla 199.
2435 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2436 Pulled from Bugzilla 817 by Wolfgang Breyha.
2438 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2440 (nb: this is the same fix as in Exim 4.80.1)
2442 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2445 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2448 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2450 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2451 advertise SMTP AUTH mechanism to us, instead of a generic
2452 protocol violation error. Also, make Exim more robust to bad
2453 data from the Dovecot auth socket.
2455 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2457 When a queue runner is handling a message, Exim first routes the
2458 recipient addresses, during which it prunes them based on the retry
2459 hints database. After that it attempts to deliver the message to
2460 any remaining recipients. It then updates the hints database using
2463 So if a recipient address works intermittently, it can get repeatedly
2464 deferred at routing time. The retry hints record remains fresh so the
2465 address never reaches the final cutoff time.
2467 This is a fairly common occurrence when a user is bumping up against
2468 their storage quota. Exim had some logic in its local delivery code
2469 to deal with this. However it did not apply to per-recipient defers
2470 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2472 This change adds a proper retry rule check during routing so that the
2473 final cutoff time is checked against the message's age. We only do
2474 this check if there is an address retry record and there is not a
2475 domain retry record; this implies that previous attempts to handle
2476 the address had the retry_use_local_parts option turned on. We use
2477 this as an approximation for the destination being like a local
2478 delivery, as in LMTP.
2480 I suspect this new check makes the old local delivery cutoff check
2481 redundant, but I have not verified this so I left the code in place.
2483 TF/02 Correct gecos expansion when From: is a prefix of the username.
2485 Test 0254 submits a message to Exim with the header
2489 When I ran the test suite under the user fanf2, Exim expanded
2490 the header to contain my full name, whereas it should have added
2491 a Resent-Sender: header. It erroneously treats any prefix of the
2492 username as equal to the username.
2494 This change corrects that bug.
2496 GF/01 DCC debug and logging tidyup
2497 Error conditions log to paniclog rather than rejectlog.
2498 Debug lines prefixed by "DCC: " to remove any ambiguity.
2500 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2502 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2503 Bug spotted by Jeremy Harris; was flawed since initial commit.
2504 Would have resulted in OCSP responses post-SNI triggering an Exim
2505 NULL dereference and crash.
2507 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2509 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2510 Bug detection, analysis and fix by Samuel Thibault.
2511 Bugzilla 1331, Debian bug #698092.
2513 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2515 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2516 Server implementation by Todd Lyons, client by JH.
2517 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2518 config variable "prdr_enable" controls whether the server
2519 advertises the facility. If the client requests PRDR a new
2520 acl_data_smtp_prdr ACL is called once for each recipient, after
2521 the body content is received and before the acl_smtp_data ACL.
2522 The client is controlled by both of: a hosts_try_prdr option
2523 on the smtp transport, and the server advertisement.
2524 Default client logging of deliveries and rejections involving
2525 PRDR are flagged with the string "PRDR".
2527 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2528 fclose(). Diagnosis by Todd Lyons.
2530 PP/17 Update configure.default to handle IPv6 localhost better.
2531 Patch by Alain Williams (plus minor tweaks).
2534 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2535 This is now consistent with GnuTLS, and is now documented: the
2536 previous undocumented portable approach to treating the option as
2537 unset was to force an expansion failure. That still works, and
2538 an empty string is now equivalent.
2540 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2541 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2542 not performing validation itself.
2544 PP/20 Added force_command boolean option to pipe transport.
2545 Patch from Nick Koston, of cPanel Inc.
2547 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2550 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2552 PP/21 Fix eximon continuous updating with timestamped log-files.
2553 Broken in a format-string cleanup in 4.80, missed when I repaired the
2554 other false fix of the same issue.
2555 Report and fix from Heiko Schlichting.
2558 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2559 Report from Prashanth Katuri.
2561 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2562 It's SecureTransport, so affects any MacOS clients which use the
2563 system-integrated TLS libraries, including email clients.
2565 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2566 using a MIME ACL for non-SMTP local injection.
2567 Report and assistance in diagnosis by Warren Baker.
2569 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2571 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2573 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2574 last id that failed so it may be referenced in subsequent ACL's.
2576 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2579 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2580 ldap library initialization, allowing self-signed CA's to be
2581 used. Also properly sets require_cert option later in code by
2582 using NULL (global ldap config) instead of ldap handle (per
2583 session). Bug diagnosis and testing by alxgomz.
2585 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2586 the src/util/ subdirectory.
2588 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2589 renamed to Transport Post Delivery Action by Jeremy Harris, as
2592 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2593 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2594 redis_servers = needs to be configured which will be used by the redis
2595 lookup. Patch from Warren Baker, of The Packet Hub.
2597 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2599 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2600 hostname or reverse DNS when processing a host list. Used suggestions
2601 from multiple comments on this bug.
2603 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2605 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2606 Missed a few lines, added it to make the runtest require no keyboard
2609 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2610 contains upper case chars. Make router use caseful_local_part.
2612 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2613 support when GnuTLS has been built with p11-kit.
2619 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2621 This, or similar/improved, will also be change PP/11 of 4.82.
2627 PP/01 Handle short writes when writing local log-files.
2628 In practice, only affects FreeBSD (8 onwards).
2629 Bugzilla 1053, with thanks to Dmitry Isaikin.
2631 NM/01 Bugzilla 949 - Documentation tweak
2633 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2636 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2638 PP/02 Implemented gsasl authenticator.
2640 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2642 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2643 `pkg-config foo` for cflags/libs.
2645 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2646 with rest of GSASL and with heimdal_gssapi.
2648 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2649 `pkg-config foo` for cflags/libs for the TLS implementation.
2651 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2652 properties get this fed in as external SSF. A number of robustness
2653 and debugging improvements to the cyrus_sasl authenticator.
2655 PP/08 cyrus_sasl server now expands the server_realm option.
2657 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2658 Patch by Jeremy Harris.
2660 PP/10 Added dbmjz lookup type.
2662 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2664 PP/12 MAIL args handles TAB as well as SP, for better interop with
2665 non-compliant senders.
2666 Analysis and variant patch by Todd Lyons.
2668 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2669 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2670 Patch from Dirk Mueller <dmueller@suse.com>
2672 PP/13 tls_peerdn now print-escaped for spool files.
2673 Observed some $tls_peerdn in wild which contained \n, which resulted
2674 in spool file corruption.
2676 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2677 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2678 or write after TLS renegotiation, which otherwise led to messages
2681 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2682 as a tracking header (ie: a signed header comes before the signature).
2683 Patch from Wolfgang Breyha.
2685 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2686 comma-sep list; embedded commas doubled.
2688 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2690 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2692 Report and patch from Dmitry Banschikov.
2694 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2695 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2696 failures. We appear to now support TLS1.1+ with Exim.
2698 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2699 lets Exim select keys and certificates based upon TLS SNI from client.
2700 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2701 before an outbound SMTP session. New log_selector, +tls_sni.
2703 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2704 NULL dereference. Report and patch from Alun Jones.
2706 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2707 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2708 Not seeing resolver debug output on NetBSD, but suspect this is a
2709 resolver implementation change.
2711 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2712 Left warnings. Added "eximon gdb" invocation mode.
2714 PP/21 Defaulting "accept_8bitmime" to true, not false.
2716 PP/22 Added -bw for inetd wait mode support.
2718 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2719 locate the relevant includes and libraries. Made this the default.
2721 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2722 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2724 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2725 This may cause build issues on older platforms.
2727 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2728 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2729 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2730 Added SNI support via GnuTLS too.
2731 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2733 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2735 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2736 Applied second patch from Janne, implementing suggestion to default
2737 multiple-strings-in-record handling to match SPF spec.
2739 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2741 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2742 read-only, out of scope).
2743 Patch from Wolfgang Breyha, report from Stuart Northfield.
2745 PP/29 Fix three issues highlighted by clang analyser static analysis.
2746 Only crash-plausible issue would require the Cambridge-specific
2747 iplookup router and a misconfiguration.
2748 Report from Marcin Mirosław.
2750 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2752 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2753 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2754 As part of this, removing so much warning spew let me fix some minor
2755 real issues in debug logging.
2757 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2758 assignment on my part. Fixed.
2760 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2761 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2762 Janne Snabb (who went above and beyond: thank you).
2764 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2765 string otherwise requires a connection and a bunch more work and it's
2766 relatively easy to get wrong. Should also expose TLS library linkage
2769 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2770 64-bit ${eval} (JH/03).
2772 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2773 GNU libc to support some of the 64-bit stuff, should not lead to
2774 conflicts. Defined before os.h is pulled in, so if a given platform
2775 needs to override this, it can.
2777 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2778 protection layer was required, which is not implemented.
2779 Bugzilla 1254, patch from Wolfgang Breyha.
2781 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2782 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2783 tls_dhparam take prime identifiers. Also unbreak combination of
2784 OpenSSL+DH_params+TLSSNI.
2786 PP/39 Disable SSLv2 by default in OpenSSL support.
2792 PP/01 Solaris build fix for Oracle's LDAP libraries.
2793 Bugzilla 1109, patch from Stephen Usher.
2795 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2797 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2800 TF/02 Fix a couple more cases where we did not log the error message
2801 when unlink() failed. See also change 4.74-TF/03.
2803 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2804 lock up or crash if it happened to be inside a call to libc when it
2805 got a SIGUSR1 from exiwhat.
2807 The SIGUSR1 handler appends the current process status to the process
2808 log which is later printed by exiwhat. It used to use the general
2809 purpose logging code to do this, but several functions it calls are
2810 not safe for signals.
2812 The new output code in the SIGUSR1 handler is specific to the process
2813 log, and simple enough that it's easy to inspect for signal safety.
2814 Removing some special cases also simplifies the general logging code.
2815 Removing the spurious timestamps from the process log simplifies
2818 TF/04 Improved ratelimit ACL condition.
2820 The /noupdate option has been deprecated in favour of /readonly which
2821 has clearer semantics. The /leaky, /strict, and /readonly update modes
2822 are mutually exclusive. The update mode is no longer included in the
2823 database key; it just determines when the database is updated. (This
2824 means that when you upgrade Exim will forget old rate measurements.)
2826 Exim now checks that the per_* options are used with an update mode that
2827 makes sense for the current ACL. For example, when Exim is processing a
2828 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2829 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2830 must specify per_mail/readonly. If you omit the update mode it defaults to
2831 /leaky where that makes sense (as before) or /readonly where required.
2833 The /noupdate option is now undocumented but still supported for
2834 backwards compatibility. It is equivalent to /readonly except that in
2835 ACLs where /readonly is required you may specify /leaky/noupdate or
2836 /strict/noupdate which are treated the same as /readonly.
2838 A useful new feature is the /count= option. This is a generalization
2839 of the per_byte option, so that you can measure the throughput of other
2840 aggregate values. For example, the per_byte option is now equivalent
2841 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2843 The per_rcpt option has been generalized using the /count= mechanism
2844 (though it's more complicated than the per_byte equivalence). When it is
2845 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2846 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2847 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2848 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2849 Note that using per_rcpt with a non-readonly update mode in more than
2850 one ACL will cause the recipients to be double-counted. (The per_mail
2851 and per_byte options don't have this problem.)
2853 The handling of very low rates has changed slightly. If the computed rate
2854 is less than the event's count (usually one) then this event is the first
2855 after a long gap. In this case the rate is set to the same as this event's
2856 count, so that the first message of a spam run is counted properly.
2858 The major new feature is a mechanism for counting the rate of unique
2859 events. The new per_addr option counts the number of different
2860 recipients that someone has sent messages to in the last time period. It
2861 behaves like per_rcpt if all the recipient addresses are different, but
2862 duplicate recipient addresses do not increase the measured rate. Like
2863 the /count= option this is a general mechanism, so the per_addr option
2864 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2865 example, measure the rate that a client uses different sender addresses
2866 with the options per_mail/unique=$sender_address. There are further
2867 details in the main documentation.
2869 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2871 TF/06 Removed a few PCRE remnants.
2873 TF/07 Automatically extract Exim's version number from tags in the git
2874 repository when doing development or release builds.
2876 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2877 Bugzilla 879. Patch from Paul Fisher.
2879 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2880 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2883 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2885 PP/05 Variable $av_failed, true if the AV scanner deferred.
2886 Bugzilla 1078. Patch from John Horne.
2888 PP/06 Stop make process more reliably on build failure.
2889 Bugzilla 1087. Patch from Heiko Schlittermann.
2891 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2892 Bugzilla 1089. Patch from Heiko Schlittermann.
2894 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2895 Bugzilla 1131. Patch from Holger Weiß.
2897 PP/09 Handle IPv6 addresses with SPF.
2898 Bugzilla 860. Patch from Wolfgang Breyha.
2900 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2902 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2905 PP/11 match_* no longer expand right-hand-side by default.
2906 New compile-time build option, EXPAND_LISTMATCH_RHS.
2907 New expansion conditions, "inlist", "inlisti".
2909 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2911 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2913 PP/14 fix log_write() format string regression from TF/03.
2914 Bugzilla 1152. Patch from Dmitry Isaikin.
2920 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2922 PP/02 Harmonised TLS library version reporting; only show if debugging.
2923 Layout now matches that introduced for other libraries in 4.74 PP/03.
2925 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2927 PP/04 New "dns_use_edns0" global option.
2929 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2932 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2933 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2935 TK/01 Updated PolarSSL code to 0.14.2.
2936 Bugzilla 1097. Patch from Andreas Metzler.
2938 PP/07 Catch divide-by-zero in ${eval:...}.
2939 Fixes bugzilla 1102.
2941 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2944 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2945 format-string attack -- SECURITY: remote arbitrary code execution.
2947 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2948 time unintentionally subject to list matching rules, letting the header
2949 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2950 arbitrary string expansion). This allowed for information disclosure.
2952 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2953 INT_MIN/-1 -- value coerced to INT_MAX.
2959 NM/01 Workaround for PCRE version dependency in version reporting
2962 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2963 This fixes portability to compilers other than gcc, notably
2964 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2966 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2967 makefiles for portability to HP-UX and POSIX correctness.
2969 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2970 Also via indented variable definition in the Makefile.
2971 (Debugging by Oliver Heesakkers).
2973 PP/02 Restore caching of spamd results with expanded spamd_address.
2974 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2976 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2977 Improves build reliability. Fix from: Frank Elsner
2979 NM/02 Fix wide character breakage in the rfc2047 coding
2980 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2982 NM/03 Allow underscore in dnslist lookups
2983 Fixes bug 1026. Patch from Graeme Fowler
2985 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2986 Code patches from Adam Ciarcinski of NetBSD.
2988 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2991 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2992 is logged, to avoid truncation. Patch from John Horne.
2994 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2995 Patch from Jakob Hirsch.
2997 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2998 SQL string expansion failure details.
2999 Patch from Andrey Oktyabrski.
3001 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3002 Patch from Simon Arlott.
3004 PP/09 New lookups functionality failed to compile on old gcc which rejects
3005 extern declarations in function scope.
3006 Patch from Oliver Fleischmann
3008 PP/10 Use sig_atomic_t for flags set from signal handlers.
3009 Check getgroups() return and improve debugging.
3010 Fixed developed for diagnosis in bug 927 (which turned out to be
3013 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3014 Patch from Mark Zealey.
3016 PP/12 Bugzilla 1056: Improved spamd server selection.
3017 Patch from Mark Zealey.
3019 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3020 Based on patch from Heiko Schlittermann.
3022 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3023 Patch from Uwe Doering, sign-off by Michael Haardt.
3025 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3026 variable declaration deep within a block. Bug and patch from
3029 PP/15 lookups-Makefile IRIX compatibility coercion.
3031 PP/16 Make DISABLE_DKIM build knob functional.
3033 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3034 Patch by Simon Arlott
3036 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3037 variable argument macros. Our copy now differs from upstream.
3043 TF/01 Failure to get a lock on a hints database can have serious
3044 consequences so log it to the panic log.
3046 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3047 controlled using the smtp_confirmation log selector.
3049 TF/03 Include the error message when we fail to unlink a spool file.
3051 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3052 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3053 for maintaining out-of-tree patches for some time.
3055 PP/01 Bugzilla 139: Documentation and portability issues.
3056 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3057 Handle per-OS dynamic-module compilation flags.
3059 PP/02 Let /dev/null have normal permissions.
3060 The 4.73 fixes were a little too stringent and complained about the
3061 permissions on /dev/null. Exempt it from some checks.
3062 Reported by Andreas M. Kirchwitz.
3064 PP/03 Report version information for many libraries, including
3065 Exim version information for dynamically loaded libraries. Created
3066 version.h, now support a version extension string for distributors
3067 who patch heavily. Dynamic module ABI change.
3069 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3070 privilege escalation vulnerability whereby the Exim run-time user
3071 can cause root to append content of the attacker's choosing to
3074 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3077 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3078 If dropping privileges for untrusted macros, we disabled normal logging
3079 on the basis that it would fail; for the Exim run-time user, this is not
3080 the case, and it resulted in successful deliveries going unlogged.
3081 Fixed. Reported by Andreas Metzler.
3087 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3088 only prepend for the Resent-* case. Fixes regression introduced in
3089 Exim 4.70 by NM/22 for Bugzilla 607.
3091 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3092 increasing numbers of administrators be bitten by this.
3094 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3096 PP/03 Bugzilla 994: added openssl_options main configuration option.
3098 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3100 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3102 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3104 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3105 without a peer certificate, leading to a segfault because of an
3106 assumption that peers always have certificates. Be a little more
3107 paranoid. Problem reported by Martin Tscholak.
3109 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3110 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3111 NB: ClamAV planning to remove STREAM in "middle of 2010".
3112 CL also introduces -bmalware, various -d+acl logging additions and
3113 more caution in buffer sizes.
3115 PP/09 Implemented reverse_ip expansion operator.
3117 PP/10 Bugzilla 937: provide a "debug" ACL control.
3119 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3121 PP/12 Bugzilla 973: Implement --version.
3123 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3125 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3127 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3129 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3130 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3131 ignore trailing whitespace.
3133 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3135 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3138 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3139 Notification from Dr Andrew Aitchison.
3141 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3142 ExtendedDetectionInfo response format.
3143 Notification from John Horne.
3145 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3148 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3149 XSL and documented dependency on system catalogs, with examples of how
3152 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3155 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3156 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3157 configuration file which is writeable by the Exim user or group.
3159 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3160 of configuration files to cover files specified with the -C option if
3161 they are going to be used with root privileges, not just the default
3164 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3165 option (effectively making it always true).
3167 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3168 files to be used while preserving root privileges.
3170 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3171 that rogue child processes cannot use them.
3173 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3174 run-time user, instead of root.
3176 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3177 Exim run-time user without dropping privileges.
3179 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3180 result string, instead of calling string_vformat() twice with the same
3183 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3184 for other users. Others should always drop root privileges if they use
3185 -C on the command line, even for a whitelisted configure file.
3187 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3189 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3195 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3196 $data_path, and $header_path variables; fixed documentation bugs and
3199 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3200 exipick to access non-standard spools, including the "frozen" queue
3203 NM/01 Bugzilla 965: Support mysql stored procedures.
3204 Patch from Alain Williams
3206 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3208 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3209 Patch from Andreas Metzler
3211 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3212 Patch from Kirill Miazine
3214 NM/05 Bugzilla 671: Added umask to procmail example.
3216 JJ/03 installed exipick 20100323.0, fixing doc bug
3218 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3219 directory. Notification and patch from Dan Rosenberg.
3221 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3223 TK/02 Improve log output when DKIM signing operation fails.
3225 MH/01 Treat the transport option dkim_domain as a colon separated
3226 list, not as a single string, and sign the message with each element,
3227 omitting multiple occurences of the same signer.
3229 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3230 Bugzilla 985, 986. Patch by Simon Arlott
3232 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3233 Patch by Simon Arlott
3235 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3236 MBX locking. Notification from Dan Rosenberg.
3242 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3244 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3246 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3248 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3250 NM/04 Bugzilla 915: Flag broken perl installation during build.
3256 TK/01 Added patch by Johannes Berg that expands the main option
3257 "spamd_address" if it starts with a dollar sign.
3259 TK/02 Write list of recipients to X-Envelope-Sender header when building
3260 the mbox-format spool file for content scanning (suggested by Jakob
3263 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3264 (http://www.dcc-servers.net/) support via dccifd. Activated by
3265 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3267 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3268 by Mark Daniel Reidel <mr@df.eu>.
3270 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3271 When building exim an external PCRE library is now needed -
3272 PCRE is a system library on the majority of modern systems.
3273 See entry on PCRE_LIBS in EDITME file.
3275 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3276 conversation. Added nologin parameter to request.
3277 Patch contributed by Kirill Miazine.
3279 TF/01 Do not log submission mode rewrites if they do not change the address.
3281 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3283 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3284 log files in place. Contributed by Roberto Lima.
3286 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3288 TF/03 Bugzilla 615: When checking the local_parts router precondition
3289 after a local_part_suffix or local_part_prefix option, Exim now
3290 does not use the address's named list lookup cache, since this
3291 contains cached lookups for the whole local part.
3293 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3294 Robert Millan. Documentation is in experimental-spec.txt.
3296 TF/04 Bugzilla 668: Fix parallel build (make -j).
3298 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3300 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3301 Patch provided by Jan Srzednicki.
3303 TF/05 Leading white space used to be stripped from $spam_report which
3304 wrecked the formatting. Now it is preserved.
3306 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3307 that they are available at delivery time.
3309 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3311 TF/08 TLS error reporting now respects the incoming_interface and
3312 incoming_port log selectors.
3314 TF/09 Produce a more useful error message if an SMTP transport's hosts
3315 setting expands to an empty string.
3317 NM/06 Bugzilla 744: EXPN did not work under TLS.
3318 Patch provided by Phil Pennock.
3320 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3321 Patch provided by Richard Godbee.
3323 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3324 acl_smtp_notquit, added index entry.
3326 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3327 Patch provided by Eugene Bujak.
3329 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3330 accept(). Patch provided by Maxim Dounin.
3332 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3333 Patch provided by Phil Pennock.
3335 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3337 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3338 Patch provided by Brad "anomie" Jorsch.
3340 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3341 Patch provided by Dean Brooks.
3343 TK/05 Add native DKIM support (does not depend on external libraries).
3345 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3346 Patch provided by Graeme Fowler.
3348 NM/16 Bugzilla 851: Documentation example syntax fix.
3350 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3352 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3355 NM/19 Bugzilla 745: TLS version reporting.
3356 Patch provided by Phil Pennock.
3358 NM/20 Bugzilla 167: bool: condition support.
3359 Patch provided by Phil Pennock.
3361 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3362 clients. Patch provided by Phil Pennock.
3364 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3365 Patch provided by Brad "anomie" Jorsch.
3367 NM/23 Bugzilla 687: Fix misparses in eximstats.
3368 Patch provided by Heiko Schlittermann.
3370 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3371 Patch provided by Heiko Schlittermann.
3373 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3374 plus update to original patch.
3376 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3378 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3379 Patch provided by David Brownlee.
3381 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3383 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3385 NM/30 Bugzilla 888: TLS documentation bugfixes.
3387 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3389 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3390 Unlike the original bugzilla I have changed all shell scripts in src tree.
3392 NM/33 Bugzilla 898: Transport filter timeout fix.
3393 Patch by Todd Rinaldo.
3395 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3396 Patch by Serge Demonchaux.
3398 NM/35 Bugzilla 39: Base64 decode bug fixes.
3399 Patch by Jakob Hirsch.
3401 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3403 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3405 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3407 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3413 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3414 ALT-N's libdkim that I have put here:
3415 http://duncanthrax.net/exim-experimental/
3417 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3418 were called 'true' and it seems that C99 defines that as a reserved
3419 keyword to be used with 'bool' variable types. That means you could
3420 not include C99-style headers which use bools without triggering
3421 build errors in sieve.c.
3423 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3424 as mailq or other aliases. Changed the --help handling significantly
3425 to do whats expected. exim_usage() emits usage/help information.
3427 SC/01 Added the -bylocaldomain option to eximstats.
3429 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3431 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3433 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3439 PH/01 Another patch from the Sieve maintainer.
3441 PH/02 When an IPv6 address is converted to a string for single-key lookup
3442 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3443 dots are used instead of colons so that keys in lsearch files need not
3444 contain colons. This was done some time before quoting was made available
3445 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3446 (notated using the quote facility) so as to distinguish them from IPv4
3447 keys. This meant that lookups for IP addresses in host lists did not work
3448 for iplsearch lookups.
3450 This has been fixed by arranging for IPv6 addresses to be expressed with
3451 colons if the lookup type is iplsearch. This is not incompatible, because
3452 previously such lookups could never work.
3454 The situation is now rather anomalous, since one *can* have colons in
3455 ordinary lsearch keys. However, making the change in all cases is
3456 incompatible and would probably break a number of configurations.
3458 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3461 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3462 conversion specification without a maximum field width, thereby enabling
3463 a rogue spamd server to cause a buffer overflow. While nobody in their
3464 right mind would setup Exim to query an untrusted spamd server, an
3465 attacker that gains access to a server running spamd could potentially
3466 exploit this vulnerability to run arbitrary code as the Exim user.
3468 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3469 $primary_hostname instead of what libspf2 thinks the hosts name is.
3471 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3472 a directory entry by the name of the lookup key. Previously, if a
3473 symlink pointed to a non-existing file or a file in a directory that
3474 Exim lacked permissions to read, a lookup for a key matching that
3475 symlink would fail. Now it is enough that a matching directory entry
3476 exists, symlink or not. (Bugzilla 503.)
3478 PH/03 The body_linecount and body_zerocount variables are now exported in the
3481 PH/04 Added the $dnslist_matched variable.
3483 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3484 This means they are set thereafter only if the connection becomes
3487 PH/06 Added the client_condition to authenticators so that some can be skipped
3488 by clients under certain conditions.
3490 PH/07 The error message for a badly-placed control=no_multiline_responses left
3491 "_responses" off the end of the name.
3493 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3495 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3496 (without spaces) instead of just copying the configuration text.
3498 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3500 PH/11 Added $max_received_linelength.
3502 PH/12 Added +ignore_defer and +include_defer to host lists.
3504 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3505 way in which PCRE > 7.0 is built.
3507 PH/14 Implemented queue_only_load_latch.
3509 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3510 MAIL command. The effect was to mangle the value on 64-bit systems.
3512 PH/16 Another patch from the Sieve maintainer.
3514 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3516 PH/18 If a system quota error occurred while trying to create the file for
3517 a maildir delivery, the message "Mailbox is full" was not appended to the
3518 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3519 only to a quota excession during the actual writing of the file.
3521 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3522 characters?) which causes problems in log lines. The DN values are now
3523 passed through string_printing() before being added to log lines.
3525 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3526 and InterBase are left for another time.)
3528 PH/21 Added message_body_newlines option.
3530 PH/22 Guard against possible overflow in moan_check_errorcopy().
3532 PH/23 POSIX allows open() to be a macro; guard against that.
3534 PH/24 If the recipient of an error message contained an @ in the local part
3535 (suitably quoted, of course), incorrect values were put in $domain and
3536 $local_part during the evaluation of errors_copy.
3542 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3543 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3546 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3547 issue a MAIL command.
3549 PH/02 In an ACL statement such as
3551 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3553 if a client was not listed at all, or was listed with a value other than
3554 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3555 the condition was not true (as it should be), so access was not denied.
3556 The bug was that the ! inversion was incorrectly passed on to the second
3557 item. This has been fixed.
3559 PH/03 Added additional dnslists conditions == and =& which are different from
3560 = and & when the dns lookup returns more than one IP address.
3562 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3563 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3565 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3566 FSYNC, which compiles an option called disable_fsync that allows for
3567 bypassing fsync(). The documentation is heavily laced with warnings.
3569 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3571 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3572 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3573 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3574 including adding "make clean"; (3) Added -fPIC when compiling the test
3575 dynamically loaded module, to get rid of a warning.
3577 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3578 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3579 = 0s. The bug is otherwise harmless.
3581 PH/07 There was a bug in the dovecot authenticator such that the value of
3582 $auth1 could be overwritten, and so not correctly preserved, after a
3583 successful authentication. This usually meant that the value preserved by
3584 the server_setid option was incorrect.
3586 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3588 PH/09 Installed PCRE release 7.0.
3590 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3591 run for batched SMTP input. It is now run at the start of every message
3592 in the batch. While fixing this I discovered that the process information
3593 (output by running exiwhat) was not always getting set for -bs and -bS
3594 input. This is fixed, and it now also says "batched" for BSMTP.
3596 PH/11 Added control=no_pipelining.
3598 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3599 patch, slightly modified), and move the expansion of helo_data till after
3600 the connection is made in the smtp transport (so it can use these
3603 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3605 PH/14 Added log_selector = +pid.
3607 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3609 PH/16 Add ${if forany and ${if forall.
3611 PH/17 Added dsn_from option to vary the From: line in DSNs.
3613 PH/18 Flush SMTP output before performing a callout, unless control =
3614 no_callout_flush is set.
3616 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3617 was true (the default) a successful delivery failed to delete the retry
3618 item, thus causing premature timeout of the address. The bug is now
3621 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3623 PH/21 Long custom messages for fakedefer and fakereject are now split up
3624 into multiline responses in the same way that messages for "deny" and
3625 other ACL rejections are.
3627 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3628 with slight modification.
3630 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3631 draft, changing the syntax and factoring some duplicate code".
3633 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3634 for deliveries of the second and subsequent messages over the same SMTP
3637 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3638 ${reduce, with only minor "tidies".
3640 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3642 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3643 expansion side effects.
3645 PH/27 When a message times out after an over-quota error from an Exim-imposed
3646 quota, the bounce message says "mailbox is full". This message was not
3647 being given when it was a system quota that was exceeded. It now should
3650 MH/03 Made $recipients available in local_scan(). local_scan() already has
3651 better access to the recipient list through recipients_list[], but
3652 $recipients can be useful in postmaster-provided expansion strings.
3654 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3655 in the case of a MAIL command with additional options following the
3656 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3657 were accidentally chopped off.
3659 PH/29 SMTP synchronization checks are implemented when a command is read -
3660 there is a check that no more input is waiting when there shouldn't be
3661 any. However, for some commands, a delay in an ACL can mean that it is
3662 some time before the response is written. In this time, more input might
3663 arrive, invalidly. So now there are extra checks after an ACL has run for
3664 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3665 pipelining has not been advertised.
3667 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3669 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3670 returned in the SMTP response when smtp_return_error_details was false.
3671 This has been fixed.
3673 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3674 instead of the C I/O that was originally supplied, because problems were
3675 reported on Solaris.
3677 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3678 Exim which did not show up earlier: it was assuming that a call to
3679 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3680 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3681 was a macro that became an assignment, so it seemed to work. This has
3682 changed to a proper function call with a void return, hence the compile
3683 error. Exim's code has been fixed.
3685 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3688 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3690 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3692 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3693 because it turns out to be too big or there is a timeout) there is no
3694 "Completed" line in the log. When some messages of this type were
3695 selected by exigrep, they were listed as "not completed". Others were
3696 picked up by some special patterns. I have improved the selection
3697 criteria to be more general.
3699 PH/38 The host_find_failed option in the manualroute router can now be set
3700 to "ignore", to completely ignore a host whose IP address cannot be
3701 found. If all hosts are ignored, the behaviour is controlled by the new
3702 host_all_ignored option.
3704 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3705 homing or because of multiple MX records with /mx) generated more than
3706 one IP address, and the following item turned out to be the local host,
3707 all the secondary addresses of the first item were incorrectly removed
3708 from the list, along with the local host and any following hosts (which
3709 is what is supposed to happen).
3711 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3712 whoever called Exim into the -H file. In the case of the daemon it was
3713 behaving confusingly. When first started, it used values for whoever
3714 started the daemon, but after a SIGHUP it used the Exim user (because it
3715 calls itself on a restart). I have changed the code so that it now always
3718 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3719 message are rejected with the same error (e.g. no authentication or bad
3720 sender address), and a DATA command is nevertheless sent (as can happen
3721 with PIPELINING or a stupid MUA), the error message that was given to the
3722 RCPT commands is included in the rejection of the DATA command. This is
3723 intended to be helpful for MUAs that show only the final error to their
3726 PH/42 Another patch from the Sieve maintainer.
3728 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3729 Eximstats - Fixed some broken HTML links and added missing column headers
3731 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3732 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3734 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3736 PH/43 Yet another patch from the Sieve maintainer.
3738 PH/44 I found a way to check for a TCP/IP connection going away before sending
3739 the response to the final '.' that terminates a message, but only in the
3740 case where the client has not sent further data following the '.'
3741 (unfortunately, this is allowed). However, in many cases there won't be
3742 any further data because there won't be any more messages to send. A call
3743 to select() can be used: if it shows that the input is "ready", there is
3744 either input waiting, or the socket has been closed. An attempt to read
3745 the next input character can distinguish the two cases. Previously, Exim
3746 would have sent an OK response which the client would never have see.
3747 This could lead to message repetition. This fix should cure that, at
3748 least in a lot of common cases.
3750 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3751 advertised in response to EHLO.
3757 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3758 fixed by 4.65/MH/01 (is this a record?) are fixed:
3760 (i) An empty string was always treated as zero by the numeric comparison
3761 operators. This behaviour has been restored.
3763 (ii) It is documented that the numeric comparison operators always treat
3764 their arguments as decimal numbers. This was broken in that numbers
3765 starting with 0 were being interpreted as octal.
3767 While fixing these problems I realized that there was another issue that
3768 hadn't been noticed. Values of message_size_limit (both the global option
3769 and the transport option) were treated as octal if they started with 0.
3770 The documentation was vague. These values are now always treated as
3771 decimal, and I will make that clear in the documentation.
3777 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3778 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3781 MH/01 Don't check that the operands of numeric comparison operators are
3782 integers when their expansion is in "skipping" mode (fixes bug
3783 introduced by 4.64-PH/07).
3785 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3786 child addresses, Exim now panics and dies. Previously, because the count
3787 is held in a short int, deliveries were likely to be lost. As such a
3788 large number of recipients for a single message is ridiculous
3789 (performance will be very, very poor), I have chosen to impose a limit
3790 rather than extend the field.
3796 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3797 leftover -K file (the existence of which was triggered by #402).
3798 While we were at it, introduced process PID as part of the -K
3799 filename. This should rule out race conditions when creating
3802 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3803 processing considerably. Previous code took too long for large mails,
3804 triggering a timeout which in turn triggers #401.
3806 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3807 in the DK code in transports.c. sendfile() is not really portable,
3808 hence the _LINUX specificness.
3810 TF/01 In the add_headers option to the mail command in an Exim filter,
3811 there was a bug that Exim would claim a syntax error in any
3812 header after the first one which had an odd number of characters
3815 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3816 callout verification, Exim cached a "reject" for the entire domain. This
3817 is correct for most verifications, but it is not correct for a recipient
3818 verification with use_sender or use_postmaster set, because in that case
3819 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3820 case of MAIL FROM:<> rejection from other early rejections (e.g.
3821 rejection of HELO). When verifying a recipient using a non-null MAIL
3822 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3823 Whatever the result of the callout, the value of the domain cache is
3824 left unchanged (for any other kind of callout, getting as far as trying
3825 RCPT means that the domain itself is ok).
3827 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3830 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3831 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3832 session, the daemon ignores ECONNECT errors and logs others; it now
3833 ignores EPIPE as well.
3835 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3836 (quoted-printable decoding).
3838 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3839 later the small subsequent patch to fix an introduced bug.
3841 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3843 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3845 PH/08 An error is now given if message_size_limit is specified negative.
3847 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3848 to be given (somewhat) arbitrary names.
3850 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3853 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3854 miscellaneous code fixes
3856 PH/10 Added the log_reject_target ACL modifier to specify where to log
3859 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3860 hostname. This is wrong, because it relates to the incoming message (and
3861 probably the interface on which it is arriving) and not to the outgoing
3862 callout (which could be using a different interface). This has been
3863 changed to use the value of the helo_data option from the smtp transport
3864 instead - this is what is used when a message is actually being sent. If
3865 there is no remote transport (possible with a router that sets up host
3866 addresses), $smtp_active_hostname is used.
3868 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3869 tweaks were necessary in order to get it to work (see also 21 below):
3870 (a) The code assumed that strncpy() returns a negative number on buffer
3871 overflow, which isn't the case. Replaced with Exim's string_format()
3873 (b) There were several signed/unsigned issues. I just did the minimum
3874 hacking in of casts. There is scope for a larger refactoring.
3875 (c) The code used strcasecmp() which is not a standard C function.
3876 Replaced with Exim's strcmpic() function.
3877 (d) The code set only $1; it now sets $auth1 as well.
3878 (e) A simple test gave the error "authentication client didn't specify
3879 service in request". It would seem that Dovecot has changed its
3880 interface. Fortunately there's a specification; I followed it and
3881 changed what the client sends and it appears to be working now.
3883 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3886 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3887 address A is aliased to B and C, where B exists and C does not. Without
3888 -v the output is "A verified" because verification stops after a
3889 successful redirection if more than one address is generated. However,
3890 with -v the child addresses are also verified. Exim was outputting "A
3891 failed to verify" and then showing the successful verification for C,
3892 with its parentage. It now outputs "B failed to verify", showing B's
3893 parentage before showing the successful verification of C.
3895 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3896 look up a TXT record in a specific list after matching in a combined
3899 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3900 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3901 they consult the DNS. I had assumed they would set it the way they
3902 wanted; and indeed my experiments on Linux seem to show that in some
3903 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3904 To be on the safe side, however, I have now made the interface to
3905 host_find_byname() similar to host_find_bydns(), with an argument
3906 containing the DNS resolver options. The host_find_byname() function now
3907 sets these options at its start, just as host_find_bydns() does. The smtp
3908 transport options dns_qualify_single and dns_search_parents are passed to
3909 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3910 of host_find_byname() use the default settings of RES_DEFNAMES
3911 (qualify_single) but not RES_DNSRCH (search_parents).
3913 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3914 spool_read_header() do less string testing, by means of a preliminary
3915 switch on the second character of optional "-foo" lines. (This is
3916 overdue, caused by the large number of possibilities that now exist.
3917 Originally there were few.) While I was there, I also converted the
3918 str(n)cmp tests so they don't re-test the leading "-" and the first
3919 character, in the hope this might squeeze out yet more improvement.
3921 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3922 flag allowing group syntax was set by the header_syntax check but not
3923 turned off, possible causing trouble later; (2) The flag was not being
3924 set at all for the header_verify test, causing "group"-style headers to
3925 be rejected. I have now set it in this case, and also caused header_
3926 verify to ignore an empty address taken from a group. While doing this, I
3927 came across some other cases where the code for allowing group syntax
3928 while scanning a header line wasn't quite right (mostly, not resetting
3929 the flag correctly in the right place). These bugs could have caused
3930 trouble for malformed header lines. I hope it is now all correct.
3932 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3933 with the "reply" argument non-NULL. The code, however (which originally
3934 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3935 but it didn't always do it. This confused somebody who was copying the
3936 code for some other use. I have removed all the tests.
3938 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3939 feature that was used to support insecure browsers during the U.S. crypto
3940 embargo. It requires special client support, and Exim is probably the
3941 only MTA that supported it -- and would never use it because real RSA is
3942 always available. This code has been removed, because it had the bad
3943 effect of slowing Exim down by computing (never used) parameters for the
3944 RSA_EXPORT functionality.
3946 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3947 authenticator to fail if there's a tab character in the incoming data
3948 (there should never be unless someone is messing about, as it's supposed
3949 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3950 if the connection is using TLS or if the remote IP is the same as the
3951 local IP, and the "valid-client-cert option" if a client certificate has
3954 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3955 authenticators. This can be used for authorization after authentication
3956 succeeds. (In the case of plaintext, it servers for both authentication
3959 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3960 if any retry times were supplied.
3962 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3963 connection, where there is no client IP address to check. In this
3964 situation, the verify now always succeeds.
3966 PH/25 Applied John Jetmore's -Mset patch.
3968 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3970 PH/27 In a string expansion for a processed (not raw) header when multiple
3971 headers of the same name were present, leading whitespace was being
3972 removed from all of them, but trailing whitespace was being removed only
3973 from the last one. Now trailing whitespace is removed from each header
3974 before concatenation. Completely empty headers in a concatenation (as
3975 before) are ignored.
3977 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3978 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3980 PH/29 [Removed. This was a change that I later backed out, and forgot to
3981 correct the ChangeLog entry (that I had efficiently created) before
3982 committing the later change.]
3984 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3985 address errors (4xx response to RCPT) over the same connection as other
3986 messages routed to the same hosts. Such deliveries are always "forced",
3987 so retry times are not inspected. This resulted in far too many retries
3988 for the affected addresses. The effect occurred only when there were more
3989 hosts than the hosts_max_try setting in the smtp transport when it had
3990 the 4xx errors. Those hosts that it had tried were not added to the list
3991 of hosts for which the message was waiting, so if all were tried, there
3992 was no problem. Two fixes have been applied:
3994 (i) If there are any address or message errors in an SMTP delivery, none
3995 of the hosts (tried or untried) are now added to the list of hosts
3996 for which the message is waiting, so the message should not be a
3997 candidate for sending over the same connection that was used for a
3998 successful delivery of some other message. This seems entirely
3999 reasonable: after all the message is NOT "waiting for some host".
4000 This is so "obvious" that I'm not sure why it wasn't done
4001 previously. Hope I haven't missed anything, but it can't do any
4002 harm, as the worst effect is to miss an optimization.
4004 (ii) If, despite (i), such a delivery is accidentally attempted, the
4005 routing retry time is respected, so at least it doesn't keep
4006 hammering the server.
4008 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4009 in ${readsocket because some servers need this prod.
4011 PH/32 Added some extra debug output when updating a wait-xxx database.
4013 PH/33 The hint "could be header name not terminated by colon", which has been
4014 given for certain expansion errors for a long time, was not being given
4015 for the ${if def:h_colon_omitted{... case.
4017 PH/34 The spec says: "With one important exception, whenever a domain list is
4018 being scanned, $domain contains the subject domain." There was at least
4019 one case where this was not true.
4021 PH/35 The error "getsockname() failed: connection reset by peer" was being
4022 written to the panic log as well as the main log, but it isn't really
4023 panic-worthy as it just means the connection died rather early on. I have
4024 removed the panic log writing for the ECONNRESET error when getsockname()
4027 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4028 runs only) independently of the message's sender address. This meant
4029 that, if the 4xx error was in fact related to the sender, a different
4030 message to the same recipient with a different sender could confuse
4031 things. In particular, this can happen when sending to a greylisting
4032 server, but other circumstances could also provoke similar problems.
4033 I have changed the default so that the retry time for these errors is now
4034 based a combination of the sender and recipient addresses. This change
4035 can be overridden by setting address_retry_include_sender=false in the
4038 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4039 remote server are returned as part of bounce messages. This was not
4040 happening for LMTP over a pipe (the lmtp transport), but now it is the
4041 same for both kinds of LMTP.
4043 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4044 in header lines that were in fact CNAMEs. This is no longer the case.
4046 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4047 and queue runs started by the daemon processed all messages. This has
4048 been fixed so that -R and -S can now usefully be given with -q<time>.
4050 PH/40 Import PCRE release 6.7 (fixes some bugs).
4052 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4054 PH/42 Give an error if -q is specified more than once.
4056 PH/43 Renamed the variables $interface_address and $interface_port as
4057 $received_ip_address and $received_port, to make it clear that these
4058 values apply to message reception, and not to the outgoing interface when
4059 a message is delivered. (The old names remain recognized, of course.)
4061 PH/44 There was no timeout on the connect() call when using a Unix domain
4062 socket in the ${readsocket expansion. There now is.
4064 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4065 be meaningful with "accept".
4067 SC/01 Eximstats V1.43
4068 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4070 SC/02 Eximstats V1.44
4071 Use a glob alias rather than an array ref in the generated
4072 parser. This improves both readability and performance.
4074 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4075 Collect SpamAssassin and rejection statistics.
4076 Don't display local sender or destination tables unless
4077 there is data to show.
4078 Added average volumes into the top table text output.
4080 SC/04 Eximstats V1.46
4081 Collect data on the number of addresses (recipients)
4082 as well as the number of messages.
4084 SC/05 Eximstats V1.47
4085 Added 'Message too big' to the list of mail rejection
4086 reasons (thanks to Marco Gaiarin).
4088 SC/06 Eximstats V1.48
4089 Mainlog lines which have GMT offsets and are too short to
4090 have a flag are now skipped.
4092 SC/07 Eximstats V1.49 (Alain Williams)
4093 Added the -emptyok flag.
4095 SC/08 Eximstats V1.50
4096 Fixes for obtaining the IP address from reject messages.
4098 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4099 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4100 whitespace changes from 4.64-PH/27
4102 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4105 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4106 are found, allow negative numbers in numeric criteria)
4108 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4110 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4111 to match changes made in 4.64-PH/43
4113 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4115 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4116 "domain" is actually an IP address, and force a failure. This locks out
4117 those revolvers/nameservers that support "A-for-A" lookups, in
4118 contravention of the specifications.
4120 PH/48 When a host name was looked up from an IP address, and the subsequent
4121 forward lookup of the name timed out, the host name was left in
4122 $sender_host_name, contrary to the specification.
4124 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4125 restricted to single-key lookups, Exim was not diagnosing an error if
4126 * or *@ was used with a query-style lookup.
4128 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4130 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4131 long ago, but noone interested enough thought of it. Let's just say that
4132 the "1.1" means that there are some new functions that weren't there at
4133 some point in the past.
4135 PH/51 Error processing for expansion failure of helo_data from an smtp
4136 transport during callout processing was broken.
4138 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4139 tested/used via the -bh/-bhc/-bs options.
4141 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4142 bug, fixed in subsequent PCRE releases).
4144 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4145 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4147 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4153 SC/01 Use a glob alias rather than an array ref in eximstats generated
4154 parser. This improves both readability and performance.
4156 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4157 Don't display local sender or destination tables in eximstats unless
4158 there is data to show.
4159 Added average volumes into the eximstats top table text output.
4161 SC/03 Collect data on the number of addresses (recipients) as well
4162 as the number of messages in eximstats.
4164 TF/01 Correct an error in the documentation for the redirect router. Exim
4165 does (usually) call initgroups() when daemonizing.
4167 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4168 with consistent privilege compared to when running as a daemon.
4170 TF/03 Note in the spec that $authenticated_id is not set for local
4171 submissions from trusted users.
4173 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4174 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4176 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4177 by adding some example configuration directives to the default
4178 configuration file. A little bit of work is required to uncomment the
4179 directives and define how usernames and passwords are checked, but
4180 there is now a framework to start from.
4182 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4183 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4184 without this. I don't know how relevant this is to other LDAP libraries.
4186 PH/02 Add the verb name to the "unknown ACL verb" error.
4188 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4190 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4192 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4193 directory not expanded when it should be if an expanded home directory
4194 was set for the address (which is overridden by the transport).
4196 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4199 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4200 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4201 because it is too late at that time, and has no effect.
4203 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4204 security issue with \' (bugzilla #107). I could not use the
4205 PQescapeStringConn() function, because it needs a PGconn value as one of
4208 PH/08 When testing addresses using -bt, indicate those final addresses that
4209 are duplicates that would not cause an additional delivery. At least one
4210 person was confused, thinking that -bt output corresponded to deliveries.
4211 (Suppressing duplicates isn't a good idea as you lose the information
4212 about possibly different redirections that led to the duplicates.)
4214 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4215 systems where poll() doesn't work, in particular OS X.
4217 PH/10 Added more information to debugging output for retry time not reached.
4219 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4220 operations in malware.c.
4222 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4225 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4226 syslog_duplication set false, log lines that would normally be written
4227 both the the main log and to the reject log were not written to syslog at
4230 PH/14 In the default configuration, change the use of "message" in ACL warn
4231 statements to "add_header".
4233 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4234 not followed by a command (e.g. "seen endif").
4236 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4237 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4240 PH/17 Added extra conditions to the default value of delay_warning_condition
4244 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4245 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4246 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4249 The Auto-Submitted: and various List- headers are standardised, whereas I
4250 don't think Precedence: ever was.
4252 PH/18 Refactored debugging code in route_finduser() to show more information,
4253 in particular, the error code if getpwnam() issues one.
4255 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4256 This is apparently needed in addition to the PH/07 change above to avoid
4257 any possible encoding problems.
4259 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4260 but not after initializing Perl.
4262 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4263 output them only if debugging. By default they are written stderr,
4264 apparently, which is not desirable.
4266 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4269 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4272 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4274 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4275 authenticated or an ident call has been made. Suppress the default
4276 values for $authenticated_id and $authenticated_sender (but permit -oMai
4277 and -oMas) when testing with -bh.
4279 PH/24 Re-jigged the order of the tests in the default configuration so that the
4280 tests for valid domains and recipients precede the DNS black list and CSA
4281 tests, on the grounds that those ones are more expensive.
4283 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4284 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4285 This bug exists in every version of Exim that I still have, right back to
4288 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4289 However, an attempt to turn on case-sensitivity in a regex key by
4290 including (?-i) didn't work because the subject string was already
4291 lowercased, and the effects were non-intuitive. It turns out that a
4292 one-line patch can be used to allow (?-i) to work as expected.
4298 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4299 other effects) broke the use of negated acl sub-conditions.
4301 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4304 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4305 "Deny" causes Exim to reject the incoming connection with a 554 error.
4306 Unfortunately, if there is a major crisis, such as a disk failure,
4307 tcp-wrappers gives "deny", whereas what one would like would be some
4308 kind of temporary error. A kludge has been added to help with this.
4309 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4310 554 error is used if errno is still zero or contains ENOENT (which occurs
4311 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4314 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4316 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4317 errors. Otherwise a message that provokes a temporary error (when other
4318 messages do not) can cause a whole host to time out.
4320 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4321 addresses were routed directly to files or pipes from a redirect router.
4322 File deliveries just didn't batch; pipe deliveries might have suffered
4325 PH/06 A failure to get a lock for a hints database would erroneously always say
4326 "Failed to get write lock", even when it was really a read lock.
4328 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4329 of 0600. This has been changed to use the value of the lockfile_mode
4330 option (which defaults to 0600).
4332 PH/08 Applied small patch from the Sieve maintainer.
4334 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4335 folder from quota calculations, a direct delivery into this folder messed
4336 up the contents of the maildirsize file. This was because the regex was
4337 used only to exclude .Trash (or whatever) when the size of the mailbox
4338 was calculated. There was no check that a delivery was happening into an
4339 excluded directory. This bug has been fixed by ignoring all quota
4340 processing for deliveries into excluded directories.
4342 PH/10 Added the maildirfolder_create_regex option to appendfile.
4348 PH/01 The code for finding all the local interface addresses on a FreeBSD
4349 system running IPv6 was broken. This may well have applied to all BSD
4350 systems, as well as to others that have similar system calls. The broken
4351 code found IPv4 interfaces correctly, but gave incorrect values for the
4352 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4353 that it would not match correctly against @[] and not recognize the IPv6
4356 PH/02 The ipliteral router was not recognizing addresses of the form user@
4357 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4359 PH/03 Added disable_ipv6.
4361 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4362 decoded form, because it is most often used to construct To: headers
4363 lines in autoreplies, and the decoded form may well be syntactically
4364 invalid. However, $reply_address has leading white space removed, and all
4365 newlines turned into spaces so that the autoreply transport does not
4368 PH/05 If group was specified without a user on a router, and no group or user
4369 was specified on a transport, the group from the router was ignored.
4371 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4372 for visible compile-time settings that can be used to change these
4373 numbers, for those that want even more. Backwards compatibility with old
4374 spool files has been maintained. However, going back to a previous Exim
4375 release will lost any variables that are in spool files.
4377 PH/07 Two small changes when running in the test harness: increase delay when
4378 passing a TCP/IP connection to a new process, in case the original
4379 process has to generate a bounce, and remove special handling of
4380 127.0.0.2 (sic), which is no longer necessary.
4382 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4383 be the same on different OS.
4385 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4388 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4389 whether --show-vars was specified or not
4391 JJ/02 exipick: Added support for new ACL variable spool format introduced
4394 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4395 syntactically invalid From: or Reply-to: line, and a filter used this to
4396 generate an autoreply, and therefore failed to obtain an address for the
4397 autoreply, Exim could try to deliver to a non-existent relative file
4398 name, causing unrelated and misleading errors. What now happens is that
4399 it logs this as a hard delivery error, but does not attempt to create a
4402 PH/11 The exinext utility has a -C option for testing purposes, but although
4403 the given file was scanned by exinext itself; it wasn't being passed on
4404 when Exim was called.
4406 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4407 an end-of-file indication when reading a command response.
4409 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4410 compiled. In many other places in Exim, IPv6 addresses are always
4411 recognized, so I have changed this. It also means that IPv4 domain
4412 literals of the form [IPV4:n.n.n.n] are now always recognized.
4414 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4415 used if the router is not running as root, for example, when verifying at
4416 ACL time, or when using -bh. The debugging output from this situation was
4417 non-existent - all you got was a failure to exec. I have made two
4420 (a) Failures to set uid/gid, the current directory, or a process leader
4421 in a subprocess such as that created by queryprogram now generate
4422 suitable debugging output when -d is set.
4424 (b) The queryprogram router detects when it is not running as root,
4425 outputs suitable debugging information if -d is set, and then runs
4426 the subprocess without attempting to change uid/gid.
4428 PH/15 Minor change to Makefile for building test_host (undocumented testing
4431 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4432 additional section of a DNS packet that returns MX or SRV records.
4433 Instead, it always explicitly searches for A/AAAA records. This avoids
4434 major problems that occur when a DNS server includes only records of one
4435 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4436 fixed another bug: if SRV records were looked up and the corresponding
4437 address records were *not* found in the additional section, the port
4438 values from the SRV records were lost.
4440 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4441 using the correct key (the original address) when searching the retry
4442 rules in order to find which one to use for generating the retry hint.
4444 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4445 adding the default one. Similarly, if it contains a Reply-To: header, the
4446 errors_reply_to option, if set, is not used.
4448 PH/19 When calculating a retry time, Exim used to measure the "time since
4449 failure" by looking at the "first failed" field in the retry record. Now
4450 it does not use this if it is later than than the arrival time of the
4451 message. Instead it uses the arrival time. This makes for better
4452 behaviour in cases where some deliveries succeed, thus re-setting the
4453 "first failed" field. An example is a quota failure for a huge message
4454 when small messages continue to be delivered. Without this change, the
4455 "time since failure" will always be short, possible causing more frequent
4456 delivery attempts for the huge message than are intended.
4457 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4459 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4460 $1, $2, $3) because the numerical variables can be reset during some
4461 expansion items (e.g. "match"), thereby losing the authentication data.
4463 PH/21 Make -bV show the size of off_t variables so that the test suite can
4464 decide whether to run tests for quotas > 2G.
4466 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4467 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4468 filecount value is greater than 2G or if a quota value is greater than 2G
4469 on a system where the size of off_t is not greater than 4, a panic error
4472 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4473 never match. The debug and -bh output now contains an explicit error
4474 message indicating a malformed IPv4 address or mask.
4476 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4477 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4478 PH/23 above applies.
4480 PH/25 Do not write to syslog when running in the test harness. The only
4481 occasion when this arises is a failure to open the main or panic logs
4482 (for which there is an explicit test).
4484 PH/26 Added the /no_tell option to "control=freeze".
4486 PH/27 If a host name lookup failed very early in a connection, for example, if
4487 the IP address matched host_lookup and the reverse lookup yielded a name
4488 that did not have a forward lookup, an error message of the form "no IP
4489 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4490 could be logged. Now it outputs the IP address instead of "NULL".
4492 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4493 allows the sender and the authenticated sender to be set when
4494 submitting a message from within Exim. Since child_open_exim() is
4495 documented for local_scan(), the new function should be too.
4497 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4498 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4499 results in an empty string is now treated as unset.
4501 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4503 PH/31 Added sender_verify_fail logging option.
4505 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4506 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4507 tidied the source and removed it altogether.
4509 PH/33 When a queue run was abandoned because the load average was too high, a
4510 log line was always written; now it is written only if the queue_run log
4511 selector is set. In addition, the log line for abandonment now contains
4512 information about the queue run such as the pid. This is always present
4513 in "start" and "stop" lines but was omitted from the "abandon" line.
4515 PH/34 Omit spaces between a header name and the colon in the error message that
4516 is given when verify = headers_syntax fails (if there are lots of them,
4517 the message gets confusing).
4519 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4520 names, as there are now some PTR records that contain slashes. This check
4521 is only to protect against broken name servers that fall over on strange
4522 characters, so the fact that it applies to all lookups doesn't matter.
4524 PH/36 Now that the new test suite is complete, we can remove some of the
4525 special code in Exim that was needed for the old test suite. For example,
4526 sorting DNS records because real resolvers return them in an arbitrary
4527 order. The new test suite's fake resolver always returns records in the
4530 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4531 bounces) except when queue_only is set, to avoid logging races between
4532 the different processes.
4534 PH/38 Panic-die if .include specifies a non-absolute path.
4536 PH/39 A tweak to the "H" retry rule from its user.
4538 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4539 a label. They prevented compilation on older perls.
4541 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4542 a warning to be raised on newish perls.
4544 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4545 on queue. Changes to match documented behaviour of showing count of
4546 messages matching specified criteria.
4548 PH/40 Changed the default ident timeout from 30s to 5s.
4550 PH/41 Added support for the use of login_cap features, on those BSD systems
4551 that have them, for controlling the resources used by pipe deliveries.
4553 PH/42 The content-scanning code uses fopen() to create files in which to put
4554 message data. Previously it was not paying any attention to the mode of
4555 the files. Exim runs with umask(0) because the rest of the code creates
4556 files with open(), and sets the required mode explicitly. Thus, these
4557 files were ending up world-writeable. This was not a big issue, because,
4558 being within the spool directory, they were not world-accessible. I have
4559 created a function called modefopen, which takes an additional mode
4560 argument. It sets umask(777), creates the file, chmods it to the required
4561 mode, then resets the umask. All the relevant calls to fopen() in the
4562 content scanning code have been changed to use this function.
4564 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4565 to 24 hours. This avoids potential overflow problems when processing G
4566 and H retry rules. I suspect nobody ever tinkers with this value.
4568 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4570 PH/45 When the plaintext authenticator is running as a client, the server's
4571 challenges are checked to ensure they are valid base64 strings. By
4572 default, the authentication attempt is cancelled if an invalid string is
4573 received. Setting client_ignore_invalid_base64 true ignores these errors.
4574 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4575 they are received. Thus, the responses can be made to depend on the
4576 challenges. If an invalid string is ignored, an empty string is placed in
4579 PH/46 Messages that are created by the autoreply transport now contains a
4580 References: header, in accordance with RFCs 2822 and 3834.
4582 PH/47 Added authenticated_sender_force to the smtp transport.
4584 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4586 PH/49 Installed latest patch from the Sieve maintainer.
4588 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4589 was also set, the appendfile transport was unnecessarily scanning a
4590 directory of message files (e.g. for maildir delivery) to find the count
4591 of files (along with the size), even though it did not need this
4592 information. It now does the scan only if it needs to find either the
4593 size of the count of files.
4595 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4597 PH/52 Two bugs concerned with error handling when the smtp transport is
4600 (i) Exim was not creating retry information for temporary errors given
4601 for individual recipients after the DATA command when the smtp transport
4602 was used in LMTP mode. This meant that they could be retried too
4603 frequently, and not timed out correctly.
4605 (ii) Exim was setting the flag that allows error details to be returned
4606 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4607 recipients that were returned after the DATA command.
4609 PH/53 This is related to PH/52, but is more general: for any failing address,
4610 when detailed error information was permitted to be returned to the
4611 sender, but the error was temporary, then after the final timeout, only
4612 "retry timeout exceeded" was returned. Now it returns the full error as
4613 well as "retry timeout exceeded".
4615 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4616 do this, and (what is worse) MTAs that accept it.
4618 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4619 will now be deprecated.
4621 PH/56 New os.c-cygwin from the Cygwin maintainer.
4623 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4624 formats (previously only available in exim formats via -bpr, -bpru,
4625 and -bpra. Now also available in native and exiqgrep formats)
4627 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4628 with very large, slow to parse queues
4630 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4632 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4634 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4635 responses to authentication challenges, though it was showing the
4636 challenges; (ii) I've removed the CR characters from the debug output for
4639 PH/58 Allow for the insertion of a newline as well as a space when a string
4640 is turned into more than one encoded-word during RFC 2047 encoding. The
4641 Sieve code now uses this.
4643 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4644 data_4xx, lost_connection, tls_required.
4646 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4647 message was being sent as an SMTP response.
4649 PH/61 Add -l and -k options to exicyclog.
4651 PH/62 When verifying, if an address was redirected to one new address, so that
4652 verification continued, and the new address failed or deferred after
4653 having set something in $address_data, the value of $address_data was not
4654 passed back to the ACL. This was different to the case when no
4655 redirection occurred. The value is now passed back in both cases.
4657 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4658 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4659 use login_cap.h, so on its own it isn't the distinguishing feature. The
4660 new name refers directly to the setclassresources() function.
4662 PH/65 Added configuration files for NetBSD3.
4664 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4666 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4667 is preferred over IPv4.
4669 PH/68 The bounce_return_message and bounce_return_body options were not being
4670 honoured for bounces generated during the reception of non-SMTP messages.
4671 In particular, this applied to messages rejected by the ACL. This bug has
4672 been fixed. However, if bounce_return_message is true and bounce_return_
4673 body is false, the headers that are returned for a non-SMTP message
4674 include only those that have been read before the error was detected.
4675 (In the case of an ACL rejection, they have all been read.)
4677 PH/69 The HTML version of the specification is now built in a directory called
4678 spec_html instead of spec.html, because the latter looks like a path with
4679 a MIME-type, and this confuses some software.
4681 PH/70 Catch two compiler warnings in sieve.c.
4683 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4684 function verify_get_ident() calls ip_connect() to connect a socket, but
4685 if the "connect()" function timed out, ip_connect() used to close the
4686 socket. However, verify_get_ident() also closes the socket later, and in
4687 between Exim writes to the log, which may get opened at this point. When
4688 the socket was closed in ip_connect(), the log could get the same file
4689 descriptor number as the socket. This naturally causes chaos. The fix is
4690 not to close the socket in ip_connect(); the socket should be closed by
4691 the function that creates it. There was only one place in the code where
4692 this was missing, in the iplookup router, which I don't think anybody now
4693 uses, but I've fixed it anyway.
4695 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4696 well as to direct DNS lookups. Otherwise the handling of names in host
4697 lists is inconsistent and therefore confusing.
4703 PH/01 Two changes to the default runtime configuration:
4705 (1) Move the checks for relay_from_hosts and authenticated clients from
4706 after to before the (commented out) DNS black list checks.
4708 (2) Add control=submission to the relay_from_hosts and authenticated
4709 clients checks, on the grounds that messages accepted by these
4710 statements are most likely to be submissions.
4712 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4714 (1) Generate an error if the third argument for the ${prvs expansion is
4717 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4720 (3) Reset the variables that are obtained from the first argument of
4721 ${prvscheck and used in the second argument before leaving the code,
4722 because their memory is reclaimed, so using them afterwards may do
4725 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4726 one (it's much easier than Tom thought :-).
4728 (5) Because of (4), we can now allow for the use of $prvscheck_result
4729 inside the third argument.
4731 PH/03 For some reason, the default setting of PATH when running a command from
4732 a pipe transport was just "/usr/bin". I have changed it to
4735 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4736 anything to be listed in the output from -bV.
4738 PH/05 When a filter generated an autoreply, the entire To: header line was
4739 quoted in the delivery log line, like this:
4741 => >A.N.Other <ano@some.domain> <original@ddress> ...
4743 This has been changed so that it extracts the operative address. There
4744 may be more than one such address. If so, they are comma-separated, like
4747 => >ano@some.domain,ona@other.domain <original@ddress> ...
4749 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4750 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4751 not being looked up in the rDNS to get a host name, Exim was showing the
4752 IP address twice in Received: lines, even though the IP addresses were
4753 identical. For example:
4755 Received: from [1.2.3.4] (helo=[1.2.3.4])
4757 However, if the real host name was known, it was omitting the HELO data
4758 if it matched the actual IP address. This has been tidied up so that it
4759 doesn't show the same IP address twice.
4761 PH/07 When both +timestamp and +memory debugging was on, the value given by
4762 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4763 called by the debug printing, thereby overwriting the timestamp buffer.
4764 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4766 PH/08 When the original message was included in an autoreply transport, it
4767 always said "this is a copy of the message, including all the headers",
4768 even if body_only or headers_only was set. It now gives an appropriate
4771 PH/09 Applied a patch from the Sieve maintainer which:
4773 o fixes some comments
4774 o adds the (disabled) notify extension core
4775 o adds some debug output for the result of if/elsif tests
4776 o points to the current vacation draft in the documentation
4777 and documents the missing references header update
4781 o fixes a bug in processing the envelope test (when testing
4782 multiple envelope elements, the last element determined the
4785 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4786 Electronic Mail") by including:
4788 Auto-submitted: auto-generated
4790 in the messages that it generates (bounce messages and others, such as
4791 warnings). In the case of bounce messages for non-SMTP messages, there was
4792 also a typo: it was using "Auto_submitted" (underscore instead of
4793 hyphen). Since every message generated by Exim is necessarily in response
4794 to another message, thes have all been changed to:
4796 Auto-Submitted: auto-replied
4798 in accordance with these statements in the RFC:
4800 The auto-replied keyword:
4802 - SHOULD be used on messages sent in direct response to another
4803 message by an automatic process,
4805 - MUST NOT be used on manually-generated messages,
4807 - MAY be used on Delivery Status Notifications (DSNs) and Message
4808 Disposition Notifications (MDNs),
4810 - MUST NOT be used on messages generated by automatic or periodic
4811 processes, except for messages which are automatic responses to
4814 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4815 to the default Received: header definition.
4817 PH/12 Added log selector acl_warn_skipped (default on).
4819 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4820 variables because (a) they are in the wrong storage pool and (b) even if
4821 they were copied, it wouldn't work properly because of the caching.
4823 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4824 checking when decoding. Apparently there are clients that generate
4825 overlong encoded strings. Why am I not surprised?
4827 PH/15 If the first argument of "${if match_address" was not empty, but did not
4828 contain an "@" character, Exim crashed. Now it writes a panic log message
4829 and treats the condition as false.
4831 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4833 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4834 extension "envelope-auth". The code is finished and in agreement with
4835 other implementations, but there is no documentation so far and in fact,
4836 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4837 not changing the active code.
4839 Print executed "if" and "elsif" statements when debugging is used. This
4840 helps a great deal to understand what a filter does.
4842 Document more things not specified clearly in RFC3028. I had all this
4843 sorted out, when out of a sudden new issues came to my mind. Oops."
4845 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4848 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4849 canonical form (as documented). However, after a host name lookup from
4850 the IP address, check_host() was doing a simple string comparison with
4851 addresses acquired from the DNS when checking that the found name did
4852 have the original IP as one of its addresses. Since any found IPv6
4853 addresses are likely to be in abbreviated form, the comparison could
4854 fail. Luckily, there already exists a function for doing the comparison
4855 by converting both addresses to binary, so now that is used instead of
4856 the text comparison.
4858 PH/20 There was another similar case to PH/19, when a complete host name was
4859 given in a host list; looking up its IP address could give an abbreviated
4860 form, whereas the current host's name might or might not be abbreviated.
4861 The same fix has been applied.
4867 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4868 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4871 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4872 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4874 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4876 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4877 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4878 Also, for "Auto-Submitted", treat anything other than "no" as
4879 non-personal, in accordance with RFC 3834. (Previously it treated
4880 anything starting "auto-" as non-personal.)
4882 TF/01 The control=submission/name=... option had a problem with syntax
4883 errors if the name included a slash character. The /name= option
4884 now slurps the rest of the string, so it can include any characters
4885 but it must come last in the list of options (after /sender_retain
4888 PH/05 Some modifications to the interface to the fake nameserver for the new
4896 TK/01 Added the "success_on_redirect" address verification option. See
4897 NewStuff for rationale and an example.
4899 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4901 PH/02 Patch to exigrep to allow it to work on syslog lines.
4903 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4904 fread() to skip over the body file's header line, because in Cygwin the
4905 header line is locked and is inaccessible.
4907 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4908 co-exist for some time) to make it clear that it is the Exim ID that is
4909 referenced, not the Message-ID: header line.
4911 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4912 string_format() function, because snprintf() does not exist on all
4915 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4916 ${stat: expansion item.
4918 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4919 protocol synchronization error", to keep the pedants happy.
4921 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4922 well as for IRIX systems, when gcc is being used. See the host.c source
4925 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4927 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4930 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4931 transport and to the smtp transport in LMTP mode.
4933 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4935 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4936 run a filter in a subprocess. This could lead to confusion in subsequent
4937 lookups in the parent process. There should also be a search_tidyup() at
4938 the end of the subprocess.
4940 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4941 only if the host matched helo_try_verify_hosts, which caused the
4942 verification to occur when the EHLO/HELO command was issued. The ACL just
4943 tested the remembered result. Now, if a previous verification attempt has
4944 not happened, "verify = helo" does it there and then.
4946 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4948 TK/03 Fix log output including CR from clamd.
4950 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4951 exist provoked a memory error which could cause a segfault.
4953 PH/15 Installed PCRE 6.2
4955 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4957 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4958 of the problem. Specifically, suggested +O2 rather than +O1 for the
4961 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4963 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4964 record, the port was not correctly logged when the outgoing_port log
4965 selector was set (it logged the transort's default port).
4967 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4968 fallback_hosts, and "hosts" in the smtp transport.
4970 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4971 host errors such as "Connection refused".
4973 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4974 authentication with radiusclient 0.4.9:
4976 - Error returned from rc_read_config was caught wrongly
4977 - Username/password not passed on to radius server due to wrong length.
4979 The presumption is that some radiusclient API changes for 4.51/PH/17
4980 were not taken care of correctly. The code is still untested by me (my
4981 Linux distribution still has 0.3.2 of radiusclient), but it was
4982 contributed by a Radius user.
4984 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4985 expanding the "port" option of the smtp transport.
4987 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4988 while reading a MIME header. Thanks to Tom Hughes for a patch.
4990 PH/24 Include config.h inside local_scan.h so that configuration settings are
4993 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4994 that in an ACL for RCPT (for example), you can examine exactly what was
4997 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4998 commands, but it was not correctly comparing the address with the actual
4999 client host address. Thus, it would show the EHLO address in Received:
5000 header lines when this was not necessary.
5002 PH/27 Added the % operator to ${eval:}.
5004 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5005 it should be ignoring failures (because with -C, for example, it has lost
5006 privilege). It wasn't ignoring creation failures other than "already
5009 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5012 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5013 because an input line was too long, either on its own, or by virtue of
5014 too many continuations, the temporary file was not being removed, and the
5015 return code was incorrect.
5017 PH/31 Missing "BOOL" in function definition in filtertest.c.
5019 PH/32 Applied Sieve patches from the maintainer.
5021 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5023 PH/33 Added "verify = not_blind".
5025 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5026 Local/Makefile (with some defaults set). These are used in built scripts
5027 such as exicyclog, but they have never been used in the exim_install
5028 script (though there are many overriding facilities there). I have
5029 arranged that the exim_install script now takes note of these two
5032 PH/35 Installed configuration files for Dragonfly.
5034 PH/36 When a locally submitted message by a trusted user did not contain a
5035 From: header, and the sender address was obtained from -f or from an SMTP
5036 MAIL command, and the trusted user did not use -F to supply a sender
5037 name, $originator_name was incorrectly used when constructing a From:
5038 header. Furthermore, $originator_name was used for submission mode
5039 messages from external hosts without From: headers in a similar way,
5040 which is clearly wrong.
5042 PH/37 Added control=suppress_local_fixups.
5044 PH/38 When log_selector = +received_sender was set, and the addition of the
5045 sender made the log line's construction buffer exactly full, or one byte
5046 less than full, an overflow happened when the terminating "\n" was
5049 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5050 when the result of a list match is failure because a DNS lookup failed.
5052 PH/40 RM_COMMAND is now used in the building process.
5054 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5055 the "build-* directories that it finds.
5057 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5058 address in a domain literal was a prefix of an interface address.
5060 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5061 when verifying a sender address, unless rewrite_headers is false.
5063 PH/44 Wrote a long comment about why errors_to addresses are verified as
5064 recipients, not senders.
5066 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5067 the ratelimit ACL was added.
5069 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5071 PH/46 Added extra information about PostgreSQL errors to the error string.
5073 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5074 avoiding the need to install special zones in a real server. This is
5075 backwards compatible; if it can't find the fake resolver, it drops back.
5076 Thus, both old and new test suites can be run.
5078 TF/02 Added util/ratelimit.pl
5080 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5081 clock is set back in time.
5083 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5084 Candler <B.Candler@pobox.com>.
5086 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5087 OK for addresses that are the result of redirections.
5089 PH/48 A number of further additions for the benefit of the new test suite,
5090 including a fake gethostbyname() that interfaces to the fake DNS resolver
5093 TF/06 The fix for widen_domains has also been applied to qualify_single and
5094 search_parents which are the other dnslookup options that can cause
5097 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5100 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5102 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5103 service" if it encountered an MX record with an empty target hostname.
5104 The message is now "an MX or SRV record indicated no SMTP service".
5106 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5107 if the DNS of the sending site is misconfigured. This is quite a
5108 common situation. This change restores the behaviour of treating a
5109 helo verification defer as a failure.
5111 PH/51 If self=fail was set on a router, the bounce message did not include the
5112 actual error message.
5118 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5120 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5121 command itself ended in error, the underlying message about the transport
5122 filter timeout was being overwritten with the pipe command error. Now the
5123 underlying error message should be appended to the second error message.
5125 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5127 PH/02 Reduce the amount of output that "make" produces by default. Full output
5128 can still be requested.
5130 PH/03 The warning log line about a condition test deferring for a "warn" verb
5131 was being output only once per connection, rather than after each
5132 occurrence (because it was using the same function as for successful
5133 "warn" verbs). This seems wrong, so I have changed it.
5135 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5136 it should not have, which might have caused a crash in the right
5137 circumstances, but probably never did.
5139 PH/04 Installed a modified version of Tony Finch's patch to make submission
5140 mode fix the return path as well as the Sender: header line, and to
5141 add a /name= option so that you can make the user's friendly name appear
5144 TF/03 Added the control = fakedefer ACL modifier.
5146 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5147 Mark Lowes for thorough testing.
5149 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5151 TK/03 Merged latest SRS patch from Miles Wilton.
5153 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5154 with the definition in sysexits.h (which is #included earlier).
5155 Fortunately, Exim does not actually use EX_OK. The code used to try to
5156 preserve the sysexits.h value, by assuming that macro definitions were
5157 scanned for macro replacements. I have been disabused of this notion,
5158 so now the code just undefines EX_OK before #including unistd.h.
5160 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5161 in the smtp transport. When a block could not be written in a single
5162 write() function, the timeout was being re-applied to each part-write.
5163 This seems wrong - if the receiver was accepting one byte at a time it
5164 would take for ever. The timeout is now adjusted when this happens. It
5165 doesn't have to be particularly precise.
5167 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5168 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5170 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5171 without a domain if the check to <postmaster@domain> fails.
5173 SC/01 Eximstats: added -xls and the ability to specify output files
5174 (patch written by Frank Heydlauf).
5176 SC/02 Eximstats: use FileHandles for outputting results.
5178 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5180 SC/04 Eximstats: fixed display of large numbers with -nvr option
5182 SC/05 Eximstats: fixed merging of reports with empty tables.
5184 SC/06 Eximstats: added the -include_original_destination flag
5186 SC/07 Eximstats: removed tabs and trailing whitespace.
5188 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5190 TK/06 MBOX spool code: Add real "From " MBOX separator line
5191 so the .eml file is really in mbox format (even though
5192 most programs do not really care). Patch from Alex Miller.
5194 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5195 The latter is generated from $received_to and is only set if the
5196 message has one envelope recipient. SA can use these headers,
5197 obviously out-of-the-box. Patch from Alex Miller.
5199 PH/08 The ${def test on a variable was returning false if the variable's
5200 value was "0", contrary to what the specification has always said!
5201 The result should be true unless the variable is empty.
5203 PH/09 The syntax error of a character other than { following "${if
5204 def:variable_name" (after optional whitespace) was not being diagnosed.
5205 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5206 accidental colon was present, for example, could give incorrect results.
5208 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5209 result is used (not including appendfile, where other changes are about
5212 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5213 This involved changing a lot of size variables from int to off_t. It
5214 should work with maildirs and everything.
5216 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5217 spamd dying while we are connected to it.
5219 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5222 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5223 function for BDB 4.3.
5225 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5227 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5228 the sources for PCRE have been re-arranged and more files are now
5231 PH/15 The code I had for printing potentially long long variables in PH/11
5232 above was not the best (it lost precision). The length of off_t variables
5233 is now inspected at build time, and an appropriate printing format (%ld
5234 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5235 to be "long long int" or "long int". This is needed for the internal
5236 formatting function string_vformat().
5238 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5239 the configuration file to be ":syslog", then the script "guesses" where
5240 the logs files are, rather than using the compiled in default. In our
5241 case the guess is not the same as the compiled default, so the script
5242 suddenly stopped working when I started to use syslog. The patch checks
5243 to see if log_file_path is "". If so, it attempts to read it from exim
5244 with no configuration file to get the compiled in version, before it
5245 falls back to the previous guessing code."
5247 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5248 implementing BATV in an Exim configuration. See NewStuff for the gory
5251 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5252 Makefile that are specific to HP-UX.
5254 PH/18 If the "use_postmaster" option was set for a recipient callout together
5255 with the "random" option, the postmaster address was used as the MAIL
5256 FROM address for the random test, but not for the subsequent recipient
5257 test. It is now used for both.
5259 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5260 patch removes a few documentation additions to RFC 3028, because the
5261 latest draft now contains them. It adds the new en;ascii-case comparator
5262 and a new error check for 8bit text in MIME parts. Comparator and
5263 require names are now matched exactly. I enabled the subaddress
5264 extension, but it is not well tested yet (read: it works for me)."
5266 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5267 rework some of the code of TK/09 above to avoid the hardwired use of
5268 "%lld" and "long long". Replaced the call to snprintf() with a call to
5271 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5272 records point to non-existent hosts", "retry timeout exceeded", and
5273 "retry time not reached for any host after a long failure period".
5275 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5276 experimental DomainKeys support:
5278 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5279 (2) On an error such as an illegally used "control", the wrong name for
5280 the control was given.
5282 These problems did NOT occur unless DomainKeys support was compiled.
5284 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5286 PH/24 Added ${if match_ip condition.
5288 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5289 databases so that it will be absolutely obvious if a crash occurs in the
5290 DB library. This is a regular occurrence (often caused by mis-matched
5293 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5294 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5295 picked up on a user's system that detects such things. There doesn't seem
5296 to be a gcc warning option for this - only an attribute that has to be
5297 put on the function's prototype. It seems that in Fedora Core 4 they have
5298 set this on a number of new functions. No doubt there will be more in due
5301 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5302 specify a transport. However, if an address that was verified by such a
5303 router was the subject of a callout, Exim crashed because it tried to
5304 read the rcpt_include_affixes from the non-existent transport. Now it
5305 just assumes that the setting of that option is false. This bug was
5306 introduced by 4.51/PH/31.
5308 PH/28 Changed -d+all to exclude +memory, because that information is very
5309 rarely of interest, but it makes the output a lot bigger. People tend to
5310 do -d+all out of habit.
5312 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5313 code in os-type was giving problems when libc.so lives in lib64, like on
5316 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5317 aren't the modern standard, and it seems that some systems' include files
5318 don't always have them. Exim was already checking for some of the newer
5319 ones like T_AAAA, and defining it itself. I've added checks for all the
5320 record types that Exim uses.
5322 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5323 not automatically generating a new one, as it is supposed to. This
5324 prevented TLS from working. If the file did exist, but contained invalid
5325 data, a new version was generated, as expected. It was only the case of a
5326 non-existent file that was broken.
5328 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5329 with a change in libdomainkeys > 0.64.
5331 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5332 from DNS. If the selector record carries the flag, it now has
5333 precedence over the domain-wide flag.
5335 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5337 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5338 the use of a port name that isn't defined in /etc/services) occurred, the
5339 message was deferred as in a normal delivery, and thus remained on the
5340 spool, instead of being failed because of the mua_wrapper setting. This
5341 is now fixed, and I tidied up some of the mua_wrapper messages at the
5344 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5345 the messages in a hash of arrays rather than using individual hashes.
5346 This is a bit cleaner and results in dramatic memory savings, albeit
5347 at a slight CPU cost.
5349 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5350 as requested by Marc Sherman.
5352 SC/10 Eximstats: added histograms for user specified patterns as requested
5355 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5357 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5358 fopen() in the content-scanning modules that did not already have it.
5364 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5365 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5367 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5369 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5371 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5372 to test Sieve filters that use "vacation".
5374 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5375 that changes the way the GnuTLS parameters are stored in the cache file.
5376 The new format can be generated externally. For backward compatibility,
5377 if the data in the cache doesn't make sense, Exim assumes it has read an
5378 old-format file, and it generates new data and writes a new file. This
5379 means that you can't go back to an older release without removing the
5382 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5383 work if there are any delivery delays because "one_time" forces the
5384 parent to be marked "delivered", so its unseen clone is never tried
5385 again. For this reason, Exim now forbids the simultaneous setting of
5388 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5389 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5390 entry if you want to know the details. The fix, however, neglected to
5391 consider the case when local delivery batching is involved. The test for
5392 "previously delivered" was not happening when checking to see if an
5393 address could be batched with a previous (undelivered) one; under
5394 certain circumstances this could lead to multiple deliveries to the same
5397 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5398 in its include files, and this causes problems building Exim.
5400 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5401 header_syntax) but Exim was just ignoring anything given after a slash.
5402 In particular, this caused confusion with an attempt to use "verify =
5403 reverse_host_lookup/defer_ok". An error is now given when options are
5404 supplied for verify items that do not have them. (Maybe reverse_host_
5405 lookup should have a defer_ok option, but that's a different point.)
5407 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5408 defined by RFC 821) to 2048, because there were problems with some AUTH
5409 commands, and RFC 1869 says the size should be increased for extended
5410 SMTP commands that take arguments.
5412 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5415 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5416 "unknown" error; now it says that the functionality isn't in the binary.
5418 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5419 an address' error message when a string expansion fails (syntax or
5420 whatever). Otherwise the password may appear in the log. Following change
5421 PH/42 below, there is no longer a chance of it appearing in a bounce
5424 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5426 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5427 of its addresses was ever tried. (Bugzilla bug #2.)
5429 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5430 the result incorrectly in the debug output. (It correctly added a newline
5431 to what was transported.)
5433 TF/01 Added $received_time.
5435 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5436 commented out examples of how to interface to a virus scanner and to
5437 SpamAssassin. Also added commented examples of av_scanner and
5438 spamd_address settings.
5440 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5441 and controls are allowed in which ACLs. There were a couple of minor
5442 errors. Some of the entries in the conditions table (which is a table of
5443 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5444 negation of where the condition IS allowed.
5446 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5448 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5449 header file does not have a version number, so I've had to invent a new
5450 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5451 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5452 radiusclient), but it was contributed by a Radius user.
5454 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5455 files or directories, for OpenSSL.
5457 PH/19 When an Exim process that is running as root has to create an Exim log
5458 file, it does so in a subprocess that runs as exim:exim so as to get the
5459 ownership right at creation (otherwise, other Exim processes might see
5460 the file with the wrong ownership). There was no test for failure of this
5461 fork() call, which would lead to the process getting stuck as it waited
5462 for a non-existent subprocess. Forks do occasionally fail when resources
5463 run out. I reviewed all the other calls to fork(); they all seem to check
5466 PH/20 When checking for unexpected SMTP input at connect time (before writing
5467 the banner), Exim was not dealing correctly with a non-positive return
5468 from the read() function. If the client had disconnected by this time,
5469 the result was a log entry for a synchronization error with an empty
5470 string after "input=" when read() returned zero. If read() returned -1
5471 (an event I could not check), uninitialized data bytes were printed.
5472 There were reports of junk text (parts of files, etc) appearing after
5475 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5477 PH/22 Added support for macro redefinition, and (re)definition in between
5478 driver and ACL definitions.
5480 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5481 forgetting to use the resulting value; it was using the unexpanded value.
5483 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5484 hadn't been configured. The fix is from Juergen Kreileder, who
5485 understands it better than I do:
5487 "Here's what I see happening with three configured cyrus_sasl
5488 authenticators configured (plain, login, cram-md5):
5490 On startup auth_cyrus_sasl_init() gets called for each of these.
5491 This means three calls to sasl_listmech() without a specified mech_list.
5492 => SASL tests which mechs of all available mechs actually work
5493 => three warnings about OTP not working
5494 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5496 With the patch, sasl_listmech() also gets called three times. But now
5497 SASL's mech_list option is set to the server_mech specified in the the
5498 authenticator. Or in other words, the answer from sasl_listmech()
5499 gets limited to just the mech you're testing for (which is different
5501 => the return list contains just 'plain' or 'login', 'cram-md5' or
5502 nothing depending on the value of ob->server_mech.
5504 I've just tested the patch: Authentication still works fine,
5505 unavailable mechs specified in the exim configuration are still
5506 caught, and the auth.log warnings about OTP are gone."
5508 PH/25 When debugging is enabled, the contents of the command line are added
5509 to the debugging output, even when log_selector=+arguments is not
5512 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5513 answer is "GNU", and only if the return is "GNU/something" is the answer
5516 PH/27 $acl_verify_message is now set immediately after the failure of a
5517 verification in an ACL, and so is available in subsequent modifiers. In
5518 particular, the message can be preserved by coding like this:
5520 warn !verify = sender
5521 set acl_m0 = $acl_verify_message
5523 Previously, $acl_verify_message was set only while expanding "message"
5524 and "log_message" when a very denied access.
5526 PH/28 Modified OS/os.c-Linux with
5528 -#ifndef OS_LOAD_AVERAGE
5529 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5531 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5532 nomenclature these days.)
5534 PH/29 Installed patch from the Sieve maintainer that adds the options
5535 sieve_useraddress and sieve_subaddress to the redirect router.
5537 PH/30 In these circumstances:
5538 . Two addresses routed to the same list of hosts;
5539 . First host does not offer TLS;
5540 . First host accepts first address;
5541 . First host gives temporary error to second address;
5542 . Second host offers TLS and a TLS session is established;
5543 . Second host accepts second address.
5544 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5545 and peerdn, if requested) that were in fact used only for the second
5548 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5549 attention to any local part prefix or suffix that was matched by the
5550 router that accepted the address. It now behaves in the same way as it
5551 does for delivery: the affixes are removed from the local part unless
5552 rcpt_include_affixes is set on the transport.
5554 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5555 timeout during the DATA phase of an incoming message.
5557 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5558 applied a patch sanctioned by the Sieve maintainer.
5560 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5561 the uid or gid is negative. A case of a negative gid caused this to be
5562 noticed. The fix allows for either to be negative.
5564 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5565 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5568 PH/36 The change PH/12 above was broken. Fixed it.
5570 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5571 the grounds that routing the same address twice would always produce the
5572 same answer. This might have been true once, but it is certainly no
5573 longer true now. Routing a child address may depend on the previous
5574 routing that produced that child. Some complicated redirection strategies
5575 went wrong when messages had multiple recipients, and made Exim's
5576 behaviour dependent on the order in which the addresses were given.
5578 I have moved the duplicate checking until after the routing is complete.
5579 Exim scans the addresses that are assigned to local and remote
5580 transports, and removes any duplicates. This means that more work will be
5581 done, as duplicates will always all be routed, but duplicates are
5582 presumably rare, so I don't expect this is of any significance.
5584 For deliveries to pipes, files, and autoreplies, the duplicate checking
5585 still happens during the routing process, since they are not going to be
5588 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5589 It corrects a timeout issue with spamd. This is Ian's comment: "The
5590 background is that sometimes spamd either never reads data from a
5591 connection it has accepted, or it never writes response data. The exiscan
5592 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5593 blindly assumes that writes won't block so it may never time out."
5595 PH/39 Allow G after quota size as well as K and M.
5597 PH/40 The value set for $authenticated_id in an authenticator may not contain
5598 binary zeroes or newlines because the value is written to log lines and
5599 to spool files. There was no check on this. Now the value is run through
5600 the string_printing() function so that such characters are converted to
5601 printable escape sequences.
5603 PH/41 $message_linecount is a new variable that contains the total number of
5604 lines in the message. Compare $body_linecount, which is the count for the
5607 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5608 bounce and delay warning messages, except in certain special cases, which
5611 (a) An SMTP error message from a remote host;
5612 (b) A message specified in a :fail: redirection;
5613 (c) A message specified in a "fail" command in a system filter;
5614 (d) A message specified in a FAIL return from the queryprogram router;
5615 (e) A message specified by the cannot_route_message router option.
5617 In these cases only, Exim does include the error details in bounce and
5618 warning messages. There are also a few cases where bland messages such
5619 as "unrouteable address" or "local delivery error" are given.
5621 PH/43 $value is now also set for the "else" part of a ${run expansion.
5623 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5624 being worked on, but at least Exim now implements the latest version to
5627 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5628 process to complete was treated as a delivery failure, a timeout while
5629 writing the message to the pipe was logged, but erroneously treated as a
5630 successful delivery. Such timeouts include transport filter timeouts. For
5631 consistency with the overall process timeout, these timeouts are now
5632 treated as errors, giving rise to delivery failures by default. However,
5633 there is now a new Boolean option for the pipe transport called
5634 timeout_defer, which, if set TRUE, converts the failures into defers for
5635 both kinds of timeout. A transport filter timeout is now identified in
5638 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5639 systems where "make" and "gmake" are different, calling "gmake" at top
5640 level broke things. I've arranged for the value of $(MAKE) to be passed
5641 from the Makefile to this script so that it can call the same version of
5645 A note about Exim versions 4.44 and 4.50
5646 ----------------------------------------
5648 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5649 changes of various kinds. As a consequence, a big documentation update was
5650 needed. This delayed the release for rather longer than seemed good, especially
5651 in the light of a couple of (minor) security issues. Therefore, the changes
5652 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5653 release. So 4.44 and 4.50 are in effect two different branches that both start
5656 I have left the 4.50 change log unchanged; it contains all the changes since
5657 4.43. The change log for 4.44 is below; many of its items are identical to
5658 those for 4.50. This seems to be the most sensible way to preserve the
5659 historical information.
5665 1. Minor wording change to the doc/README.SIEVE file.
5667 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5668 computation of the current number of files was incorrect.
5670 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5671 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5674 4. Give more explanation in the error message when the command for a transport
5675 filter fails to execute.
5677 5. There are several places where Exim runs a non-Exim command in a
5678 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5679 was being done only for the command run by the queryprogram router. It is
5680 now done for all such subprocesses. The other cases are: ${run, transport
5681 filters, and the commands run by the lmtp and pipe transports.
5683 6. Added CONFIGURE_GROUP build-time option.
5685 7. Some older OS have a limit of 256 on the maximum number of file
5686 descriptors. Exim was using setrlimit() to set 1000 as a large value
5687 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5688 systems. I've change it so that if it can't get 1000, it tries for 256.
5690 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5691 was an oversight, and furthermore, ever since the addition of extra
5692 controls (e.g. 4.43/32), the checks on when to allow different forms of
5693 "control" were broken. There should now be diagnostics for all cases when a
5694 control that does not make sense is encountered.
5696 9. Added the /retain_sender option to "control=submission".
5698 10. $recipients is now available in the predata ACL (oversight).
5700 11. Tidy the search cache before the fork to do a delivery from a message
5701 received from the command line. Otherwise the child will trigger a lookup
5702 failure and thereby defer the delivery if it tries to use (for example) a
5703 cached ldap connection that the parent has called unbind on.
5705 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5706 of $address_data from the recipient verification was clobbered by the
5707 sender verification.
5709 13. The value of address_data from a sender verification is now available in
5710 $sender_address_data in subsequent conditions in the ACL statement.
5712 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5714 15. Added a new option "connect=<time>" to callout options, to set a different
5717 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5718 was its contents. (It was OK if the option was not defined at all.)
5720 17. A "Completed" log line is now written for messages that are removed from
5721 the spool by the -Mrm option.
5723 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5724 information about exactly what failed.
5726 19. Added -dd to debug only the daemon process.
5728 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5729 handles timeouts, both on the server side and network timeouts. Renamed the
5730 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5732 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5733 It is now set to "smtps".
5735 22. $host_address is now set to the target address during the checking of
5736 ignore_target_hosts.
5738 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5739 being passed; this would have caused $sender_host_name to have been used if
5740 matching the list had actually called for a host name (not very likely,
5741 since this list is usually IP addresses). A host name is now passed as
5744 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5745 code that specifies a non-restarting handler (typically sigaction() in
5746 modern systems) in an attempt to fix a rare and obscure crash bug.
5748 25. Narrowed the window for a race in the daemon that could cause it to ignore
5749 SIGCHLD signals. This is not a major problem, because they are used only to
5750 wake it up if nothing else does.
5752 26. A malformed maildirsize file could cause Exim to calculate negative values
5753 for the mailbox size or file count. Odd effects could occur as a result.
5754 The maildirsize information is now recalculated if the size or filecount
5757 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5758 support for a long time. Removed HAVE_SYS_VFS_H.
5760 28. Installed the latest version of exipick from John Jetmore.
5762 29. In an address list, if the pattern was not a regular expression, an empty
5763 subject address (from a bounce message) matched only if the pattern was an
5764 empty string. Non-empty patterns were not even tested. This was the wrong
5765 because it is perfectly reasonable to use an empty address as part of a
5766 database query. An empty address is now tested by patterns that are
5767 lookups. However, all the other forms of pattern expect the subject to
5768 contain a local part and a domain, and therefore, for them, an empty
5769 address still always fails if the pattern is not itself empty.
5771 30. Exim went into a mad DNS loop when attempting to do a callout where the
5772 host was specified on an smtp transport, and looking it up yielded more
5773 than one IP address.
5775 31. Re-factored the code for checking spool and log partition space into a
5776 function that finds that data and another that does the check. The former
5777 is then used to implement four new variables: $spool_space, $log_space,
5778 $spool_inodes, and $log_inodes.
5780 32. The RFC2047 encoding function was originally intended for short strings
5781 such as real names; it was not keeping to the 75-character limit for
5782 encoded words that the RFC imposes. It now respects the limit, and
5783 generates multiple encoded words if necessary. To be on the safe side, I
5784 have increased the buffer size for the ${rfc2047: expansion operator from
5787 33. It is now permitted to omit both strings after an "if" condition; if the
5788 condition is true, the result is "true". As before, when the second string
5789 is omitted, a false condition yields an empty string. This makes it less
5790 cumbersome to write custom ACL and router conditions.
5792 34. Failure to deliver a bounce message always caused it to be frozen, even if
5793 there was an errors_to setting on the router. The errors_to setting is now
5796 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5797 canonical form (fully expanded) before being placed in
5798 $sender_host_address.
5800 36. The table in the code that translates DNS record types into text (T_A to
5801 "A" for instance) was missing entries for NS and CNAME. It is just possible
5802 that this could have caused confusion if both these types were looked up
5803 for the same domain, because the text type is used as part of Exim's
5804 per-process caching. But the chance of anyone hitting this buglet seems
5807 37. The dnsdb lookup has been extended in a number of ways.
5809 (1) There is a new type, "zns", which walks up the domain tree until it
5810 finds some nameserver records. It should be used with care.
5812 (2) There is a new type, "mxh", which is like "mx" except that it returns
5813 just the host names, not the priorities.
5815 (3) It is now possible to give a list of domains (or IP addresses) to be
5816 looked up. The behaviour when one of the lookups defers can be
5817 controlled by a keyword.
5819 (4) It is now possible to specify the separator character for use when
5820 multiple records are returned.
5822 38. The dnslists ACL condition has been extended: it is now possible to supply
5823 a list of IP addresses and/or domains to be looked up in a particular DNS
5826 39. Added log_selector=+queue_time_overall.
5828 40. When running the queue in the test harness, wait just a tad after forking a
5829 delivery process, to get repeatability of debugging output.
5831 41. Include certificate and key file names in error message when GnuTLS fails
5832 to set them up, because the GnuTLS error message doesn't include the name
5833 of the failing file when there is a problem reading it.
5835 42. Allow both -bf and -bF in the same test run.
5837 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5839 44. The "Exiscan patch" is now merged into the mainline Exim source.
5841 45. Sometimes the final signoff response after QUIT could fail to get
5842 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5843 < 0 before doing a fflush(). This bug looks as though it goes back to the
5844 introduction of TLS in release 3.20, but "sometimes" must have been rare
5845 because the tests only now provoked it.
5847 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5848 (this can affect the format of dates).
5850 47. exim_tidydb, when checking for the continued existence of a message for
5851 which it has found a message-specific retry record, was not finding
5852 messages that were in split spool directories. Consequently, it was
5853 deleting retry records that should have stayed in existence.
5855 48. Steve fixed some bugs in eximstats.
5857 49. The SPA authentication driver was not abandoning authentication and moving
5858 on to the next authenticator when an expansion was forced to fail,
5859 contradicting the general specification for all authenticators. Instead it
5860 was generating a temporary error. It now behaves as specified.
5862 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5863 (the order specifies the preference for clients). The order is now AES256,
5864 AES128, 3DES, ARCFOUR128.
5866 51. Small patch to Sieve code - explicitly set From: when generating an
5869 52. Exim crashed if a remote delivery caused a very long error message to be
5870 recorded - for instance if somebody sent an entire SpamAssassin report back
5871 as a large number of 550 error lines. This bug was coincidentally fixed by
5872 increasing the size of one of Exim's internal buffers (big_buffer) that
5873 happened as part of the Exiscan merge. However, to be on the safe side, I
5874 have made the code more robust (and fixed the comments that describe what
5877 53. Now that there can be additional text after "Completed" in log lines (if
5878 the queue_time_overall log selector is set), a one-byte patch to exigrep
5879 was needed to allow it to recognize "Completed" as not the last thing in
5882 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5883 patch that reportedly fixes this has been added. I am not expert enough to
5884 create a test for it. This is what the patch creator wrote:
5886 "I found a little strange behaviour of ldap code when working with
5887 Windows 2003 AD Domain, where users was placed in more than one
5888 Organization Units. When I tried to give exim partial DN, the exit code
5889 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5890 But simultaneously result of request was absolutely normal ldap result,
5891 so I produce this patch..."
5893 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5894 REFERENCE, so I have modified the code to exclude the patch when that macro
5897 55. Some experimental protocols are using DNS PTR records for new purposes. The
5898 keys for these records are domain names, not reversed IP addresses. The
5899 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5900 leaves it alone. Component reversal etc. now happens only for IP addresses.
5903 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5905 57. Double the size of the debug message buffer (to 2048) so that more of very
5906 long debug lines gets shown.
5908 58. The exicyclog utility now does better if the number of log files to keep
5909 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5911 59. Two changes related to the smtp_active_hostname option:
5913 (1) $smtp_active_hostname is now available as a variable.
5914 (2) The default for smtp_banner uses $smtp_active_hostname instead
5915 of $primary_hostname.
5917 60. The host_aton() function is supposed to be passed a string that is known
5918 to be a valid IP address. However, in the case of IPv6 addresses, it was
5919 not checking this. This is a hostage to fortune. Exim now panics and dies
5920 if the condition is not met. A case was found where this could be provoked
5921 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5922 components; fortuitously, this particular loophole had already been fixed
5923 by change 4.50/55 above.
5925 If there are any other similar loopholes, the new check in host_aton()
5926 itself should stop them being exploited. The report I received stated that
5927 data on the command line could provoke the exploit when Exim was running as
5928 exim, but did not say which command line option was involved. All I could
5929 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5930 running as the user.
5933 61. There was a buffer overflow vulnerability in the SPA authentication code
5934 (which came originally from the Samba project). I have added a test to the
5935 spa_base64_to_bits() function which I hope fixes it.
5938 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5939 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5941 63. The daemon start-up calls getloadavg() while still root for those OS that
5942 need the first call to be done as root, but it missed one case: when
5943 deliver_queue_load_max is set with deliver_drop_privilege. This is
5944 necessary for the benefit of the queue runner, because there is no re-exec
5945 when deliver_drop_privilege is set.
5947 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5948 This has been fixed.
5950 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5951 was in use, was not putting the data itself into the right store pool;
5952 consequently, it could be overwritten for a subsequent message in the same
5953 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5956 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5958 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5959 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5960 the function were treating the return as a boolean value, which happened to
5961 work because 0=false and not-0=true, but is not correct code.
5963 68. The host_aton() function was not handling scoped IPv6 addresses (those
5964 with, for example, "%eth0" on the end) correctly.
5966 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5967 negated items (that is, ~something) in unsigned ints. Some compilers
5968 apparently mutter when there is no cast.
5970 70. If an address verification called from an ACL failed, and did not produce a
5971 user-specific message (i.e. there was only a "system" message), nothing was
5972 put in $acl_verify_message. In this situation, it now puts the system
5975 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5976 session; change 4.31/43 added the unwanted input to the log line - except
5977 that it did not do this in the start of session case. It now does.
5979 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5980 This is wrong and can cause the other end to generate a synchronization
5981 error if it is another Exim or anything else that does the synchronization
5982 check. A QUIT command is no longer sent after a timeout.
5984 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5985 during host lookups.
5987 74. The defer_ok option of callout verification was not working if it was used
5988 when verifying addresses in header lines, that is, for this case:
5990 verify = header_sender/callout=defer_ok
5992 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5993 those file descriptors could be used for SMTP connections. If anything
5994 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5995 could be sent to the SMTP client, causing chaos. The daemon now opens
5996 stdin, stdout, and stderr to /dev/null when it puts itself into the
5999 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6000 log by default. The user can override this with suitable Perl magic.
6002 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6003 the log message when discard triggers, was not working for the DATA ACL or
6004 for the non-SMTP ACL.
6006 78. Error message wording change in sieve.c.
6008 79. If smtp_accept_max_per_host was set, the number of connections could be
6009 restricted to fewer than expected, because the daemon was trying to set up
6010 a new connection before checking whether the processes handling previous
6011 connections had finished. The check for completed processes is now done
6012 earlier. On busy systems, this bug wouldn't be noticed because something
6013 else would have woken the daemon, and it would have reaped the completed
6016 80. If a message was submitted locally by a user whose login name contained one
6017 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6018 It caused a spool format error. I have fixed the spool reading code. A
6019 related problem was that the "from" clause in the Received: line became
6020 illegal because of the space(s). It is now covered by ${quote_local_part.
6022 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6025 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6027 83. Patch from Sieve maintainer to fix unterminated string problem in
6028 "vacation" handling.
6030 84. Some minor changes to the Linux configuration files to help with other
6031 OS variants using glibc.
6033 85. One more patch for Sieve to update vacation handling to latest spec.
6036 ----------------------------------------------------
6037 See the note above about the 4.44 and 4.50 releases.
6038 ----------------------------------------------------
6044 1. Change 4.43/35 introduced a bug that caused file counts to be
6045 incorrectly computed when quota_filecount was set in an appendfile
6048 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6049 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6052 3. Give more explanation in the error message when the command for a transport
6053 filter fails to execute.
6055 4. There are several places where Exim runs a non-Exim command in a
6056 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6057 was being done only for the command run by the queryprogram router. It is
6058 now done for all such subprocesses. The other cases are: ${run, transport
6059 filters, and the commands run by the lmtp and pipe transports.
6061 5. Some older OS have a limit of 256 on the maximum number of file
6062 descriptors. Exim was using setrlimit() to set 1000 as a large value
6063 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6064 systems. I've change it so that if it can't get 1000, it tries for 256.
6066 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6067 was an oversight, and furthermore, ever since the addition of extra
6068 controls (e.g. 4.43/32), the checks on when to allow different forms of
6069 "control" were broken. There should now be diagnostics for all cases when a
6070 control that does not make sense is encountered.
6072 7. $recipients is now available in the predata ACL (oversight).
6074 8. Tidy the search cache before the fork to do a delivery from a message
6075 received from the command line. Otherwise the child will trigger a lookup
6076 failure and thereby defer the delivery if it tries to use (for example) a
6077 cached ldap connection that the parent has called unbind on.
6079 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6080 of $address_data from the recipient verification was clobbered by the
6081 sender verification.
6083 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6084 was its contents. (It was OK if the option was not defined at all.)
6086 11. A "Completed" log line is now written for messages that are removed from
6087 the spool by the -Mrm option.
6089 12. $host_address is now set to the target address during the checking of
6090 ignore_target_hosts.
6092 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6093 being passed; this would have caused $sender_host_name to have been used if
6094 matching the list had actually called for a host name (not very likely,
6095 since this list is usually IP addresses). A host name is now passed as
6098 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6099 code that specifies a non-restarting handler (typically sigaction() in
6100 modern systems) in an attempt to fix a rare and obscure crash bug.
6102 15. Narrowed the window for a race in the daemon that could cause it to ignore
6103 SIGCHLD signals. This is not a major problem, because they are used only to
6104 wake it up if nothing else does.
6106 16. A malformed maildirsize file could cause Exim to calculate negative values
6107 for the mailbox size or file count. Odd effects could occur as a result.
6108 The maildirsize information is now recalculated if the size or filecount
6111 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6112 support for a long time. Removed HAVE_SYS_VFS_H.
6114 18. Updated exipick to current release from John Jetmore.
6116 19. Allow an empty sender to be matched against a lookup in an address list.
6117 Previously the only cases considered were a regular expression, or an
6120 20. Exim went into a mad DNS lookup loop when doing a callout where the
6121 host was specified on the transport, if the DNS lookup yielded more than
6124 21. The RFC2047 encoding function was originally intended for short strings
6125 such as real names; it was not keeping to the 75-character limit for
6126 encoded words that the RFC imposes. It now respects the limit, and
6127 generates multiple encoded words if necessary. To be on the safe side, I
6128 have increased the buffer size for the ${rfc2047: expansion operator from
6131 22. Failure to deliver a bounce message always caused it to be frozen, even if
6132 there was an errors_to setting on the router. The errors_to setting is now
6135 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6136 canonical form (fully expanded) before being placed in
6137 $sender_host_address.
6139 24. Updated eximstats to version 1.33
6141 25. Include certificate and key file names in error message when GnuTLS fails
6142 to set them up, because the GnuTLS error message doesn't include the name
6143 of the failing file when there is a problem reading it.
6145 26. Expand error message when OpenSSL has problems setting up cert/key files.
6148 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6149 (this can affect the format of dates).
6151 28. exim_tidydb, when checking for the continued existence of a message for
6152 which it has found a message-specific retry record, was not finding
6153 messages that were in split spool directories. Consequently, it was
6154 deleting retry records that should have stayed in existence.
6156 29. eximstats updated to version 1.35
6157 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6158 1.35 - bugfix such that pie charts by volume are generated correctly
6160 30. The SPA authentication driver was not abandoning authentication and moving
6161 on to the next authenticator when an expansion was forced to fail,
6162 contradicting the general specification for all authenticators. Instead it
6163 was generating a temporary error. It now behaves as specified.
6165 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6166 (the order specifies the preference for clients). The order is now AES256,
6167 AES128, 3DES, ARCFOUR128.
6169 31. Small patch to Sieve code - explicitly set From: when generating an
6172 32. Exim crashed if a remote delivery caused a very long error message to be
6173 recorded - for instance if somebody sent an entire SpamAssassin report back
6174 as a large number of 550 error lines. This bug was coincidentally fixed by
6175 increasing the size of one of Exim's internal buffers (big_buffer) that
6176 happened as part of the Exiscan merge. However, to be on the safe side, I
6177 have made the code more robust (and fixed the comments that describe what
6180 33. Some experimental protocols are using DNS PTR records for new purposes. The
6181 keys for these records are domain names, not reversed IP addresses. The
6182 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6183 leaves it alone. Component reversal etc. now happens only for IP addresses.
6186 34. The host_aton() function is supposed to be passed a string that is known
6187 to be a valid IP address. However, in the case of IPv6 addresses, it was
6188 not checking this. This is a hostage to fortune. Exim now panics and dies
6189 if the condition is not met. A case was found where this could be provoked
6190 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6191 components; fortuitously, this particular loophole had already been fixed
6192 by change 4.50/55 or 4.44/33 above.
6194 If there are any other similar loopholes, the new check in host_aton()
6195 itself should stop them being exploited. The report I received stated that
6196 data on the command line could provoke the exploit when Exim was running as
6197 exim, but did not say which command line option was involved. All I could
6198 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6199 running as the user.
6202 35. There was a buffer overflow vulnerability in the SPA authentication code
6203 (which came originally from the Samba project). I have added a test to the
6204 spa_base64_to_bits() function which I hope fixes it.
6207 36. The daemon start-up calls getloadavg() while still root for those OS that
6208 need the first call to be done as root, but it missed one case: when
6209 deliver_queue_load_max is set with deliver_drop_privilege. This is
6210 necessary for the benefit of the queue runner, because there is no re-exec
6211 when deliver_drop_privilege is set.
6213 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6214 was in use, was not putting the data itself into the right store pool;
6215 consequently, it could be overwritten for a subsequent message in the same
6216 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6219 38. Sometimes the final signoff response after QUIT could fail to get
6220 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6221 < 0 before doing a fflush(). This bug looks as though it goes back to the
6222 introduction of TLS in release 3.20, but "sometimes" must have been rare
6223 because the tests only now provoked it.
6229 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6230 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6231 Now it returns an integer. A number of calls were still expecting a T/F
6232 return. Fortuitously, in all cases, the tests worked in OK situations,
6233 which is the norm. However, things would have gone wrong on any write
6234 failures on the smtp file descriptor. This function is used when sending
6235 messages over SMTP and also when doing verify callouts.
6237 2. When Exim is called to do synchronous delivery of a locally submitted
6238 message (the -odf or -odi options), it no longer closes stderr before doing
6241 3. Implemented the mua_wrapper option.
6243 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6245 5. Implemented the functions header_remove(), header_testname(),
6246 header_add_at_position(), and receive_remove_recipient(), and exported them
6249 6. If an ACL "warn" statement specified the addition of headers, Exim already
6250 inserted X-ACL-Warn: at the start if there was no header name. However, it
6251 was not making this test for the second and subsequent header lines if
6252 there were newlines in the string. This meant that an invalid header could
6253 be inserted if Exim was badly configured.
6255 7. Allow an ACL "warn" statement to add header lines at the start or after all
6256 the Received: headers, as well as at the end.
6258 8. Added the rcpt_4xx retry error code.
6260 9. Added postmaster_mailfrom=xxx to callout verification option.
6262 10. Added mailfrom=xxxx to the callout verification option, for verify=
6265 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6266 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6268 12. Inserted some casts to stop certain compilers warning when using pointer
6269 differences as field lengths or precisions in printf-type calls (mostly
6270 affecting debugging statements).
6272 13. Added optional readline() support for -be (dynamically loaded).
6274 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6275 same clock tick as a message's arrival, so that its received time was the
6276 same as the "first fail" time on the retry record, and that message
6277 remained on the queue past the ultimate address timeout, every queue runner
6278 would try a delivery (because it was past the ultimate address timeout) but
6279 after another failure, the ultimate address timeout, which should have then
6280 bounced the address, did not kick in. This was a "< instead of <=" error;
6281 in most cases the first failure would have been in the next clock tick
6282 after the received time, and all would be well.
6284 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6285 being recognized when the domain list was tested by the match_domain
6286 condition in an expansion string.
6288 16. Added the ${str2b64: operator.
6290 17. Exim was always calling setrlimit() to set a large limit for the number of
6291 processes, without checking whether the existing limit was already
6292 adequate. (It did check for the limit on file descriptors.) Furthermore,
6293 errors from getrlimit() and setrlimit() were being ignored. Now they are
6294 logged to the main and panic logs, but Exim does carry on, to try to do its
6295 job under whatever limits there are.
6297 18. Imported PCRE 5.0.
6299 19. Trivial typo in log message " temporarily refused connection" (the leading
6302 20. If the log selector return_path_on_delivery was set and an address was
6303 redirected to /dev/null, the delivery process crashed because it assumed
6304 that a return path would always be set for a "successful" delivery. In this
6305 case, the whole delivery is bypassed as an optimization, and therefore no
6308 21. Internal re-arrangement: the function for sending a challenge and reading
6309 a response while authentication was assuming a zero-terminated challenge
6310 string. It's now changed to take a pointer and a length, to allow for
6311 binary data in such strings.
6313 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6315 23. Exim was not respecting finduser_retries when seeking the login of the
6316 uid under which it was called; it was always trying 10 times. (The default
6317 setting of finduser_retries is zero.) Also, it was sleeping after the final
6318 failure, which is pointless.
6320 24. Implemented tls_on_connect_ports.
6322 25. Implemented acl_smtp_predata.
6324 26. If the domain in control=submission is set empty, Exim assumes that the
6325 authenticated id is a complete email address when it generates From: or
6326 Sender: header lines.
6328 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6329 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6330 chown and chgrp in /bin and hostname in /usr/bin.
6332 28. Exim was keeping the "process log" file open after each use, just as it
6333 does for the main log. This opens the possibility of it remaining open for
6334 long periods when the USR1 signal hits a daemon. Occasional processlog
6335 errors were reported, that could have been caused by this. Anyway, it seems
6336 much more sensible not to leave this file open at all, so that is what now
6339 29. The long-running daemon process does not normally write to the log once it
6340 has entered its main loop, and it closes the log before doing so. This is
6341 so that log files can straightforwardly be renamed and moved. However,
6342 there are a couple of unusual error situations where the daemon does write
6343 log entries, and I had neglected to close the log afterwards.
6345 30. The text of an SMTP error response that was received during a remote
6346 delivery was being truncated at 512 bytes. This is too short for some of
6347 the long messages that one sometimes sees. I've increased the limit to
6350 31. It is now possible to make retry rules that apply only when a message has a
6351 specific sender, in particular, an empty sender.
6353 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6354 it possible to be selective about when SMTP synchronization is enforced.
6356 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6358 32. Implemented hosts_connection_nolog.
6360 33. Added an ACL for QUIT.
6362 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6365 35. Added mailbox_size and mailbox_filecount to appendfile.
6367 36. Added control = no_multiline_responses to ACLs.
6369 37. There was a bug in the logic of the code that waits for the clock to tick
6370 in the case where the clock went backwards by a substantial amount such
6371 that the microsecond fraction of "now" was more than the microsecond
6372 fraction of "then" (but the whole seconds number was less).
6374 38. Added support for the libradius Radius client library this is found on
6375 FreeBSD (previously only the radiusclient library was supported).
6381 1. When certain lookups returned multiple values in the form name=value, the
6382 quoting of the values was not always being done properly. Specifically:
6383 (a) If the value started with a double quote, but contained no whitespace,
6385 (b) If the value contained whitespace other than a space character (i.e.
6386 tabs or newlines or carriage returns) it was not quoted.
6387 This fix has been applied to the mysql and pgsql lookups by writing a
6388 separate quoting function and calling it from the lookup code. The fix
6389 should probably also be applied to nisplus, ibase and oracle lookups, but
6390 since I cannot test any of those, I have not disturbed their existing code.
6392 2. A hit in the callout cache for a specific address caused a log line with no
6393 reason for rejecting RCPT. Now it says "Previous (cached) callout
6394 verification failure".
6396 3. There was an off-by-one bug in the queryprogram router. An over-long
6397 return line was truncated at 256 instead of 255 characters, thereby
6398 overflowing its buffer with the terminating zero. As well as fixing this, I
6399 have increased the buffer size to 1024 (and made a note to document this).
6401 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6402 when Exim is waiting for an SMTP response from a remote server, Exim
6403 restarts its select() call on the socket, thereby resetting its timeout.
6404 This is not a problem when such interrupts are rare. Somebody set up a cron
6405 job to run exiwhat every 2 minutes, which is less than the normal select()
6406 timeout (5 or 10 minutes). This meant that the select() timeout never
6407 kicked in because it was always reset. I have fixed this by comparing the
6408 time when an interrupt arrives with the time at the start of the first call
6409 to select(). If more time than the timeout has elapsed, the interrupt is
6410 treated as a timeout.
6412 5. Some internal re-factoring in preparation for the addition of Sieve
6413 extensions (by MH). In particular, the "personal" test is moved to a
6414 separate function, and given an option for scanning Cc: and Bcc: (which is
6415 not set for Exim filters).
6417 6. When Exim created an email address using the login of the caller as the
6418 local part (e.g. when creating a From: or Sender: header line), it was not
6419 quoting the local part when it contained special characters such as @.
6421 7. Installed new OpenBSD configuration files.
6423 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6424 try to make them clearer.
6426 9. Callout options, other than the timeout value, were being ignored when
6427 verifying sender addresses in header lines. For example, when using
6429 verify = header_sender/callout=no_cache
6431 the cache was (incorrectly) being used.
6433 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6434 only the Cygwin environment.
6436 11. When return_path_on_delivery was set as a log selector, if different remote
6437 addresses in the same message used different return paths and parallel
6438 remote delivery occurred, the wrong values would sometimes be logged.
6439 (Whenever a remote delivery process finished, the return path value from
6440 the most recently started remote delivery process was logged.)
6442 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6443 lines when AUTH and/or TLS are in use. This is the "received protocol"
6444 field. Exim used to use "asmtp" for authenticated SMTP, without any
6445 indication (in the protocol name) for TLS use. Now it follows the RFC and
6446 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6447 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6448 names appear in log lines as well as in Received: header lines.
6450 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6451 extensions, and comparison tests, and to fix some bugs.
6453 14. Changes to the "personal" filter test:
6455 (1) The test was buggy in that it was just doing the equivalent of
6456 "contains" tests on header lines. For example, if a user's address was
6457 anne@some.where, the "personal" test would incorrectly be true for
6459 To: susanne@some.where
6461 This test is now done by extracting each address from the header in turn,
6462 and checking the entire address. Other tests that are part of "personal"
6463 are now done using regular expressions (for example, to check local parts
6464 of addresses in From: header lines).
6466 (2) The list of non-personal local parts in From: addresses has been
6467 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6468 taken from the Sieve specification recommendations.
6470 (3) If the message contains any header line starting with "List-" it is
6471 treated as non-personal.
6473 (4) The test for "circular" in the Subject: header line has been removed
6474 because it now seems ill-conceived.
6476 15. Minor typos in src/EDITME comments corrected.
6478 16. Installed latest exipick from John Jetmore.
6480 17. If headers_add on a router specified a text string that was too long for
6481 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6482 of string_sprintf() is now avoided.
6484 18. $message_body_size was not set (it was always zero) when running the DATA
6485 ACL and the local_scan() function.
6487 19. For the "mail" command in an Exim filter, no default was being set for
6488 the once_repeat time, causing a random time value to be used if "once" was
6489 specified. (If the value happened to be <= 0, no repeat happened.) The
6490 default is now 0s, meaning "never repeat". The "vacation" command was OK
6491 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6492 (I found it when inspecting the code).
6494 20. There is now an overall timeout for performing a callout verification. It
6495 defaults to 4 times the callout timeout, which applies to individual SMTP
6496 commands during the callout. The overall timeout applies when there is more
6497 than one host that can be tried. The timeout is checked before trying the
6498 next host. This prevents very long delays if there are a large number of
6499 hosts and all are timing out (e.g. when the network connections are timing
6500 out). The value of the overall timeout can be changed by specifying an
6501 additional sub-option for "callout", called "maxwait". For example:
6503 verify = sender/callout=5s,maxwait=20s
6505 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6506 to the end before writing, but this should make it even safer).
6508 22. Exim was forgetting that it had advertised PIPELINING for the second and
6509 subsequent messages on an SMTP connection. It was also not resetting its
6510 memory on STARTTLS and an internal HELO.
6512 23. When Exim logs an SMTP synchronization error within a session, it now
6513 records whether PIPELINING has been advertised or not.
6515 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6516 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6517 rather than long int.
6519 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6521 26. Added the never_mail option to autoreply.
6527 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6528 crash if the getsockname() call failed; this can happen if a connection is
6529 closed very soon after it is established. The problem was simply in the
6530 order in which certain operations were done, causing Exim to try to write
6531 to the SMTP stream before it had set up the file descriptor. The bug has
6532 been fixed by making things happen in the correct order.
6538 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6539 before the connection was closed, thus losing the rejection response.
6541 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6542 some early Solaris releases, but causes trouble in current releases where
6543 socklen_t is defined.
6545 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6548 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6551 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6552 that releasing the top of it at the end releases what was used for sub-
6553 expansions (unless the block got too big). However, discard this block if
6554 the first thing is a variable or header, so that we can use its block when
6555 it is dynamic (useful for very large $message_headers, for example).
6557 6. Lookups now cache *every* query, not just the most recent. A new, separate
6558 store pool is used for this. It can be recovered when all lookup caches are
6559 flushed. Lookups now release memory at the end of their result strings.
6560 This has involved some general refactoring of the lookup sources.
6562 7. Some code has been added to the store_xxx() functions to reduce the amount
6563 of flapping under certain conditions.
6565 8. log_incoming_interface used to affect only the <= reception log lines. Now
6566 it causes the local interface and port to be added to several more SMTP log
6567 lines, for example "SMTP connection from", and rejection lines.
6569 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6571 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6573 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6574 could be overwritten at the end of the current message (or the start of a
6575 new message if it was set in a HELO ACL). The value is now preserved for
6576 the duration of the SMTP connection.
6578 12. If a transport had a headers_rewrite setting, and a matching header line
6579 contained an unqualified address, that address was qualified, even if it
6580 did not match any rewriting rules. The underlying bug was that the values
6581 of the flags that permit the existence of unqualified sender and recipient
6582 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6583 non-local messages, and by -bnq for local messages) were not being
6584 preserved with the message after it was received.
6586 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6587 "next input=" as part of the text comprising the host identity instead of
6588 the correct text. The code was using the same buffer for two different
6589 strings. However, depending on which order the printing function evaluated
6590 its arguments, the bug did not always show up. Under Linux, for example, my
6591 test suite worked just fine.
6593 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6594 doesn't work with some older versions of Perl. It has been changed to "my",
6595 which in any case is probably the better facility to use.
6597 15. A really picky compiler found some instances of statements for creating
6598 error messages that either had too many or two few arguments for the format
6601 16. The size of the buffer for calls to the DNS resolver has been increased
6602 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6603 for addresses that have a lot of PTR records. This alleviates a problem; it
6604 does not fully solve it.
6606 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6607 buffer now truncates the list and logs the incident, which is the same
6608 action as happens when Exim is looking up a host name and its aliases.
6609 Previously in this situation something unpredictable would happen;
6610 sometimes it was "internal error: store_reset failed".
6612 18. If a server dropped the connection unexpectedly when an Exim client was
6613 using GnuTLS and trying to read a response, the client delivery process
6614 crashed while trying to generate an error log message.
6616 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6617 string, for example:
6619 warn message = H1: something\nH2: something
6621 the text was added as a single header line from Exim's point of view
6622 though it ended up OK in the delivered message. However, searching for the
6623 second and subsequent header lines using $h_h2: did not work. This has been
6624 fixed. Similarly, if a system filter added multiple headers in this way,
6625 the routers could not see them.
6627 20. Expanded the error message when iplsearch is called with an invalid key to
6628 suggest using net-iplsearch in a host list.
6630 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6631 ACLs are no longer actually imposed (and a message to that effect is
6634 22. If a "gecos" field in a passwd entry contained escaped characters, in
6635 particular, if it contained a \" sequence, Exim got it wrong when building
6636 a From: or a Sender: header from that name. A second bug also caused
6637 incorrect handling when an unquoted " was present following a character
6638 that needed quoting.
6640 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6641 was not being matched caselessly.
6643 24. Arranged for all hyphens in the exim.8 source to be escaped with
6646 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6647 properly. Recipient callout cache records were still being keyed to include
6648 the sender, even when use_sender was set false. This led to far more
6649 callouts that were necessary. The sender is no longer included in the key
6650 when use_sender is false.
6652 26. Added "control = submission" modifier to ACLs.
6654 27. Added the ${base62d: operator to decode base 62 numbers.
6656 28. dnsdb lookups can now access SRV records.
6658 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6659 the configuration file.
6661 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6662 (-v) mode. This makes the output for a verbose queue run more intelligible.
6664 31. Added a use_postmaster feature to recipient callouts.
6666 32. Added the $body_zerocount variable, containing the number of binary zero
6667 bytes in the message body.
6669 33. The time of last modification of the "new" subdirectory is now used as the
6670 "mailbox time last read" when there is a quota error for a maildir
6673 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6675 35. Added +ignore_unknown as a special item in host lists.
6677 36. Code for decoding IPv6 addresses in host lists is now included, even if
6678 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6679 address was recognized as an IP address, but was then not correctly decoded
6680 into binary, causing unexpected and incorrect effects when compared with
6687 1. Very minor rewording of debugging text in manualroute to say "list of
6688 hosts" instead of "hostlist".
6690 2. If verify=header_syntax was set, and a header line with an unqualified
6691 address (no domain) and a large number of spaces between the end of the
6692 name and the colon was received, the reception process suffered a buffer
6693 overflow, and (when I tested it) crashed. This was caused by some obsolete
6694 code that should have been removed. The fix is to remove it!
6696 3. When running in the test harness, delay a bit after writing a bounce
6697 message to get a bit more predictability in the log output.
6699 4. Added a call to search_tidyup() just before forking a reception process. In
6700 theory, someone could use a lookup in the expansion of smtp_accept_max_
6701 per_host which, without the tidyup, could leave open a database connection.
6703 5. Added the variables $recipient_data and $sender_data which get set from a
6704 lookup success in an ACL "recipients" or "senders" condition, or a router
6705 "senders" option, similar to $domain_data and $local_part_data.
6707 6. Moved the writing of debug_print from before to after the "senders" test
6710 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6711 problems for message scanning, either using a data ACL, or using
6712 local_scan() because the Received: header was not generated till after they
6713 were called (in order to set the time as the time of reception completion).
6714 I have revised the way this works. The header is now generated after the
6715 body is received, but before the ACL or local_scan() are called. After they
6716 are run, the timestamp in the header is updated.
6722 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6723 before starting a queue runner without re-exec. This happened only when
6724 deliver_drop_privilege was set or when the Exim user was set to root. The
6725 effect of the bug was that timeouts during subsequent deliveries caused
6726 crashes instead of being properly handled. The handler is now left at its
6727 default (and expected) setting.
6729 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6730 message, again when deliver_drop_privilege is set or Exim is run as root.
6731 The bug described in (1) was not present in this case, but the tidying up
6732 of the other signals was missing. I have made the two cases consistent.
6734 3. The ignore_target_hosts setting on a manualroute router was being ignored
6735 for hosts that were looked up using the /MX notation.
6737 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6740 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6741 operated on the sender address. After changing the $sender_address to <>
6742 for the sender address verify, Exim was re-instated it as the original
6743 (before rewriting) address, but remembering that it had rewritten it, so it
6744 wasn't rewriting it again. This bug also had the effect of breaking the
6745 sender address verification caching when the sender address was rewritten.
6747 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6748 This has been changed so that if the ip literal address matches
6749 ignore_target_hosts, the router declines.
6751 7. Added expansion conditions match_domain, match_address, and match_local_
6752 part (NOT match_host).
6754 8. The placeholder for the Received: header didn't have a length field set.
6756 9. Added code to Exim itself and to exim_lock to test for a specific race
6757 condition that could lead to file corruption when using MBX delivery. The
6758 issue is with the lockfile that is created in /tmp. If this file is removed
6759 after a process has opened it but before that process has acquired a lock,
6760 there is the potential for a second process to recreate the file and also
6761 acquire a lock. This could lead to two Exim processes writing to the file
6762 at the same time. The added code performs the same test as UW imapd; it
6763 checks after acquiring the lock that its file descriptor still refers to
6764 the same named file.
6766 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6767 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6770 11. Added the smtp_active_hostname option. If used, this will typically be made
6771 to depend on the incoming interface address. Because $interface_address is
6772 not set up until the daemon has forked a reception process, error responses
6773 that can happen earlier (such as "too many connections") no longer contain
6776 12. If an expansion in a condition on a "warn" statement fails because a lookup
6777 defers, the "warn" statement is abandoned, and the next ACL statement is
6778 processed. Previously this caused the whole ACL to be aborted.
6780 13. Added the iplsearch lookup type.
6782 14. Added ident_timeout as a log selector.
6784 15. Added tls_certificate_verified as a log selector.
6786 16. Added a global option tls_require_ciphers (compare the smtp transport
6787 option of the same name). This controls incoming TLS connections.
6789 17. I finally figured out how to make tls_require_ciphers do a similar thing
6790 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6791 before starting the TLS session.
6793 18. Tabs are now shown as \t in -bP output.
6795 19. If the log selector return_path_on_delivery was set, Exim crashed when
6796 bouncing a message because it had too many Received: header lines.
6798 20. If two routers both had headers_remove settings, and the first one included
6799 a superfluous trailing colon, the final name in the first list and the
6800 first name in the second list were incorrectly joined into one item (with a
6801 colon in the middle).
6807 1. Added -C and -D options to the exinext utility, mainly to make it easier
6808 to include in the automated testing, but these could be helpful when
6809 multiple configurations are in use.
6811 2. The exinext utility was not formatting the output nicely when there was
6812 an alternate port involved in the retry record key, nor when there was a
6813 message id as well (for retries that were specific to a specific message
6814 and a specific host). It was also confused by IPv6 addresses, because of
6815 the additional colons they contain. I have fixed the IPv4 problem, and
6816 patched it up to do a reasonable job for IPv6.
6818 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6819 delivery, the log line now contains "pipelined" if PIPELINING was used.
6821 4. An SMTP transport process used to panic and die if the bind() call to set
6822 an explicit outgoing interface failed. This has been changed; it is now
6823 treated in the same way as a connect() failure.
6825 5. A reference to $sender_host_name in the part of a conditional expansion
6826 that was being skipped was still causing a DNS lookup. This no longer
6829 6. The def: expansion condition was not recognizing references to header lines
6830 that used bh_ and bheader_.
6832 7. Added the _cache feature to named lists.
6834 8. The code for checking quota_filecount in the appendfile transport was
6835 allowing one more file than it should have been.
6837 9. For compatibility with Sendmail, the command line option
6845 and sets the incoming protocol and host name (for trusted callers). The
6846 host name and its colon can be omitted when only the protocol is to be set.
6847 Note the Exim already has two private options, -pd and -ps, that refer to
6848 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6849 "s", but I don't think that's a major issue.
6851 10. A number of refactoring changes to the code, none of which should affect
6854 (a) The number of logging options was getting close to filling up the
6855 32-bit word that was used as a bit map. I have split them into two classes:
6856 those that are passed in the argument to log_write(), and those that are
6857 only ever tested independently outside of that function. These are now in
6858 separate 32-bit words, so there is plenty of room for expansion again.
6859 There is no change in the user interface or the logging behaviour.
6861 (b) When building, for example, log lines, the code previously used a
6862 macro that called string_cat() twice, in order to add two strings. This is
6863 not really sufficiently general. Furthermore, there was one instance where
6864 it was actually wrong because one of the argument was used twice, and in
6865 one call a function was used. (As it happened, calling the function twice
6866 did not affect the overall behaviour.) The macro has been replaced by a
6867 function that can join an arbitrary number of extra strings onto a growing
6870 (c) The code for expansion conditions now uses a table and a binary chop
6871 instead of a serial search (which was left over from when there were very
6872 few conditions). Also, it now recognizes conditions like "pam" even when
6873 the relevant support is not compiled in: a suitably worded error message is
6874 given if an attempt is made to use such a condition.
6876 11. Added ${time_interval:xxxxx}.
6878 12. A bug was causing one of the ddress fields not to be passed back correctly
6879 from remote delivery subprocesses. The field in question was not being
6880 subsequently used, so this caused to problems in practice.
6882 13. Added new log selectors queue_time and deliver_time.
6884 14. Might have fixed a bug in maildirsizefile handling that threw up
6885 "unexpected character" debug warnings, and recalculated the data
6886 unnecessarily. In any case, I expanded the warning message to give more
6889 15. Added the message "Restricted characters in address" to the statements in
6890 the default ACL that block characters like @ and % in local parts.
6892 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6893 Three changes have been made:
6895 (a) There was a serious bug; a negative response to MAIL caused the whole
6896 recipient domain to be cached as invalid, thereby blocking all messages
6897 to all local parts at the same domain, from all senders. This bug has
6898 been fixed. The domain is no longer cached after a negative response to
6899 MAIL if the sender used is not empty.
6901 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6904 (c) A new callout option, "use_sender" has been added for people who want
6905 the modified behaviour.
6911 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6914 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6915 indeed breaks things for older releases.
6917 3. Added additional logging to the case where there is a problem reading data
6918 from a filter that is running in a subprocess using a pipe, in order to
6919 try to track down a specific problem.
6921 4. Testing facility fudge: when running in the test harness and attempting
6922 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6923 getting "No route to host". Convert this to a timeout.
6925 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6928 6. Some OS don't have socklen_t but use size_t instead. This affects the
6929 fifth argument of getsockopt() amongst other things. This is now
6930 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6931 can be set for individual OS. I have set it for SunOS5, OSF1, and
6932 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6933 some earlier ones do not.
6935 7. Change 4.30/15 was not doing the test caselessly.
6937 8. The standard form for an IPv6 address literal was being rejected by address
6938 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6939 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6940 this, as well as the form without the "IPv6" on the front (but only when
6941 address literals are enabled, of course).
6943 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6945 10. Exim crashed if a message with an empty sender address specified by -f
6946 encountered a router with an errors_to setting. This could be provoked only
6947 by a command such as
6951 where an empty string was supplied; "<>" did not hit this bug.
6953 11. Installed PCRE release 4.5.
6955 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6956 remained set. It is now erased.
6958 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6959 times from message ids (which are base 36 rather than the normal 62).
6961 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6962 were being counted as actual protocol errors, and logged if the log
6963 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6964 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6965 and DATA following a set of rejected RCPTs do not count as protocol errors.
6966 In other words, Exim assumes they were pipelined, though this may not
6967 actually be the case. Of course, in all cases the client gets an
6968 appropriate error code.
6970 15. If a lookup fails in an ACL condition, a message about the failure may
6971 be available; it is used if testing the ACL cannot continue, because most
6972 such messages specify what the cause of the deferral is. However, some
6973 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6974 that caused an old message to be retained and used if a later statement
6975 caused a defer, replacing the real cause of the deferral.
6977 16. If an IP address had so many PTR records that the DNS lookup buffer
6978 was not large enough to hold them, Exim could crash while trying to process
6979 the truncated data. It now detects and logs this case.
6981 17. Further to 4.21/58, another change has been made: if (and only if) the
6982 first line of a message (the first header line) ends with CRLF, a bare LF
6983 in a subsequent header line has a space inserted after it, so as not to
6984 terminate the header.
6986 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6987 unnecessarily, used atoi() instead of strtol(), and didn't check the
6988 termination when getting file sizes from file names by regex.
6990 19. Completely re-implemented the support for maildirsize files, in the light
6991 of a number of problems with the previous contributed implementation
6992 (4.30/29). In particular:
6994 . If the quota is zero, the maildirsize file is maintained, but no quota is
6997 . If the maildir directory does not exist, it is created before any attempt
6998 to write a maildirsize file.
7000 . The quota value in the file is just a cache; if the quota is changed in
7001 the transport, the new value overrides.
7003 . A regular expression is available for excluding directories from the
7006 20. The autoreply transport checks the characters in options that define the
7007 message's headers; it allows continued headers, but it was checking with
7008 isspace() after an embedded newline instead of explicitly looking for a
7011 21. If all the "regular" hosts to which an address was routed had passed their
7012 expiry times, and had not reached their retry times, the address was
7013 bounced, even if fallback hosts were defined. Now Exim should go on to try
7016 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7017 equivalent code in the SMTP transport. Some hosts send humungous responses
7018 to HELO/EHLO, more than 1024 it seems.
7020 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7021 gives compiler warnings in some environments. I've now done it "properly",
7024 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7025 (because of problems with the built-in one) was declared to return uschar *
7026 instead of char *, causing compiler failure.
7028 25. Fixed a file descriptor leak when processing alias/forward files.
7030 26. Fixed a minor format string issue in dbfn.c.
7032 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7034 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7035 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7038 29. When a sender address is verified, it is cached, to save repeating the test
7039 when there is more than one recipient in a message. However, when the
7040 verification involves a callout, it is possible for different callout
7041 options to be set for different recipients. It is too complicated to keep
7042 track of this in the cache, so now Exim always runs a verification when a
7043 callout is required, relying on the callout cache for the optimization.
7044 The overhead is duplication of the address routing, but this should not be
7047 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7048 to be verified with callout=postmaster, and the main callout worked but the
7049 postmaster check failed, the verification correctly failed. However, if a
7050 subsequent RCPT command asked for sender verification *without* the
7051 postmaster check, incorrect caching caused this verification also to fail,
7054 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7055 it was not caching the DNS options (qualify_single, search_parents) that
7056 were used when the lookup failed. A subsequent lookup with different
7057 options therefore always gave the same answer, though there were cases
7058 where it should not have. (Example: a "domains = !$mx_any" option on a
7059 dnslookup router: the "domains" option is always processed without any
7060 widening, but the router might have qualify_single set.) Now Exim uses the
7061 cached value only when the same options are set.
7063 32. Added John Jetmore's "exipick" utility to the distribution.
7065 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7066 than a timeout (e.g. a certificate is required, and is not provided), an
7067 Exim server now closes the connection immediately. Previously it waited for
7068 the client to close - but if the client is SSL, it seems that they each
7069 wait for each other, leading to a delay before one of them times out.
7071 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7072 maintained 0.8.x compatibility because I don't think many are using it, and
7073 it is clearly obsolete.
7075 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7078 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7079 client certificate was expired. A simple patch fixes this, though I don't
7080 understand the full logic of why the verify callback is called multiple
7083 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7084 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7085 which causes problems with some clients (such as the Certicom SSL Plus
7086 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7087 disables the coutermeasure allowing Eudora to connect."
7089 38. Exim was not checking that a write() to a log file succeeded. This could
7090 lead to Bad Things if a log got too big, in particular if it hit a file
7091 size limit. Exim now panics and dies if it cannot write to a log file, just
7092 as it does if it cannot open a log file.
7094 39. Modified OS/Makefile-Linux so that it now contains
7096 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7098 The two -D definitions ensure that Exim is compiled with large file
7099 support, which makes it possible to handle log files that are bigger than
7102 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7103 instance) a domain was checked against a named list that involved a lookup,
7104 causing $domain_data to be set, then another domain was checked against the
7105 same list, then the first domain was re-checked, the value of $domain_data
7106 after the final check could be wrong. In particular, if the second check
7107 failed, it could be set empty. This bug probably also applied to
7110 41. The strip_trailing_dot option was not being applied to the address given
7111 with the -f command-line option.
7113 42. The code for reading a message's header from the spool was incrementing
7114 $received_count, but never initializing it. This meant that the value was
7115 incorrect (doubled) while delivering a message in the same process in which
7116 it was received. In the most common configuration of Exim, this never
7117 happens - a fresh exec is done - but it can happen when
7118 deliver_drop_privilege is set.
7120 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7121 it now includes up to 150 characters of the unexpected data in the log
7124 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7125 and building data strings. The size of both of these buffers was 10 000
7126 bytes - far larger than anybody would *ever* want, thought I. Needless to
7127 say, somebody hit the limit. I have increased the maximum line length to
7128 20 000 and the maximum data length of concatenated lines to 100 000. I have
7129 also fixed two bugs, because there was no checking on these buffers. Tsk,
7130 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7131 buffer is too small.
7133 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7134 lsearch lookups. Now it does.
7136 46. When parsing a route_list item in a manualroute router, a fixed-length
7137 buffer was used for the list of hosts. I made this 1024 bytes long,
7138 thinking that nobody would ever have a list of hosts that long. Wrong.
7139 Somebody had a whole pile of complicated expansion conditions, and the
7140 string was silently truncated, leading to an expansion error. It turns out
7141 that it is easier to change to an unlimited length (owing to other changes
7142 that have happened since this code was originally written) than to build
7143 structure for giving a limitation error. The length of the item that
7144 expands into the list of hosts is now unlimited.
7146 47. The lsearch lookup could not handle data where the length of text line was
7147 more than 4095 characters. Such lines were truncated, leading to shortened
7148 data being returned. It should now handle lines of any length.
7150 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7151 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7154 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7155 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7156 no longer happens in comment lines. A list of replacements is now placed
7157 at the head of all of the source files, except those whose only change is
7158 to replace PERL_COMMAND in the very first #! line.
7160 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7161 messages on the queue, with a bottom-up merge sort, using code contributed
7162 by Michael Haardt. This should make operations like -bp somewhat faster on
7163 large queues. It won't affect queue runners, except when queue_run_in_order
7166 51. Installed eximstats 1.31 in the distribution.
7168 52. Added support for SRV lookups to the dnslookup router.
7170 53. If an ACL referred to $message_body or $message_body_end, the value was not
7171 reset for any messages that followed in the same SMTP session.
7173 54. The store-handling optimization for building very long strings was not
7174 differentiating between the different store pools. I don't think this
7175 actually made any difference in practice, but I've tidied it.
7177 55. While running the routers to verify a sender address, $sender_address
7178 was still set to the sender address. This is wrong, because when routing to
7179 send a bounce to the sender, it would be empty. Therefore, I have changed
7180 it so that, while verifying a sender address, $sender_address is set to <>.
7181 (There is no change to what happens when verifying a recipient address.)
7183 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7184 target A or AAAA records (if not already returned) without resetting the
7185 qualify_single or search_parents options of the DNS resolver. These are
7186 inappropriate in this case because the targets of MX and SRV records must
7187 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7188 target that, when qualified, matched something in the local domain. These
7189 two options are now turned off when doing these lookups.
7191 57. It seems that at least some releases of Reiserfs (which does not have the
7192 concept of a fixed number of inodes) returns zero and not -1 for the
7193 number of available inodes. This interacted badly with check_spool_inodes,
7194 which assumed that -1 was the "no such thing" setting. What I have done is
7195 to check that the total number of inodes is greater than zero before doing
7196 the test of how many are available.
7198 58. When a "warn" ACL statement has a log_message modifier, the message is
7199 remembered, and not repeated. This is to avoid a lot of repetition when a
7200 message has many recipients that cause the same warning to be written.
7201 However, Exim was preserving the list of already written lines for an
7202 entire SMTP session, which doesn't seem right. The memory is now reset if a
7203 new message is started.
7205 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7206 debugging output unless log_rewrite was also set.
7208 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7209 of (int)(handle) when we know that handle contains (void *)(-1).
7211 61. The Exim daemon panic-logs an error return when it closes the incoming
7212 connection. However "connection reset by peer" seems to be common, and
7213 isn't really an error worthy of noting specially, so that particular error
7216 62. When Exim is trying to find all the local interfaces, it used to panic and
7217 die if the ioctl to get the interface flags failed. However, it seems that
7218 on at least one OS (Solaris 9) it is possible to have an interface that is
7219 included in the list of interfaces, but for which you get a failure error
7220 for this call. This happens when the interface is not "plumbed" into a
7221 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7222 failure of the "get flags" call assumes that the interface is down.
7224 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7225 makes life easier for people who are doing arithmetic on fields extracted
7226 from dates, where you often get leading zeros that should not be
7227 interpreted as octal.
7229 64. Added qualify_domain to the redirect router, to override the global
7232 65. If a pathologically long header line contained very many addresses (the
7233 report of this problem mentioned 10 000) and each of them was rewritten,
7234 Exim could use up a very large amount of memory. (It kept on making new
7235 copies of the header line as it rewrote, and never released the old ones.)
7236 At the expense of a bit more processing, the header rewriting function has
7237 been changed so that it no longer eats memory in this way.
7239 66. The generation of the Received: header has been moved from the time that a
7240 message starts to be received, to the time that it finishes. The timestamp
7241 in the Received: header should now be very close to that of the <= log
7242 line. There are two side-effects of this change:
7244 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7245 logged header lines no longer include the local Received: line, because
7246 it has not yet been created. The same applies to a copy of the message
7247 that is returned to a non-SMTP sender when a message is rejected.
7249 (b) When a filter file is tested using -bf, no additional Received: header
7250 is added to the test message. After some thought, I decided that this
7253 This change does not affect the value of $received_for. It is still set
7254 after address rewriting, but before local_scan() is called.
7256 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7258 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7259 gave an unhelpful panic error message, and a defer error. I have managed to
7260 change this behaviour so that it now rejects any supplied certificate,
7261 which seems right, as the list of acceptable certificates is empty.
7263 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7264 gave an unhelpful defer error. I have not managed to make this reject any
7265 supplied certificates, but the error message it gives is "no certificate
7266 supplied", which is not helpful.
7268 70. exigrep's output now also includes lines that are not associated with any
7269 message, but which match the given pattern. Implemented by a patch from
7270 Martin Sluka, which also tidied up the Perl a bit.
7272 71. Recipient callout verification, like sender verification, was using <> in
7273 the MAIL FROM command. This isn't really the right thing, since the actual
7274 sender may affect whether the remote host accepts the recipient or not. I
7275 have changed it to use the actual sender in the callout; this means that
7276 the cache record is now keyed on a recipient/sender pair, not just the
7277 recipient address. There doesn't seem to be a real danger of callout loops,
7278 since a callout by the remote host to check the sender would use <>.
7279 [SEE ABOVE: changed after hitting problems.]
7281 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7282 temporary errors. However, in the case of such a code being given after
7283 the end of a data transmission (i.e. after ".") Exim was failing to write
7284 a retry record for the message. (Yes, there was some broken host that was
7285 actually sending 8xx at this point.)
7287 73. An unknown lookup type in a host list could cause Exim to panic-die when
7288 the list was checked. (An example that provoked this was putting <; in the
7289 middle of a list instead of at the start.) If this happened during a DATA
7290 ACL check, a -D file could be left lying around. This kind of configuration
7291 error no longer causes Exim to die; instead it causes a defer error. The
7292 incident is still logged to the main and panic logs.
7294 74. Buglet left over from Exim 3 conversion. The message "too many messages
7295 in one connection" was written to the rejectlog but not the mainlog, except
7296 when address rewriting (yes!) was being logged.
7298 75. Added write_rejectlog option.
7300 76. When a system filter was run not as root (that is, when system_filter_user
7301 was set), the values of the $n variables were not being returned to the
7302 main process; thus, they were not subsequently available in the $sn
7305 77. Added +return_path_on_delivery log selector.
7307 78. A connection timeout was being treated differently from recipients deferred
7308 when testing hosts_max_try with a message that was older than the host's
7309 retry timeout. (The host should not be counted, thus allowing all hosts to
7310 be tried at least once before bouncing.) This may have been the cause of an
7311 occasionally reported bug whereby a message would remain on the queue
7312 longer than the retry timeout, but would be bounced if a delivery was
7313 forced. I say "may" because I never totally pinned down the problem;
7314 setting up timeout/retry tests is difficult. See also the next item.
7316 79. The ultimate address timeout was not being applied to errors that involved
7317 a combination of host plus message (for example, a timeout on a MAIL
7318 command). When an address resolved to a number of possible hosts, and they
7319 were not all tried for each delivery (e.g. because of hosts_max_try), a
7320 message could remain on the queue longer than the retry timeout.
7322 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7325 81. Fixed an obscure SMTP outgoing bug which required at least the following
7326 conditions: (a) there was another message waiting for the same server;
7327 (b) the server returned 5xx to all RCPT commands in the first message so
7328 that the message was not completed; (c) the server dropped the connection
7329 or gave a negative response to the RSET that Exim sends to abort the
7330 transaction. The observed case was a dropped connection after DATA that had
7331 been sent in pipelining mode. That is, the server had advertised PIPELINING
7332 but was not implementing it correctly. The effect of the bug was incorrect
7333 behaviour, such as trying another host, and this could lead to a crash.
7339 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7340 and daemon.c were passed as pointers to ints; they should have been
7341 pointers to socklen_t variables (which are typically unsigned ints).
7343 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7346 3. Fixed a really odd bug that affected only the testing scheme; patching a
7347 certain fixed string in the binary changed the value of another string that
7348 happened to be identical to the end of the original first string.
7350 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7351 name", it returns that address as the IP address. On some operating
7352 systems (e.g. Solaris), it also passes back the IP address string as the
7353 "host name". However, on others (e.g. Linux), it passes back an empty
7354 string. Exim wasn't checking for this, and was changing the host name to an
7355 empty string, assuming it had been canonicalized.
7357 5. Although rare, it is permitted to have more than one PTR record for a given
7358 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7359 all the names associated with an address, because they do in Solaris.
7360 However, it seems that they do not in Linux for data that comes from the
7361 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7362 I found this out when I moved to a new Linux workstation and tried to run
7363 the Exim test suite.
7365 To get round this problem I have changed the code so that it now does its
7366 own call to the DNS to look up PTR records when searching for a host name.
7367 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7368 addresses that are only in /etc/hosts are still found.
7370 This behaviour is, however, controlled by an option called host_lookup_
7371 order, which defaults to "bydns:byaddr". If people want to use the other
7372 order, or indeed, just use one or the other means of lookup, they can
7373 specify it in this variable.
7375 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7376 some operating systems, this comes back from gethostbyaddr() as an empty
7377 string, and this is what Exim used to test for. However, it seems that in
7378 other systems, "." is yielded. Exim now tests for this case too.
7380 7. The values of check_spool_space and check_log_space are now held internally
7381 as a number of kilobytes instead of an absolute number of bytes. If a
7382 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7383 kilobyte. This means that much larger values can be stored.
7385 8. Exim monitor: an attempt to get the action menu when not actually pointing
7386 at a message produces an empty menu entitled "No message selected". This
7387 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7388 no entries in it ("Shell widget menu has zero width and/or height"). So I
7389 have added a single, blank menu entry in this case.
7391 9. Added ${quote_local_part.
7393 10. MIME decoding is now applied to the contents of Subject: header lines when
7396 11. Now that a reference to $sender_host_address automatically causes a reverse
7397 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7398 host lookup before query-style lookups in lists that might use this
7399 variable. This has therefore been abolished, and the "net-" prefix is no
7400 longer necessary for query-style lookups.
7402 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7403 have been a typo for LFLAGS, so it has been changed.
7405 13. The install script calls Exim with "-C /dev/null" in order to find the
7406 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7407 to be output. However, since Exim outputs its version number before the
7408 error, it didn't break the script. It just looked ugly. I fixed this by
7409 always allowing "-C /dev/null" if the caller is root.
7411 14. Ignore overlarge ACL variable number when reading spool file - insurance
7412 against a later release with more variables having written the file.
7414 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7415 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7416 this, as well as the form without the "IPv6" on the front.
7418 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7419 OS/Makefile-Darwin file.
7421 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7422 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7424 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7425 variable. However, some OS use ints for inodes. I've added cast to long int
7426 to get rid of the compiler warning.
7428 19. I had forgotten to lock out "/../" in configuration file names when
7429 ALT_CONFIG_PREFIX was set.
7431 20. Routers used for verification do not need to specify transports. However,
7432 if such a router generated a host list, and callout was configured, Exim
7433 crashed, because it could not find a port number from the (non-existent)
7434 transport. It now assumes port 25 in this circumstance.
7436 21. Added the -t option to exigrep.
7438 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7439 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7440 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7441 them is set, LOOKUP_LSEARCH is forced.
7443 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7445 24. Added sender and host information to the "rejected by local_scan()" log
7446 line; previously there was no indication of these.
7448 25. Added .include_if_exists.
7450 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7451 Linux. It turns out that not all file systems support this. Apparently some
7452 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7453 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7454 not supported on the file descriptor, is now ignored when Exim is trying to
7455 sync a directory. This applies only to Linux.
7457 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7459 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7460 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7463 29. Added support for maildirsize files from supplied patch (modified a bit).
7465 30. The use of :fail: followed by an empty string could lead Exim to respond to
7466 sender verification failures with (e.g.):
7468 550 Verification failed for <xxx>
7469 550 Sender verify failed
7471 where the first response line was missing the '-' that indicates it is not
7472 the final line of the response.
7474 31. The loop for finding the name of the user that called Exim had a hardwired
7475 limit of 10; it now uses the value of finduser_retries, which is used for
7476 all other user lookups.
7478 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7481 33. Exim was neglecting to zero errno before one call of strtol() when
7482 expanding a string and expecting an integer value. On some systems this
7483 resulted in spurious "integer overflow" errors. Also, it was casting the
7484 result into an int without checking.
7486 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7487 did not work. The code looks as if it has *never* worked, though it appears
7488 to have been documented since at least release 1.62. I have made it work.
7490 35. The "timeout_DNS" error in retry rules, also documented since at least
7491 1.62, also never worked. As it isn't clear exactly what this means, and
7492 clearly it isn't a major issue, I have abolished the feature by treating it
7493 as "timeout", and writing a warning to the main and panic logs.
7495 36. The display of retry rules for -brt wasn't always showing the error code
7498 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7499 timeout_connect_A, timeout_connect_MX.
7501 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7502 to the empty sender.
7504 39. The daemon was not analysing the content of -oX till after it had closed
7505 stderr and disconnected from the controlling terminal. This meant that any
7506 syntax errors were only noted on the panic log, and the return code from
7507 the command was 0. By re-arranging the code a little, I've made the
7508 decoding happen first, so such errors now appear on stderr, and the return
7509 code is 1. However, the actual setting up of the sockets still happens in
7510 the disconnected process, so errors there are still only recorded on the
7513 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7514 connections (as happens on some IP stacks) was logged at start up time as
7515 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7516 it from "IPv6 and IPv4", which means that two separate sockets are being
7519 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7520 says whether AF_INET or AF_INET6 was passed as an argument.
7522 42. Exiwhat output was messed up when time zones were included in log
7525 43. Exiwhat now gives more information about the daemon's listening ports,
7526 and whether -tls-on-connect was used.
7528 44. The "port" option of the smtp transport is now expanded.
7530 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7531 silently ignored. Now an error message is written to the main and panic
7534 46. There's a new ACL modifier called "logwrite" which writes to a log file
7535 as soon as it is encountered.
7537 47. Added $local_user_uid and $local_user_gid at routing time.
7539 48. Exim crashed when trying to verify a sender address that was being
7542 49. Exim was recognizing only a space character after ".include". It now also
7543 recognizes a tab character.
7545 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7546 extracting the relevant information from the specification. The man page no
7547 longer contains scrambled data for the -d option, and I've added a section
7548 at the front about calling Exim under different names.
7550 51. Added "extra_headers" argument to the "mail" command in filter files.
7552 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7555 53. Installed eximstats 1.29.
7557 54. Added transport_filter_timeout as a generic transport option.
7559 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7560 Cc: header lines. This was required by RFC 822, but it not required by RFC
7563 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7564 incoming messages that did not have them. Now it does so only if the
7565 message originates locally, that is, if there is no associated remote host
7566 address. When Resent- header lines are present, this applies to the Resent-
7567 lines rather than the non-Resent- lines.
7569 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7570 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7572 58. Messages for configuration errors now include the name of the main
7573 configuration file - useful now that there may be more than one file in a
7574 list (.included file names were always shown).
7576 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7577 for those rare installations that do not start the daemon as root or run it
7578 setuid root. I've cut out the call to initgroups() if the daemon is not
7581 60. The Exim user and group can now be bound into the binary as text strings
7582 that are looked up at the start of Exim's processing.
7584 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7586 62. Added $mailstore_basename variable.
7588 63. Installed patch to sieve.c from Michael Haardt.
7590 64. When Exim failed to open the panic log after failing to open the main log,
7591 the original message it was trying to log was written to stderr and debug
7592 output, but if they were not available (the usual case in production), it
7593 was lost. Now it is written to syslog before the two lines that record the
7594 failures to open the logs.
7596 65. Users' Exim filters run in subprocesses under the user's uid. It is
7597 possible for a "deliver" command or an alias in a "personal" command to
7598 provoke an address rewrite. If logging of address rewriting is configured,
7599 this fails because the process is not running as root or exim. There may be
7600 a better way of dealing with this, but for the moment (because 4.30 needs
7601 to be released), I have disabled address rewrite logging when running a
7602 filter in a non-root, non-exim process.
7608 1. The buildconfig auxiliary program wasn't quoting the value set for
7609 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7610 not defined. This bug was masked in 4.22 by the effect that was fixed in
7613 2. Some messages that were rejected after a message id was allocated were
7614 shown as "incomplete" by exigrep. It no longer does this for messages that
7615 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7617 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7618 have allow_domain_literals set, the ID did not get logged in the <= line.
7619 Domain literals are now always recognized in Message-ID: header lines.
7621 4. The first argument for a ${extract expansion item is the key name or field
7622 number. Leading and trailing spaces in this item were not being ignored,
7623 causing some misleading effects.
7625 5. When deliver_drop_privilege was set, single queue runner processes started
7626 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7627 same command in the process it spins off) were not dropping privilege.
7629 6. When the daemon running as "exim" started a queue runner, it always
7630 re-executed Exim in the spun-off process. This is a waste of effort when
7631 deliver_drop_privilege is set. The new process now just calls the
7632 queue-runner function directly.
7638 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7641 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7642 ignored. Though the use of -oP was forcing the writing of a pid file, it
7643 was always written to the default place.
7645 3. If the message "no IP address found for host xxxx" is generated during
7646 incoming verification, it is now followed by identification of the incoming
7647 connection (so you can more easily find what provoked it).
7649 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7651 5. Added some features to "harden" Exim a bit more against certain attacks:
7653 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7654 be put in Local/Makefile. This is like the never_users runtime option,
7655 but it cannot be overridden. The default setting is "root".
7657 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7658 prefix string with which any file named in a -C command line option
7661 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7662 is retained for -C and -D only if the caller of Exim is root. Without
7663 it, the exim user may also use -C and -D and retain privilege.
7665 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7666 command line option is disabled.
7668 6. Macro names set by the -D option must start with an upper case letter, just
7669 like macro names defined in the configuration file.
7671 7. Added "dereference=" facility to LDAP.
7673 8. Two instances of the typo "uknown" in the source files are fixed.
7675 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7676 the Configure-Makefile script screwed up while processing it.
7678 10. Incorporated PCRE 4.4.
7680 11. The SMTP synchronization check was not operating right at the start of an
7681 SMTP session. For example, it could not catch a HELO sent before the client
7682 waited for the greeting. There is now a check for outstanding input at the
7683 point when the greeting is written. Because of the duplex, asynchronous
7684 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7685 way, but not yet received, when the check is performed.
7687 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7688 on TCP/IP sockets, because this apparently causes some broken clients to
7691 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7692 unchanged) from the Cygwin maintainer.
7694 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7695 was supported instead of testing for mbx. Effectively a typo.
7697 15. The spa authenticator server code was not checking that the input it
7698 received was valid base64.
7700 16. The debug output line for the "set" modifier in ACLs was not showing the
7701 name of the variable that was being set.
7703 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7705 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7706 lookup on its own. Something else was needed to trigger the lookup. For
7707 example, a match in host_lookup or the need for a host name in a host list.
7708 Now, if $sender_host_name is referenced and the host name has not yet been
7709 looked up, a lookup is performed. If the lookup fails, the variable remains
7710 empty, and $host_lookup_failed is set to "1".
7712 19. Added "eqi" as a case-independent comparison operator.
7714 20. The saslauthd authentication condition could segfault if neither service
7715 nor realm was specified.
7717 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7718 error message that was logged was misleading, and incoming SMTP
7719 connections were dropped. The message is now more accurate, and temporary
7720 errors are given to SMTP connections.
7722 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7723 (except RSET) with a 421 error, until QUIT is received. However, it was
7724 failing to send a response to QUIT.
7726 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7727 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7728 ACL is now run after the helo_try_verify_hosts code.
7730 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7731 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7732 case-independent, so other case variants are also recognized). Apparently
7733 some systems use these upper case variants.
7735 25. If more than two messages were waiting for the same host, and a transport
7736 filter was specified for the transport, Exim sent two messages over the
7737 same TCP/IP connection, and then failed with "socket operation on non-
7738 socket" when it tried to send the third.
7740 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7742 27. The extern definition of crypt16() in expand.c was not being excluded when
7743 the OS had its own crypt16() function.
7745 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7746 as a preferred synonym for return_size_limit, both as an option and as an
7749 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7751 30. Changed the default configuration ACL to relax the local part checking rule
7752 for addresses that are not in any local domains. For these addresses,
7753 slashes and pipe symbols are allowed within local parts, but the sequence
7754 /../ is explicitly forbidden.
7756 31. SPA server authentication was not clearing the challenge buffer before
7759 32. log_message in a "warn" ACL statement was writing to the reject log as
7760 well as to the main log, which contradicts the documentation and doesn't
7761 seem right (because no rejection is happening). So I have stopped it.
7763 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7764 However, I am unable to do any testing of this.
7766 34. Fixed an infelicity in the appendfile transport. When checking directories
7767 for a mailbox, to see if any needed to be created, it was accidentally
7768 using path names with one or more superfluous leading slashes; tracing
7769 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7771 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7772 contents are added to the log line that is written for every discarded
7773 recipient. (Previously a log_message setting was ignored.)
7775 36. The ${quote: operator now quotes the string if it is empty.
7777 37. The install script runs exim in order to find its version number. If for
7778 some reason other than non-existence or emptiness, which it checks, it
7779 could not run './exim', it was installing it with an empty version number,
7780 i.e. as "exim-". This error state is now caught, and the installation is
7783 38. An argument was missing from the function that creates an error message
7784 when Exim fails to connect to the socket for saslauthd authentication.
7785 This could cause Exim to crash, or give a corrupted message.
7787 39. Added isip, isip4, and isip6 to ${if conditions.
7789 40. The ACL variables $acl_xx are now saved with the message, and can be
7790 accessed later in routers, transports, and filters.
7792 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7793 strings in the file are not string-expanded.
7795 42. If a MAIL command specified a SIZE value that was too large to fit into an
7796 int variable, the check against message_size_limit failed. Such values are
7797 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7798 day this will have to be increased, but I don't think I want to be around
7799 when emails are that large.
7806 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7807 iconv() is not standard in FreeBSD.
7809 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7810 IPv6 enabled. The observed symptom was a segmentation fault on return from
7811 the function os_common_find_running_interfaces() in src/os.c.
7813 3. In the check_special_case() function in daemon.c I had used "errno" as an
7814 argument name, which causes warnings on some systems. This was basically a
7815 typo, since it was named "eno" in the comments!
7817 4. The code that waits for the clock to tick (at a resolution of some fraction
7818 of a second) so as to ensure message-id uniqueness was always waiting for
7819 at least one whole tick, when it could have waited for less. [This is
7820 almost certainly not relevant at current processor speeds, where it is
7821 unlikely to ever wait at all. But we try to future-proof.]
7823 5. The function that sleeps for a time interval that includes fractions of a
7824 second contained a race. It did not block SIGALRM between setting the
7825 timer, and suspending (a couple of lines later). If the interval was short
7826 and the sigsuspend() was delayed until after it had expired, the suspension
7827 never ended. On busy systems this could lead to processes getting stuck for
7830 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7831 process, before it forks any delivery processes. The open lookup caching
7832 mechanism meant that the open file or database connection was passed into
7833 the delivery process. The problem was that delivery processes always tidy
7834 up cached lookup data. This could cause a problem for the next delivery
7835 process started by the queue runner, because the external queue runner
7836 process does not know about the closure. So the next delivery process
7837 still has data in the lookup cache. In the case of a file lookup, there was
7838 no problem because closing a file descriptor in a subprocess doesn't affect
7839 the parent. However, if the lookup was caching a connection to a database,
7840 the connection was closed, and the second delivery process was likely to
7841 see errors such as "PGSQL: query failed: server closed the connection
7842 unexpectedly". The problem has been fixed by closing all cached lookups
7843 in a queue runner before running a delivery process.
7845 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7846 seem to have the "const" qualifier which it has on other OS. I've
7849 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7850 *of the same type* (client or server) with the same public name that an
7851 error should be diagnosed.
7853 9. When Exim looked up a host name for an IP address, but failed to find the
7854 original IP address when looking up the host name (a safety check), it
7855 output the message "<ip address> does not match any IP for NULL", which was
7856 confusing, to say the least. The bug was that the host name should have
7857 appeared instead of "NULL".
7859 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7860 user that is built into the binary, and the -C or -D options is used, root
7861 privilege is dropped before the configuration file is read. In addition,
7862 logging is switched to stderr instead of the normal log files. If the
7863 configuration then re-defines the Exim user, the unprivileged environment
7864 is probably not what is expected, so Exim logs a panic warning message (but
7867 However, if deliver_drop_privilege is set, the unprivileged state may well
7868 be exactly what is intended, so the warning has been cut out in that case,
7869 and Exim is allowed to try to write to its normal log files.
7875 1. smtp_return_error_details was not giving details for temporary sender
7876 or receiver verification errors.
7878 2. Diagnose a configuration error if two authenticators have the same public
7881 3. Exim used not to create the message log file for a message until the first
7882 delivery attempt. This could be confusing when incoming messages were held
7883 for policy or load reasons. The message log file is now created at the time
7884 the message is received, and an initial "Received" line is written to it.
7886 4. The automatically generated man page for command line options had a minor
7887 bug that caused no ill effects; however, a more serious problem was that
7888 the procedure for building the man page automatically didn't always
7889 operate. Consequently, release 4.20 contains an out-of-date version. This
7890 shouldn't happen again.
7892 5. When building Exim with embedded Perl support, the script that builds the
7893 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7894 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7896 6. The freeze_tell option was not being used for messages that were frozen on
7897 arrival, either by an ACL or by local_scan().
7899 7. Added the smtp_incomplete_transaction log selector.
7901 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7902 was accepting AUTH without a new EHLO.
7904 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7905 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7908 10. Logging of TCP/IP connections (when configured) now happens in the main
7909 daemon process instead of the child process, so that the TCP/IP connection
7910 count is more accurate (but it can never be perfect).
7912 11. The use of "drop" in a nested ACL was not being handled correctly in the
7913 outer ACL. Now, if condition failure induced by the nested "drop" causes
7914 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7915 or "require"), the connection is dropped.
7917 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7918 that yield "discard" can now be used with an "accept" or a "discard" verb,
7919 but an error is generated for any others (because I can't see a useful way
7920 to define what should happen).
7922 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7923 are now processed in the same way as lines in the Exim configuration file.
7924 In particular, continuation lines are supported.
7926 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7928 15. Added -ti meaning -t -i.
7930 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7931 domains, and warn by logging if others are found.
7933 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7934 returned from GIFCONF when Exim is trying to find the list of interfaces on
7935 a host. The code in os.c has been modified to copy each ifreq to an aligned
7936 structure in all cases.
7938 Also, in some cases, the returned ifreq's were being copied to a 'struct
7939 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7940 means the last couple of bytes of an IPv6 address could be chopped if the
7941 ifreq contained only a normal sockaddr (14 bytes storage).
7943 18. Named domain lists were not supported in the hosts_treat_as_local option.
7944 An entry such as +xxxx was not recognized, and was treated as a literal
7947 19. Ensure that header lines added by a DATA ACL are included in the reject log
7948 if the ACL subsequently rejects the message.
7950 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7951 MD5 (which is deprecated).
7953 21. When testing a filter file using -bf, Exim was writing a message when it
7954 took the sender from a "From " line in the message, but it was not doing so
7955 when it took $return_path from a Return-Path: header line. It now does.
7957 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7958 with a valid header line field name (a series of printing characters
7959 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7961 23. Changed "disc" in the source to "disk" to conform to the documentation and
7962 the book and for uniformity.
7964 24. Ignore Sendmail's -Ooption=value command line item.
7966 25. When execve() failed while trying to run a command in a pipe transport,
7967 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7968 could be confused with a return value of 69 from the command itself. This
7969 has been changed to 127, the value the shell returns if it is asked to run
7970 a non-existent command. The wording for the related log line suggests a
7971 non-existent command as the problem.
7973 26. If received_header_text expands to an empty string, do not add a Received:
7974 header line to the message. (Well, it adds a token one on the spool, but
7975 marks it "old" so that it doesn't get used or transmitted.)
7977 27. Installed eximstats 1.28 (addition of -nt option).
7979 28. There was no check for failure on the call to getsockname() in the daemon
7980 code. This can fail if there is a shortage of resources on the system, with
7981 ENOMEM, for example. A temporary error is now given on failure.
7983 29. Contrary to the C standard, it seems that in some environments, the
7984 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7985 program. Exim now does this explicitly; it affects the formatting of
7986 timestamps using strftime().
7988 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7989 complaints. I've now initialized all the variables, to avoid this.
7991 32. Header lines added by a system filter were not being "seen" during
7992 transport-time rewrites.
7994 33. The info_callback() function passed to OpenSSL is set up with type void
7995 (*)(SSL *, int, int), as described somewhere. However, when calling the
7996 function (actually a macro) that sets it up, the type void(*)() is
7997 expected. I've put in a cast to prevent warnings from picky compilers.
7999 34. If a DNS black list lookup found a CNAME record, but there were no A
8000 records associated with the domain it pointed at, Exim crashed.
8002 35. If a DNS black list lookup returned more than one A record, Exim ignored
8003 all but the first. It now scans all returned addresses if a particular IP
8004 value is being sought. In this situation, the contents of the
8005 $dnslist_value variable are a list of all the addresses, separated by a
8008 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8009 to accept a host name and all its aliases if the forward lookup for any of
8010 them yielded the IP address of the incoming connection. Now it accepts only
8011 those names whose forward lookup yields the correct IP address. Any other
8012 names are discarded. This closes a loophole whereby a rogue DNS
8013 administrator could create reverse DNS records to break through a
8014 wildcarded host restriction in an ACL.
8016 37. If a user filter or a system filter that ran in a subprocess used any of
8017 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8018 the wrong values were passed to the pipe command ($thisaddress had the
8019 value of $0, $0 had the value of $1, etc). This bug was introduced by
8020 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8022 38. Improved the line breaking for long SMTP error messages from ACLs.
8023 Previously, if there was no break point between 40 and 75 characters, Exim
8024 left the rest of the message alone. Two changes have been made: (a) I've
8025 reduced the minimum length to 35 characters; (b) if it can't find a break
8026 point between 35 and 75 characters, it looks ahead and uses the first one
8027 that it finds. This may give the occasional overlong line, but at least the
8028 remaining text gets split now.
8030 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8031 file descriptors might be low, and that setting 1000 would always raise it.
8032 It turns out that in some environments, the limit is already over 1000 and
8033 that lowering it causes trouble. So now Exim takes care not to decrease it.
8035 40. When delivering a message, the value of $return_path is set to $sender_
8036 address at the start of routing (routers may change the value). By an
8037 oversight, this default was not being set up when an address was tested by
8038 -bt or -bv, which affected the outcome if any router or filter referred to
8041 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8042 log only when "message" or "log_message" are set. However, if one of the
8043 conditions was an address verification, or a call to a nested ACL, the
8044 messages generated by the underlying test were being passed through. This
8045 no longer happens. The underlying message is available in $acl_verify_
8046 message for both "message" and "log_message" expansions, so it can be
8047 passed through if needed.
8049 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8050 new expansion $bh_ to give the encoded byte string without charset
8051 translation. Translation happens only if iconv() is available; HAVE_ICONV
8052 indicates this at build time. HEADERS_CHARSET gives the charset to
8053 translate to; headers_charset can change it in the configuration, and
8054 "headers charset" can change it in an individual filter file.
8056 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8057 that creates RFC 2047 encoded "words" labels them as that charset instead
8058 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8059 expansion operator; (ii) when Exim creates a From: line for a local
8060 message; (iii) when a header line is rewritten to include a "phrase" part.
8062 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8063 buggy, causing it to skip the first lines of messages whose message ID
8064 ended in 'D'. This would not have bitten before Exim release 4.14, because
8065 message IDs were unlikely to end in 'D' before then. The effect was to have
8066 incorrect size information for certain domains.
8068 45. #include "config.h" was missing at the start of the crypt16.c module. This
8069 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8072 46. If there was a timeout during a "random" callout check, Exim treated it as
8073 a failure of the random address, and carried on sending RSET and the real
8074 address. If the delay was just some slowness somewhere, the response to the
8075 original RCPT would be taken as a response to RSET and so on, causing
8076 mayhem of various kinds.
8078 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8079 when I implemented it. It didn't allow for the fact that some option values
8080 may legitimately be negative (e.g. size_addition), and it didn't even do
8081 the right test for positive values.
8083 48. Domain names in DNS records are case-independent. Exim always looks them up
8084 in lower case. Some resolvers return domain names in exactly the case they
8085 appear in the zone file, that is, they may contain uppercase letters. Not
8086 all resolvers do this - some return always lower case. Exim was treating a
8087 change of case by a resolver as a change of domain, similar to a widening
8088 of a domain abbreviation. This triggered its re-routing code and so it was
8089 trying to route what was effectively the same domain again. This normally
8090 caused routing to fail (because the router wouldn't handle the domain
8091 twice). Now Exim checks for this case specially, and just changes the
8092 casing of the domain that it ultimately uses when it transmits the message
8095 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8098 50. If a filter generated a file delivery with a non-absolute name (possible if
8099 no home directory exists for the router), the forbid_file option was not
8102 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8103 the existing equality matching.
8105 52. Exim was using ints instead of ino_t variables in some places where it was
8106 dealing with inode numbers.
8108 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8109 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8110 called TMPDIR, and if it finds it is different, it changes its value.
8112 54. The smtp_printf() function is now made available to local_scan() so
8113 additional output lines can be written before returning. There is also an
8114 smtp_fflush() function to enable the detection of a dropped connection.
8115 The variables smtp_input and smtp_batched_input are exported to
8118 55. Changed the default runtime configuration: the message "Unknown user"
8119 has been removed from the ACL, and instead placed on the localuser router,
8120 using the cannot_route_message feature. This means that any verification
8121 failures that generate their own messages won't get overridden. Similarly,
8122 the "Unrouteable address" message that was in the ACL for unverifiable
8123 relay addresses has also been removed.
8125 56. Added hosts_avoid_esmtp to the smtp transport.
8127 57. The exicyclog script was not checking for the esoteric option
8128 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8129 will work only if exicyclog is run under the appropriate euid.
8131 58. Following a discussion on the list, the rules by which Exim recognises line
8132 endings on incoming messages have been changed. The -dropcr and drop_cr
8133 options are now no-ops, retained only for backwards compatibility. The
8134 following line terminators are recognized: LF CRLF CR. However, special
8135 processing applies to CR:
8137 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8138 nor a local message in the state where . is a terminator.
8140 (ii) If a bare CR is encountered in a header line, an extra space is added
8141 after the line terminator so as not to end the header. The reasoning
8142 behind this is that bare CRs in header lines are most likely either
8143 to be mistakes, or people trying to play silly games.
8145 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8146 was being incorrectly given as 18 bytes larger than it should have been.
8147 This is a VOB (very old bug).
8149 60. This may never have affected anything current, but just in case it has:
8150 When the local host is found other than at the start of a list of hosts,
8151 the local host, those with the same MX, and any that follow, are discarded.
8152 When the list in question was part of a longer list of hosts, the following
8153 hosts (not currently being processed) were also being discarded. This no
8154 longer happens. I'm not sure if this situation could ever has previously
8157 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8160 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8161 header. This is something that is recommended in a new Internet Draft, and
8162 is something that is documented as being done by Sendmail. There are two
8163 possible values. For messages generated by the autoreply transport, Exim
8166 Auto-Submitted: auto-replied
8168 whereas for all other generated messages (e.g. bounces) it adds
8170 Auto-Submitted: auto-generated
8172 63. The "personal" condition in filters now includes a test for the
8173 Auto-Submitted: header. If it contains the string "auto-" the message it
8174 not considered personal.
8176 64. Added rcpt_include_affixes as a generic transport option.
8178 65. Added queue_only_override (default true).
8180 66. Added the syslog_duplication option.
8182 67. If what should have been the first header line of a message consisted of
8183 a space followed by a colon, Exim was mis-interpreting it as a header line.
8184 It isn't of course - it is syntactically invalid and should therefore be
8185 treated as the start of the message body. The misbehaviour could have
8186 caused a number of strange effects, including loss of data in subsequent
8187 header lines, and spool format errors.
8189 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8190 client host had authenticated. This control can now be exercised by an ACL
8191 for more flexibility.
8193 69. By default, callouts do not happen when testing with -bh. There is now a
8194 variant, -bhc, which does actually run the callout code, including
8195 consulting and updating the callout cache.
8197 70. Added support for saslauthd authentication, courtesy of Alexander
8200 71. If statvfs() failed on the spool or log directories while checking their
8201 size for availability, Exim confusingly gave the error "space shortage".
8202 Furthermore, in debugging mode it crashed with a floating point exception.
8203 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8204 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8205 serious problem, Exim now writes to the main and panic logs when this
8206 happens, with details of the failure. It then refuses to accept the
8207 incoming message, giving the message "spool directory problem" or "log
8208 directory problem" with a 421 code for SMTP messages.
8210 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8211 0, 1, and 2 exist, because some OS complain for execs without them (see
8212 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8213 descriptors. However, the code omitted to check that the open succeeded,
8214 causing mysterious errors if for some reason the permissions on /dev/null
8215 got screwed. Now Exim writes a message to the main and panic logs, and
8216 bombs out if it can't open /dev/null.
8218 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8219 interact so that it is all more flexible. It is supposed to remain
8220 backwards compatible. Also added extra_local_interfaces.
8222 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8223 to bomb out with an assertion failure - to the client this appears as a
8224 connection drop. This problem occurs in the part of the code that was taken
8225 from the Samba project. Fortunately, the assertion is in a very simple
8226 function, so I have fixed this by reproducing the function inline in the
8227 one place where it is called, and arranging for authentication to fail
8228 instead of killing the process with assert().
8230 75. The SPA client code was not working when the server requested OEM rather
8231 than Unicode encoding.
8233 76. Added code to make require_files with a specific uid setting more usable in
8234 the case where statting the file as root fails - usually a non-root-mounted
8235 NFS file system. When this happens and the failure is EACCES, Exim now
8236 forks a subprocess and does the per-uid checking as the relevant uid.
8238 77. Added process_log_path.
8240 78. If log_file_path was not explicitly set, a setting of check_log_space or
8241 check_log_inodes was ignored.
8243 79. If a space check for the spool or log partitions fails, the incident is now
8244 logged. Of course, in the latter case the data may get lost...
8246 80. Added the %p formatting code to string_format() so that it can be used to
8247 print addresses in debug_print(). Adjusted all the address printing in the
8248 debugging in store.c to use %p rather than %d.
8250 81. There was a concern that a line of code in smtp_in.c could overflow a
8251 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8252 initially expressed, the concern was not well-founded, because trailing
8253 spaces are removed early. However, if the trailing spaces were followed by
8254 a NULL, they did not get removed, so the overflow was possible. Two fixes
8257 (a) I re-wrote the offending code in a cleaner fashion.
8258 (b) If an incoming SMTP command contains a NULL character, it is rejected
8261 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8262 runs initgroups(), so that if the Exim user is in any additional groups,
8263 they will be used during message reception.
8269 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob