1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes, when delivery is immediate. Previously debugging
74 stopped any time Exim re-execs.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
84 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
90 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
91 related applications. Previously an "H" was used where available info
92 says that "M" should be, so change to match.
94 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
95 as arguments, so an implementation trying to copy these into a local
96 buffer was taking a taint-enforcement trap. Fix by using dynamically
97 created buffers. Similar fix for radius expansion condition.
99 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
100 reasonable, eg. to count headers. Fix by using dynamically created
101 buffers rather than a local. Do similar fixes for ACL actions "dcc",
102 "log_reject_target", "malware" and "spam"; the arguments are expanded
103 so could be handling tainted values.
105 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
106 broken the (no-op) support for this sendmail command. Restore it
107 to doing nothing, silently, and returning good status.
109 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
110 record path was given (or the default used) without a leading directory
111 path, an error occurred on trying to open it. Use the transport's working
114 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
115 transport. Previously it was the name on the DNS A-record; use instead
116 the head of the CNAME chain leading there (if there is one). This seems
117 to align better with RFC 6125.
119 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
120 smtp_accept_max_per_host allocated resources which were not released
121 when the limit was exceeded. This eventually crashed the daemon. Fix
122 by adding a release action in that path.
124 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
125 expanded; previously using tainted values was rejected. Fix by using
126 dynamically-created buffers.
128 JH/09 Relax restrictions on ACL verify condition needing access to message
129 headers. Previously they were only permitted in data and non-smtp ACLs;
130 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
131 not_blind, header_sender and header_names_ascii verification.
133 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
134 Previously a macro used one argument twice; when called with the
135 argument as an expression having side-effects, incorrect operation
136 resulted. Use an inlineable function.
138 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
139 held open for a verify callout. Previously this wan not accounted for
140 and a corrupt onward SMTP conversation resulted.
142 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
143 Previously they were embedded naively in the constructed address; when
144 needed, strip the quoting and quote the entire local_part.
145 Also make the inbound_srs expansion condition handle quoting.
147 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
148 excluded, not matching the documentation.
150 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
151 was given for the sqlite_dbfile a trap resulted.
153 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
154 "name" argument resulted in a trap. There is no reason to disallow such;
155 this was a coding error.
157 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
158 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
159 spent suspended, ignoring the POSIX definition. Previously we assumed
160 it did and a constant offset from real time could be used as a correction.
161 Change to using the same clock source for the start-of-message and the
162 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
163 exists, just to get a clock slightly more aligned to reality.
165 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
166 RFC says it is optional some validators care. The missing char was not
167 intended but triggered by a line-wrap alignment. Discovery and fix by
168 Guillaume Outters, hacked on by JH.
170 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
171 name being quoted was tainted a trap would be taken. Fix by using
172 dynamically created buffers. The routine could have been called by a
173 rewrite with the "h" flag, by using the "-F" command-line option, or
174 by using a "name=" option on a control=submission ACL modifier.
176 JH/19 SPF: change the Authentication-Results expansion component to give
177 smtp.helo when the sender domain is empty. Previously it gave
180 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
181 not in 127.0.0.0/8 to help in spotting list domains taken over by a
182 domain-parking registrar.
184 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
185 Previously when a whitespace character was specified it was not inserted
186 after removing the newline.
188 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
189 the domain part of the recipient address. This overrides any tls_sni
190 option set, which was previously used.
192 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
195 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
196 is_tainted() had an off-by-one error in the overenthusiastic direction.
197 Find and fix by Gavan. Although NetBSD is not a supported platform for
198 4.94 this bug could affect other platforms.
200 PP/01 Fix default prime selection to be consistent.
201 One path used ike23 still, instead of exim.dev.20160529.3; now both
202 execution flows will use the same DH primes (currently
203 exim.dev.20160529.3).
205 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
206 option in smtp transport, to match the documentation. Previously
207 verification was not being done in some cases where it should have been.
209 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
210 than one server was defined and depending on the platform memory layout
211 details, an internal consistency trap could be hit while walking the list
214 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
215 files. The value is used by the authresults expansion item. Previously
216 if this was used in a router or transport, a crash could result.
218 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
219 have a sufficient incidence of EINTR returns from select that an
220 interaction with other operations done by the main daemon loop exposed
221 a bug in the error-handling. This was benign apart from the log
224 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
225 for consistency with delivered "=>" and failed "**" lines. While we're
226 there, handle PRX and TFO.
228 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
229 applied. This resulted, if any header-line rewrite rules were configured,
230 in a panic-log triggerable by sending a message with a long address in
231 a header. Fix by increasing the arbitrary limit to larger than a single
232 (dewrapped) 5322 header line maximum size.
234 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
235 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
236 RFC 6648 which deprecates X- options in protocols as a general practice.
237 Changeover between the implementations is handled by the mechanisms
240 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
241 a remote which uses callout/hold. Previously the local was queued.
243 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
246 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
247 log line. The asterisk is now consistently appended to the remote IP
248 (and port, if given), and will also be provided on defer and fail log
249 lines. Previously it could be placed on the local IP if that was being
250 logged, and was only provided on delivery lines.
252 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
254 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
255 authenticator client_send option. Previously the next char, after a pair
256 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
257 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
258 documentation. There is still no way to get a leading ^ immediately
259 after a NUL (ie. for the password of a PLAIN method authenticator.
261 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
262 files. For bad sizes read, delete the record and whine to paniclog.
264 JH/38 When logging an AUTH failure, as server, do not include sensitive
265 information. Previously, the credentials would be included if given
266 as part of the AUTH command line and an ACL denied authentication.
268 JH/39 Bug 2691: fix $local_part_data. When the matching list element
269 referred to a file, bad data was returned. This likely also affected
272 JH/40 The gsasl authenticator now supports caching of the salted password
273 generated by the client-side implementation. This required the addition
274 of a new variable: $auth4.
276 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
277 left undeleted; the attempt to re-create it then failed - resulting in
278 the usual "SIGHUP tp have daemon reload configuration" to not work.
279 This affected any platform not supporting "abstract" Unix-domain
280 sockets (i.e. not Linux).
282 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
283 recipients" response to RCPT in a later response, with a 250. The
284 previous coding assumed this would not happen, and under PIPELINING
285 would result in both lost and duplicate recipients for a message.
287 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
288 Previously the weighting was incorrectly applied. Similar fix for socks
289 proxies. Found and fixed by Heiko Schlichting.
291 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
292 not handle sub-lists included using the +namedlist syntax. While
293 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
294 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
297 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
298 Previously the permanent pool was used, so the sore could not be freed.
299 This meant a connection with many messages would use continually-growing
302 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
303 per-pool so as not to waste too much space. Previously a constant size
304 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
305 DOS attacks harder. The cost is wasted memory use in the larger blocks.
307 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
308 same space repeatedly, and a smaller process footprint.
310 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
311 for GnuTLS. Previously the name started "%s" which, while not a bug,
312 looked as if if might be one.
314 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
315 "continued connection") make the $proxy_* variables available. Previously
316 the information was not passed across the exec() call for subsequent
317 transport executions. This also mean that the log lines for the
318 messages can show the proxy information.
320 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
321 receive time. With modern systems the difference is significant.
322 The historical behaviour can be restored by disabling (a new) log_selector
323 "queue_time_exclusive".
325 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
326 content) were specifically tested for. Now, also cover expansions
327 resulting in ACL names and inline ACL content.
329 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
330 making it unusable in complex expressions.
332 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
333 fail with a taint-check on the spoolfile name, and leave the message
336 HS/01 Enforce absolute PID file path name.
338 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
340 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
342 PP/02 Bug 2643: Correct TLS DH constants.
343 A missing NUL termination in our code-generation tool had led to some
344 incorrect Diffie-Hellman constants in the Exim source.
345 Reported by kylon94, code-gen tool fix by Simon Arlott.
347 PP/03 Impose security length checks on various command-line options.
348 Fixes CVE-2020-SPRSS reported by Qualys.
350 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
351 better. Reported by Qualys.
353 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
354 providing a particularly obnoxious sender full name.
357 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
359 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
362 PP/08 Change default for recipients_max from unlimited to 50,000.
364 PP/09 Fix security issue with too many recipients on a message (to remove a
365 known security problem if someone does set recipients_max to unlimited,
366 or if local additions add to the recipient list).
367 Fixes CVE-2020-RCPTL reported by Qualys.
369 PP/10 Fix security issue in SMTP verb option parsing
370 Fixes CVE-2020-EXOPT reported by Qualys.
372 PP/11 Fix security issue in BDAT state confusion.
373 Ensure we reset known-good where we know we need to not be reading BDAT
374 data, as a general case fix, and move the places where we switch to BDAT
375 mode until after various protocol state checks.
376 Fixes CVE-2020-BDATA reported by Qualys.
378 HS/03 Die on "/../" in msglog file names
380 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
381 the Exim runtime user are allowed to create files.
383 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
386 QS/03 When reading the output from interpreted forward files we do not
387 pass the pipe between the parent and the interpreting process to
388 executed child processes (if any).
390 QS/04 Always die if requested from internal logging, even is logging is
393 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
394 the API; compilation noo longer completes with DMARC support included.
395 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
396 been reported on other platforms.
398 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
400 JH/56 Make the majority of info read from config files readonly, for defence-in-
401 depth against exploits. Suggestion by Qualys.
402 Not supported on Solaris 10.
404 JH/57 Fix control=fakreject for a custom message containing tainted data.
405 Previously this resulted in a log complaint, due to a re-expansion present
406 since fakereject was originally introduced.
408 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
409 modifier was given, a loop resulted.
411 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
412 pipelined SMTP command followed the BDAT LAST then it would be
413 incorrectly treated as part of the message body, causing a verification
416 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
417 require looser validation rules than those for 821-level addresses,
418 which only permit IP addresses.
424 JH/01 Avoid costly startup code when not strictly needed. This reduces time
425 for some exim process initialisations. It does mean that the logging
426 of TLS configuration problems is only done for the daemon startup.
428 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
430 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
431 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
434 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
435 DKIM signing is being done. Previously a transport_filter would always
436 disable CHUNKING, falling back to traditional DATA.
438 JH/05 Regard command-line recipients as tainted.
440 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
442 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
443 PAM library frees one of the arguments given to it, despite the
444 documentation. Therefore a plain malloc must be used.
446 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
447 on-stack buffers were used, resulting in a taint trap when DSN information
448 copied from a received message was written into the buffer.
450 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
451 the ordering of its ARC headers. This caused a crash.
453 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
454 a new record was being constructed with information from the peer, a trap
457 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
458 installation would get error messages from DMARC verify, when it hit the
459 nonexistent file indicated by the default. Distros wanting DMARC enabled
460 should both provide the file and set the option.
461 Also enforce no DMARC verification for command-line sourced messages.
463 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
464 could, depending on the platform, hang at the STARTTLS response.
466 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
467 message on a connection. Previously if one message had ARC headers and
468 the following one did not, a crash could result when adding an
469 Authentication-Results: header.
471 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
472 Exim main code and Exim-related utities. The introduction of taint
473 tracking also did many adjustments to string handling. Since then, eximon
474 frequently terminated with an assert failure.
476 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
477 check for 452 responses. This slightly helps the inefficieny of doing
478 a large alias-expansion into a recipient-limited target. The max_rcpt
479 transport option still applies (and at the current default, will override
480 the new feature). The check is done for either cause of synch, and forces
481 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
482 connection. The new facility is not tunable at this time.
484 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
485 library live data was being used, so the results became garbage. Make
486 copies while it is still usable.
488 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
489 on delivery deferred (==) and failed (**) lines (if a delivery was
490 attemtped). Previously it was only on completion (=>) lines.
492 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
493 for the expansion of the server_scram_iter and server_scram_salt options.
495 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
496 are now specifically given a NO_DATA response without hitting the system
497 resolver. The library goes on to do the now-standard TXT lookup.
498 Use of dnsdb lookups is not affected.
500 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
501 only retrieve the errormessage once. Previously two calls to dlerror()
502 were used, and the second one (for mainlog/paniclog) retrieved null
505 JH/20 Taint checking: disallow use of tainted data for
506 - the appendfile transport file and directory options
507 - the pipe transport command
508 - the autoreply transport file, log and once options
509 - file names used by the redirect router (including filter files)
511 - paths used by single-key lookups
512 Previously this was permitted.
514 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
515 adjusted the size of a major service buffer; this failed because the
516 buffer was in use at the time. Change to a compile-time increase in the
517 buffer size, when this authenticator is compiled into exim.
519 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
520 previous fast-mode was untenable in the face of glibs using mmap to
521 support larger malloc requests.
523 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
524 New values supported, if defined on system where compiled:
525 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
526 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
528 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
529 running a limited number of proceses in parallel, a benefit is gained. The
530 amount varies with the platform hardware and load. The use of the option
531 queue_run_in_order means we cannot do this, as ordering becomes
534 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
535 had introduced a string-copy (for ensuring NUL-termination) which was not
536 appropriate for that case, which can include embedded NUL bytes in the
537 block of data. Investigation showed the copy to actually be needless, the
538 data being length-specified.
540 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
541 done during a receiving connection, and both used TLS, global info was
542 used rather than per-connection info for tracking the state of data
543 queued for transmission. This could result in a connection hang.
545 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
546 Previously, when delivering serveral messages down a single connection
547 only the first would provide a SIZE. This was due to the size information
548 not being properly tracked.
550 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
551 TAI (at 37 seconds currently), pretend to be in UTC for time-related
552 expansion and logging. Previously, spurious values such as a future
553 minute could be seen.
555 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
556 it could crash from a null-deref. This could also affect the
557 ${addresses: } operator and ${readsock } item.
559 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
560 message following a mime one, the variable was not reset.
562 JH/30 When an pipelined-connect fails at the first response, assume incorrect
563 cached capability (perhaps the peer reneged?) and immediately retry in
566 JH/31 Fix spurious detection of timeout while writing to transport filter.
568 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
569 an attempt to copy the string was made before checking it.
571 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
572 taint of the lookup key was maintained; we now regard the presence in the
573 filesystem as sufficient validation.
575 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
576 argument is supplied.
578 JH/35 The dsearch lookup now requires that the directory is an absolute path.
579 Previously this was not checked, and nonempty relative paths made an
580 access under Exim's current working directory.
582 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
583 Previously no event was raised.
585 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
586 parameter supplied by the sender MAIL FROM command. Previously it was
587 ignored, and only the check_spool_space option value for the required
590 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
591 the size of the signing public-key. Previously it was instead giving
592 the size of the signature hash.
594 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
595 the default. See the (new) dkim_verify_min_keysizes option.
597 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
598 and an ACL use a lookup for checking either the local_part or domain,
599 stale data could be accessed. Ensure that variable references are
600 dropped between messages.
602 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
603 by the client was not checked as pointing within response data before
604 being used. A malicious client could thus cause an out-of-bounds read and
605 possibly gain authentication. Fix by adding the check.
607 JH/42 Internationalisation: change the default for downconversion in the smtp
608 transport to be "if needed". Previously it was "as previously set" for
609 the message, which usually meant "if needed" for message-submission but
610 "no" for everything else. However, MTAs have been seen using SMTPUTF8
611 even when the envelope addresses did not need it, resulting in forwarding
612 failures to non-supporting MTAs. A downconvert in such cases will be
613 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
614 transport. The change does mean that addresses needing conversion will
615 be converted when previously a delivery failure would occur.
617 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
618 response was received it would be used unchecked in a fail-DSN, violating
619 standards on line-length limits. Truncate if needed.
621 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
622 doesn't work. (Additionally add a new main config option to configure the
629 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
630 side, to decode a TLS 1.3 packet capture.
632 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
633 Previously the default library behaviour applied, sending two, each in
636 JH/03 Debug output for ACL now gives the config file name and line number for
639 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
641 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
643 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
644 buffer overrun for (non-chunking) other transports.
646 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
647 TLS1.3, means that a server rejecting a client certificate is not visible
648 to the client until the first read of encrypted data (typically the
649 response to EHLO). Add detection for that case and treat it as a failed
650 TLS connection attempt, so that the normal retry-in-clear can work (if
651 suitably configured).
653 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
654 and/or domain. Found and fixed by Jason Betts.
656 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
657 configuration). If a CNAME target was not a wellformed name pattern, a
660 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
661 the OS reports them interleaved with other addresses.
663 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
664 used both for input and for a verify callout, both encrypted, SMTP
665 responses being sent by the server could be lost. This resulted in
666 dropped connections and sometimes bounces generated by a peer sending
669 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
670 string. Previously it was possible to cause undefined behaviour in a
671 library routine (usually a crash). Found by "zerons".
673 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
676 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
677 API was removed, so update to use the newer ones.
679 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
680 any timeout set, is taking a long time. Previously we would hang on to a
681 rotated logfile "forever" if the input was arriving with long gaps
682 (a previous attempt to fix addressed lack, for a long time, of initial
685 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
686 shared (NFS) environment. The length of the tempfile name is now
687 4 + 16 ("hdr.$message_exim_id") which might break on file
688 systems which restrict the file name length to lower values.
691 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
692 shared (NFS) environment.
694 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
695 did for all versions <4.90). Notably -M, -m, --invert, -I may be
698 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
699 on some platforms for bit 31.
701 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
702 to changes apparently associated with TLS1.3 handling some of the APIs
703 previously used were either nonfunctional or inappropriate. Strings
704 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
705 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
706 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
707 This affects log line X= elements, the $tls_{in,out}_cipher variables,
708 and the use of specific cipher names in the encrypted= ACL condition.
710 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
712 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
713 verification result was not updated unless hosts_require_ocsp applied.
715 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
716 queue_list_requires_admin set to false, non-admin users were denied the
719 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
720 directory-of-certs mode. Previously they were advertised despite the
723 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
724 A single TCP connection by a client will now hold a TLS connection open
725 for multiple message deliveries, by default. Previously the default was to
728 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
729 default. If built with the facility, DANE will be used. The facility
730 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
732 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
733 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
734 must be defined and you must still, unless you define DISABLE_TLS, manage
735 the the include-dir and library-file requirements that go with that
736 choice. Non-TLS builds are still supported.
738 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
741 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
742 default. If the platform supports and has the facility enabled, it will
743 be requested on all coneections.
745 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
746 controlled by the build-time option SUPPORT_PIPE_CONNECT.
748 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
750 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
751 success-DSN messages. Previously the From: header was always the default
752 one for these; the option was ignored.
754 JH/28 Fix the timeout on smtp response to apply to the whole response.
755 Previously it was reset for every read, so a teergrubing peer sending
756 single bytes within the time limit could extend the connection for a
757 long time. Credit to Qualsys Security Advisory Team for the discovery.
759 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
760 delivery address, which leaked information of the results of local
761 forwarding. Change to the original envelope recipient address, per
764 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
765 requested. Previously not bounce was generated and a log entry of
766 error ignored was made.
768 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
770 JH/32 Introduce a general tainting mechanism for values read from the input
771 channel, and values derived from them. Refuse to expand any tainted
772 values, to catch one form of exploit.
774 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
775 was unused and the unexpanded text used for the test. Found and
776 fixed by Ruben Jenster.
778 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
779 an attempt to use a TLS library read routine dereffed a nul pointer,
782 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
783 them in our smtp response.
785 JH/36 Have the general router option retry_use_local_part default to true when
786 any of the restrictive preconditions are set (to anything). Previously it
787 was only for check_local user. The change removes one item of manual
788 configuration which is required for proper retries when a remote router
789 handles a subset of addresses for a domain.
791 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
792 link count into consideration.
794 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
795 caused the extension of big_buffer, the following lines were ignored.
797 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
798 accordance with RFC 2308. Previously there was no expiry, so a longlived
799 receive process (eg. due to ACL delays) versus a short SOA value could
802 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
804 JH/39 Promote DMARC support to mainline.
806 JH/40 Bug 2452: Add a References: header to DSNs.
808 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
809 parameters. The relevant library call is documented as "Deprecated: This
810 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
811 3.6.0, DH parameters are negotiated following RFC7919."
813 HS/06 Change the default of dnssec_request_domains to "*"
815 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
816 carried on and emitted a BDAT command, even when PIPELINING was not
819 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
820 buffer was used for the filename, resulting in a trap when tainted
821 arguments (eg. $domain) were used.
823 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
824 recommended to avoid a possible server-load attack. The feature can be
825 re-enabled via the openssl_options main cofiguration option.
827 JH/45 local_scan API: documented the current smtp_printf() call. This changed
828 for version 4.90 - adding a "more data" boolean to the arguments.
829 Bumped the ABI version number also, this having been missed previously;
830 release versions 4.90 to 4.92.3 inclusive were effectively broken in
831 respect of usage of smtp_printf() by either local_scan code or libraries
832 accessed via the ${dlfunc } expansion item. Both will need coding
833 adjustment for any calls to smtp_printf() to match the new function
834 signature; a FALSE value for the new argument is always safe.
836 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
837 the file-offset (which the Linux syscall does, and exim expects); this
838 resulted in an indefinite loop.
840 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
841 to do ARC verification. The Authentication-Results: header line added
842 by the configuration then had no ARC item.
844 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
845 an incorrect mode was used for reading the result, resulting in it being
852 JH/01 Remove code calling the customisable local_scan function, unless a new
853 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
855 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
856 non-signal-safe functions being used.
858 JH/03 Bug 2269: When presented with a received message having a stupidly large
859 number of DKIM-Signature headers, disable DKIM verification to avoid
860 a resource-consumption attack. The limit is set at twenty.
862 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
863 report of oldest_pass in ${authres } in consequence, and separate out
864 some descriptions of reasons for verification fail.
866 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
867 files in the spool were present and unlocked. A queue-runner could spot
868 them, resulting in a duplicate delivery. Fix that by doing the unlock
869 after the unlink. Investigation by Tim Stewart. Take the opportunity to
870 add more error-checking on spoolfile handling while that code is being
873 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
874 No known attacks, no CVE, this is defensive hardening.
876 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
877 a queue-runner could start a delivery while other operations were ongoing.
878 Cutthrough delivery was a common victim, resulting in duplicate delivery.
879 Found and investigated by Tim Stewart. Fix by using the open message data
880 file handle rather than opening another, and not locally closing it (which
881 releases a lock) for that case, while creating the temporary .eml format
882 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
884 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
885 $sender_verify_failure/$recipient_verify_failure to "random".
887 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
890 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
891 Previously this would segfault.
893 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
896 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
897 like zero, since the resolver should be doing this for us, But we need one
898 as a CNAME but no MX presence gets the CNAME returned; we need to check
899 that doesn't point to an MX to declare it "no MX returned" rather than
900 "error, loop". A new main option is added so the older capability of
901 following some limited number of chain links is maintained.
903 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
905 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
906 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
907 now-deprecated earlier definitions used only specified the range up to TLS
908 1.2 (in the older-version library docs).
910 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
912 JH/15 Rework TLS client-side context management. Stop using a global, and
913 explicitly pass a context around. This enables future use of TLS for
914 connections to service-daemons (eg. malware scanning) while a client smtp
915 connection is using TLS; with cutthrough connections this is quite likely.
917 JH/16 Fix ARC verification to do AS checks in reverse order.
919 JH/17 Support a "tls" option on the ${readsocket } expansion item.
921 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
922 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
923 Previously the "utf8" would be re-prepended for every additional message.
925 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
926 Previously thery were accepted, resulting in issues when attempting to
927 forward messages to a non-supporting MTA.
929 PP/02 Let -n work with printing macros too, not just options.
931 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
932 one parent address was copied, and bogus data was used at delivery-logging
933 time. Either a crash (after delivery) or bogus log data could result.
934 Discovery and analysis by Tim Stewart.
936 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
937 Previously if the string ended mid-character, we did not insert the
938 promised '?' replacement.
940 PP/04 Documentation: current string operators work on bytes, not codepoints.
942 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
943 should pack well giving a smaller memory footprint so better caching and
944 therefore performance. Group the declarations where this can't be done so
945 that the byte-sized flag variables are not interspersed among pointer
946 variables, giving a better chance of good packing by the compiler.
948 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
949 non-null, to avoid issues with sites running BATV. Previously reports were
950 sent with an empty envelope sender so looked like bounces.
952 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
953 The ignore_error flag wasn't being returned from the filter subprocess so
954 was not set for later routers. Investigation and fix by Matthias Kurz.
956 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
957 and a msg:complete for the whole, when a message is manually removed using
958 -Mrm. Developement by Matthias Kurz, hacked on by JH.
960 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
961 a "Gnu special" function, asprintf() in the DB utility binary builds; I
962 hope that is portable enough.
964 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
965 requiring a known-CA anchor certificate; make it now rely entirely on the
966 TLSA as an anchor. Checking the name on the leaf cert against the name
967 on the A-record for the host is still done for TA (but not for EE mode).
969 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
970 included in delivery lines for non-proxied connections, when compiled with
971 SUPPORT_SOCKS and running with proxy logging enabled.
973 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
974 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
975 move the existing event to fire before the normal logging of message
976 failure so that custom logging is bracketed by normal logging.
978 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
979 msg:fail:internal event. Developement by Matthias Kurz.
981 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
982 far too small for todays use of crypto signatures stored there. Go all
983 the way to the max DNS message size of 64kB, even though this might be
984 overmuch for IOT constrained device use.
986 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
987 copy a string over itself. The library routine is documented as not
988 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
990 JH/32 For main options check_spool_space and check_inode_space, where the
991 platform supports 64b integers, support more than the previous 2^31 kB
992 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
993 the previous G, M, k.
995 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
996 $authenticated_fail_id variable on authentication failure. Previously
999 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1000 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1001 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1002 GNUTLS_SEC_PARAM_MEDIUM.
1004 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1005 side. Previously we would continue as if no SNI had been received.
1007 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1008 "<" character, which should be a list-separator specification, we walked
1009 off past the nul-terimation.
1011 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1012 causes) even when the retry time is not yet met. Previously they were
1013 not, meaning that when (say) an account was over-quota and temp-rejecting,
1014 and multiple senders' messages were queued, only one sender would get
1015 notified on each configured delay_warning cycle.
1017 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1019 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1020 cipher-suites, an error can be left on the stack even for a succeeding
1021 accept; this results in impossible error messages when a later operation
1024 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1025 return error codes indicating retry. Under TLS1.3 this becomes required.
1027 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1028 it only wrote the new authenticators, resulting in a lack of tracking of
1029 peer changes of ESMTP extensions until the next cache flush.
1031 JH/41 Fix the loop reading a message header line to check for integer overflow,
1032 and more-often against header_maxsize. Previously a crafted message could
1033 induce a crash of the recive process; now the message is cleanly rejected.
1035 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1036 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1042 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1043 When redis_servers is set to a list of > 1 element, and the Redis servers
1044 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1045 case of MOVED into a DEFER case instead, thus moving the query onto the
1046 next server in the list. For a cluster of N elements, all N servers must
1047 be defined in redis_servers.
1049 GF/02 Catch and remove uninitialized value warning in exiqsumm
1050 Check for existence of @ARGV before looking at $ARGV[0]
1052 JH/01 Replace the store_release() internal interface with store_newblock(),
1053 which internalises the check required to safely use the old one, plus
1054 the allocate and data copy operations duplicated in both (!) of the
1055 extant use locations.
1057 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1058 modifier. This matches the restriction on the commandline.
1060 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1061 Previously only the last row was returned.
1063 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1064 we assumed that tags in the header were well-formed, and parsed the
1065 element content after inspecting only the first char of the tag.
1066 Assumptions at that stage could crash the receive process on malformed
1069 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1070 While running the DKIM ACL we operate on the Permanent memory pool so that
1071 variables created with "set" persist to the DATA ACL. Also (at any time)
1072 DNS lookups that fail create cache records using the Permanent pool. But
1073 expansions release any allocations made on the current pool - so a dnsdb
1074 lookup expansion done in the DKIM ACL releases the memory used for the
1075 DNS negative-cache, and bad things result. Solution is to switch to the
1076 Main pool for expansions.
1077 While we're in that code, add checks on the DNS cache during store_reset,
1078 active in the testsuite.
1079 Problem spotted, and debugging aided, by Wolfgang Breyha.
1081 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1082 When none of the hosts presented to a transport match an already-open
1083 connection, close it and proceed with the list. Previously we would
1084 queue the message. Spotted by Lena with Yahoo, probably involving
1087 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1088 Previously a spurious "250 OK id=" response was appended to the proper
1091 JH/08 The "support for" informational output now, which built with Content
1092 Scanning support, has a line for the malware scanner interfaces compiled
1093 in. Interface can be individually included or not at build time.
1095 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1096 by the template makefile "src/EDITME". The "STREAM" support for an older
1097 ClamAV interface method is removed.
1099 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1100 rows affected is given instead).
1102 JH/11 The runtime Berkeley DB library version is now additionally output by
1103 "exim -d -bV". Previously only the compile-time version was shown.
1105 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1106 SMTP connection. Previously, when one had more recipients than the
1107 first, an abortive onward connection was made. Move to full support for
1108 multiple onward connections in sequence, handling cutthrough connection
1109 for all multi-message initiating connections.
1111 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1112 routers. Previously, a multi-recipient message would fail to match the
1113 onward-connection opened for the first recipient, and cause its closure.
1115 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1116 a timeout on read on a GnuTLS initiating connection, resulting in the
1117 initiating connection being dropped. This mattered most when the callout
1118 was marked defer_ok. Fix to keep the two timeout-detection methods
1121 JH/15 Relax results from ACL control request to enable cutthrough, in
1122 unsupported situations, from error to silently (except under debug)
1123 ignoring. This covers use with PRDR, frozen messages, queue-only and
1126 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1128 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1129 metadata, resulting in a crash in free().
1131 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1132 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1133 Broken also in d185889f4, with init system revamp.
1135 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1136 Previously we abruptly closed the connection after reading a malware-
1137 found indication; now we go on to read the "scan ok" response line,
1140 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1141 ACL. Previously, a crash would result.
1143 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1144 macro text after a replacement (previously it was only once per line) and
1145 by skipping builtin macros when searching for an uppercase lead character.
1147 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1148 for the build is renamed.
1150 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1151 was allocated for every new TLS startup, meaning one per message. Fix
1152 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1154 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1155 reported the original. Fix to report (as far as possible) the ACL
1156 result replacing the original.
1158 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1159 OpenSSL. Certificate information is loaded for every new TLS startup,
1160 and the resources needed to be freed.
1162 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1164 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1167 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1168 DATA response info to the (existing) per-recipient response info for
1169 the "C=" log element. It can have useful tracking info from the
1170 destination system. Patch from Simon Arlott.
1172 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1173 length value. Previously this would segfault.
1175 HS/02 Support Avast multiline protoocol, this allows passing flags to
1176 newer versions of the scanner.
1178 JH/28 Ensure that variables possibly set during message acceptance are marked
1179 dead before release of memory in the daemon loop. This stops complaints
1180 about them when the debug_store option is enabled. Discovered specifically
1181 for sender_rate_period, but applies to a whole set of variables.
1182 Do the same for the queue-runner and queue-list loops, for variables set
1183 from spool message files. Do the same for the SMTP per-message loop, for
1184 certain variables indirectly set in ACL operations.
1186 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1187 as a multi-recipient message from a mailinglist manager). The coding had
1188 an arbitrary cutoff number of characters while checking for more input;
1189 enforced by writing a NUL into the buffer. This corrupted long / fast
1190 input. The problem was exposed more widely when more pipelineing of SMTP
1191 responses was introduced, and one Exim system was feeding another.
1192 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1193 receiving system, and refused recipients seen by the sending system
1194 (propating to people being dropped from mailing lists).
1195 Discovered and pinpointed by David Carter.
1197 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1198 replaced by the ${authresults } expansion.
1200 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1202 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1203 allows proper process termination in container environments.
1205 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1206 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1208 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1209 and "err_perm", deprecated since 4.83 when the RFC-defined words
1210 "temperror" and "permerror" were introduced.
1212 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1213 transport-filters or DKIM-signing. The restriction was lost in the
1214 consolidation of verify-callout and delivery SMTP handling.
1215 Extend the restriction to also cover ARC-signing.
1217 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1218 in defer=pass mode supply a 450 to the initiator. Previously the message
1221 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1222 tls_require_ciphers is used as before.
1224 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1225 "pass_unscanned". Only tmpfails from the scanner are written to
1226 the paniclog, as they may require admin intervention (permission
1227 denied, license issues). Other scanner errors (like decompression
1228 bombs) do not cause a paniclog entry.
1230 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1231 Previously it was possible to log spurious information in receive log
1234 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1235 triggered odd behaviour from Outlook Express clients.
1237 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1240 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1241 since the IETF WG has not yet settled on that versus the original
1242 "bare" representation.
1244 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1245 Previously the millisecond value corrupted the output.
1246 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1247 corrupted the output.
1253 JH/01 Rework error string handling in TLS interface so that the caller in
1254 more cases is responsible for logging. This permits library-sourced
1255 string to be attached to addresses during delivery, and collapses
1256 pairs of long lines into single ones.
1258 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1259 during configuration. Wildcards are allowed and expanded.
1261 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1262 This permits better logging.
1264 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1265 do not close it down and have the child transport start it up again on
1266 the passed-on TCP connection. Instead, proxy the child (and any
1267 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1268 affected: the continued delivery log lines do not have any DNSSEC, TLS
1269 Certificate or OCSP information. TLS cipher information is still logged.
1271 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1272 identical IP addresses on different listening ports. Will also affect
1275 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1276 add noisy ifdef guards to special-case this sillyness.
1277 Patch from Bernd Kuhls.
1279 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1280 than 255 are no longer allowed.
1282 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1283 Disable the session-cache too, which might reduce our load. Since we
1284 currrectly use a new context for every connection, both as server and
1285 client, there is no benefit for these.
1286 GnuTLS appears to not support tickets server-side by default (we don't
1287 call gnutls_session_ticket_enable_server()) but client side is enabled
1288 by default on recent versions (3.1.3 +) unless the PFS priority string
1291 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1292 <https://reproducible-builds.org/specs/source-date-epoch/>.
1294 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1295 the check for any unsuccessful recipients did not notice the limit, and
1296 erroneously found still-pending ones.
1298 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1299 MSG_MORE. Only in-clear (not on TLS connections).
1301 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1302 transport-filter is involved we can buffer the headers in memory for
1303 creating the signature, and read the spool data file once for the
1304 signature and again for transmission.
1306 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1307 4.77 as the kernel support then wasn't solid, having issues in 64bit
1308 mode. Now, it's been long enough. Add support for FreeBSD also.
1310 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1311 case where the routing stage had gathered several addresses to send to
1312 a host before calling the transport for the first, we previously failed
1313 to close down TLS in the old transport process before passing the TCP
1314 connection to the new process. The new one sent a STARTTLS command
1315 which naturally failed, giving a failed delivery and bloating the retry
1316 database. Investigation and fix prototype from Wolfgang Breyha.
1318 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1319 false-negatives in the check that the sender had not preempted a response
1320 or prompt from Exim (running as a server), due to that code's lack of
1321 awareness of the SMTP input buffering.
1323 PP/04 Add commandline_checks_require_admin option.
1324 Exim drops privileges sanely, various checks such as -be aren't a
1325 security problem, as long as you trust local users with access to their
1326 own account. When invoked by services which pass untrusted data to
1327 Exim, this might be an issue. Set this option in main configuration
1328 AND make fixes to the calling application, such as using `--` to stop
1331 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1332 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1333 responses to those, into a single TLS record each way (this usually means
1334 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1337 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1338 affects you only if you're dancing at the edge of the param size limits.
1339 If you are, and this message makes sense to you, then: raise the
1340 configured limit or use OpenSSL 1.1. Nothing we can do for older
1343 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1344 cmdline element to get the documented default one. Previously it was
1347 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1348 are unsuable for non-message-specific reasons - eg. connection timeout,
1349 banner-time rejection.
1351 JH/16 Fix logging of delivery remote port, when specified by router, under
1354 PP/06 Repair manualroute's ability to take options in any order, even if one
1355 is the name of a transport.
1358 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1360 JH/17 Change the list-building routines interface to use the expanding-string
1361 triplet model, for better allocation and copying behaviour.
1363 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1364 Previously it was constructed the first time a possibly-matching string
1365 was met in the configuration file input during startup; now it is done
1368 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1369 compatible one, to avoid the (poorly documented) possibility of a config
1370 file in the working directory redirecting the DB files, possibly correpting
1371 some existing file. CVE-2017-10140 assigned for BDB.
1373 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1374 cache-hot. Previously, although the result was properly cached, the
1375 initial verify call returned a defer.
1377 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1378 the main verify for receipient in uncached-mode.
1380 JH/22 Retire historical build files to an "unsupported" subdir. These are
1381 defined as "ones for which we have no current evidence of testing".
1383 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1384 if present. Previously it was ignored.
1386 JH/24 Start using specified-initialisers in C structure init coding. This is
1387 a C99 feature (it's 2017, so now considered safe).
1389 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1390 if was a fixed-sized field and bitmask ops via macros; it is now more
1393 PP/07 GitHub PR 56: Apply MariaDB build fix.
1394 Patch provided by Jaroslav Škarvada.
1396 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1397 during Coverity cleanups [4.87 JH/47]
1398 Diagnosis and fix provided by Michael Fischer v. Mollard.
1400 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1401 the right size to place the terminating semicolon on its own folded
1402 line, the header hash was calculated to an incorrect value thanks to
1403 the (relaxed) space the fold became.
1405 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1406 and confused the parent.
1408 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1409 which could crash as a result. This could lead to undeliverable messages.
1411 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1414 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1415 matters on fast-turnover and PID-randomising systems, which were getting
1416 out-of-order delivery.
1418 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1419 a possibly-overlapping copy. The symptom was that "Remote host closed
1420 connection in response to HELO" was logged instead of the actual 4xx
1423 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1424 Previously only that bufferd was discarded, resulting in SYMTP command
1427 JH/32 DKIM: when a message has multiple signatures matching an identity given
1428 in dkim_verify_signers, run the dkim acl once for each. Previously only
1429 one run was done. Bug 2189.
1431 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1432 "panic the current process" to "deliberately defer". The panic log is
1433 still written with the problem list name; the mail and reject logs now
1434 get a temp-reject line for the message that was being handled, saying
1435 something like "domains check lookup or other defer". The SMTP 451
1436 message is still "Temporary local problem".
1438 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1439 A crafted sequence of BDAT commands could result in in-use memory beeing
1440 freed. CVE-2017-16943.
1442 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1443 from SMTP input. Previously it was always done; now only done for DATA
1444 and not BDAT commands. CVE-2017-16944.
1446 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1447 to the message (such as an overlong header line). Previously this was
1448 not done and we did not exit BDAT mode. Followon from the previous item
1449 though a different problem.
1455 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1456 than -2003 did; needs libidn2 in addition to libidn.
1458 JH/02 The path option on a pipe transport is now expanded before use.
1460 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1461 Patch provided by "Björn", documentation fix added too.
1463 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1464 missing a wire-to-host endian conversion.
1466 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1467 close after a BDAT command line could be taken as a following command,
1468 giving a synch failure. Fix by only checking for synch immediately
1469 before acknowledging the chunk.
1471 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1472 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1473 macro. Patches provided by Josh Soref.
1475 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1476 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1477 by IANA as a service extension. However, John Klensin suggests that we
1480 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1481 the dkim code may be unix-mode line endings rather than smtp wire-format
1482 CRLF, so prepend a CR to any bare LF.
1484 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1485 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1487 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1488 the first were themselves being wrongly included in the feed into dkim
1489 processing; with most chunk sizes in use this resulted in an incorrect
1490 body hash calculated value.
1492 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1493 DKIM signature block, for verification. Although advised against by
1494 standards it is specifically not ruled illegal.
1496 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1498 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1499 missing a body hash (the bh= tag).
1501 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1502 It seems that HAProxy sends the Proxy Protocol information in clear and
1503 only then does a TLS startup, so do the same.
1505 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1506 TCP connections (such as for Spamd) unless the daemon successfully set
1507 Fast Open mode on its listening sockets. This fixes breakage seen on
1508 too-old kernels or those not configured for Fast Open, at the cost of
1509 requiring both directions being enabled for TFO, and TFO never being used
1510 by non-daemon-related Exim processes.
1512 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1513 endings, at least on the first header line. Try to canonify any that get
1514 past that check, despite the cost.
1516 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1517 now limited to an arbitrary five deep, while parsing addresses with the
1518 strip_excess_angle_brackets option enabled.
1520 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1521 instead leave the unprompted TLS handshake in socket buffer for the
1522 TLS library to consume.
1524 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1526 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1528 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1529 region in one operation, for speed, and this leaves assigned pointers
1530 dangling. Add checks run only under the testsuite which checks all
1531 variables at a store-reset and panics on a dangling pointer; add code
1532 explicitly nulling out all the variables discovered. Fixes one known
1533 bug: a transport crash, where a dangling pointer for $sending_ip_address
1534 originally assigned in a verify callout, is re-used.
1536 PP/06 Drop '.' from @INC in various Perl scripts.
1538 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1540 PP/08 Reduce a number of compilation warnings under clang; building with
1541 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1542 should be warning-free.
1544 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1546 HS/01 Fix portability problems introduced by PP/08 for platforms where
1547 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1549 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1550 chunk. This allows us to accept broken chunked messages. We need a more
1551 general solution here.
1553 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1554 already-broken messages in the queue.
1556 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1558 JH/19 Fix reference counting bug in routing-generated-address tracking.
1564 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1565 supports it and a size is available (ie. the sending peer gave us one).
1567 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1568 years of being deprecated). The replacements are the ACLs
1569 acl_smtp_mime and acl_not_smtp_mime.
1571 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1572 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1573 or even an in-clear connection were permitted. Now, if the host lookup
1574 was dnssec and dane was requested then the host is only used if the
1575 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1576 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1577 if one fails this test.
1578 This means that a poorly-configured remote DNS will make it incommunicado;
1579 but it protects against a DNS-interception attack on it.
1581 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1582 non-noisy when a race steals the message being considered.
1584 JH/05 If main configuration option tls_certificate is unset, generate a
1585 self-signed certificate for inbound TLS connections.
1587 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1588 in rewrites and routers.
1590 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1591 and logged a warning sing 4.83; now they are a configuration file error.
1593 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1594 (lacking @domain). Apply the same qualification processing as RCPT.
1596 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1598 JH/10 Support ${sha256:} applied to a string (as well as the previous
1601 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1602 a cutthrough deliver is pending, as we always want to make a connection.
1603 This also avoids re-routing the message when later placing the cutthrough
1604 connection after a verify cache hit.
1605 Do not update it with the verify result either.
1607 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1608 when routing results in more than one destination address.
1610 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1611 signing (which inhibits the cutthrough capability). Previously only
1612 the presence of an option was tested; now an expansion evaluating as
1613 empty is permissible (obviously it should depend only on data available
1614 when the cutthrough connection is made).
1616 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1617 the relevant preceding SMTP command did not note the pipelining mode.
1619 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1620 Previously they were not counted.
1622 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1623 as one having no matching records. Previously we deferred the message
1624 that needed the lookup.
1626 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1627 distinguished as "(=".
1629 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1630 for missing MX records. Previously it only worked for missing A records.
1632 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1634 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1635 after the data-go-ahead and data-ack. Patch from Jason Betts.
1637 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1638 even for a "none" policy. Patch from Tony Meyer.
1640 JH/22 Fix continued use of a connection for further deliveries. If a port was
1641 specified by a router, it must also match for the delivery to be
1644 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1645 When one of the recipients of a message was unsuitable for the connection
1646 (has no matching addresses), we lost track of needing to mark it
1647 deferred. As a result mail would be lost.
1649 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1651 JH/25 Decoding ACL controls is now done using a binary search; the source code
1652 takes up less space and should be simpler to maintain. Merge the ACL
1653 condition decode tables also, with similar effect.
1655 JH/26 Fix problem with one_time used on a redirect router which returned the
1656 parent address unchanged. A retry would see the parent address marked as
1657 delivered, so not attempt the (identical) child. As a result mail would
1660 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1661 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1662 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1665 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1666 Main config options check_{log,spool}_{inodes,space} are now
1667 100 inodes, 10MB unless set otherwise in the configuration.
1669 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1670 Previously it only applied to the main-section connection policy
1673 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1675 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1676 by me. Added RFC7919 DH primes as an alternative.
1678 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1679 are not in the system include path.
1681 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1682 GnuTLS, when a session startup failed (eg because the client disconnected)
1683 Exim did stdio operations after fclose. This was exposed by a recent
1684 change which nulled out the file handle after the fclose.
1686 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1687 signed directly by the cert-signing cert, rather than an intermediate
1688 OCSP-signing cert. This is the model used by LetsEncrypt.
1690 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1692 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1693 an incoming connection.
1695 HS/02 Bug 1802: Do not half-close the connection after sending a request
1698 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1699 fallback to "prime256v1".
1701 JH/34 SECURITY: Use proper copy of DATA command in error message.
1702 Could leak key material. Remotely exploitable. CVE-2016-9963.
1708 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1709 and 3.4.4 - once the server is enabled to respond to an OCSP request
1710 it does even when not requested, resulting in a stapling non-aware
1711 client dropping the TLS connection.
1713 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1714 support variable-length bit vectors. No functional change.
1716 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1717 The I= interface field on outgoing lines is now after the H= remote
1718 host field, same as incoming lines. There is a separate
1719 outgoing_interface log selector which allows you to disable the
1722 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1723 If not running log_selector +smtp_connection the mainlog would be held
1724 open indefinitely after a "too many connections" event, including to a
1725 deleted file after a log rotate. Leave the per net connection logging
1726 leaving it open for efficiency as that will be quickly detected by the
1727 check on the next write.
1729 HS/01 Bug 1671: Fix post transport crash.
1730 Processing the wait-<transport> messages could crash the delivery
1731 process if the message IDs didn't exist for some reason. When
1732 using 'split_spool_directory=yes' the construction of the spool
1733 file name failed already, exposing the same netto behaviour.
1735 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1736 mime_regex ACL conditions.
1738 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1739 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1740 to HELO, local diagnostic string.
1742 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1743 log line to debug. Even when configured with a tls authenticator many
1744 client connections are expected to not authenticate in this way, so
1745 an authenticate fail is not an error.
1747 HS/02 Add the Exim version string to the process info. This way exiwhat
1748 gives some more detail about the running daemon.
1750 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1751 matter for fast-change records such as DNSBLs.
1753 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1754 retry record, even if constant. There may be multiple transports with
1755 different interface settings and the retry behaviour needs to be kept
1758 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1760 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1762 JH/10 Bug 840: fix log_defer_output option of pipe transport
1764 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1765 hosts_randomize. This matters a lot when combined with mua_wrapper.
1767 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1768 ${quote_pgsql:<string>} operator.
1770 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1772 JH/14 Change header file name for freeradius-client. Relevant if compiling
1773 with Radius support; from the Gentoo tree and checked under Fedora.
1775 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1777 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1778 option transport_filter, meaning no filtering.
1780 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1782 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1783 defaults to "*" (all hosts). The variable is now available when not built
1784 with TLS, default unset, mainly to enable keeping the testsuite sane.
1785 If a server certificate is not supplied (via tls_certificate) an error is
1786 logged, and clients will find TLS connections fail on startup. Presumably
1787 they will retry in-clear.
1788 Packagers of Exim are strongly encouraged to create a server certificate
1789 at installation time.
1791 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1792 with the $config_file variable.
1794 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1795 in transport context, after the attempt, and per-recipient. The latter type
1796 is per host attempted. The event data is the error message, and the errno
1797 information encodes the lookup type (A vs. MX) used for the (first) host,
1798 and the trailing two digits of the smtp 4xx response.
1800 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1801 to write to mainlog (or rejectlog, paniclog) in the window between file
1802 creation and permissions/ownership being changed. Particularly affects
1803 installations where exicyclog is run as root, rather than exim user;
1804 result is that the running daemon panics and dies.
1806 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1808 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1809 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1810 "pri" and "weight". Note that the previous implicit priority given by the
1811 list order is no longer honoured.
1813 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1814 for DKIM processing.
1816 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1817 by defining SUPPORT_SOCKS.
1819 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1820 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1821 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1822 variables are renamed to proxy_{local,external}_{address,port}.
1824 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1825 for a build by defining SUPPORT_I18N
1827 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1828 of the query string, and make ${quote_redis:} do that quoting.
1830 JH/29 Move Events support from Experimental to mainline, enabled by default
1831 and removable for a build by defining DISABLE_EVENT.
1833 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1835 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1836 cached by the daemon.
1838 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1839 by defining LOOKUP_REDIS. The libhiredis library is required.
1841 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1842 keys are given for lookup.
1844 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1845 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1846 only supported when built with TLS support. The PolarSSL SHA routines
1847 are still used when the TLS library is too old for convenient support.
1849 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1850 openssl_options), for security. OpenSSL forces this from version 1.1.0
1851 server-side so match that on older versions.
1853 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1854 allocation for $value could be released as the expansion processing
1855 concluded, but leaving the global pointer active for it.
1857 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1858 and to use the domains and local_parts ACL conditions.
1860 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1861 incorrectly not doubled on cutthrough transmission, hence seen as a
1862 body-termination at the receiving system - resulting in truncated mails.
1863 Commonly the sender saw a TCP-level error, and retransmitted the message
1864 via the normal store-and-forward channel. This could result in duplicates
1865 received - but deduplicating mailstores were liable to retain only the
1866 initial truncated version.
1868 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1870 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1872 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1873 we're in there, support oversigning also; bug 1309.
1875 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1877 HS/04 Add support for keep_environment and add_environment options.
1879 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1880 either intentional arithmetic overflow during PRNG, or testing config-
1883 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1884 delivery resulted in actual delivery. Cancel cutthrough before DATA
1887 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1888 rejects a recipient: pass the reject to the originator.
1890 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1891 Many were false-positives and ignorable, but it's worth fixing the
1894 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1895 for the new environment-manipulation done at startup. Move the routines
1896 from being local to tls.c to being global via the os.c file.
1898 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1899 an extract embedded as result-arg for a map, the first arg for extract
1900 is unavailable so we cannot tell if this is a numbered or keyed
1901 extraction. Accept either.
1907 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1910 JH/02 The smtp transport option "multi_domain" is now expanded.
1912 JH/03 The smtp transport now requests PRDR by default, if the server offers
1915 JH/04 Certificate name checking on server certificates, when exim is a client,
1916 is now done by default. The transport option tls_verify_cert_hostnames
1917 can be used to disable this per-host. The build option
1918 EXPERIMENTAL_CERTNAMES is withdrawn.
1920 JH/05 The value of the tls_verify_certificates smtp transport and main options
1921 default to the word "system" to access the system default CA bundle.
1922 For GnuTLS, only version 3.0.20 or later.
1924 JH/06 Verification of the server certificate for a TLS connection is now tried
1925 (but not required) by default. The verification status is now logged by
1926 default, for both outbound TLS and client-certificate supplying inbound
1929 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1932 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1933 Status Notification (bounce) messages are now MIME format per RFC 3464.
1934 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1935 under the control of the dsn_advertise_hosts option, and routers may
1936 have a dsn_lasthop option.
1938 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1939 default, modifiable by a malware= option. The list separator for
1940 the options can now be changed in the usual way. Bug 68.
1942 JH/10 The smtp_receive_timeout main option is now expanded before use.
1944 JH/11 The incoming_interface log option now also enables logging of the
1945 local interface on delivery outgoing connections.
1947 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1948 if the interface and destination host and port all match.
1950 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1953 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1954 Patch from Andrew Lewis.
1956 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1957 now supports optional time-restrictions, weighting, and priority
1958 modifiers per server. Patch originally by <rommer@active.by>.
1960 JH/16 The spamd_address main option now supports a mixed list of local
1961 and remote servers. Remote servers can be IPv6 addresses, and
1962 specify a port-range.
1964 JH/17 Bug 68: The spamd_address main option now supports an optional
1965 timeout value per server.
1967 JH/18 Bug 1581: Router and transport options headers_add/remove can
1968 now have the list separator specified.
1970 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1973 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1976 JH/21 Support for the A6 type of dns record is withdrawn.
1978 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1979 rather than the verbs used.
1981 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1982 from 255 to 1024 chars.
1984 JH/24 Verification callouts now attempt to use TLS by default.
1986 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1987 are generic router options now. The defaults didn't change.
1989 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1990 Original patch from Alexander Shikoff, worked over by JH.
1992 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1993 files if ZCAT_COMMAND is not executable.
1995 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1997 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1999 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2000 Normally benign, it bites when the pair was led to by a CNAME;
2001 modern usage is to not canonicalize the domain to a CNAME target
2002 (and we were inconsistent anyway for A-only vs AAAA+A).
2004 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2006 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2007 when evaluating $sender_host_dnssec.
2009 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2010 $sender_helo_dnssec variable.
2012 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2014 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2016 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2018 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2019 documented as working, but never had. Support all but $spam_report.
2021 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2022 added for tls authenticator.
2024 HS/03 Add perl_taintmode main config option
2030 TL/01 When running the test suite, the README says that variables such as
2031 no_msglog_check are global and can be placed anywhere in a specific
2032 test's script, however it was observed that placement needed to be near
2033 the beginning for it to behave that way. Changed the runtest perl
2034 script to read through the entire script once to detect and set these
2035 variables, reset to the beginning of the script, and then run through
2036 the script parsing/test process like normal.
2038 TL/02 The BSD's have an arc4random API. One of the functions to induce
2039 adding randomness was arc4random_stir(), but it has been removed in
2040 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2041 function when detected.
2043 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2044 cause callback expansion.
2046 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2047 syntax errors in an expansion can be treated as a string instead of
2048 logging or causing an error, due to the internal use of bool_lax
2049 instead of bool when processing it.
2051 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2052 server certificates when making smtp deliveries.
2054 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2056 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2058 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2060 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2061 Merged patch from Sebastian Wiedenroth.
2063 JH/05 Fix results-pipe from transport process. Several recipients, combined
2064 with certificate use, exposed issues where response data items split
2065 over buffer boundaries were not parsed properly. This eventually
2066 resulted in duplicates being sent. This issue only became common enough
2067 to notice due to the introduction of connection certificate information,
2068 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2070 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2071 size buffer was used, resulting in syntax errors when an expansion
2074 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2075 version 3.3.6 or later.
2077 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2078 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2079 both become "event_action", the variables become $event_name, $event_data
2080 and $event_defer_errno. There is a new variable $verify_mode, usable in
2081 routers, transports and related events. The tls:cert event is now also
2082 raised for inbound connections, if the main configuration event_action
2085 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2086 early OCSP support, but no stapling (appears to be less than 1.0.0).
2088 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2089 server certificate names available under the smtp transport option
2090 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2093 JH/10 Time-related extraction expansions from certificates now use the main
2094 option "timezone" setting for output formatting, and are consistent
2095 between OpenSSL and GnuTLS compilations. Bug 1541.
2097 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2098 encoded parameter in the incoming message. Bug 1558.
2100 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2101 include certificate info, eximon was claiming there were spoolfile
2104 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2106 JH/14 Log delivery-related information more consistently, using the sequence
2107 "H=<name> [<ip>]" wherever possible.
2109 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2110 are problematic for Debian distribution, omit them from the release
2113 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2115 JH/16 Fix string representation of time values on 64bit time_t architectures.
2118 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2119 output list separator was used.
2124 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2125 checkers that were complaining about end of non-void function with no
2128 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2129 This was a regression introduced in 4.83 by another bugfix.
2131 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2133 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2134 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2140 TF/01 Correctly close the server side of TLS when forking for delivery.
2142 When a message was received over SMTP with TLS, Exim failed to clear up
2143 the incoming connection properly after forking off the child process to
2144 deliver the message. In some situations the subsequent outgoing
2145 delivery connection happened to have the same fd number as the incoming
2146 connection previously had. Exim would try to use TLS and fail, logging
2147 a "Bad file descriptor" error.
2149 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2150 utilities have not been installed.
2152 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2153 temporary space as the ACL may create new global variables.
2155 TL/01 LDAP support uses per connection or global context settings, depending
2156 upon the detected version of the libraries at build time.
2158 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2159 to extract and use the src ip:port in logging and expansions as if it
2160 were a direct connection from the outside internet. PPv2 support was
2161 updated based on HAProxy spec change in May 2014.
2163 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2165 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2166 Properly escape header and check for NULL return.
2168 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2171 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2173 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2174 characters in header names, implemented as a verify condition.
2175 Contributed by Michael Fischer v. Mollard.
2177 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2178 results permerror and temperror. Previous values are deprecated but
2179 still accepted. In a future release, err_perm and err_temp will be
2180 completely removed, which will be a backward incompatibility if the
2181 ACL tests for either of these two old results. Patch contributed by
2182 user bes-internal on the mailing list.
2184 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2186 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2187 selectors, in both main and reject logs.
2189 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2192 JH/07 Add malware type "sock" for talking to simple daemon.
2194 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2196 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2197 routers/transports under cutthrough routing.
2199 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2200 numbers. Touch up "bool" conditional to keep the same definition.
2202 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2204 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2206 JH/12 Expand items in router/transport headers_add or headers_remove lists
2207 individually rather than the list as a whole. Bug 1452.
2209 Required for reasonable handling of multiple headers_ options when
2210 they may be empty; requires that headers_remove items with embedded
2211 colons must have them doubled (or the list-separator changed).
2213 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2214 view the policy declared in the DMARC record. Currently, $dmarc_status
2215 is a combined value of both the record presence and the result of the
2218 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2220 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2221 dnslookup router and the smtp transport (applying to the forward
2224 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2225 of ldap servers used for a specific lookup. Patch provided by Heiko
2228 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2229 New variable $lookup_dnssec_authenticated for observability.
2231 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2232 Patch submitted by Lars Timman.
2234 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2236 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2237 Requires trusted mode and valid format message id, aborts otherwise.
2238 Patch contributed by Heiko Schlichting.
2240 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2241 certextract with support for various fields. Bug 1358.
2243 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2244 is requested by default, modifiable by smtp transport option
2247 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2248 operate on certificate variables to give certificate fingerprints
2249 Also new ${sha256:cert_variable}.
2251 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2253 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2254 Christian Aistleitner.
2256 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2258 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2259 file. Patch from Wolfgang Breyha.
2261 JH/25 Expand the coverage of the delivery $host and $host_address to
2262 client authenticators run in verify callout. Bug 1476.
2264 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2265 align with daemon_smtp_ports. Bug 72.
2267 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2268 support and error reporting did not work properly.
2270 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2271 and is readable. Patch from Andrew Colin Kissa.
2273 TL/14 Enhance documentation of ${run expansion and how it parses the
2274 commandline after expansion, particularly in the case when an
2275 unquoted variable expansion results in an empty value.
2277 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2279 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2280 by renaming away from it. Addresses GH issue 15, reported by
2283 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2285 TL/15 SECURITY: prevent double expansion in math comparison functions
2286 (can expand unsanitized data). Not remotely exploitable.
2293 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2295 PP/02 Make -n do something, by making it not do something.
2296 When combined with -bP, the name of an option is not output.
2298 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2301 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2302 $sender_host_name and config options to manage this, and basic check
2305 PP/05 DSCP support for outbound connections and control modifier for inbound.
2307 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2308 (Only plugin which currently uses this is kerberos4, which nobody should
2309 be using, but we should make it available and other future plugins might
2310 conceivably use it, even though it would break NAT; stuff *should* be
2311 using channel bindings instead).
2313 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2314 name; added for Sendmail compatibility; requires admin caller.
2315 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2316 just ignore it); requires trusted caller.
2317 Also parse but ignore: -Ac -Am -X<logfile>
2320 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2322 TL/02 Add +smtp_confirmation as a default logging option.
2324 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2325 Patch by Magnus Holmgren from 2007-02-20.
2327 TL/04 Bugzilla 1281 - Spec typo.
2328 Bugzilla 1283 - Spec typo.
2329 Bugzilla 1290 - Spec grammar fixes.
2331 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2333 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2335 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2336 mailing list by by Dmitry Isaikin.
2338 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2340 JH/02 Support "G" suffix to numbers in ${if comparisons.
2342 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2344 NM/01 Bugzilla 1197 - Spec typo
2345 Bugzilla 1196 - Spec examples corrections
2347 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2349 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2350 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2353 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2354 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2356 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2357 "acl {{name}{arg}...}", and optional args on acl condition
2360 JH/05 Permit multiple router/transport headers_add/remove lines.
2362 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2364 JH/07 Avoid using a waiting database for a single-message-only transport.
2365 Performance patch from Paul Fisher. Bugzilla 1262.
2367 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2370 JH/09 Add $headers_added variable, with content from use of ACL modifier
2371 add_header (but not yet added to the message). Bugzilla 199.
2373 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2374 Pulled from Bugzilla 817 by Wolfgang Breyha.
2376 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2378 (nb: this is the same fix as in Exim 4.80.1)
2380 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2383 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2386 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2388 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2389 advertise SMTP AUTH mechanism to us, instead of a generic
2390 protocol violation error. Also, make Exim more robust to bad
2391 data from the Dovecot auth socket.
2393 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2395 When a queue runner is handling a message, Exim first routes the
2396 recipient addresses, during which it prunes them based on the retry
2397 hints database. After that it attempts to deliver the message to
2398 any remaining recipients. It then updates the hints database using
2401 So if a recipient address works intermittently, it can get repeatedly
2402 deferred at routing time. The retry hints record remains fresh so the
2403 address never reaches the final cutoff time.
2405 This is a fairly common occurrence when a user is bumping up against
2406 their storage quota. Exim had some logic in its local delivery code
2407 to deal with this. However it did not apply to per-recipient defers
2408 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2410 This change adds a proper retry rule check during routing so that the
2411 final cutoff time is checked against the message's age. We only do
2412 this check if there is an address retry record and there is not a
2413 domain retry record; this implies that previous attempts to handle
2414 the address had the retry_use_local_parts option turned on. We use
2415 this as an approximation for the destination being like a local
2416 delivery, as in LMTP.
2418 I suspect this new check makes the old local delivery cutoff check
2419 redundant, but I have not verified this so I left the code in place.
2421 TF/02 Correct gecos expansion when From: is a prefix of the username.
2423 Test 0254 submits a message to Exim with the header
2427 When I ran the test suite under the user fanf2, Exim expanded
2428 the header to contain my full name, whereas it should have added
2429 a Resent-Sender: header. It erroneously treats any prefix of the
2430 username as equal to the username.
2432 This change corrects that bug.
2434 GF/01 DCC debug and logging tidyup
2435 Error conditions log to paniclog rather than rejectlog.
2436 Debug lines prefixed by "DCC: " to remove any ambiguity.
2438 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2440 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2441 Bug spotted by Jeremy Harris; was flawed since initial commit.
2442 Would have resulted in OCSP responses post-SNI triggering an Exim
2443 NULL dereference and crash.
2445 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2447 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2448 Bug detection, analysis and fix by Samuel Thibault.
2449 Bugzilla 1331, Debian bug #698092.
2451 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2453 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2454 Server implementation by Todd Lyons, client by JH.
2455 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2456 config variable "prdr_enable" controls whether the server
2457 advertises the facility. If the client requests PRDR a new
2458 acl_data_smtp_prdr ACL is called once for each recipient, after
2459 the body content is received and before the acl_smtp_data ACL.
2460 The client is controlled by both of: a hosts_try_prdr option
2461 on the smtp transport, and the server advertisement.
2462 Default client logging of deliveries and rejections involving
2463 PRDR are flagged with the string "PRDR".
2465 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2466 fclose(). Diagnosis by Todd Lyons.
2468 PP/17 Update configure.default to handle IPv6 localhost better.
2469 Patch by Alain Williams (plus minor tweaks).
2472 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2473 This is now consistent with GnuTLS, and is now documented: the
2474 previous undocumented portable approach to treating the option as
2475 unset was to force an expansion failure. That still works, and
2476 an empty string is now equivalent.
2478 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2479 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2480 not performing validation itself.
2482 PP/20 Added force_command boolean option to pipe transport.
2483 Patch from Nick Koston, of cPanel Inc.
2485 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2488 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2490 PP/21 Fix eximon continuous updating with timestamped log-files.
2491 Broken in a format-string cleanup in 4.80, missed when I repaired the
2492 other false fix of the same issue.
2493 Report and fix from Heiko Schlichting.
2496 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2497 Report from Prashanth Katuri.
2499 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2500 It's SecureTransport, so affects any MacOS clients which use the
2501 system-integrated TLS libraries, including email clients.
2503 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2504 using a MIME ACL for non-SMTP local injection.
2505 Report and assistance in diagnosis by Warren Baker.
2507 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2509 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2511 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2512 last id that failed so it may be referenced in subsequent ACL's.
2514 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2517 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2518 ldap library initialization, allowing self-signed CA's to be
2519 used. Also properly sets require_cert option later in code by
2520 using NULL (global ldap config) instead of ldap handle (per
2521 session). Bug diagnosis and testing by alxgomz.
2523 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2524 the src/util/ subdirectory.
2526 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2527 renamed to Transport Post Delivery Action by Jeremy Harris, as
2530 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2531 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2532 redis_servers = needs to be configured which will be used by the redis
2533 lookup. Patch from Warren Baker, of The Packet Hub.
2535 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2537 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2538 hostname or reverse DNS when processing a host list. Used suggestions
2539 from multiple comments on this bug.
2541 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2543 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2544 Missed a few lines, added it to make the runtest require no keyboard
2547 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2548 contains upper case chars. Make router use caseful_local_part.
2550 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2551 support when GnuTLS has been built with p11-kit.
2557 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2559 This, or similar/improved, will also be change PP/11 of 4.82.
2565 PP/01 Handle short writes when writing local log-files.
2566 In practice, only affects FreeBSD (8 onwards).
2567 Bugzilla 1053, with thanks to Dmitry Isaikin.
2569 NM/01 Bugzilla 949 - Documentation tweak
2571 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2574 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2576 PP/02 Implemented gsasl authenticator.
2578 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2580 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2581 `pkg-config foo` for cflags/libs.
2583 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2584 with rest of GSASL and with heimdal_gssapi.
2586 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2587 `pkg-config foo` for cflags/libs for the TLS implementation.
2589 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2590 properties get this fed in as external SSF. A number of robustness
2591 and debugging improvements to the cyrus_sasl authenticator.
2593 PP/08 cyrus_sasl server now expands the server_realm option.
2595 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2596 Patch by Jeremy Harris.
2598 PP/10 Added dbmjz lookup type.
2600 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2602 PP/12 MAIL args handles TAB as well as SP, for better interop with
2603 non-compliant senders.
2604 Analysis and variant patch by Todd Lyons.
2606 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2607 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2608 Patch from Dirk Mueller <dmueller@suse.com>
2610 PP/13 tls_peerdn now print-escaped for spool files.
2611 Observed some $tls_peerdn in wild which contained \n, which resulted
2612 in spool file corruption.
2614 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2615 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2616 or write after TLS renegotiation, which otherwise led to messages
2619 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2620 as a tracking header (ie: a signed header comes before the signature).
2621 Patch from Wolfgang Breyha.
2623 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2624 comma-sep list; embedded commas doubled.
2626 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2628 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2630 Report and patch from Dmitry Banschikov.
2632 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2633 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2634 failures. We appear to now support TLS1.1+ with Exim.
2636 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2637 lets Exim select keys and certificates based upon TLS SNI from client.
2638 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2639 before an outbound SMTP session. New log_selector, +tls_sni.
2641 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2642 NULL dereference. Report and patch from Alun Jones.
2644 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2645 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2646 Not seeing resolver debug output on NetBSD, but suspect this is a
2647 resolver implementation change.
2649 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2650 Left warnings. Added "eximon gdb" invocation mode.
2652 PP/21 Defaulting "accept_8bitmime" to true, not false.
2654 PP/22 Added -bw for inetd wait mode support.
2656 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2657 locate the relevant includes and libraries. Made this the default.
2659 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2660 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2662 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2663 This may cause build issues on older platforms.
2665 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2666 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2667 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2668 Added SNI support via GnuTLS too.
2669 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2671 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2673 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2674 Applied second patch from Janne, implementing suggestion to default
2675 multiple-strings-in-record handling to match SPF spec.
2677 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2679 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2680 read-only, out of scope).
2681 Patch from Wolfgang Breyha, report from Stuart Northfield.
2683 PP/29 Fix three issues highlighted by clang analyser static analysis.
2684 Only crash-plausible issue would require the Cambridge-specific
2685 iplookup router and a misconfiguration.
2686 Report from Marcin Mirosław.
2688 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2690 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2691 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2692 As part of this, removing so much warning spew let me fix some minor
2693 real issues in debug logging.
2695 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2696 assignment on my part. Fixed.
2698 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2699 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2700 Janne Snabb (who went above and beyond: thank you).
2702 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2703 string otherwise requires a connection and a bunch more work and it's
2704 relatively easy to get wrong. Should also expose TLS library linkage
2707 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2708 64-bit ${eval} (JH/03).
2710 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2711 GNU libc to support some of the 64-bit stuff, should not lead to
2712 conflicts. Defined before os.h is pulled in, so if a given platform
2713 needs to override this, it can.
2715 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2716 protection layer was required, which is not implemented.
2717 Bugzilla 1254, patch from Wolfgang Breyha.
2719 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2720 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2721 tls_dhparam take prime identifiers. Also unbreak combination of
2722 OpenSSL+DH_params+TLSSNI.
2724 PP/39 Disable SSLv2 by default in OpenSSL support.
2730 PP/01 Solaris build fix for Oracle's LDAP libraries.
2731 Bugzilla 1109, patch from Stephen Usher.
2733 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2735 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2738 TF/02 Fix a couple more cases where we did not log the error message
2739 when unlink() failed. See also change 4.74-TF/03.
2741 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2742 lock up or crash if it happened to be inside a call to libc when it
2743 got a SIGUSR1 from exiwhat.
2745 The SIGUSR1 handler appends the current process status to the process
2746 log which is later printed by exiwhat. It used to use the general
2747 purpose logging code to do this, but several functions it calls are
2748 not safe for signals.
2750 The new output code in the SIGUSR1 handler is specific to the process
2751 log, and simple enough that it's easy to inspect for signal safety.
2752 Removing some special cases also simplifies the general logging code.
2753 Removing the spurious timestamps from the process log simplifies
2756 TF/04 Improved ratelimit ACL condition.
2758 The /noupdate option has been deprecated in favour of /readonly which
2759 has clearer semantics. The /leaky, /strict, and /readonly update modes
2760 are mutually exclusive. The update mode is no longer included in the
2761 database key; it just determines when the database is updated. (This
2762 means that when you upgrade Exim will forget old rate measurements.)
2764 Exim now checks that the per_* options are used with an update mode that
2765 makes sense for the current ACL. For example, when Exim is processing a
2766 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2767 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2768 must specify per_mail/readonly. If you omit the update mode it defaults to
2769 /leaky where that makes sense (as before) or /readonly where required.
2771 The /noupdate option is now undocumented but still supported for
2772 backwards compatibility. It is equivalent to /readonly except that in
2773 ACLs where /readonly is required you may specify /leaky/noupdate or
2774 /strict/noupdate which are treated the same as /readonly.
2776 A useful new feature is the /count= option. This is a generalization
2777 of the per_byte option, so that you can measure the throughput of other
2778 aggregate values. For example, the per_byte option is now equivalent
2779 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2781 The per_rcpt option has been generalized using the /count= mechanism
2782 (though it's more complicated than the per_byte equivalence). When it is
2783 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2784 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2785 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2786 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2787 Note that using per_rcpt with a non-readonly update mode in more than
2788 one ACL will cause the recipients to be double-counted. (The per_mail
2789 and per_byte options don't have this problem.)
2791 The handling of very low rates has changed slightly. If the computed rate
2792 is less than the event's count (usually one) then this event is the first
2793 after a long gap. In this case the rate is set to the same as this event's
2794 count, so that the first message of a spam run is counted properly.
2796 The major new feature is a mechanism for counting the rate of unique
2797 events. The new per_addr option counts the number of different
2798 recipients that someone has sent messages to in the last time period. It
2799 behaves like per_rcpt if all the recipient addresses are different, but
2800 duplicate recipient addresses do not increase the measured rate. Like
2801 the /count= option this is a general mechanism, so the per_addr option
2802 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2803 example, measure the rate that a client uses different sender addresses
2804 with the options per_mail/unique=$sender_address. There are further
2805 details in the main documentation.
2807 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2809 TF/06 Removed a few PCRE remnants.
2811 TF/07 Automatically extract Exim's version number from tags in the git
2812 repository when doing development or release builds.
2814 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2815 Bugzilla 879. Patch from Paul Fisher.
2817 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2818 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2821 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2823 PP/05 Variable $av_failed, true if the AV scanner deferred.
2824 Bugzilla 1078. Patch from John Horne.
2826 PP/06 Stop make process more reliably on build failure.
2827 Bugzilla 1087. Patch from Heiko Schlittermann.
2829 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2830 Bugzilla 1089. Patch from Heiko Schlittermann.
2832 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2833 Bugzilla 1131. Patch from Holger Weiß.
2835 PP/09 Handle IPv6 addresses with SPF.
2836 Bugzilla 860. Patch from Wolfgang Breyha.
2838 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2840 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2843 PP/11 match_* no longer expand right-hand-side by default.
2844 New compile-time build option, EXPAND_LISTMATCH_RHS.
2845 New expansion conditions, "inlist", "inlisti".
2847 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2849 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2851 PP/14 fix log_write() format string regression from TF/03.
2852 Bugzilla 1152. Patch from Dmitry Isaikin.
2858 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2860 PP/02 Harmonised TLS library version reporting; only show if debugging.
2861 Layout now matches that introduced for other libraries in 4.74 PP/03.
2863 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2865 PP/04 New "dns_use_edns0" global option.
2867 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2870 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2871 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2873 TK/01 Updated PolarSSL code to 0.14.2.
2874 Bugzilla 1097. Patch from Andreas Metzler.
2876 PP/07 Catch divide-by-zero in ${eval:...}.
2877 Fixes bugzilla 1102.
2879 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2882 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2883 format-string attack -- SECURITY: remote arbitrary code execution.
2885 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2886 time unintentionally subject to list matching rules, letting the header
2887 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2888 arbitrary string expansion). This allowed for information disclosure.
2890 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2891 INT_MIN/-1 -- value coerced to INT_MAX.
2897 NM/01 Workaround for PCRE version dependency in version reporting
2900 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2901 This fixes portability to compilers other than gcc, notably
2902 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2904 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2905 makefiles for portability to HP-UX and POSIX correctness.
2907 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2908 Also via indented variable definition in the Makefile.
2909 (Debugging by Oliver Heesakkers).
2911 PP/02 Restore caching of spamd results with expanded spamd_address.
2912 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2914 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2915 Improves build reliability. Fix from: Frank Elsner
2917 NM/02 Fix wide character breakage in the rfc2047 coding
2918 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2920 NM/03 Allow underscore in dnslist lookups
2921 Fixes bug 1026. Patch from Graeme Fowler
2923 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2924 Code patches from Adam Ciarcinski of NetBSD.
2926 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2929 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2930 is logged, to avoid truncation. Patch from John Horne.
2932 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2933 Patch from Jakob Hirsch.
2935 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2936 SQL string expansion failure details.
2937 Patch from Andrey Oktyabrski.
2939 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2940 Patch from Simon Arlott.
2942 PP/09 New lookups functionality failed to compile on old gcc which rejects
2943 extern declarations in function scope.
2944 Patch from Oliver Fleischmann
2946 PP/10 Use sig_atomic_t for flags set from signal handlers.
2947 Check getgroups() return and improve debugging.
2948 Fixed developed for diagnosis in bug 927 (which turned out to be
2951 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2952 Patch from Mark Zealey.
2954 PP/12 Bugzilla 1056: Improved spamd server selection.
2955 Patch from Mark Zealey.
2957 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2958 Based on patch from Heiko Schlittermann.
2960 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2961 Patch from Uwe Doering, sign-off by Michael Haardt.
2963 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2964 variable declaration deep within a block. Bug and patch from
2967 PP/15 lookups-Makefile IRIX compatibility coercion.
2969 PP/16 Make DISABLE_DKIM build knob functional.
2971 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2972 Patch by Simon Arlott
2974 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2975 variable argument macros. Our copy now differs from upstream.
2981 TF/01 Failure to get a lock on a hints database can have serious
2982 consequences so log it to the panic log.
2984 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2985 controlled using the smtp_confirmation log selector.
2987 TF/03 Include the error message when we fail to unlink a spool file.
2989 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2990 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2991 for maintaining out-of-tree patches for some time.
2993 PP/01 Bugzilla 139: Documentation and portability issues.
2994 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2995 Handle per-OS dynamic-module compilation flags.
2997 PP/02 Let /dev/null have normal permissions.
2998 The 4.73 fixes were a little too stringent and complained about the
2999 permissions on /dev/null. Exempt it from some checks.
3000 Reported by Andreas M. Kirchwitz.
3002 PP/03 Report version information for many libraries, including
3003 Exim version information for dynamically loaded libraries. Created
3004 version.h, now support a version extension string for distributors
3005 who patch heavily. Dynamic module ABI change.
3007 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3008 privilege escalation vulnerability whereby the Exim run-time user
3009 can cause root to append content of the attacker's choosing to
3012 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3015 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3016 If dropping privileges for untrusted macros, we disabled normal logging
3017 on the basis that it would fail; for the Exim run-time user, this is not
3018 the case, and it resulted in successful deliveries going unlogged.
3019 Fixed. Reported by Andreas Metzler.
3025 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3026 only prepend for the Resent-* case. Fixes regression introduced in
3027 Exim 4.70 by NM/22 for Bugzilla 607.
3029 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3030 increasing numbers of administrators be bitten by this.
3032 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3034 PP/03 Bugzilla 994: added openssl_options main configuration option.
3036 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3038 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3040 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3042 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3043 without a peer certificate, leading to a segfault because of an
3044 assumption that peers always have certificates. Be a little more
3045 paranoid. Problem reported by Martin Tscholak.
3047 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3048 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3049 NB: ClamAV planning to remove STREAM in "middle of 2010".
3050 CL also introduces -bmalware, various -d+acl logging additions and
3051 more caution in buffer sizes.
3053 PP/09 Implemented reverse_ip expansion operator.
3055 PP/10 Bugzilla 937: provide a "debug" ACL control.
3057 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3059 PP/12 Bugzilla 973: Implement --version.
3061 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3063 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3065 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3067 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3068 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3069 ignore trailing whitespace.
3071 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3073 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3076 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3077 Notification from Dr Andrew Aitchison.
3079 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3080 ExtendedDetectionInfo response format.
3081 Notification from John Horne.
3083 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3086 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3087 XSL and documented dependency on system catalogs, with examples of how
3090 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3093 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3094 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3095 configuration file which is writeable by the Exim user or group.
3097 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3098 of configuration files to cover files specified with the -C option if
3099 they are going to be used with root privileges, not just the default
3102 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3103 option (effectively making it always true).
3105 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3106 files to be used while preserving root privileges.
3108 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3109 that rogue child processes cannot use them.
3111 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3112 run-time user, instead of root.
3114 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3115 Exim run-time user without dropping privileges.
3117 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3118 result string, instead of calling string_vformat() twice with the same
3121 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3122 for other users. Others should always drop root privileges if they use
3123 -C on the command line, even for a whitelisted configure file.
3125 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3127 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3133 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3134 $data_path, and $header_path variables; fixed documentation bugs and
3137 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3138 exipick to access non-standard spools, including the "frozen" queue
3141 NM/01 Bugzilla 965: Support mysql stored procedures.
3142 Patch from Alain Williams
3144 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3146 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3147 Patch from Andreas Metzler
3149 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3150 Patch from Kirill Miazine
3152 NM/05 Bugzilla 671: Added umask to procmail example.
3154 JJ/03 installed exipick 20100323.0, fixing doc bug
3156 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3157 directory. Notification and patch from Dan Rosenberg.
3159 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3161 TK/02 Improve log output when DKIM signing operation fails.
3163 MH/01 Treat the transport option dkim_domain as a colon separated
3164 list, not as a single string, and sign the message with each element,
3165 omitting multiple occurences of the same signer.
3167 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3168 Bugzilla 985, 986. Patch by Simon Arlott
3170 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3171 Patch by Simon Arlott
3173 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3174 MBX locking. Notification from Dan Rosenberg.
3180 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3182 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3184 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3186 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3188 NM/04 Bugzilla 915: Flag broken perl installation during build.
3194 TK/01 Added patch by Johannes Berg that expands the main option
3195 "spamd_address" if it starts with a dollar sign.
3197 TK/02 Write list of recipients to X-Envelope-Sender header when building
3198 the mbox-format spool file for content scanning (suggested by Jakob
3201 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3202 (http://www.dcc-servers.net/) support via dccifd. Activated by
3203 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3205 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3206 by Mark Daniel Reidel <mr@df.eu>.
3208 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3209 When building exim an external PCRE library is now needed -
3210 PCRE is a system library on the majority of modern systems.
3211 See entry on PCRE_LIBS in EDITME file.
3213 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3214 conversation. Added nologin parameter to request.
3215 Patch contributed by Kirill Miazine.
3217 TF/01 Do not log submission mode rewrites if they do not change the address.
3219 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3221 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3222 log files in place. Contributed by Roberto Lima.
3224 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3226 TF/03 Bugzilla 615: When checking the local_parts router precondition
3227 after a local_part_suffix or local_part_prefix option, Exim now
3228 does not use the address's named list lookup cache, since this
3229 contains cached lookups for the whole local part.
3231 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3232 Robert Millan. Documentation is in experimental-spec.txt.
3234 TF/04 Bugzilla 668: Fix parallel build (make -j).
3236 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3238 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3239 Patch provided by Jan Srzednicki.
3241 TF/05 Leading white space used to be stripped from $spam_report which
3242 wrecked the formatting. Now it is preserved.
3244 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3245 that they are available at delivery time.
3247 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3249 TF/08 TLS error reporting now respects the incoming_interface and
3250 incoming_port log selectors.
3252 TF/09 Produce a more useful error message if an SMTP transport's hosts
3253 setting expands to an empty string.
3255 NM/06 Bugzilla 744: EXPN did not work under TLS.
3256 Patch provided by Phil Pennock.
3258 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3259 Patch provided by Richard Godbee.
3261 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3262 acl_smtp_notquit, added index entry.
3264 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3265 Patch provided by Eugene Bujak.
3267 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3268 accept(). Patch provided by Maxim Dounin.
3270 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3271 Patch provided by Phil Pennock.
3273 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3275 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3276 Patch provided by Brad "anomie" Jorsch.
3278 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3279 Patch provided by Dean Brooks.
3281 TK/05 Add native DKIM support (does not depend on external libraries).
3283 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3284 Patch provided by Graeme Fowler.
3286 NM/16 Bugzilla 851: Documentation example syntax fix.
3288 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3290 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3293 NM/19 Bugzilla 745: TLS version reporting.
3294 Patch provided by Phil Pennock.
3296 NM/20 Bugzilla 167: bool: condition support.
3297 Patch provided by Phil Pennock.
3299 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3300 clients. Patch provided by Phil Pennock.
3302 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3303 Patch provided by Brad "anomie" Jorsch.
3305 NM/23 Bugzilla 687: Fix misparses in eximstats.
3306 Patch provided by Heiko Schlittermann.
3308 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3309 Patch provided by Heiko Schlittermann.
3311 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3312 plus update to original patch.
3314 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3316 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3317 Patch provided by David Brownlee.
3319 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3321 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3323 NM/30 Bugzilla 888: TLS documentation bugfixes.
3325 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3327 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3328 Unlike the original bugzilla I have changed all shell scripts in src tree.
3330 NM/33 Bugzilla 898: Transport filter timeout fix.
3331 Patch by Todd Rinaldo.
3333 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3334 Patch by Serge Demonchaux.
3336 NM/35 Bugzilla 39: Base64 decode bug fixes.
3337 Patch by Jakob Hirsch.
3339 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3341 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3343 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3345 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3351 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3352 ALT-N's libdkim that I have put here:
3353 http://duncanthrax.net/exim-experimental/
3355 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3356 were called 'true' and it seems that C99 defines that as a reserved
3357 keyword to be used with 'bool' variable types. That means you could
3358 not include C99-style headers which use bools without triggering
3359 build errors in sieve.c.
3361 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3362 as mailq or other aliases. Changed the --help handling significantly
3363 to do whats expected. exim_usage() emits usage/help information.
3365 SC/01 Added the -bylocaldomain option to eximstats.
3367 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3369 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3371 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3377 PH/01 Another patch from the Sieve maintainer.
3379 PH/02 When an IPv6 address is converted to a string for single-key lookup
3380 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3381 dots are used instead of colons so that keys in lsearch files need not
3382 contain colons. This was done some time before quoting was made available
3383 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3384 (notated using the quote facility) so as to distinguish them from IPv4
3385 keys. This meant that lookups for IP addresses in host lists did not work
3386 for iplsearch lookups.
3388 This has been fixed by arranging for IPv6 addresses to be expressed with
3389 colons if the lookup type is iplsearch. This is not incompatible, because
3390 previously such lookups could never work.
3392 The situation is now rather anomalous, since one *can* have colons in
3393 ordinary lsearch keys. However, making the change in all cases is
3394 incompatible and would probably break a number of configurations.
3396 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3399 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3400 conversion specification without a maximum field width, thereby enabling
3401 a rogue spamd server to cause a buffer overflow. While nobody in their
3402 right mind would setup Exim to query an untrusted spamd server, an
3403 attacker that gains access to a server running spamd could potentially
3404 exploit this vulnerability to run arbitrary code as the Exim user.
3406 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3407 $primary_hostname instead of what libspf2 thinks the hosts name is.
3409 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3410 a directory entry by the name of the lookup key. Previously, if a
3411 symlink pointed to a non-existing file or a file in a directory that
3412 Exim lacked permissions to read, a lookup for a key matching that
3413 symlink would fail. Now it is enough that a matching directory entry
3414 exists, symlink or not. (Bugzilla 503.)
3416 PH/03 The body_linecount and body_zerocount variables are now exported in the
3419 PH/04 Added the $dnslist_matched variable.
3421 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3422 This means they are set thereafter only if the connection becomes
3425 PH/06 Added the client_condition to authenticators so that some can be skipped
3426 by clients under certain conditions.
3428 PH/07 The error message for a badly-placed control=no_multiline_responses left
3429 "_responses" off the end of the name.
3431 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3433 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3434 (without spaces) instead of just copying the configuration text.
3436 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3438 PH/11 Added $max_received_linelength.
3440 PH/12 Added +ignore_defer and +include_defer to host lists.
3442 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3443 way in which PCRE > 7.0 is built.
3445 PH/14 Implemented queue_only_load_latch.
3447 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3448 MAIL command. The effect was to mangle the value on 64-bit systems.
3450 PH/16 Another patch from the Sieve maintainer.
3452 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3454 PH/18 If a system quota error occurred while trying to create the file for
3455 a maildir delivery, the message "Mailbox is full" was not appended to the
3456 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3457 only to a quota excession during the actual writing of the file.
3459 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3460 characters?) which causes problems in log lines. The DN values are now
3461 passed through string_printing() before being added to log lines.
3463 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3464 and InterBase are left for another time.)
3466 PH/21 Added message_body_newlines option.
3468 PH/22 Guard against possible overflow in moan_check_errorcopy().
3470 PH/23 POSIX allows open() to be a macro; guard against that.
3472 PH/24 If the recipient of an error message contained an @ in the local part
3473 (suitably quoted, of course), incorrect values were put in $domain and
3474 $local_part during the evaluation of errors_copy.
3480 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3481 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3484 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3485 issue a MAIL command.
3487 PH/02 In an ACL statement such as
3489 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3491 if a client was not listed at all, or was listed with a value other than
3492 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3493 the condition was not true (as it should be), so access was not denied.
3494 The bug was that the ! inversion was incorrectly passed on to the second
3495 item. This has been fixed.
3497 PH/03 Added additional dnslists conditions == and =& which are different from
3498 = and & when the dns lookup returns more than one IP address.
3500 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3501 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3503 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3504 FSYNC, which compiles an option called disable_fsync that allows for
3505 bypassing fsync(). The documentation is heavily laced with warnings.
3507 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3509 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3510 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3511 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3512 including adding "make clean"; (3) Added -fPIC when compiling the test
3513 dynamically loaded module, to get rid of a warning.
3515 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3516 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3517 = 0s. The bug is otherwise harmless.
3519 PH/07 There was a bug in the dovecot authenticator such that the value of
3520 $auth1 could be overwritten, and so not correctly preserved, after a
3521 successful authentication. This usually meant that the value preserved by
3522 the server_setid option was incorrect.
3524 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3526 PH/09 Installed PCRE release 7.0.
3528 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3529 run for batched SMTP input. It is now run at the start of every message
3530 in the batch. While fixing this I discovered that the process information
3531 (output by running exiwhat) was not always getting set for -bs and -bS
3532 input. This is fixed, and it now also says "batched" for BSMTP.
3534 PH/11 Added control=no_pipelining.
3536 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3537 patch, slightly modified), and move the expansion of helo_data till after
3538 the connection is made in the smtp transport (so it can use these
3541 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3543 PH/14 Added log_selector = +pid.
3545 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3547 PH/16 Add ${if forany and ${if forall.
3549 PH/17 Added dsn_from option to vary the From: line in DSNs.
3551 PH/18 Flush SMTP output before performing a callout, unless control =
3552 no_callout_flush is set.
3554 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3555 was true (the default) a successful delivery failed to delete the retry
3556 item, thus causing premature timeout of the address. The bug is now
3559 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3561 PH/21 Long custom messages for fakedefer and fakereject are now split up
3562 into multiline responses in the same way that messages for "deny" and
3563 other ACL rejections are.
3565 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3566 with slight modification.
3568 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3569 draft, changing the syntax and factoring some duplicate code".
3571 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3572 for deliveries of the second and subsequent messages over the same SMTP
3575 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3576 ${reduce, with only minor "tidies".
3578 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3580 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3581 expansion side effects.
3583 PH/27 When a message times out after an over-quota error from an Exim-imposed
3584 quota, the bounce message says "mailbox is full". This message was not
3585 being given when it was a system quota that was exceeded. It now should
3588 MH/03 Made $recipients available in local_scan(). local_scan() already has
3589 better access to the recipient list through recipients_list[], but
3590 $recipients can be useful in postmaster-provided expansion strings.
3592 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3593 in the case of a MAIL command with additional options following the
3594 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3595 were accidentally chopped off.
3597 PH/29 SMTP synchronization checks are implemented when a command is read -
3598 there is a check that no more input is waiting when there shouldn't be
3599 any. However, for some commands, a delay in an ACL can mean that it is
3600 some time before the response is written. In this time, more input might
3601 arrive, invalidly. So now there are extra checks after an ACL has run for
3602 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3603 pipelining has not been advertised.
3605 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3607 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3608 returned in the SMTP response when smtp_return_error_details was false.
3609 This has been fixed.
3611 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3612 instead of the C I/O that was originally supplied, because problems were
3613 reported on Solaris.
3615 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3616 Exim which did not show up earlier: it was assuming that a call to
3617 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3618 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3619 was a macro that became an assignment, so it seemed to work. This has
3620 changed to a proper function call with a void return, hence the compile
3621 error. Exim's code has been fixed.
3623 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3626 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3628 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3630 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3631 because it turns out to be too big or there is a timeout) there is no
3632 "Completed" line in the log. When some messages of this type were
3633 selected by exigrep, they were listed as "not completed". Others were
3634 picked up by some special patterns. I have improved the selection
3635 criteria to be more general.
3637 PH/38 The host_find_failed option in the manualroute router can now be set
3638 to "ignore", to completely ignore a host whose IP address cannot be
3639 found. If all hosts are ignored, the behaviour is controlled by the new
3640 host_all_ignored option.
3642 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3643 homing or because of multiple MX records with /mx) generated more than
3644 one IP address, and the following item turned out to be the local host,
3645 all the secondary addresses of the first item were incorrectly removed
3646 from the list, along with the local host and any following hosts (which
3647 is what is supposed to happen).
3649 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3650 whoever called Exim into the -H file. In the case of the daemon it was
3651 behaving confusingly. When first started, it used values for whoever
3652 started the daemon, but after a SIGHUP it used the Exim user (because it
3653 calls itself on a restart). I have changed the code so that it now always
3656 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3657 message are rejected with the same error (e.g. no authentication or bad
3658 sender address), and a DATA command is nevertheless sent (as can happen
3659 with PIPELINING or a stupid MUA), the error message that was given to the
3660 RCPT commands is included in the rejection of the DATA command. This is
3661 intended to be helpful for MUAs that show only the final error to their
3664 PH/42 Another patch from the Sieve maintainer.
3666 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3667 Eximstats - Fixed some broken HTML links and added missing column headers
3669 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3670 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3672 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3674 PH/43 Yet another patch from the Sieve maintainer.
3676 PH/44 I found a way to check for a TCP/IP connection going away before sending
3677 the response to the final '.' that terminates a message, but only in the
3678 case where the client has not sent further data following the '.'
3679 (unfortunately, this is allowed). However, in many cases there won't be
3680 any further data because there won't be any more messages to send. A call
3681 to select() can be used: if it shows that the input is "ready", there is
3682 either input waiting, or the socket has been closed. An attempt to read
3683 the next input character can distinguish the two cases. Previously, Exim
3684 would have sent an OK response which the client would never have see.
3685 This could lead to message repetition. This fix should cure that, at
3686 least in a lot of common cases.
3688 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3689 advertised in response to EHLO.
3695 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3696 fixed by 4.65/MH/01 (is this a record?) are fixed:
3698 (i) An empty string was always treated as zero by the numeric comparison
3699 operators. This behaviour has been restored.
3701 (ii) It is documented that the numeric comparison operators always treat
3702 their arguments as decimal numbers. This was broken in that numbers
3703 starting with 0 were being interpreted as octal.
3705 While fixing these problems I realized that there was another issue that
3706 hadn't been noticed. Values of message_size_limit (both the global option
3707 and the transport option) were treated as octal if they started with 0.
3708 The documentation was vague. These values are now always treated as
3709 decimal, and I will make that clear in the documentation.
3715 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3716 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3719 MH/01 Don't check that the operands of numeric comparison operators are
3720 integers when their expansion is in "skipping" mode (fixes bug
3721 introduced by 4.64-PH/07).
3723 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3724 child addresses, Exim now panics and dies. Previously, because the count
3725 is held in a short int, deliveries were likely to be lost. As such a
3726 large number of recipients for a single message is ridiculous
3727 (performance will be very, very poor), I have chosen to impose a limit
3728 rather than extend the field.
3734 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3735 leftover -K file (the existence of which was triggered by #402).
3736 While we were at it, introduced process PID as part of the -K
3737 filename. This should rule out race conditions when creating
3740 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3741 processing considerably. Previous code took too long for large mails,
3742 triggering a timeout which in turn triggers #401.
3744 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3745 in the DK code in transports.c. sendfile() is not really portable,
3746 hence the _LINUX specificness.
3748 TF/01 In the add_headers option to the mail command in an Exim filter,
3749 there was a bug that Exim would claim a syntax error in any
3750 header after the first one which had an odd number of characters
3753 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3754 callout verification, Exim cached a "reject" for the entire domain. This
3755 is correct for most verifications, but it is not correct for a recipient
3756 verification with use_sender or use_postmaster set, because in that case
3757 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3758 case of MAIL FROM:<> rejection from other early rejections (e.g.
3759 rejection of HELO). When verifying a recipient using a non-null MAIL
3760 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3761 Whatever the result of the callout, the value of the domain cache is
3762 left unchanged (for any other kind of callout, getting as far as trying
3763 RCPT means that the domain itself is ok).
3765 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3768 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3769 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3770 session, the daemon ignores ECONNECT errors and logs others; it now
3771 ignores EPIPE as well.
3773 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3774 (quoted-printable decoding).
3776 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3777 later the small subsequent patch to fix an introduced bug.
3779 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3781 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3783 PH/08 An error is now given if message_size_limit is specified negative.
3785 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3786 to be given (somewhat) arbitrary names.
3788 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3791 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3792 miscellaneous code fixes
3794 PH/10 Added the log_reject_target ACL modifier to specify where to log
3797 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3798 hostname. This is wrong, because it relates to the incoming message (and
3799 probably the interface on which it is arriving) and not to the outgoing
3800 callout (which could be using a different interface). This has been
3801 changed to use the value of the helo_data option from the smtp transport
3802 instead - this is what is used when a message is actually being sent. If
3803 there is no remote transport (possible with a router that sets up host
3804 addresses), $smtp_active_hostname is used.
3806 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3807 tweaks were necessary in order to get it to work (see also 21 below):
3808 (a) The code assumed that strncpy() returns a negative number on buffer
3809 overflow, which isn't the case. Replaced with Exim's string_format()
3811 (b) There were several signed/unsigned issues. I just did the minimum
3812 hacking in of casts. There is scope for a larger refactoring.
3813 (c) The code used strcasecmp() which is not a standard C function.
3814 Replaced with Exim's strcmpic() function.
3815 (d) The code set only $1; it now sets $auth1 as well.
3816 (e) A simple test gave the error "authentication client didn't specify
3817 service in request". It would seem that Dovecot has changed its
3818 interface. Fortunately there's a specification; I followed it and
3819 changed what the client sends and it appears to be working now.
3821 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3824 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3825 address A is aliased to B and C, where B exists and C does not. Without
3826 -v the output is "A verified" because verification stops after a
3827 successful redirection if more than one address is generated. However,
3828 with -v the child addresses are also verified. Exim was outputting "A
3829 failed to verify" and then showing the successful verification for C,
3830 with its parentage. It now outputs "B failed to verify", showing B's
3831 parentage before showing the successful verification of C.
3833 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3834 look up a TXT record in a specific list after matching in a combined
3837 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3838 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3839 they consult the DNS. I had assumed they would set it the way they
3840 wanted; and indeed my experiments on Linux seem to show that in some
3841 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3842 To be on the safe side, however, I have now made the interface to
3843 host_find_byname() similar to host_find_bydns(), with an argument
3844 containing the DNS resolver options. The host_find_byname() function now
3845 sets these options at its start, just as host_find_bydns() does. The smtp
3846 transport options dns_qualify_single and dns_search_parents are passed to
3847 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3848 of host_find_byname() use the default settings of RES_DEFNAMES
3849 (qualify_single) but not RES_DNSRCH (search_parents).
3851 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3852 spool_read_header() do less string testing, by means of a preliminary
3853 switch on the second character of optional "-foo" lines. (This is
3854 overdue, caused by the large number of possibilities that now exist.
3855 Originally there were few.) While I was there, I also converted the
3856 str(n)cmp tests so they don't re-test the leading "-" and the first
3857 character, in the hope this might squeeze out yet more improvement.
3859 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3860 flag allowing group syntax was set by the header_syntax check but not
3861 turned off, possible causing trouble later; (2) The flag was not being
3862 set at all for the header_verify test, causing "group"-style headers to
3863 be rejected. I have now set it in this case, and also caused header_
3864 verify to ignore an empty address taken from a group. While doing this, I
3865 came across some other cases where the code for allowing group syntax
3866 while scanning a header line wasn't quite right (mostly, not resetting
3867 the flag correctly in the right place). These bugs could have caused
3868 trouble for malformed header lines. I hope it is now all correct.
3870 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3871 with the "reply" argument non-NULL. The code, however (which originally
3872 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3873 but it didn't always do it. This confused somebody who was copying the
3874 code for some other use. I have removed all the tests.
3876 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3877 feature that was used to support insecure browsers during the U.S. crypto
3878 embargo. It requires special client support, and Exim is probably the
3879 only MTA that supported it -- and would never use it because real RSA is
3880 always available. This code has been removed, because it had the bad
3881 effect of slowing Exim down by computing (never used) parameters for the
3882 RSA_EXPORT functionality.
3884 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3885 authenticator to fail if there's a tab character in the incoming data
3886 (there should never be unless someone is messing about, as it's supposed
3887 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3888 if the connection is using TLS or if the remote IP is the same as the
3889 local IP, and the "valid-client-cert option" if a client certificate has
3892 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3893 authenticators. This can be used for authorization after authentication
3894 succeeds. (In the case of plaintext, it servers for both authentication
3897 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3898 if any retry times were supplied.
3900 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3901 connection, where there is no client IP address to check. In this
3902 situation, the verify now always succeeds.
3904 PH/25 Applied John Jetmore's -Mset patch.
3906 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3908 PH/27 In a string expansion for a processed (not raw) header when multiple
3909 headers of the same name were present, leading whitespace was being
3910 removed from all of them, but trailing whitespace was being removed only
3911 from the last one. Now trailing whitespace is removed from each header
3912 before concatenation. Completely empty headers in a concatenation (as
3913 before) are ignored.
3915 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3916 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3918 PH/29 [Removed. This was a change that I later backed out, and forgot to
3919 correct the ChangeLog entry (that I had efficiently created) before
3920 committing the later change.]
3922 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3923 address errors (4xx response to RCPT) over the same connection as other
3924 messages routed to the same hosts. Such deliveries are always "forced",
3925 so retry times are not inspected. This resulted in far too many retries
3926 for the affected addresses. The effect occurred only when there were more
3927 hosts than the hosts_max_try setting in the smtp transport when it had
3928 the 4xx errors. Those hosts that it had tried were not added to the list
3929 of hosts for which the message was waiting, so if all were tried, there
3930 was no problem. Two fixes have been applied:
3932 (i) If there are any address or message errors in an SMTP delivery, none
3933 of the hosts (tried or untried) are now added to the list of hosts
3934 for which the message is waiting, so the message should not be a
3935 candidate for sending over the same connection that was used for a
3936 successful delivery of some other message. This seems entirely
3937 reasonable: after all the message is NOT "waiting for some host".
3938 This is so "obvious" that I'm not sure why it wasn't done
3939 previously. Hope I haven't missed anything, but it can't do any
3940 harm, as the worst effect is to miss an optimization.
3942 (ii) If, despite (i), such a delivery is accidentally attempted, the
3943 routing retry time is respected, so at least it doesn't keep
3944 hammering the server.
3946 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3947 in ${readsocket because some servers need this prod.
3949 PH/32 Added some extra debug output when updating a wait-xxx database.
3951 PH/33 The hint "could be header name not terminated by colon", which has been
3952 given for certain expansion errors for a long time, was not being given
3953 for the ${if def:h_colon_omitted{... case.
3955 PH/34 The spec says: "With one important exception, whenever a domain list is
3956 being scanned, $domain contains the subject domain." There was at least
3957 one case where this was not true.
3959 PH/35 The error "getsockname() failed: connection reset by peer" was being
3960 written to the panic log as well as the main log, but it isn't really
3961 panic-worthy as it just means the connection died rather early on. I have
3962 removed the panic log writing for the ECONNRESET error when getsockname()
3965 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3966 runs only) independently of the message's sender address. This meant
3967 that, if the 4xx error was in fact related to the sender, a different
3968 message to the same recipient with a different sender could confuse
3969 things. In particular, this can happen when sending to a greylisting
3970 server, but other circumstances could also provoke similar problems.
3971 I have changed the default so that the retry time for these errors is now
3972 based a combination of the sender and recipient addresses. This change
3973 can be overridden by setting address_retry_include_sender=false in the
3976 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3977 remote server are returned as part of bounce messages. This was not
3978 happening for LMTP over a pipe (the lmtp transport), but now it is the
3979 same for both kinds of LMTP.
3981 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3982 in header lines that were in fact CNAMEs. This is no longer the case.
3984 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3985 and queue runs started by the daemon processed all messages. This has
3986 been fixed so that -R and -S can now usefully be given with -q<time>.
3988 PH/40 Import PCRE release 6.7 (fixes some bugs).
3990 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3992 PH/42 Give an error if -q is specified more than once.
3994 PH/43 Renamed the variables $interface_address and $interface_port as
3995 $received_ip_address and $received_port, to make it clear that these
3996 values apply to message reception, and not to the outgoing interface when
3997 a message is delivered. (The old names remain recognized, of course.)
3999 PH/44 There was no timeout on the connect() call when using a Unix domain
4000 socket in the ${readsocket expansion. There now is.
4002 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4003 be meaningful with "accept".
4005 SC/01 Eximstats V1.43
4006 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4008 SC/02 Eximstats V1.44
4009 Use a glob alias rather than an array ref in the generated
4010 parser. This improves both readability and performance.
4012 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4013 Collect SpamAssassin and rejection statistics.
4014 Don't display local sender or destination tables unless
4015 there is data to show.
4016 Added average volumes into the top table text output.
4018 SC/04 Eximstats V1.46
4019 Collect data on the number of addresses (recipients)
4020 as well as the number of messages.
4022 SC/05 Eximstats V1.47
4023 Added 'Message too big' to the list of mail rejection
4024 reasons (thanks to Marco Gaiarin).
4026 SC/06 Eximstats V1.48
4027 Mainlog lines which have GMT offsets and are too short to
4028 have a flag are now skipped.
4030 SC/07 Eximstats V1.49 (Alain Williams)
4031 Added the -emptyok flag.
4033 SC/08 Eximstats V1.50
4034 Fixes for obtaining the IP address from reject messages.
4036 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4037 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4038 whitespace changes from 4.64-PH/27
4040 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4043 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4044 are found, allow negative numbers in numeric criteria)
4046 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4048 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4049 to match changes made in 4.64-PH/43
4051 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4053 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4054 "domain" is actually an IP address, and force a failure. This locks out
4055 those revolvers/nameservers that support "A-for-A" lookups, in
4056 contravention of the specifications.
4058 PH/48 When a host name was looked up from an IP address, and the subsequent
4059 forward lookup of the name timed out, the host name was left in
4060 $sender_host_name, contrary to the specification.
4062 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4063 restricted to single-key lookups, Exim was not diagnosing an error if
4064 * or *@ was used with a query-style lookup.
4066 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4068 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4069 long ago, but noone interested enough thought of it. Let's just say that
4070 the "1.1" means that there are some new functions that weren't there at
4071 some point in the past.
4073 PH/51 Error processing for expansion failure of helo_data from an smtp
4074 transport during callout processing was broken.
4076 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4077 tested/used via the -bh/-bhc/-bs options.
4079 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4080 bug, fixed in subsequent PCRE releases).
4082 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4083 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4085 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4091 SC/01 Use a glob alias rather than an array ref in eximstats generated
4092 parser. This improves both readability and performance.
4094 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4095 Don't display local sender or destination tables in eximstats unless
4096 there is data to show.
4097 Added average volumes into the eximstats top table text output.
4099 SC/03 Collect data on the number of addresses (recipients) as well
4100 as the number of messages in eximstats.
4102 TF/01 Correct an error in the documentation for the redirect router. Exim
4103 does (usually) call initgroups() when daemonizing.
4105 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4106 with consistent privilege compared to when running as a daemon.
4108 TF/03 Note in the spec that $authenticated_id is not set for local
4109 submissions from trusted users.
4111 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4112 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4114 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4115 by adding some example configuration directives to the default
4116 configuration file. A little bit of work is required to uncomment the
4117 directives and define how usernames and passwords are checked, but
4118 there is now a framework to start from.
4120 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4121 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4122 without this. I don't know how relevant this is to other LDAP libraries.
4124 PH/02 Add the verb name to the "unknown ACL verb" error.
4126 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4128 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4130 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4131 directory not expanded when it should be if an expanded home directory
4132 was set for the address (which is overridden by the transport).
4134 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4137 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4138 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4139 because it is too late at that time, and has no effect.
4141 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4142 security issue with \' (bugzilla #107). I could not use the
4143 PQescapeStringConn() function, because it needs a PGconn value as one of
4146 PH/08 When testing addresses using -bt, indicate those final addresses that
4147 are duplicates that would not cause an additional delivery. At least one
4148 person was confused, thinking that -bt output corresponded to deliveries.
4149 (Suppressing duplicates isn't a good idea as you lose the information
4150 about possibly different redirections that led to the duplicates.)
4152 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4153 systems where poll() doesn't work, in particular OS X.
4155 PH/10 Added more information to debugging output for retry time not reached.
4157 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4158 operations in malware.c.
4160 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4163 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4164 syslog_duplication set false, log lines that would normally be written
4165 both the the main log and to the reject log were not written to syslog at
4168 PH/14 In the default configuration, change the use of "message" in ACL warn
4169 statements to "add_header".
4171 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4172 not followed by a command (e.g. "seen endif").
4174 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4175 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4178 PH/17 Added extra conditions to the default value of delay_warning_condition
4182 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4183 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4184 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4187 The Auto-Submitted: and various List- headers are standardised, whereas I
4188 don't think Precedence: ever was.
4190 PH/18 Refactored debugging code in route_finduser() to show more information,
4191 in particular, the error code if getpwnam() issues one.
4193 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4194 This is apparently needed in addition to the PH/07 change above to avoid
4195 any possible encoding problems.
4197 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4198 but not after initializing Perl.
4200 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4201 output them only if debugging. By default they are written stderr,
4202 apparently, which is not desirable.
4204 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4207 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4210 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4212 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4213 authenticated or an ident call has been made. Suppress the default
4214 values for $authenticated_id and $authenticated_sender (but permit -oMai
4215 and -oMas) when testing with -bh.
4217 PH/24 Re-jigged the order of the tests in the default configuration so that the
4218 tests for valid domains and recipients precede the DNS black list and CSA
4219 tests, on the grounds that those ones are more expensive.
4221 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4222 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4223 This bug exists in every version of Exim that I still have, right back to
4226 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4227 However, an attempt to turn on case-sensitivity in a regex key by
4228 including (?-i) didn't work because the subject string was already
4229 lowercased, and the effects were non-intuitive. It turns out that a
4230 one-line patch can be used to allow (?-i) to work as expected.
4236 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4237 other effects) broke the use of negated acl sub-conditions.
4239 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4242 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4243 "Deny" causes Exim to reject the incoming connection with a 554 error.
4244 Unfortunately, if there is a major crisis, such as a disk failure,
4245 tcp-wrappers gives "deny", whereas what one would like would be some
4246 kind of temporary error. A kludge has been added to help with this.
4247 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4248 554 error is used if errno is still zero or contains ENOENT (which occurs
4249 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4252 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4254 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4255 errors. Otherwise a message that provokes a temporary error (when other
4256 messages do not) can cause a whole host to time out.
4258 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4259 addresses were routed directly to files or pipes from a redirect router.
4260 File deliveries just didn't batch; pipe deliveries might have suffered
4263 PH/06 A failure to get a lock for a hints database would erroneously always say
4264 "Failed to get write lock", even when it was really a read lock.
4266 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4267 of 0600. This has been changed to use the value of the lockfile_mode
4268 option (which defaults to 0600).
4270 PH/08 Applied small patch from the Sieve maintainer.
4272 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4273 folder from quota calculations, a direct delivery into this folder messed
4274 up the contents of the maildirsize file. This was because the regex was
4275 used only to exclude .Trash (or whatever) when the size of the mailbox
4276 was calculated. There was no check that a delivery was happening into an
4277 excluded directory. This bug has been fixed by ignoring all quota
4278 processing for deliveries into excluded directories.
4280 PH/10 Added the maildirfolder_create_regex option to appendfile.
4286 PH/01 The code for finding all the local interface addresses on a FreeBSD
4287 system running IPv6 was broken. This may well have applied to all BSD
4288 systems, as well as to others that have similar system calls. The broken
4289 code found IPv4 interfaces correctly, but gave incorrect values for the
4290 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4291 that it would not match correctly against @[] and not recognize the IPv6
4294 PH/02 The ipliteral router was not recognizing addresses of the form user@
4295 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4297 PH/03 Added disable_ipv6.
4299 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4300 decoded form, because it is most often used to construct To: headers
4301 lines in autoreplies, and the decoded form may well be syntactically
4302 invalid. However, $reply_address has leading white space removed, and all
4303 newlines turned into spaces so that the autoreply transport does not
4306 PH/05 If group was specified without a user on a router, and no group or user
4307 was specified on a transport, the group from the router was ignored.
4309 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4310 for visible compile-time settings that can be used to change these
4311 numbers, for those that want even more. Backwards compatibility with old
4312 spool files has been maintained. However, going back to a previous Exim
4313 release will lost any variables that are in spool files.
4315 PH/07 Two small changes when running in the test harness: increase delay when
4316 passing a TCP/IP connection to a new process, in case the original
4317 process has to generate a bounce, and remove special handling of
4318 127.0.0.2 (sic), which is no longer necessary.
4320 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4321 be the same on different OS.
4323 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4326 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4327 whether --show-vars was specified or not
4329 JJ/02 exipick: Added support for new ACL variable spool format introduced
4332 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4333 syntactically invalid From: or Reply-to: line, and a filter used this to
4334 generate an autoreply, and therefore failed to obtain an address for the
4335 autoreply, Exim could try to deliver to a non-existent relative file
4336 name, causing unrelated and misleading errors. What now happens is that
4337 it logs this as a hard delivery error, but does not attempt to create a
4340 PH/11 The exinext utility has a -C option for testing purposes, but although
4341 the given file was scanned by exinext itself; it wasn't being passed on
4342 when Exim was called.
4344 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4345 an end-of-file indication when reading a command response.
4347 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4348 compiled. In many other places in Exim, IPv6 addresses are always
4349 recognized, so I have changed this. It also means that IPv4 domain
4350 literals of the form [IPV4:n.n.n.n] are now always recognized.
4352 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4353 used if the router is not running as root, for example, when verifying at
4354 ACL time, or when using -bh. The debugging output from this situation was
4355 non-existent - all you got was a failure to exec. I have made two
4358 (a) Failures to set uid/gid, the current directory, or a process leader
4359 in a subprocess such as that created by queryprogram now generate
4360 suitable debugging output when -d is set.
4362 (b) The queryprogram router detects when it is not running as root,
4363 outputs suitable debugging information if -d is set, and then runs
4364 the subprocess without attempting to change uid/gid.
4366 PH/15 Minor change to Makefile for building test_host (undocumented testing
4369 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4370 additional section of a DNS packet that returns MX or SRV records.
4371 Instead, it always explicitly searches for A/AAAA records. This avoids
4372 major problems that occur when a DNS server includes only records of one
4373 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4374 fixed another bug: if SRV records were looked up and the corresponding
4375 address records were *not* found in the additional section, the port
4376 values from the SRV records were lost.
4378 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4379 using the correct key (the original address) when searching the retry
4380 rules in order to find which one to use for generating the retry hint.
4382 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4383 adding the default one. Similarly, if it contains a Reply-To: header, the
4384 errors_reply_to option, if set, is not used.
4386 PH/19 When calculating a retry time, Exim used to measure the "time since
4387 failure" by looking at the "first failed" field in the retry record. Now
4388 it does not use this if it is later than than the arrival time of the
4389 message. Instead it uses the arrival time. This makes for better
4390 behaviour in cases where some deliveries succeed, thus re-setting the
4391 "first failed" field. An example is a quota failure for a huge message
4392 when small messages continue to be delivered. Without this change, the
4393 "time since failure" will always be short, possible causing more frequent
4394 delivery attempts for the huge message than are intended.
4395 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4397 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4398 $1, $2, $3) because the numerical variables can be reset during some
4399 expansion items (e.g. "match"), thereby losing the authentication data.
4401 PH/21 Make -bV show the size of off_t variables so that the test suite can
4402 decide whether to run tests for quotas > 2G.
4404 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4405 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4406 filecount value is greater than 2G or if a quota value is greater than 2G
4407 on a system where the size of off_t is not greater than 4, a panic error
4410 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4411 never match. The debug and -bh output now contains an explicit error
4412 message indicating a malformed IPv4 address or mask.
4414 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4415 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4416 PH/23 above applies.
4418 PH/25 Do not write to syslog when running in the test harness. The only
4419 occasion when this arises is a failure to open the main or panic logs
4420 (for which there is an explicit test).
4422 PH/26 Added the /no_tell option to "control=freeze".
4424 PH/27 If a host name lookup failed very early in a connection, for example, if
4425 the IP address matched host_lookup and the reverse lookup yielded a name
4426 that did not have a forward lookup, an error message of the form "no IP
4427 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4428 could be logged. Now it outputs the IP address instead of "NULL".
4430 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4431 allows the sender and the authenticated sender to be set when
4432 submitting a message from within Exim. Since child_open_exim() is
4433 documented for local_scan(), the new function should be too.
4435 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4436 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4437 results in an empty string is now treated as unset.
4439 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4441 PH/31 Added sender_verify_fail logging option.
4443 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4444 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4445 tidied the source and removed it altogether.
4447 PH/33 When a queue run was abandoned because the load average was too high, a
4448 log line was always written; now it is written only if the queue_run log
4449 selector is set. In addition, the log line for abandonment now contains
4450 information about the queue run such as the pid. This is always present
4451 in "start" and "stop" lines but was omitted from the "abandon" line.
4453 PH/34 Omit spaces between a header name and the colon in the error message that
4454 is given when verify = headers_syntax fails (if there are lots of them,
4455 the message gets confusing).
4457 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4458 names, as there are now some PTR records that contain slashes. This check
4459 is only to protect against broken name servers that fall over on strange
4460 characters, so the fact that it applies to all lookups doesn't matter.
4462 PH/36 Now that the new test suite is complete, we can remove some of the
4463 special code in Exim that was needed for the old test suite. For example,
4464 sorting DNS records because real resolvers return them in an arbitrary
4465 order. The new test suite's fake resolver always returns records in the
4468 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4469 bounces) except when queue_only is set, to avoid logging races between
4470 the different processes.
4472 PH/38 Panic-die if .include specifies a non-absolute path.
4474 PH/39 A tweak to the "H" retry rule from its user.
4476 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4477 a label. They prevented compilation on older perls.
4479 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4480 a warning to be raised on newish perls.
4482 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4483 on queue. Changes to match documented behaviour of showing count of
4484 messages matching specified criteria.
4486 PH/40 Changed the default ident timeout from 30s to 5s.
4488 PH/41 Added support for the use of login_cap features, on those BSD systems
4489 that have them, for controlling the resources used by pipe deliveries.
4491 PH/42 The content-scanning code uses fopen() to create files in which to put
4492 message data. Previously it was not paying any attention to the mode of
4493 the files. Exim runs with umask(0) because the rest of the code creates
4494 files with open(), and sets the required mode explicitly. Thus, these
4495 files were ending up world-writeable. This was not a big issue, because,
4496 being within the spool directory, they were not world-accessible. I have
4497 created a function called modefopen, which takes an additional mode
4498 argument. It sets umask(777), creates the file, chmods it to the required
4499 mode, then resets the umask. All the relevant calls to fopen() in the
4500 content scanning code have been changed to use this function.
4502 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4503 to 24 hours. This avoids potential overflow problems when processing G
4504 and H retry rules. I suspect nobody ever tinkers with this value.
4506 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4508 PH/45 When the plaintext authenticator is running as a client, the server's
4509 challenges are checked to ensure they are valid base64 strings. By
4510 default, the authentication attempt is cancelled if an invalid string is
4511 received. Setting client_ignore_invalid_base64 true ignores these errors.
4512 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4513 they are received. Thus, the responses can be made to depend on the
4514 challenges. If an invalid string is ignored, an empty string is placed in
4517 PH/46 Messages that are created by the autoreply transport now contains a
4518 References: header, in accordance with RFCs 2822 and 3834.
4520 PH/47 Added authenticated_sender_force to the smtp transport.
4522 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4524 PH/49 Installed latest patch from the Sieve maintainer.
4526 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4527 was also set, the appendfile transport was unnecessarily scanning a
4528 directory of message files (e.g. for maildir delivery) to find the count
4529 of files (along with the size), even though it did not need this
4530 information. It now does the scan only if it needs to find either the
4531 size of the count of files.
4533 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4535 PH/52 Two bugs concerned with error handling when the smtp transport is
4538 (i) Exim was not creating retry information for temporary errors given
4539 for individual recipients after the DATA command when the smtp transport
4540 was used in LMTP mode. This meant that they could be retried too
4541 frequently, and not timed out correctly.
4543 (ii) Exim was setting the flag that allows error details to be returned
4544 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4545 recipients that were returned after the DATA command.
4547 PH/53 This is related to PH/52, but is more general: for any failing address,
4548 when detailed error information was permitted to be returned to the
4549 sender, but the error was temporary, then after the final timeout, only
4550 "retry timeout exceeded" was returned. Now it returns the full error as
4551 well as "retry timeout exceeded".
4553 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4554 do this, and (what is worse) MTAs that accept it.
4556 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4557 will now be deprecated.
4559 PH/56 New os.c-cygwin from the Cygwin maintainer.
4561 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4562 formats (previously only available in exim formats via -bpr, -bpru,
4563 and -bpra. Now also available in native and exiqgrep formats)
4565 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4566 with very large, slow to parse queues
4568 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4570 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4572 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4573 responses to authentication challenges, though it was showing the
4574 challenges; (ii) I've removed the CR characters from the debug output for
4577 PH/58 Allow for the insertion of a newline as well as a space when a string
4578 is turned into more than one encoded-word during RFC 2047 encoding. The
4579 Sieve code now uses this.
4581 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4582 data_4xx, lost_connection, tls_required.
4584 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4585 message was being sent as an SMTP response.
4587 PH/61 Add -l and -k options to exicyclog.
4589 PH/62 When verifying, if an address was redirected to one new address, so that
4590 verification continued, and the new address failed or deferred after
4591 having set something in $address_data, the value of $address_data was not
4592 passed back to the ACL. This was different to the case when no
4593 redirection occurred. The value is now passed back in both cases.
4595 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4596 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4597 use login_cap.h, so on its own it isn't the distinguishing feature. The
4598 new name refers directly to the setclassresources() function.
4600 PH/65 Added configuration files for NetBSD3.
4602 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4604 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4605 is preferred over IPv4.
4607 PH/68 The bounce_return_message and bounce_return_body options were not being
4608 honoured for bounces generated during the reception of non-SMTP messages.
4609 In particular, this applied to messages rejected by the ACL. This bug has
4610 been fixed. However, if bounce_return_message is true and bounce_return_
4611 body is false, the headers that are returned for a non-SMTP message
4612 include only those that have been read before the error was detected.
4613 (In the case of an ACL rejection, they have all been read.)
4615 PH/69 The HTML version of the specification is now built in a directory called
4616 spec_html instead of spec.html, because the latter looks like a path with
4617 a MIME-type, and this confuses some software.
4619 PH/70 Catch two compiler warnings in sieve.c.
4621 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4622 function verify_get_ident() calls ip_connect() to connect a socket, but
4623 if the "connect()" function timed out, ip_connect() used to close the
4624 socket. However, verify_get_ident() also closes the socket later, and in
4625 between Exim writes to the log, which may get opened at this point. When
4626 the socket was closed in ip_connect(), the log could get the same file
4627 descriptor number as the socket. This naturally causes chaos. The fix is
4628 not to close the socket in ip_connect(); the socket should be closed by
4629 the function that creates it. There was only one place in the code where
4630 this was missing, in the iplookup router, which I don't think anybody now
4631 uses, but I've fixed it anyway.
4633 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4634 well as to direct DNS lookups. Otherwise the handling of names in host
4635 lists is inconsistent and therefore confusing.
4641 PH/01 Two changes to the default runtime configuration:
4643 (1) Move the checks for relay_from_hosts and authenticated clients from
4644 after to before the (commented out) DNS black list checks.
4646 (2) Add control=submission to the relay_from_hosts and authenticated
4647 clients checks, on the grounds that messages accepted by these
4648 statements are most likely to be submissions.
4650 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4652 (1) Generate an error if the third argument for the ${prvs expansion is
4655 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4658 (3) Reset the variables that are obtained from the first argument of
4659 ${prvscheck and used in the second argument before leaving the code,
4660 because their memory is reclaimed, so using them afterwards may do
4663 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4664 one (it's much easier than Tom thought :-).
4666 (5) Because of (4), we can now allow for the use of $prvscheck_result
4667 inside the third argument.
4669 PH/03 For some reason, the default setting of PATH when running a command from
4670 a pipe transport was just "/usr/bin". I have changed it to
4673 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4674 anything to be listed in the output from -bV.
4676 PH/05 When a filter generated an autoreply, the entire To: header line was
4677 quoted in the delivery log line, like this:
4679 => >A.N.Other <ano@some.domain> <original@ddress> ...
4681 This has been changed so that it extracts the operative address. There
4682 may be more than one such address. If so, they are comma-separated, like
4685 => >ano@some.domain,ona@other.domain <original@ddress> ...
4687 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4688 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4689 not being looked up in the rDNS to get a host name, Exim was showing the
4690 IP address twice in Received: lines, even though the IP addresses were
4691 identical. For example:
4693 Received: from [1.2.3.4] (helo=[1.2.3.4])
4695 However, if the real host name was known, it was omitting the HELO data
4696 if it matched the actual IP address. This has been tidied up so that it
4697 doesn't show the same IP address twice.
4699 PH/07 When both +timestamp and +memory debugging was on, the value given by
4700 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4701 called by the debug printing, thereby overwriting the timestamp buffer.
4702 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4704 PH/08 When the original message was included in an autoreply transport, it
4705 always said "this is a copy of the message, including all the headers",
4706 even if body_only or headers_only was set. It now gives an appropriate
4709 PH/09 Applied a patch from the Sieve maintainer which:
4711 o fixes some comments
4712 o adds the (disabled) notify extension core
4713 o adds some debug output for the result of if/elsif tests
4714 o points to the current vacation draft in the documentation
4715 and documents the missing references header update
4719 o fixes a bug in processing the envelope test (when testing
4720 multiple envelope elements, the last element determined the
4723 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4724 Electronic Mail") by including:
4726 Auto-submitted: auto-generated
4728 in the messages that it generates (bounce messages and others, such as
4729 warnings). In the case of bounce messages for non-SMTP messages, there was
4730 also a typo: it was using "Auto_submitted" (underscore instead of
4731 hyphen). Since every message generated by Exim is necessarily in response
4732 to another message, thes have all been changed to:
4734 Auto-Submitted: auto-replied
4736 in accordance with these statements in the RFC:
4738 The auto-replied keyword:
4740 - SHOULD be used on messages sent in direct response to another
4741 message by an automatic process,
4743 - MUST NOT be used on manually-generated messages,
4745 - MAY be used on Delivery Status Notifications (DSNs) and Message
4746 Disposition Notifications (MDNs),
4748 - MUST NOT be used on messages generated by automatic or periodic
4749 processes, except for messages which are automatic responses to
4752 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4753 to the default Received: header definition.
4755 PH/12 Added log selector acl_warn_skipped (default on).
4757 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4758 variables because (a) they are in the wrong storage pool and (b) even if
4759 they were copied, it wouldn't work properly because of the caching.
4761 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4762 checking when decoding. Apparently there are clients that generate
4763 overlong encoded strings. Why am I not surprised?
4765 PH/15 If the first argument of "${if match_address" was not empty, but did not
4766 contain an "@" character, Exim crashed. Now it writes a panic log message
4767 and treats the condition as false.
4769 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4771 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4772 extension "envelope-auth". The code is finished and in agreement with
4773 other implementations, but there is no documentation so far and in fact,
4774 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4775 not changing the active code.
4777 Print executed "if" and "elsif" statements when debugging is used. This
4778 helps a great deal to understand what a filter does.
4780 Document more things not specified clearly in RFC3028. I had all this
4781 sorted out, when out of a sudden new issues came to my mind. Oops."
4783 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4786 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4787 canonical form (as documented). However, after a host name lookup from
4788 the IP address, check_host() was doing a simple string comparison with
4789 addresses acquired from the DNS when checking that the found name did
4790 have the original IP as one of its addresses. Since any found IPv6
4791 addresses are likely to be in abbreviated form, the comparison could
4792 fail. Luckily, there already exists a function for doing the comparison
4793 by converting both addresses to binary, so now that is used instead of
4794 the text comparison.
4796 PH/20 There was another similar case to PH/19, when a complete host name was
4797 given in a host list; looking up its IP address could give an abbreviated
4798 form, whereas the current host's name might or might not be abbreviated.
4799 The same fix has been applied.
4805 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4806 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4809 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4810 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4812 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4814 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4815 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4816 Also, for "Auto-Submitted", treat anything other than "no" as
4817 non-personal, in accordance with RFC 3834. (Previously it treated
4818 anything starting "auto-" as non-personal.)
4820 TF/01 The control=submission/name=... option had a problem with syntax
4821 errors if the name included a slash character. The /name= option
4822 now slurps the rest of the string, so it can include any characters
4823 but it must come last in the list of options (after /sender_retain
4826 PH/05 Some modifications to the interface to the fake nameserver for the new
4834 TK/01 Added the "success_on_redirect" address verification option. See
4835 NewStuff for rationale and an example.
4837 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4839 PH/02 Patch to exigrep to allow it to work on syslog lines.
4841 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4842 fread() to skip over the body file's header line, because in Cygwin the
4843 header line is locked and is inaccessible.
4845 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4846 co-exist for some time) to make it clear that it is the Exim ID that is
4847 referenced, not the Message-ID: header line.
4849 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4850 string_format() function, because snprintf() does not exist on all
4853 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4854 ${stat: expansion item.
4856 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4857 protocol synchronization error", to keep the pedants happy.
4859 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4860 well as for IRIX systems, when gcc is being used. See the host.c source
4863 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4865 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4868 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4869 transport and to the smtp transport in LMTP mode.
4871 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4873 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4874 run a filter in a subprocess. This could lead to confusion in subsequent
4875 lookups in the parent process. There should also be a search_tidyup() at
4876 the end of the subprocess.
4878 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4879 only if the host matched helo_try_verify_hosts, which caused the
4880 verification to occur when the EHLO/HELO command was issued. The ACL just
4881 tested the remembered result. Now, if a previous verification attempt has
4882 not happened, "verify = helo" does it there and then.
4884 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4886 TK/03 Fix log output including CR from clamd.
4888 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4889 exist provoked a memory error which could cause a segfault.
4891 PH/15 Installed PCRE 6.2
4893 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4895 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4896 of the problem. Specifically, suggested +O2 rather than +O1 for the
4899 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4901 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4902 record, the port was not correctly logged when the outgoing_port log
4903 selector was set (it logged the transort's default port).
4905 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4906 fallback_hosts, and "hosts" in the smtp transport.
4908 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4909 host errors such as "Connection refused".
4911 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4912 authentication with radiusclient 0.4.9:
4914 - Error returned from rc_read_config was caught wrongly
4915 - Username/password not passed on to radius server due to wrong length.
4917 The presumption is that some radiusclient API changes for 4.51/PH/17
4918 were not taken care of correctly. The code is still untested by me (my
4919 Linux distribution still has 0.3.2 of radiusclient), but it was
4920 contributed by a Radius user.
4922 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4923 expanding the "port" option of the smtp transport.
4925 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4926 while reading a MIME header. Thanks to Tom Hughes for a patch.
4928 PH/24 Include config.h inside local_scan.h so that configuration settings are
4931 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4932 that in an ACL for RCPT (for example), you can examine exactly what was
4935 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4936 commands, but it was not correctly comparing the address with the actual
4937 client host address. Thus, it would show the EHLO address in Received:
4938 header lines when this was not necessary.
4940 PH/27 Added the % operator to ${eval:}.
4942 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4943 it should be ignoring failures (because with -C, for example, it has lost
4944 privilege). It wasn't ignoring creation failures other than "already
4947 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4950 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4951 because an input line was too long, either on its own, or by virtue of
4952 too many continuations, the temporary file was not being removed, and the
4953 return code was incorrect.
4955 PH/31 Missing "BOOL" in function definition in filtertest.c.
4957 PH/32 Applied Sieve patches from the maintainer.
4959 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4961 PH/33 Added "verify = not_blind".
4963 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4964 Local/Makefile (with some defaults set). These are used in built scripts
4965 such as exicyclog, but they have never been used in the exim_install
4966 script (though there are many overriding facilities there). I have
4967 arranged that the exim_install script now takes note of these two
4970 PH/35 Installed configuration files for Dragonfly.
4972 PH/36 When a locally submitted message by a trusted user did not contain a
4973 From: header, and the sender address was obtained from -f or from an SMTP
4974 MAIL command, and the trusted user did not use -F to supply a sender
4975 name, $originator_name was incorrectly used when constructing a From:
4976 header. Furthermore, $originator_name was used for submission mode
4977 messages from external hosts without From: headers in a similar way,
4978 which is clearly wrong.
4980 PH/37 Added control=suppress_local_fixups.
4982 PH/38 When log_selector = +received_sender was set, and the addition of the
4983 sender made the log line's construction buffer exactly full, or one byte
4984 less than full, an overflow happened when the terminating "\n" was
4987 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4988 when the result of a list match is failure because a DNS lookup failed.
4990 PH/40 RM_COMMAND is now used in the building process.
4992 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4993 the "build-* directories that it finds.
4995 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4996 address in a domain literal was a prefix of an interface address.
4998 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4999 when verifying a sender address, unless rewrite_headers is false.
5001 PH/44 Wrote a long comment about why errors_to addresses are verified as
5002 recipients, not senders.
5004 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5005 the ratelimit ACL was added.
5007 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5009 PH/46 Added extra information about PostgreSQL errors to the error string.
5011 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5012 avoiding the need to install special zones in a real server. This is
5013 backwards compatible; if it can't find the fake resolver, it drops back.
5014 Thus, both old and new test suites can be run.
5016 TF/02 Added util/ratelimit.pl
5018 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5019 clock is set back in time.
5021 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5022 Candler <B.Candler@pobox.com>.
5024 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5025 OK for addresses that are the result of redirections.
5027 PH/48 A number of further additions for the benefit of the new test suite,
5028 including a fake gethostbyname() that interfaces to the fake DNS resolver
5031 TF/06 The fix for widen_domains has also been applied to qualify_single and
5032 search_parents which are the other dnslookup options that can cause
5035 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5038 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5040 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5041 service" if it encountered an MX record with an empty target hostname.
5042 The message is now "an MX or SRV record indicated no SMTP service".
5044 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5045 if the DNS of the sending site is misconfigured. This is quite a
5046 common situation. This change restores the behaviour of treating a
5047 helo verification defer as a failure.
5049 PH/51 If self=fail was set on a router, the bounce message did not include the
5050 actual error message.
5056 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5058 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5059 command itself ended in error, the underlying message about the transport
5060 filter timeout was being overwritten with the pipe command error. Now the
5061 underlying error message should be appended to the second error message.
5063 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5065 PH/02 Reduce the amount of output that "make" produces by default. Full output
5066 can still be requested.
5068 PH/03 The warning log line about a condition test deferring for a "warn" verb
5069 was being output only once per connection, rather than after each
5070 occurrence (because it was using the same function as for successful
5071 "warn" verbs). This seems wrong, so I have changed it.
5073 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5074 it should not have, which might have caused a crash in the right
5075 circumstances, but probably never did.
5077 PH/04 Installed a modified version of Tony Finch's patch to make submission
5078 mode fix the return path as well as the Sender: header line, and to
5079 add a /name= option so that you can make the user's friendly name appear
5082 TF/03 Added the control = fakedefer ACL modifier.
5084 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5085 Mark Lowes for thorough testing.
5087 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5089 TK/03 Merged latest SRS patch from Miles Wilton.
5091 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5092 with the definition in sysexits.h (which is #included earlier).
5093 Fortunately, Exim does not actually use EX_OK. The code used to try to
5094 preserve the sysexits.h value, by assuming that macro definitions were
5095 scanned for macro replacements. I have been disabused of this notion,
5096 so now the code just undefines EX_OK before #including unistd.h.
5098 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5099 in the smtp transport. When a block could not be written in a single
5100 write() function, the timeout was being re-applied to each part-write.
5101 This seems wrong - if the receiver was accepting one byte at a time it
5102 would take for ever. The timeout is now adjusted when this happens. It
5103 doesn't have to be particularly precise.
5105 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5106 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5108 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5109 without a domain if the check to <postmaster@domain> fails.
5111 SC/01 Eximstats: added -xls and the ability to specify output files
5112 (patch written by Frank Heydlauf).
5114 SC/02 Eximstats: use FileHandles for outputting results.
5116 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5118 SC/04 Eximstats: fixed display of large numbers with -nvr option
5120 SC/05 Eximstats: fixed merging of reports with empty tables.
5122 SC/06 Eximstats: added the -include_original_destination flag
5124 SC/07 Eximstats: removed tabs and trailing whitespace.
5126 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5128 TK/06 MBOX spool code: Add real "From " MBOX separator line
5129 so the .eml file is really in mbox format (even though
5130 most programs do not really care). Patch from Alex Miller.
5132 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5133 The latter is generated from $received_to and is only set if the
5134 message has one envelope recipient. SA can use these headers,
5135 obviously out-of-the-box. Patch from Alex Miller.
5137 PH/08 The ${def test on a variable was returning false if the variable's
5138 value was "0", contrary to what the specification has always said!
5139 The result should be true unless the variable is empty.
5141 PH/09 The syntax error of a character other than { following "${if
5142 def:variable_name" (after optional whitespace) was not being diagnosed.
5143 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5144 accidental colon was present, for example, could give incorrect results.
5146 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5147 result is used (not including appendfile, where other changes are about
5150 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5151 This involved changing a lot of size variables from int to off_t. It
5152 should work with maildirs and everything.
5154 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5155 spamd dying while we are connected to it.
5157 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5160 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5161 function for BDB 4.3.
5163 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5165 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5166 the sources for PCRE have been re-arranged and more files are now
5169 PH/15 The code I had for printing potentially long long variables in PH/11
5170 above was not the best (it lost precision). The length of off_t variables
5171 is now inspected at build time, and an appropriate printing format (%ld
5172 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5173 to be "long long int" or "long int". This is needed for the internal
5174 formatting function string_vformat().
5176 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5177 the configuration file to be ":syslog", then the script "guesses" where
5178 the logs files are, rather than using the compiled in default. In our
5179 case the guess is not the same as the compiled default, so the script
5180 suddenly stopped working when I started to use syslog. The patch checks
5181 to see if log_file_path is "". If so, it attempts to read it from exim
5182 with no configuration file to get the compiled in version, before it
5183 falls back to the previous guessing code."
5185 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5186 implementing BATV in an Exim configuration. See NewStuff for the gory
5189 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5190 Makefile that are specific to HP-UX.
5192 PH/18 If the "use_postmaster" option was set for a recipient callout together
5193 with the "random" option, the postmaster address was used as the MAIL
5194 FROM address for the random test, but not for the subsequent recipient
5195 test. It is now used for both.
5197 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5198 patch removes a few documentation additions to RFC 3028, because the
5199 latest draft now contains them. It adds the new en;ascii-case comparator
5200 and a new error check for 8bit text in MIME parts. Comparator and
5201 require names are now matched exactly. I enabled the subaddress
5202 extension, but it is not well tested yet (read: it works for me)."
5204 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5205 rework some of the code of TK/09 above to avoid the hardwired use of
5206 "%lld" and "long long". Replaced the call to snprintf() with a call to
5209 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5210 records point to non-existent hosts", "retry timeout exceeded", and
5211 "retry time not reached for any host after a long failure period".
5213 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5214 experimental DomainKeys support:
5216 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5217 (2) On an error such as an illegally used "control", the wrong name for
5218 the control was given.
5220 These problems did NOT occur unless DomainKeys support was compiled.
5222 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5224 PH/24 Added ${if match_ip condition.
5226 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5227 databases so that it will be absolutely obvious if a crash occurs in the
5228 DB library. This is a regular occurrence (often caused by mis-matched
5231 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5232 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5233 picked up on a user's system that detects such things. There doesn't seem
5234 to be a gcc warning option for this - only an attribute that has to be
5235 put on the function's prototype. It seems that in Fedora Core 4 they have
5236 set this on a number of new functions. No doubt there will be more in due
5239 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5240 specify a transport. However, if an address that was verified by such a
5241 router was the subject of a callout, Exim crashed because it tried to
5242 read the rcpt_include_affixes from the non-existent transport. Now it
5243 just assumes that the setting of that option is false. This bug was
5244 introduced by 4.51/PH/31.
5246 PH/28 Changed -d+all to exclude +memory, because that information is very
5247 rarely of interest, but it makes the output a lot bigger. People tend to
5248 do -d+all out of habit.
5250 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5251 code in os-type was giving problems when libc.so lives in lib64, like on
5254 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5255 aren't the modern standard, and it seems that some systems' include files
5256 don't always have them. Exim was already checking for some of the newer
5257 ones like T_AAAA, and defining it itself. I've added checks for all the
5258 record types that Exim uses.
5260 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5261 not automatically generating a new one, as it is supposed to. This
5262 prevented TLS from working. If the file did exist, but contained invalid
5263 data, a new version was generated, as expected. It was only the case of a
5264 non-existent file that was broken.
5266 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5267 with a change in libdomainkeys > 0.64.
5269 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5270 from DNS. If the selector record carries the flag, it now has
5271 precedence over the domain-wide flag.
5273 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5275 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5276 the use of a port name that isn't defined in /etc/services) occurred, the
5277 message was deferred as in a normal delivery, and thus remained on the
5278 spool, instead of being failed because of the mua_wrapper setting. This
5279 is now fixed, and I tidied up some of the mua_wrapper messages at the
5282 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5283 the messages in a hash of arrays rather than using individual hashes.
5284 This is a bit cleaner and results in dramatic memory savings, albeit
5285 at a slight CPU cost.
5287 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5288 as requested by Marc Sherman.
5290 SC/10 Eximstats: added histograms for user specified patterns as requested
5293 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5295 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5296 fopen() in the content-scanning modules that did not already have it.
5302 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5303 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5305 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5307 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5309 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5310 to test Sieve filters that use "vacation".
5312 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5313 that changes the way the GnuTLS parameters are stored in the cache file.
5314 The new format can be generated externally. For backward compatibility,
5315 if the data in the cache doesn't make sense, Exim assumes it has read an
5316 old-format file, and it generates new data and writes a new file. This
5317 means that you can't go back to an older release without removing the
5320 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5321 work if there are any delivery delays because "one_time" forces the
5322 parent to be marked "delivered", so its unseen clone is never tried
5323 again. For this reason, Exim now forbids the simultaneous setting of
5326 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5327 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5328 entry if you want to know the details. The fix, however, neglected to
5329 consider the case when local delivery batching is involved. The test for
5330 "previously delivered" was not happening when checking to see if an
5331 address could be batched with a previous (undelivered) one; under
5332 certain circumstances this could lead to multiple deliveries to the same
5335 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5336 in its include files, and this causes problems building Exim.
5338 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5339 header_syntax) but Exim was just ignoring anything given after a slash.
5340 In particular, this caused confusion with an attempt to use "verify =
5341 reverse_host_lookup/defer_ok". An error is now given when options are
5342 supplied for verify items that do not have them. (Maybe reverse_host_
5343 lookup should have a defer_ok option, but that's a different point.)
5345 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5346 defined by RFC 821) to 2048, because there were problems with some AUTH
5347 commands, and RFC 1869 says the size should be increased for extended
5348 SMTP commands that take arguments.
5350 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5353 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5354 "unknown" error; now it says that the functionality isn't in the binary.
5356 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5357 an address' error message when a string expansion fails (syntax or
5358 whatever). Otherwise the password may appear in the log. Following change
5359 PH/42 below, there is no longer a chance of it appearing in a bounce
5362 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5364 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5365 of its addresses was ever tried. (Bugzilla bug #2.)
5367 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5368 the result incorrectly in the debug output. (It correctly added a newline
5369 to what was transported.)
5371 TF/01 Added $received_time.
5373 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5374 commented out examples of how to interface to a virus scanner and to
5375 SpamAssassin. Also added commented examples of av_scanner and
5376 spamd_address settings.
5378 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5379 and controls are allowed in which ACLs. There were a couple of minor
5380 errors. Some of the entries in the conditions table (which is a table of
5381 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5382 negation of where the condition IS allowed.
5384 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5386 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5387 header file does not have a version number, so I've had to invent a new
5388 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5389 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5390 radiusclient), but it was contributed by a Radius user.
5392 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5393 files or directories, for OpenSSL.
5395 PH/19 When an Exim process that is running as root has to create an Exim log
5396 file, it does so in a subprocess that runs as exim:exim so as to get the
5397 ownership right at creation (otherwise, other Exim processes might see
5398 the file with the wrong ownership). There was no test for failure of this
5399 fork() call, which would lead to the process getting stuck as it waited
5400 for a non-existent subprocess. Forks do occasionally fail when resources
5401 run out. I reviewed all the other calls to fork(); they all seem to check
5404 PH/20 When checking for unexpected SMTP input at connect time (before writing
5405 the banner), Exim was not dealing correctly with a non-positive return
5406 from the read() function. If the client had disconnected by this time,
5407 the result was a log entry for a synchronization error with an empty
5408 string after "input=" when read() returned zero. If read() returned -1
5409 (an event I could not check), uninitialized data bytes were printed.
5410 There were reports of junk text (parts of files, etc) appearing after
5413 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5415 PH/22 Added support for macro redefinition, and (re)definition in between
5416 driver and ACL definitions.
5418 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5419 forgetting to use the resulting value; it was using the unexpanded value.
5421 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5422 hadn't been configured. The fix is from Juergen Kreileder, who
5423 understands it better than I do:
5425 "Here's what I see happening with three configured cyrus_sasl
5426 authenticators configured (plain, login, cram-md5):
5428 On startup auth_cyrus_sasl_init() gets called for each of these.
5429 This means three calls to sasl_listmech() without a specified mech_list.
5430 => SASL tests which mechs of all available mechs actually work
5431 => three warnings about OTP not working
5432 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5434 With the patch, sasl_listmech() also gets called three times. But now
5435 SASL's mech_list option is set to the server_mech specified in the the
5436 authenticator. Or in other words, the answer from sasl_listmech()
5437 gets limited to just the mech you're testing for (which is different
5439 => the return list contains just 'plain' or 'login', 'cram-md5' or
5440 nothing depending on the value of ob->server_mech.
5442 I've just tested the patch: Authentication still works fine,
5443 unavailable mechs specified in the exim configuration are still
5444 caught, and the auth.log warnings about OTP are gone."
5446 PH/25 When debugging is enabled, the contents of the command line are added
5447 to the debugging output, even when log_selector=+arguments is not
5450 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5451 answer is "GNU", and only if the return is "GNU/something" is the answer
5454 PH/27 $acl_verify_message is now set immediately after the failure of a
5455 verification in an ACL, and so is available in subsequent modifiers. In
5456 particular, the message can be preserved by coding like this:
5458 warn !verify = sender
5459 set acl_m0 = $acl_verify_message
5461 Previously, $acl_verify_message was set only while expanding "message"
5462 and "log_message" when a very denied access.
5464 PH/28 Modified OS/os.c-Linux with
5466 -#ifndef OS_LOAD_AVERAGE
5467 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5469 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5470 nomenclature these days.)
5472 PH/29 Installed patch from the Sieve maintainer that adds the options
5473 sieve_useraddress and sieve_subaddress to the redirect router.
5475 PH/30 In these circumstances:
5476 . Two addresses routed to the same list of hosts;
5477 . First host does not offer TLS;
5478 . First host accepts first address;
5479 . First host gives temporary error to second address;
5480 . Second host offers TLS and a TLS session is established;
5481 . Second host accepts second address.
5482 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5483 and peerdn, if requested) that were in fact used only for the second
5486 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5487 attention to any local part prefix or suffix that was matched by the
5488 router that accepted the address. It now behaves in the same way as it
5489 does for delivery: the affixes are removed from the local part unless
5490 rcpt_include_affixes is set on the transport.
5492 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5493 timeout during the DATA phase of an incoming message.
5495 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5496 applied a patch sanctioned by the Sieve maintainer.
5498 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5499 the uid or gid is negative. A case of a negative gid caused this to be
5500 noticed. The fix allows for either to be negative.
5502 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5503 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5506 PH/36 The change PH/12 above was broken. Fixed it.
5508 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5509 the grounds that routing the same address twice would always produce the
5510 same answer. This might have been true once, but it is certainly no
5511 longer true now. Routing a child address may depend on the previous
5512 routing that produced that child. Some complicated redirection strategies
5513 went wrong when messages had multiple recipients, and made Exim's
5514 behaviour dependent on the order in which the addresses were given.
5516 I have moved the duplicate checking until after the routing is complete.
5517 Exim scans the addresses that are assigned to local and remote
5518 transports, and removes any duplicates. This means that more work will be
5519 done, as duplicates will always all be routed, but duplicates are
5520 presumably rare, so I don't expect this is of any significance.
5522 For deliveries to pipes, files, and autoreplies, the duplicate checking
5523 still happens during the routing process, since they are not going to be
5526 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5527 It corrects a timeout issue with spamd. This is Ian's comment: "The
5528 background is that sometimes spamd either never reads data from a
5529 connection it has accepted, or it never writes response data. The exiscan
5530 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5531 blindly assumes that writes won't block so it may never time out."
5533 PH/39 Allow G after quota size as well as K and M.
5535 PH/40 The value set for $authenticated_id in an authenticator may not contain
5536 binary zeroes or newlines because the value is written to log lines and
5537 to spool files. There was no check on this. Now the value is run through
5538 the string_printing() function so that such characters are converted to
5539 printable escape sequences.
5541 PH/41 $message_linecount is a new variable that contains the total number of
5542 lines in the message. Compare $body_linecount, which is the count for the
5545 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5546 bounce and delay warning messages, except in certain special cases, which
5549 (a) An SMTP error message from a remote host;
5550 (b) A message specified in a :fail: redirection;
5551 (c) A message specified in a "fail" command in a system filter;
5552 (d) A message specified in a FAIL return from the queryprogram router;
5553 (e) A message specified by the cannot_route_message router option.
5555 In these cases only, Exim does include the error details in bounce and
5556 warning messages. There are also a few cases where bland messages such
5557 as "unrouteable address" or "local delivery error" are given.
5559 PH/43 $value is now also set for the "else" part of a ${run expansion.
5561 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5562 being worked on, but at least Exim now implements the latest version to
5565 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5566 process to complete was treated as a delivery failure, a timeout while
5567 writing the message to the pipe was logged, but erroneously treated as a
5568 successful delivery. Such timeouts include transport filter timeouts. For
5569 consistency with the overall process timeout, these timeouts are now
5570 treated as errors, giving rise to delivery failures by default. However,
5571 there is now a new Boolean option for the pipe transport called
5572 timeout_defer, which, if set TRUE, converts the failures into defers for
5573 both kinds of timeout. A transport filter timeout is now identified in
5576 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5577 systems where "make" and "gmake" are different, calling "gmake" at top
5578 level broke things. I've arranged for the value of $(MAKE) to be passed
5579 from the Makefile to this script so that it can call the same version of
5583 A note about Exim versions 4.44 and 4.50
5584 ----------------------------------------
5586 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5587 changes of various kinds. As a consequence, a big documentation update was
5588 needed. This delayed the release for rather longer than seemed good, especially
5589 in the light of a couple of (minor) security issues. Therefore, the changes
5590 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5591 release. So 4.44 and 4.50 are in effect two different branches that both start
5594 I have left the 4.50 change log unchanged; it contains all the changes since
5595 4.43. The change log for 4.44 is below; many of its items are identical to
5596 those for 4.50. This seems to be the most sensible way to preserve the
5597 historical information.
5603 1. Minor wording change to the doc/README.SIEVE file.
5605 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5606 computation of the current number of files was incorrect.
5608 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5609 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5612 4. Give more explanation in the error message when the command for a transport
5613 filter fails to execute.
5615 5. There are several places where Exim runs a non-Exim command in a
5616 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5617 was being done only for the command run by the queryprogram router. It is
5618 now done for all such subprocesses. The other cases are: ${run, transport
5619 filters, and the commands run by the lmtp and pipe transports.
5621 6. Added CONFIGURE_GROUP build-time option.
5623 7. Some older OS have a limit of 256 on the maximum number of file
5624 descriptors. Exim was using setrlimit() to set 1000 as a large value
5625 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5626 systems. I've change it so that if it can't get 1000, it tries for 256.
5628 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5629 was an oversight, and furthermore, ever since the addition of extra
5630 controls (e.g. 4.43/32), the checks on when to allow different forms of
5631 "control" were broken. There should now be diagnostics for all cases when a
5632 control that does not make sense is encountered.
5634 9. Added the /retain_sender option to "control=submission".
5636 10. $recipients is now available in the predata ACL (oversight).
5638 11. Tidy the search cache before the fork to do a delivery from a message
5639 received from the command line. Otherwise the child will trigger a lookup
5640 failure and thereby defer the delivery if it tries to use (for example) a
5641 cached ldap connection that the parent has called unbind on.
5643 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5644 of $address_data from the recipient verification was clobbered by the
5645 sender verification.
5647 13. The value of address_data from a sender verification is now available in
5648 $sender_address_data in subsequent conditions in the ACL statement.
5650 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5652 15. Added a new option "connect=<time>" to callout options, to set a different
5655 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5656 was its contents. (It was OK if the option was not defined at all.)
5658 17. A "Completed" log line is now written for messages that are removed from
5659 the spool by the -Mrm option.
5661 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5662 information about exactly what failed.
5664 19. Added -dd to debug only the daemon process.
5666 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5667 handles timeouts, both on the server side and network timeouts. Renamed the
5668 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5670 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5671 It is now set to "smtps".
5673 22. $host_address is now set to the target address during the checking of
5674 ignore_target_hosts.
5676 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5677 being passed; this would have caused $sender_host_name to have been used if
5678 matching the list had actually called for a host name (not very likely,
5679 since this list is usually IP addresses). A host name is now passed as
5682 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5683 code that specifies a non-restarting handler (typically sigaction() in
5684 modern systems) in an attempt to fix a rare and obscure crash bug.
5686 25. Narrowed the window for a race in the daemon that could cause it to ignore
5687 SIGCHLD signals. This is not a major problem, because they are used only to
5688 wake it up if nothing else does.
5690 26. A malformed maildirsize file could cause Exim to calculate negative values
5691 for the mailbox size or file count. Odd effects could occur as a result.
5692 The maildirsize information is now recalculated if the size or filecount
5695 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5696 support for a long time. Removed HAVE_SYS_VFS_H.
5698 28. Installed the latest version of exipick from John Jetmore.
5700 29. In an address list, if the pattern was not a regular expression, an empty
5701 subject address (from a bounce message) matched only if the pattern was an
5702 empty string. Non-empty patterns were not even tested. This was the wrong
5703 because it is perfectly reasonable to use an empty address as part of a
5704 database query. An empty address is now tested by patterns that are
5705 lookups. However, all the other forms of pattern expect the subject to
5706 contain a local part and a domain, and therefore, for them, an empty
5707 address still always fails if the pattern is not itself empty.
5709 30. Exim went into a mad DNS loop when attempting to do a callout where the
5710 host was specified on an smtp transport, and looking it up yielded more
5711 than one IP address.
5713 31. Re-factored the code for checking spool and log partition space into a
5714 function that finds that data and another that does the check. The former
5715 is then used to implement four new variables: $spool_space, $log_space,
5716 $spool_inodes, and $log_inodes.
5718 32. The RFC2047 encoding function was originally intended for short strings
5719 such as real names; it was not keeping to the 75-character limit for
5720 encoded words that the RFC imposes. It now respects the limit, and
5721 generates multiple encoded words if necessary. To be on the safe side, I
5722 have increased the buffer size for the ${rfc2047: expansion operator from
5725 33. It is now permitted to omit both strings after an "if" condition; if the
5726 condition is true, the result is "true". As before, when the second string
5727 is omitted, a false condition yields an empty string. This makes it less
5728 cumbersome to write custom ACL and router conditions.
5730 34. Failure to deliver a bounce message always caused it to be frozen, even if
5731 there was an errors_to setting on the router. The errors_to setting is now
5734 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5735 canonical form (fully expanded) before being placed in
5736 $sender_host_address.
5738 36. The table in the code that translates DNS record types into text (T_A to
5739 "A" for instance) was missing entries for NS and CNAME. It is just possible
5740 that this could have caused confusion if both these types were looked up
5741 for the same domain, because the text type is used as part of Exim's
5742 per-process caching. But the chance of anyone hitting this buglet seems
5745 37. The dnsdb lookup has been extended in a number of ways.
5747 (1) There is a new type, "zns", which walks up the domain tree until it
5748 finds some nameserver records. It should be used with care.
5750 (2) There is a new type, "mxh", which is like "mx" except that it returns
5751 just the host names, not the priorities.
5753 (3) It is now possible to give a list of domains (or IP addresses) to be
5754 looked up. The behaviour when one of the lookups defers can be
5755 controlled by a keyword.
5757 (4) It is now possible to specify the separator character for use when
5758 multiple records are returned.
5760 38. The dnslists ACL condition has been extended: it is now possible to supply
5761 a list of IP addresses and/or domains to be looked up in a particular DNS
5764 39. Added log_selector=+queue_time_overall.
5766 40. When running the queue in the test harness, wait just a tad after forking a
5767 delivery process, to get repeatability of debugging output.
5769 41. Include certificate and key file names in error message when GnuTLS fails
5770 to set them up, because the GnuTLS error message doesn't include the name
5771 of the failing file when there is a problem reading it.
5773 42. Allow both -bf and -bF in the same test run.
5775 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5777 44. The "Exiscan patch" is now merged into the mainline Exim source.
5779 45. Sometimes the final signoff response after QUIT could fail to get
5780 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5781 < 0 before doing a fflush(). This bug looks as though it goes back to the
5782 introduction of TLS in release 3.20, but "sometimes" must have been rare
5783 because the tests only now provoked it.
5785 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5786 (this can affect the format of dates).
5788 47. exim_tidydb, when checking for the continued existence of a message for
5789 which it has found a message-specific retry record, was not finding
5790 messages that were in split spool directories. Consequently, it was
5791 deleting retry records that should have stayed in existence.
5793 48. Steve fixed some bugs in eximstats.
5795 49. The SPA authentication driver was not abandoning authentication and moving
5796 on to the next authenticator when an expansion was forced to fail,
5797 contradicting the general specification for all authenticators. Instead it
5798 was generating a temporary error. It now behaves as specified.
5800 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5801 (the order specifies the preference for clients). The order is now AES256,
5802 AES128, 3DES, ARCFOUR128.
5804 51. Small patch to Sieve code - explicitly set From: when generating an
5807 52. Exim crashed if a remote delivery caused a very long error message to be
5808 recorded - for instance if somebody sent an entire SpamAssassin report back
5809 as a large number of 550 error lines. This bug was coincidentally fixed by
5810 increasing the size of one of Exim's internal buffers (big_buffer) that
5811 happened as part of the Exiscan merge. However, to be on the safe side, I
5812 have made the code more robust (and fixed the comments that describe what
5815 53. Now that there can be additional text after "Completed" in log lines (if
5816 the queue_time_overall log selector is set), a one-byte patch to exigrep
5817 was needed to allow it to recognize "Completed" as not the last thing in
5820 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5821 patch that reportedly fixes this has been added. I am not expert enough to
5822 create a test for it. This is what the patch creator wrote:
5824 "I found a little strange behaviour of ldap code when working with
5825 Windows 2003 AD Domain, where users was placed in more than one
5826 Organization Units. When I tried to give exim partial DN, the exit code
5827 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5828 But simultaneously result of request was absolutely normal ldap result,
5829 so I produce this patch..."
5831 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5832 REFERENCE, so I have modified the code to exclude the patch when that macro
5835 55. Some experimental protocols are using DNS PTR records for new purposes. The
5836 keys for these records are domain names, not reversed IP addresses. The
5837 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5838 leaves it alone. Component reversal etc. now happens only for IP addresses.
5841 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5843 57. Double the size of the debug message buffer (to 2048) so that more of very
5844 long debug lines gets shown.
5846 58. The exicyclog utility now does better if the number of log files to keep
5847 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5849 59. Two changes related to the smtp_active_hostname option:
5851 (1) $smtp_active_hostname is now available as a variable.
5852 (2) The default for smtp_banner uses $smtp_active_hostname instead
5853 of $primary_hostname.
5855 60. The host_aton() function is supposed to be passed a string that is known
5856 to be a valid IP address. However, in the case of IPv6 addresses, it was
5857 not checking this. This is a hostage to fortune. Exim now panics and dies
5858 if the condition is not met. A case was found where this could be provoked
5859 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5860 components; fortuitously, this particular loophole had already been fixed
5861 by change 4.50/55 above.
5863 If there are any other similar loopholes, the new check in host_aton()
5864 itself should stop them being exploited. The report I received stated that
5865 data on the command line could provoke the exploit when Exim was running as
5866 exim, but did not say which command line option was involved. All I could
5867 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5868 running as the user.
5871 61. There was a buffer overflow vulnerability in the SPA authentication code
5872 (which came originally from the Samba project). I have added a test to the
5873 spa_base64_to_bits() function which I hope fixes it.
5876 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5877 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5879 63. The daemon start-up calls getloadavg() while still root for those OS that
5880 need the first call to be done as root, but it missed one case: when
5881 deliver_queue_load_max is set with deliver_drop_privilege. This is
5882 necessary for the benefit of the queue runner, because there is no re-exec
5883 when deliver_drop_privilege is set.
5885 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5886 This has been fixed.
5888 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5889 was in use, was not putting the data itself into the right store pool;
5890 consequently, it could be overwritten for a subsequent message in the same
5891 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5894 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5896 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5897 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5898 the function were treating the return as a boolean value, which happened to
5899 work because 0=false and not-0=true, but is not correct code.
5901 68. The host_aton() function was not handling scoped IPv6 addresses (those
5902 with, for example, "%eth0" on the end) correctly.
5904 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5905 negated items (that is, ~something) in unsigned ints. Some compilers
5906 apparently mutter when there is no cast.
5908 70. If an address verification called from an ACL failed, and did not produce a
5909 user-specific message (i.e. there was only a "system" message), nothing was
5910 put in $acl_verify_message. In this situation, it now puts the system
5913 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5914 session; change 4.31/43 added the unwanted input to the log line - except
5915 that it did not do this in the start of session case. It now does.
5917 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5918 This is wrong and can cause the other end to generate a synchronization
5919 error if it is another Exim or anything else that does the synchronization
5920 check. A QUIT command is no longer sent after a timeout.
5922 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5923 during host lookups.
5925 74. The defer_ok option of callout verification was not working if it was used
5926 when verifying addresses in header lines, that is, for this case:
5928 verify = header_sender/callout=defer_ok
5930 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5931 those file descriptors could be used for SMTP connections. If anything
5932 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5933 could be sent to the SMTP client, causing chaos. The daemon now opens
5934 stdin, stdout, and stderr to /dev/null when it puts itself into the
5937 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5938 log by default. The user can override this with suitable Perl magic.
5940 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5941 the log message when discard triggers, was not working for the DATA ACL or
5942 for the non-SMTP ACL.
5944 78. Error message wording change in sieve.c.
5946 79. If smtp_accept_max_per_host was set, the number of connections could be
5947 restricted to fewer than expected, because the daemon was trying to set up
5948 a new connection before checking whether the processes handling previous
5949 connections had finished. The check for completed processes is now done
5950 earlier. On busy systems, this bug wouldn't be noticed because something
5951 else would have woken the daemon, and it would have reaped the completed
5954 80. If a message was submitted locally by a user whose login name contained one
5955 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5956 It caused a spool format error. I have fixed the spool reading code. A
5957 related problem was that the "from" clause in the Received: line became
5958 illegal because of the space(s). It is now covered by ${quote_local_part.
5960 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5963 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5965 83. Patch from Sieve maintainer to fix unterminated string problem in
5966 "vacation" handling.
5968 84. Some minor changes to the Linux configuration files to help with other
5969 OS variants using glibc.
5971 85. One more patch for Sieve to update vacation handling to latest spec.
5974 ----------------------------------------------------
5975 See the note above about the 4.44 and 4.50 releases.
5976 ----------------------------------------------------
5982 1. Change 4.43/35 introduced a bug that caused file counts to be
5983 incorrectly computed when quota_filecount was set in an appendfile
5986 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5987 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5990 3. Give more explanation in the error message when the command for a transport
5991 filter fails to execute.
5993 4. There are several places where Exim runs a non-Exim command in a
5994 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5995 was being done only for the command run by the queryprogram router. It is
5996 now done for all such subprocesses. The other cases are: ${run, transport
5997 filters, and the commands run by the lmtp and pipe transports.
5999 5. Some older OS have a limit of 256 on the maximum number of file
6000 descriptors. Exim was using setrlimit() to set 1000 as a large value
6001 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6002 systems. I've change it so that if it can't get 1000, it tries for 256.
6004 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6005 was an oversight, and furthermore, ever since the addition of extra
6006 controls (e.g. 4.43/32), the checks on when to allow different forms of
6007 "control" were broken. There should now be diagnostics for all cases when a
6008 control that does not make sense is encountered.
6010 7. $recipients is now available in the predata ACL (oversight).
6012 8. Tidy the search cache before the fork to do a delivery from a message
6013 received from the command line. Otherwise the child will trigger a lookup
6014 failure and thereby defer the delivery if it tries to use (for example) a
6015 cached ldap connection that the parent has called unbind on.
6017 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6018 of $address_data from the recipient verification was clobbered by the
6019 sender verification.
6021 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6022 was its contents. (It was OK if the option was not defined at all.)
6024 11. A "Completed" log line is now written for messages that are removed from
6025 the spool by the -Mrm option.
6027 12. $host_address is now set to the target address during the checking of
6028 ignore_target_hosts.
6030 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6031 being passed; this would have caused $sender_host_name to have been used if
6032 matching the list had actually called for a host name (not very likely,
6033 since this list is usually IP addresses). A host name is now passed as
6036 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6037 code that specifies a non-restarting handler (typically sigaction() in
6038 modern systems) in an attempt to fix a rare and obscure crash bug.
6040 15. Narrowed the window for a race in the daemon that could cause it to ignore
6041 SIGCHLD signals. This is not a major problem, because they are used only to
6042 wake it up if nothing else does.
6044 16. A malformed maildirsize file could cause Exim to calculate negative values
6045 for the mailbox size or file count. Odd effects could occur as a result.
6046 The maildirsize information is now recalculated if the size or filecount
6049 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6050 support for a long time. Removed HAVE_SYS_VFS_H.
6052 18. Updated exipick to current release from John Jetmore.
6054 19. Allow an empty sender to be matched against a lookup in an address list.
6055 Previously the only cases considered were a regular expression, or an
6058 20. Exim went into a mad DNS lookup loop when doing a callout where the
6059 host was specified on the transport, if the DNS lookup yielded more than
6062 21. The RFC2047 encoding function was originally intended for short strings
6063 such as real names; it was not keeping to the 75-character limit for
6064 encoded words that the RFC imposes. It now respects the limit, and
6065 generates multiple encoded words if necessary. To be on the safe side, I
6066 have increased the buffer size for the ${rfc2047: expansion operator from
6069 22. Failure to deliver a bounce message always caused it to be frozen, even if
6070 there was an errors_to setting on the router. The errors_to setting is now
6073 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6074 canonical form (fully expanded) before being placed in
6075 $sender_host_address.
6077 24. Updated eximstats to version 1.33
6079 25. Include certificate and key file names in error message when GnuTLS fails
6080 to set them up, because the GnuTLS error message doesn't include the name
6081 of the failing file when there is a problem reading it.
6083 26. Expand error message when OpenSSL has problems setting up cert/key files.
6086 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6087 (this can affect the format of dates).
6089 28. exim_tidydb, when checking for the continued existence of a message for
6090 which it has found a message-specific retry record, was not finding
6091 messages that were in split spool directories. Consequently, it was
6092 deleting retry records that should have stayed in existence.
6094 29. eximstats updated to version 1.35
6095 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6096 1.35 - bugfix such that pie charts by volume are generated correctly
6098 30. The SPA authentication driver was not abandoning authentication and moving
6099 on to the next authenticator when an expansion was forced to fail,
6100 contradicting the general specification for all authenticators. Instead it
6101 was generating a temporary error. It now behaves as specified.
6103 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6104 (the order specifies the preference for clients). The order is now AES256,
6105 AES128, 3DES, ARCFOUR128.
6107 31. Small patch to Sieve code - explicitly set From: when generating an
6110 32. Exim crashed if a remote delivery caused a very long error message to be
6111 recorded - for instance if somebody sent an entire SpamAssassin report back
6112 as a large number of 550 error lines. This bug was coincidentally fixed by
6113 increasing the size of one of Exim's internal buffers (big_buffer) that
6114 happened as part of the Exiscan merge. However, to be on the safe side, I
6115 have made the code more robust (and fixed the comments that describe what
6118 33. Some experimental protocols are using DNS PTR records for new purposes. The
6119 keys for these records are domain names, not reversed IP addresses. The
6120 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6121 leaves it alone. Component reversal etc. now happens only for IP addresses.
6124 34. The host_aton() function is supposed to be passed a string that is known
6125 to be a valid IP address. However, in the case of IPv6 addresses, it was
6126 not checking this. This is a hostage to fortune. Exim now panics and dies
6127 if the condition is not met. A case was found where this could be provoked
6128 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6129 components; fortuitously, this particular loophole had already been fixed
6130 by change 4.50/55 or 4.44/33 above.
6132 If there are any other similar loopholes, the new check in host_aton()
6133 itself should stop them being exploited. The report I received stated that
6134 data on the command line could provoke the exploit when Exim was running as
6135 exim, but did not say which command line option was involved. All I could
6136 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6137 running as the user.
6140 35. There was a buffer overflow vulnerability in the SPA authentication code
6141 (which came originally from the Samba project). I have added a test to the
6142 spa_base64_to_bits() function which I hope fixes it.
6145 36. The daemon start-up calls getloadavg() while still root for those OS that
6146 need the first call to be done as root, but it missed one case: when
6147 deliver_queue_load_max is set with deliver_drop_privilege. This is
6148 necessary for the benefit of the queue runner, because there is no re-exec
6149 when deliver_drop_privilege is set.
6151 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6152 was in use, was not putting the data itself into the right store pool;
6153 consequently, it could be overwritten for a subsequent message in the same
6154 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6157 38. Sometimes the final signoff response after QUIT could fail to get
6158 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6159 < 0 before doing a fflush(). This bug looks as though it goes back to the
6160 introduction of TLS in release 3.20, but "sometimes" must have been rare
6161 because the tests only now provoked it.
6167 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6168 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6169 Now it returns an integer. A number of calls were still expecting a T/F
6170 return. Fortuitously, in all cases, the tests worked in OK situations,
6171 which is the norm. However, things would have gone wrong on any write
6172 failures on the smtp file descriptor. This function is used when sending
6173 messages over SMTP and also when doing verify callouts.
6175 2. When Exim is called to do synchronous delivery of a locally submitted
6176 message (the -odf or -odi options), it no longer closes stderr before doing
6179 3. Implemented the mua_wrapper option.
6181 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6183 5. Implemented the functions header_remove(), header_testname(),
6184 header_add_at_position(), and receive_remove_recipient(), and exported them
6187 6. If an ACL "warn" statement specified the addition of headers, Exim already
6188 inserted X-ACL-Warn: at the start if there was no header name. However, it
6189 was not making this test for the second and subsequent header lines if
6190 there were newlines in the string. This meant that an invalid header could
6191 be inserted if Exim was badly configured.
6193 7. Allow an ACL "warn" statement to add header lines at the start or after all
6194 the Received: headers, as well as at the end.
6196 8. Added the rcpt_4xx retry error code.
6198 9. Added postmaster_mailfrom=xxx to callout verification option.
6200 10. Added mailfrom=xxxx to the callout verification option, for verify=
6203 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6204 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6206 12. Inserted some casts to stop certain compilers warning when using pointer
6207 differences as field lengths or precisions in printf-type calls (mostly
6208 affecting debugging statements).
6210 13. Added optional readline() support for -be (dynamically loaded).
6212 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6213 same clock tick as a message's arrival, so that its received time was the
6214 same as the "first fail" time on the retry record, and that message
6215 remained on the queue past the ultimate address timeout, every queue runner
6216 would try a delivery (because it was past the ultimate address timeout) but
6217 after another failure, the ultimate address timeout, which should have then
6218 bounced the address, did not kick in. This was a "< instead of <=" error;
6219 in most cases the first failure would have been in the next clock tick
6220 after the received time, and all would be well.
6222 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6223 being recognized when the domain list was tested by the match_domain
6224 condition in an expansion string.
6226 16. Added the ${str2b64: operator.
6228 17. Exim was always calling setrlimit() to set a large limit for the number of
6229 processes, without checking whether the existing limit was already
6230 adequate. (It did check for the limit on file descriptors.) Furthermore,
6231 errors from getrlimit() and setrlimit() were being ignored. Now they are
6232 logged to the main and panic logs, but Exim does carry on, to try to do its
6233 job under whatever limits there are.
6235 18. Imported PCRE 5.0.
6237 19. Trivial typo in log message " temporarily refused connection" (the leading
6240 20. If the log selector return_path_on_delivery was set and an address was
6241 redirected to /dev/null, the delivery process crashed because it assumed
6242 that a return path would always be set for a "successful" delivery. In this
6243 case, the whole delivery is bypassed as an optimization, and therefore no
6246 21. Internal re-arrangement: the function for sending a challenge and reading
6247 a response while authentication was assuming a zero-terminated challenge
6248 string. It's now changed to take a pointer and a length, to allow for
6249 binary data in such strings.
6251 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6253 23. Exim was not respecting finduser_retries when seeking the login of the
6254 uid under which it was called; it was always trying 10 times. (The default
6255 setting of finduser_retries is zero.) Also, it was sleeping after the final
6256 failure, which is pointless.
6258 24. Implemented tls_on_connect_ports.
6260 25. Implemented acl_smtp_predata.
6262 26. If the domain in control=submission is set empty, Exim assumes that the
6263 authenticated id is a complete email address when it generates From: or
6264 Sender: header lines.
6266 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6267 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6268 chown and chgrp in /bin and hostname in /usr/bin.
6270 28. Exim was keeping the "process log" file open after each use, just as it
6271 does for the main log. This opens the possibility of it remaining open for
6272 long periods when the USR1 signal hits a daemon. Occasional processlog
6273 errors were reported, that could have been caused by this. Anyway, it seems
6274 much more sensible not to leave this file open at all, so that is what now
6277 29. The long-running daemon process does not normally write to the log once it
6278 has entered its main loop, and it closes the log before doing so. This is
6279 so that log files can straightforwardly be renamed and moved. However,
6280 there are a couple of unusual error situations where the daemon does write
6281 log entries, and I had neglected to close the log afterwards.
6283 30. The text of an SMTP error response that was received during a remote
6284 delivery was being truncated at 512 bytes. This is too short for some of
6285 the long messages that one sometimes sees. I've increased the limit to
6288 31. It is now possible to make retry rules that apply only when a message has a
6289 specific sender, in particular, an empty sender.
6291 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6292 it possible to be selective about when SMTP synchronization is enforced.
6294 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6296 32. Implemented hosts_connection_nolog.
6298 33. Added an ACL for QUIT.
6300 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6303 35. Added mailbox_size and mailbox_filecount to appendfile.
6305 36. Added control = no_multiline_responses to ACLs.
6307 37. There was a bug in the logic of the code that waits for the clock to tick
6308 in the case where the clock went backwards by a substantial amount such
6309 that the microsecond fraction of "now" was more than the microsecond
6310 fraction of "then" (but the whole seconds number was less).
6312 38. Added support for the libradius Radius client library this is found on
6313 FreeBSD (previously only the radiusclient library was supported).
6319 1. When certain lookups returned multiple values in the form name=value, the
6320 quoting of the values was not always being done properly. Specifically:
6321 (a) If the value started with a double quote, but contained no whitespace,
6323 (b) If the value contained whitespace other than a space character (i.e.
6324 tabs or newlines or carriage returns) it was not quoted.
6325 This fix has been applied to the mysql and pgsql lookups by writing a
6326 separate quoting function and calling it from the lookup code. The fix
6327 should probably also be applied to nisplus, ibase and oracle lookups, but
6328 since I cannot test any of those, I have not disturbed their existing code.
6330 2. A hit in the callout cache for a specific address caused a log line with no
6331 reason for rejecting RCPT. Now it says "Previous (cached) callout
6332 verification failure".
6334 3. There was an off-by-one bug in the queryprogram router. An over-long
6335 return line was truncated at 256 instead of 255 characters, thereby
6336 overflowing its buffer with the terminating zero. As well as fixing this, I
6337 have increased the buffer size to 1024 (and made a note to document this).
6339 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6340 when Exim is waiting for an SMTP response from a remote server, Exim
6341 restarts its select() call on the socket, thereby resetting its timeout.
6342 This is not a problem when such interrupts are rare. Somebody set up a cron
6343 job to run exiwhat every 2 minutes, which is less than the normal select()
6344 timeout (5 or 10 minutes). This meant that the select() timeout never
6345 kicked in because it was always reset. I have fixed this by comparing the
6346 time when an interrupt arrives with the time at the start of the first call
6347 to select(). If more time than the timeout has elapsed, the interrupt is
6348 treated as a timeout.
6350 5. Some internal re-factoring in preparation for the addition of Sieve
6351 extensions (by MH). In particular, the "personal" test is moved to a
6352 separate function, and given an option for scanning Cc: and Bcc: (which is
6353 not set for Exim filters).
6355 6. When Exim created an email address using the login of the caller as the
6356 local part (e.g. when creating a From: or Sender: header line), it was not
6357 quoting the local part when it contained special characters such as @.
6359 7. Installed new OpenBSD configuration files.
6361 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6362 try to make them clearer.
6364 9. Callout options, other than the timeout value, were being ignored when
6365 verifying sender addresses in header lines. For example, when using
6367 verify = header_sender/callout=no_cache
6369 the cache was (incorrectly) being used.
6371 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6372 only the Cygwin environment.
6374 11. When return_path_on_delivery was set as a log selector, if different remote
6375 addresses in the same message used different return paths and parallel
6376 remote delivery occurred, the wrong values would sometimes be logged.
6377 (Whenever a remote delivery process finished, the return path value from
6378 the most recently started remote delivery process was logged.)
6380 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6381 lines when AUTH and/or TLS are in use. This is the "received protocol"
6382 field. Exim used to use "asmtp" for authenticated SMTP, without any
6383 indication (in the protocol name) for TLS use. Now it follows the RFC and
6384 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6385 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6386 names appear in log lines as well as in Received: header lines.
6388 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6389 extensions, and comparison tests, and to fix some bugs.
6391 14. Changes to the "personal" filter test:
6393 (1) The test was buggy in that it was just doing the equivalent of
6394 "contains" tests on header lines. For example, if a user's address was
6395 anne@some.where, the "personal" test would incorrectly be true for
6397 To: susanne@some.where
6399 This test is now done by extracting each address from the header in turn,
6400 and checking the entire address. Other tests that are part of "personal"
6401 are now done using regular expressions (for example, to check local parts
6402 of addresses in From: header lines).
6404 (2) The list of non-personal local parts in From: addresses has been
6405 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6406 taken from the Sieve specification recommendations.
6408 (3) If the message contains any header line starting with "List-" it is
6409 treated as non-personal.
6411 (4) The test for "circular" in the Subject: header line has been removed
6412 because it now seems ill-conceived.
6414 15. Minor typos in src/EDITME comments corrected.
6416 16. Installed latest exipick from John Jetmore.
6418 17. If headers_add on a router specified a text string that was too long for
6419 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6420 of string_sprintf() is now avoided.
6422 18. $message_body_size was not set (it was always zero) when running the DATA
6423 ACL and the local_scan() function.
6425 19. For the "mail" command in an Exim filter, no default was being set for
6426 the once_repeat time, causing a random time value to be used if "once" was
6427 specified. (If the value happened to be <= 0, no repeat happened.) The
6428 default is now 0s, meaning "never repeat". The "vacation" command was OK
6429 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6430 (I found it when inspecting the code).
6432 20. There is now an overall timeout for performing a callout verification. It
6433 defaults to 4 times the callout timeout, which applies to individual SMTP
6434 commands during the callout. The overall timeout applies when there is more
6435 than one host that can be tried. The timeout is checked before trying the
6436 next host. This prevents very long delays if there are a large number of
6437 hosts and all are timing out (e.g. when the network connections are timing
6438 out). The value of the overall timeout can be changed by specifying an
6439 additional sub-option for "callout", called "maxwait". For example:
6441 verify = sender/callout=5s,maxwait=20s
6443 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6444 to the end before writing, but this should make it even safer).
6446 22. Exim was forgetting that it had advertised PIPELINING for the second and
6447 subsequent messages on an SMTP connection. It was also not resetting its
6448 memory on STARTTLS and an internal HELO.
6450 23. When Exim logs an SMTP synchronization error within a session, it now
6451 records whether PIPELINING has been advertised or not.
6453 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6454 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6455 rather than long int.
6457 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6459 26. Added the never_mail option to autoreply.
6465 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6466 crash if the getsockname() call failed; this can happen if a connection is
6467 closed very soon after it is established. The problem was simply in the
6468 order in which certain operations were done, causing Exim to try to write
6469 to the SMTP stream before it had set up the file descriptor. The bug has
6470 been fixed by making things happen in the correct order.
6476 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6477 before the connection was closed, thus losing the rejection response.
6479 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6480 some early Solaris releases, but causes trouble in current releases where
6481 socklen_t is defined.
6483 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6486 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6489 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6490 that releasing the top of it at the end releases what was used for sub-
6491 expansions (unless the block got too big). However, discard this block if
6492 the first thing is a variable or header, so that we can use its block when
6493 it is dynamic (useful for very large $message_headers, for example).
6495 6. Lookups now cache *every* query, not just the most recent. A new, separate
6496 store pool is used for this. It can be recovered when all lookup caches are
6497 flushed. Lookups now release memory at the end of their result strings.
6498 This has involved some general refactoring of the lookup sources.
6500 7. Some code has been added to the store_xxx() functions to reduce the amount
6501 of flapping under certain conditions.
6503 8. log_incoming_interface used to affect only the <= reception log lines. Now
6504 it causes the local interface and port to be added to several more SMTP log
6505 lines, for example "SMTP connection from", and rejection lines.
6507 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6509 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6511 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6512 could be overwritten at the end of the current message (or the start of a
6513 new message if it was set in a HELO ACL). The value is now preserved for
6514 the duration of the SMTP connection.
6516 12. If a transport had a headers_rewrite setting, and a matching header line
6517 contained an unqualified address, that address was qualified, even if it
6518 did not match any rewriting rules. The underlying bug was that the values
6519 of the flags that permit the existence of unqualified sender and recipient
6520 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6521 non-local messages, and by -bnq for local messages) were not being
6522 preserved with the message after it was received.
6524 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6525 "next input=" as part of the text comprising the host identity instead of
6526 the correct text. The code was using the same buffer for two different
6527 strings. However, depending on which order the printing function evaluated
6528 its arguments, the bug did not always show up. Under Linux, for example, my
6529 test suite worked just fine.
6531 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6532 doesn't work with some older versions of Perl. It has been changed to "my",
6533 which in any case is probably the better facility to use.
6535 15. A really picky compiler found some instances of statements for creating
6536 error messages that either had too many or two few arguments for the format
6539 16. The size of the buffer for calls to the DNS resolver has been increased
6540 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6541 for addresses that have a lot of PTR records. This alleviates a problem; it
6542 does not fully solve it.
6544 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6545 buffer now truncates the list and logs the incident, which is the same
6546 action as happens when Exim is looking up a host name and its aliases.
6547 Previously in this situation something unpredictable would happen;
6548 sometimes it was "internal error: store_reset failed".
6550 18. If a server dropped the connection unexpectedly when an Exim client was
6551 using GnuTLS and trying to read a response, the client delivery process
6552 crashed while trying to generate an error log message.
6554 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6555 string, for example:
6557 warn message = H1: something\nH2: something
6559 the text was added as a single header line from Exim's point of view
6560 though it ended up OK in the delivered message. However, searching for the
6561 second and subsequent header lines using $h_h2: did not work. This has been
6562 fixed. Similarly, if a system filter added multiple headers in this way,
6563 the routers could not see them.
6565 20. Expanded the error message when iplsearch is called with an invalid key to
6566 suggest using net-iplsearch in a host list.
6568 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6569 ACLs are no longer actually imposed (and a message to that effect is
6572 22. If a "gecos" field in a passwd entry contained escaped characters, in
6573 particular, if it contained a \" sequence, Exim got it wrong when building
6574 a From: or a Sender: header from that name. A second bug also caused
6575 incorrect handling when an unquoted " was present following a character
6576 that needed quoting.
6578 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6579 was not being matched caselessly.
6581 24. Arranged for all hyphens in the exim.8 source to be escaped with
6584 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6585 properly. Recipient callout cache records were still being keyed to include
6586 the sender, even when use_sender was set false. This led to far more
6587 callouts that were necessary. The sender is no longer included in the key
6588 when use_sender is false.
6590 26. Added "control = submission" modifier to ACLs.
6592 27. Added the ${base62d: operator to decode base 62 numbers.
6594 28. dnsdb lookups can now access SRV records.
6596 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6597 the configuration file.
6599 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6600 (-v) mode. This makes the output for a verbose queue run more intelligible.
6602 31. Added a use_postmaster feature to recipient callouts.
6604 32. Added the $body_zerocount variable, containing the number of binary zero
6605 bytes in the message body.
6607 33. The time of last modification of the "new" subdirectory is now used as the
6608 "mailbox time last read" when there is a quota error for a maildir
6611 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6613 35. Added +ignore_unknown as a special item in host lists.
6615 36. Code for decoding IPv6 addresses in host lists is now included, even if
6616 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6617 address was recognized as an IP address, but was then not correctly decoded
6618 into binary, causing unexpected and incorrect effects when compared with
6625 1. Very minor rewording of debugging text in manualroute to say "list of
6626 hosts" instead of "hostlist".
6628 2. If verify=header_syntax was set, and a header line with an unqualified
6629 address (no domain) and a large number of spaces between the end of the
6630 name and the colon was received, the reception process suffered a buffer
6631 overflow, and (when I tested it) crashed. This was caused by some obsolete
6632 code that should have been removed. The fix is to remove it!
6634 3. When running in the test harness, delay a bit after writing a bounce
6635 message to get a bit more predictability in the log output.
6637 4. Added a call to search_tidyup() just before forking a reception process. In
6638 theory, someone could use a lookup in the expansion of smtp_accept_max_
6639 per_host which, without the tidyup, could leave open a database connection.
6641 5. Added the variables $recipient_data and $sender_data which get set from a
6642 lookup success in an ACL "recipients" or "senders" condition, or a router
6643 "senders" option, similar to $domain_data and $local_part_data.
6645 6. Moved the writing of debug_print from before to after the "senders" test
6648 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6649 problems for message scanning, either using a data ACL, or using
6650 local_scan() because the Received: header was not generated till after they
6651 were called (in order to set the time as the time of reception completion).
6652 I have revised the way this works. The header is now generated after the
6653 body is received, but before the ACL or local_scan() are called. After they
6654 are run, the timestamp in the header is updated.
6660 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6661 before starting a queue runner without re-exec. This happened only when
6662 deliver_drop_privilege was set or when the Exim user was set to root. The
6663 effect of the bug was that timeouts during subsequent deliveries caused
6664 crashes instead of being properly handled. The handler is now left at its
6665 default (and expected) setting.
6667 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6668 message, again when deliver_drop_privilege is set or Exim is run as root.
6669 The bug described in (1) was not present in this case, but the tidying up
6670 of the other signals was missing. I have made the two cases consistent.
6672 3. The ignore_target_hosts setting on a manualroute router was being ignored
6673 for hosts that were looked up using the /MX notation.
6675 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6678 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6679 operated on the sender address. After changing the $sender_address to <>
6680 for the sender address verify, Exim was re-instated it as the original
6681 (before rewriting) address, but remembering that it had rewritten it, so it
6682 wasn't rewriting it again. This bug also had the effect of breaking the
6683 sender address verification caching when the sender address was rewritten.
6685 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6686 This has been changed so that if the ip literal address matches
6687 ignore_target_hosts, the router declines.
6689 7. Added expansion conditions match_domain, match_address, and match_local_
6690 part (NOT match_host).
6692 8. The placeholder for the Received: header didn't have a length field set.
6694 9. Added code to Exim itself and to exim_lock to test for a specific race
6695 condition that could lead to file corruption when using MBX delivery. The
6696 issue is with the lockfile that is created in /tmp. If this file is removed
6697 after a process has opened it but before that process has acquired a lock,
6698 there is the potential for a second process to recreate the file and also
6699 acquire a lock. This could lead to two Exim processes writing to the file
6700 at the same time. The added code performs the same test as UW imapd; it
6701 checks after acquiring the lock that its file descriptor still refers to
6702 the same named file.
6704 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6705 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6708 11. Added the smtp_active_hostname option. If used, this will typically be made
6709 to depend on the incoming interface address. Because $interface_address is
6710 not set up until the daemon has forked a reception process, error responses
6711 that can happen earlier (such as "too many connections") no longer contain
6714 12. If an expansion in a condition on a "warn" statement fails because a lookup
6715 defers, the "warn" statement is abandoned, and the next ACL statement is
6716 processed. Previously this caused the whole ACL to be aborted.
6718 13. Added the iplsearch lookup type.
6720 14. Added ident_timeout as a log selector.
6722 15. Added tls_certificate_verified as a log selector.
6724 16. Added a global option tls_require_ciphers (compare the smtp transport
6725 option of the same name). This controls incoming TLS connections.
6727 17. I finally figured out how to make tls_require_ciphers do a similar thing
6728 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6729 before starting the TLS session.
6731 18. Tabs are now shown as \t in -bP output.
6733 19. If the log selector return_path_on_delivery was set, Exim crashed when
6734 bouncing a message because it had too many Received: header lines.
6736 20. If two routers both had headers_remove settings, and the first one included
6737 a superfluous trailing colon, the final name in the first list and the
6738 first name in the second list were incorrectly joined into one item (with a
6739 colon in the middle).
6745 1. Added -C and -D options to the exinext utility, mainly to make it easier
6746 to include in the automated testing, but these could be helpful when
6747 multiple configurations are in use.
6749 2. The exinext utility was not formatting the output nicely when there was
6750 an alternate port involved in the retry record key, nor when there was a
6751 message id as well (for retries that were specific to a specific message
6752 and a specific host). It was also confused by IPv6 addresses, because of
6753 the additional colons they contain. I have fixed the IPv4 problem, and
6754 patched it up to do a reasonable job for IPv6.
6756 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6757 delivery, the log line now contains "pipelined" if PIPELINING was used.
6759 4. An SMTP transport process used to panic and die if the bind() call to set
6760 an explicit outgoing interface failed. This has been changed; it is now
6761 treated in the same way as a connect() failure.
6763 5. A reference to $sender_host_name in the part of a conditional expansion
6764 that was being skipped was still causing a DNS lookup. This no longer
6767 6. The def: expansion condition was not recognizing references to header lines
6768 that used bh_ and bheader_.
6770 7. Added the _cache feature to named lists.
6772 8. The code for checking quota_filecount in the appendfile transport was
6773 allowing one more file than it should have been.
6775 9. For compatibility with Sendmail, the command line option
6783 and sets the incoming protocol and host name (for trusted callers). The
6784 host name and its colon can be omitted when only the protocol is to be set.
6785 Note the Exim already has two private options, -pd and -ps, that refer to
6786 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6787 "s", but I don't think that's a major issue.
6789 10. A number of refactoring changes to the code, none of which should affect
6792 (a) The number of logging options was getting close to filling up the
6793 32-bit word that was used as a bit map. I have split them into two classes:
6794 those that are passed in the argument to log_write(), and those that are
6795 only ever tested independently outside of that function. These are now in
6796 separate 32-bit words, so there is plenty of room for expansion again.
6797 There is no change in the user interface or the logging behaviour.
6799 (b) When building, for example, log lines, the code previously used a
6800 macro that called string_cat() twice, in order to add two strings. This is
6801 not really sufficiently general. Furthermore, there was one instance where
6802 it was actually wrong because one of the argument was used twice, and in
6803 one call a function was used. (As it happened, calling the function twice
6804 did not affect the overall behaviour.) The macro has been replaced by a
6805 function that can join an arbitrary number of extra strings onto a growing
6808 (c) The code for expansion conditions now uses a table and a binary chop
6809 instead of a serial search (which was left over from when there were very
6810 few conditions). Also, it now recognizes conditions like "pam" even when
6811 the relevant support is not compiled in: a suitably worded error message is
6812 given if an attempt is made to use such a condition.
6814 11. Added ${time_interval:xxxxx}.
6816 12. A bug was causing one of the ddress fields not to be passed back correctly
6817 from remote delivery subprocesses. The field in question was not being
6818 subsequently used, so this caused to problems in practice.
6820 13. Added new log selectors queue_time and deliver_time.
6822 14. Might have fixed a bug in maildirsizefile handling that threw up
6823 "unexpected character" debug warnings, and recalculated the data
6824 unnecessarily. In any case, I expanded the warning message to give more
6827 15. Added the message "Restricted characters in address" to the statements in
6828 the default ACL that block characters like @ and % in local parts.
6830 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6831 Three changes have been made:
6833 (a) There was a serious bug; a negative response to MAIL caused the whole
6834 recipient domain to be cached as invalid, thereby blocking all messages
6835 to all local parts at the same domain, from all senders. This bug has
6836 been fixed. The domain is no longer cached after a negative response to
6837 MAIL if the sender used is not empty.
6839 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6842 (c) A new callout option, "use_sender" has been added for people who want
6843 the modified behaviour.
6849 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6852 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6853 indeed breaks things for older releases.
6855 3. Added additional logging to the case where there is a problem reading data
6856 from a filter that is running in a subprocess using a pipe, in order to
6857 try to track down a specific problem.
6859 4. Testing facility fudge: when running in the test harness and attempting
6860 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6861 getting "No route to host". Convert this to a timeout.
6863 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6866 6. Some OS don't have socklen_t but use size_t instead. This affects the
6867 fifth argument of getsockopt() amongst other things. This is now
6868 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6869 can be set for individual OS. I have set it for SunOS5, OSF1, and
6870 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6871 some earlier ones do not.
6873 7. Change 4.30/15 was not doing the test caselessly.
6875 8. The standard form for an IPv6 address literal was being rejected by address
6876 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6877 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6878 this, as well as the form without the "IPv6" on the front (but only when
6879 address literals are enabled, of course).
6881 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6883 10. Exim crashed if a message with an empty sender address specified by -f
6884 encountered a router with an errors_to setting. This could be provoked only
6885 by a command such as
6889 where an empty string was supplied; "<>" did not hit this bug.
6891 11. Installed PCRE release 4.5.
6893 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6894 remained set. It is now erased.
6896 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6897 times from message ids (which are base 36 rather than the normal 62).
6899 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6900 were being counted as actual protocol errors, and logged if the log
6901 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6902 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6903 and DATA following a set of rejected RCPTs do not count as protocol errors.
6904 In other words, Exim assumes they were pipelined, though this may not
6905 actually be the case. Of course, in all cases the client gets an
6906 appropriate error code.
6908 15. If a lookup fails in an ACL condition, a message about the failure may
6909 be available; it is used if testing the ACL cannot continue, because most
6910 such messages specify what the cause of the deferral is. However, some
6911 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6912 that caused an old message to be retained and used if a later statement
6913 caused a defer, replacing the real cause of the deferral.
6915 16. If an IP address had so many PTR records that the DNS lookup buffer
6916 was not large enough to hold them, Exim could crash while trying to process
6917 the truncated data. It now detects and logs this case.
6919 17. Further to 4.21/58, another change has been made: if (and only if) the
6920 first line of a message (the first header line) ends with CRLF, a bare LF
6921 in a subsequent header line has a space inserted after it, so as not to
6922 terminate the header.
6924 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6925 unnecessarily, used atoi() instead of strtol(), and didn't check the
6926 termination when getting file sizes from file names by regex.
6928 19. Completely re-implemented the support for maildirsize files, in the light
6929 of a number of problems with the previous contributed implementation
6930 (4.30/29). In particular:
6932 . If the quota is zero, the maildirsize file is maintained, but no quota is
6935 . If the maildir directory does not exist, it is created before any attempt
6936 to write a maildirsize file.
6938 . The quota value in the file is just a cache; if the quota is changed in
6939 the transport, the new value overrides.
6941 . A regular expression is available for excluding directories from the
6944 20. The autoreply transport checks the characters in options that define the
6945 message's headers; it allows continued headers, but it was checking with
6946 isspace() after an embedded newline instead of explicitly looking for a
6949 21. If all the "regular" hosts to which an address was routed had passed their
6950 expiry times, and had not reached their retry times, the address was
6951 bounced, even if fallback hosts were defined. Now Exim should go on to try
6954 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6955 equivalent code in the SMTP transport. Some hosts send humungous responses
6956 to HELO/EHLO, more than 1024 it seems.
6958 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6959 gives compiler warnings in some environments. I've now done it "properly",
6962 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6963 (because of problems with the built-in one) was declared to return uschar *
6964 instead of char *, causing compiler failure.
6966 25. Fixed a file descriptor leak when processing alias/forward files.
6968 26. Fixed a minor format string issue in dbfn.c.
6970 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6972 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6973 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6976 29. When a sender address is verified, it is cached, to save repeating the test
6977 when there is more than one recipient in a message. However, when the
6978 verification involves a callout, it is possible for different callout
6979 options to be set for different recipients. It is too complicated to keep
6980 track of this in the cache, so now Exim always runs a verification when a
6981 callout is required, relying on the callout cache for the optimization.
6982 The overhead is duplication of the address routing, but this should not be
6985 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6986 to be verified with callout=postmaster, and the main callout worked but the
6987 postmaster check failed, the verification correctly failed. However, if a
6988 subsequent RCPT command asked for sender verification *without* the
6989 postmaster check, incorrect caching caused this verification also to fail,
6992 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6993 it was not caching the DNS options (qualify_single, search_parents) that
6994 were used when the lookup failed. A subsequent lookup with different
6995 options therefore always gave the same answer, though there were cases
6996 where it should not have. (Example: a "domains = !$mx_any" option on a
6997 dnslookup router: the "domains" option is always processed without any
6998 widening, but the router might have qualify_single set.) Now Exim uses the
6999 cached value only when the same options are set.
7001 32. Added John Jetmore's "exipick" utility to the distribution.
7003 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7004 than a timeout (e.g. a certificate is required, and is not provided), an
7005 Exim server now closes the connection immediately. Previously it waited for
7006 the client to close - but if the client is SSL, it seems that they each
7007 wait for each other, leading to a delay before one of them times out.
7009 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7010 maintained 0.8.x compatibility because I don't think many are using it, and
7011 it is clearly obsolete.
7013 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7016 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7017 client certificate was expired. A simple patch fixes this, though I don't
7018 understand the full logic of why the verify callback is called multiple
7021 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7022 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7023 which causes problems with some clients (such as the Certicom SSL Plus
7024 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7025 disables the coutermeasure allowing Eudora to connect."
7027 38. Exim was not checking that a write() to a log file succeeded. This could
7028 lead to Bad Things if a log got too big, in particular if it hit a file
7029 size limit. Exim now panics and dies if it cannot write to a log file, just
7030 as it does if it cannot open a log file.
7032 39. Modified OS/Makefile-Linux so that it now contains
7034 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7036 The two -D definitions ensure that Exim is compiled with large file
7037 support, which makes it possible to handle log files that are bigger than
7040 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7041 instance) a domain was checked against a named list that involved a lookup,
7042 causing $domain_data to be set, then another domain was checked against the
7043 same list, then the first domain was re-checked, the value of $domain_data
7044 after the final check could be wrong. In particular, if the second check
7045 failed, it could be set empty. This bug probably also applied to
7048 41. The strip_trailing_dot option was not being applied to the address given
7049 with the -f command-line option.
7051 42. The code for reading a message's header from the spool was incrementing
7052 $received_count, but never initializing it. This meant that the value was
7053 incorrect (doubled) while delivering a message in the same process in which
7054 it was received. In the most common configuration of Exim, this never
7055 happens - a fresh exec is done - but it can happen when
7056 deliver_drop_privilege is set.
7058 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7059 it now includes up to 150 characters of the unexpected data in the log
7062 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7063 and building data strings. The size of both of these buffers was 10 000
7064 bytes - far larger than anybody would *ever* want, thought I. Needless to
7065 say, somebody hit the limit. I have increased the maximum line length to
7066 20 000 and the maximum data length of concatenated lines to 100 000. I have
7067 also fixed two bugs, because there was no checking on these buffers. Tsk,
7068 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7069 buffer is too small.
7071 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7072 lsearch lookups. Now it does.
7074 46. When parsing a route_list item in a manualroute router, a fixed-length
7075 buffer was used for the list of hosts. I made this 1024 bytes long,
7076 thinking that nobody would ever have a list of hosts that long. Wrong.
7077 Somebody had a whole pile of complicated expansion conditions, and the
7078 string was silently truncated, leading to an expansion error. It turns out
7079 that it is easier to change to an unlimited length (owing to other changes
7080 that have happened since this code was originally written) than to build
7081 structure for giving a limitation error. The length of the item that
7082 expands into the list of hosts is now unlimited.
7084 47. The lsearch lookup could not handle data where the length of text line was
7085 more than 4095 characters. Such lines were truncated, leading to shortened
7086 data being returned. It should now handle lines of any length.
7088 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7089 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7092 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7093 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7094 no longer happens in comment lines. A list of replacements is now placed
7095 at the head of all of the source files, except those whose only change is
7096 to replace PERL_COMMAND in the very first #! line.
7098 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7099 messages on the queue, with a bottom-up merge sort, using code contributed
7100 by Michael Haardt. This should make operations like -bp somewhat faster on
7101 large queues. It won't affect queue runners, except when queue_run_in_order
7104 51. Installed eximstats 1.31 in the distribution.
7106 52. Added support for SRV lookups to the dnslookup router.
7108 53. If an ACL referred to $message_body or $message_body_end, the value was not
7109 reset for any messages that followed in the same SMTP session.
7111 54. The store-handling optimization for building very long strings was not
7112 differentiating between the different store pools. I don't think this
7113 actually made any difference in practice, but I've tidied it.
7115 55. While running the routers to verify a sender address, $sender_address
7116 was still set to the sender address. This is wrong, because when routing to
7117 send a bounce to the sender, it would be empty. Therefore, I have changed
7118 it so that, while verifying a sender address, $sender_address is set to <>.
7119 (There is no change to what happens when verifying a recipient address.)
7121 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7122 target A or AAAA records (if not already returned) without resetting the
7123 qualify_single or search_parents options of the DNS resolver. These are
7124 inappropriate in this case because the targets of MX and SRV records must
7125 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7126 target that, when qualified, matched something in the local domain. These
7127 two options are now turned off when doing these lookups.
7129 57. It seems that at least some releases of Reiserfs (which does not have the
7130 concept of a fixed number of inodes) returns zero and not -1 for the
7131 number of available inodes. This interacted badly with check_spool_inodes,
7132 which assumed that -1 was the "no such thing" setting. What I have done is
7133 to check that the total number of inodes is greater than zero before doing
7134 the test of how many are available.
7136 58. When a "warn" ACL statement has a log_message modifier, the message is
7137 remembered, and not repeated. This is to avoid a lot of repetition when a
7138 message has many recipients that cause the same warning to be written.
7139 However, Exim was preserving the list of already written lines for an
7140 entire SMTP session, which doesn't seem right. The memory is now reset if a
7141 new message is started.
7143 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7144 debugging output unless log_rewrite was also set.
7146 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7147 of (int)(handle) when we know that handle contains (void *)(-1).
7149 61. The Exim daemon panic-logs an error return when it closes the incoming
7150 connection. However "connection reset by peer" seems to be common, and
7151 isn't really an error worthy of noting specially, so that particular error
7154 62. When Exim is trying to find all the local interfaces, it used to panic and
7155 die if the ioctl to get the interface flags failed. However, it seems that
7156 on at least one OS (Solaris 9) it is possible to have an interface that is
7157 included in the list of interfaces, but for which you get a failure error
7158 for this call. This happens when the interface is not "plumbed" into a
7159 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7160 failure of the "get flags" call assumes that the interface is down.
7162 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7163 makes life easier for people who are doing arithmetic on fields extracted
7164 from dates, where you often get leading zeros that should not be
7165 interpreted as octal.
7167 64. Added qualify_domain to the redirect router, to override the global
7170 65. If a pathologically long header line contained very many addresses (the
7171 report of this problem mentioned 10 000) and each of them was rewritten,
7172 Exim could use up a very large amount of memory. (It kept on making new
7173 copies of the header line as it rewrote, and never released the old ones.)
7174 At the expense of a bit more processing, the header rewriting function has
7175 been changed so that it no longer eats memory in this way.
7177 66. The generation of the Received: header has been moved from the time that a
7178 message starts to be received, to the time that it finishes. The timestamp
7179 in the Received: header should now be very close to that of the <= log
7180 line. There are two side-effects of this change:
7182 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7183 logged header lines no longer include the local Received: line, because
7184 it has not yet been created. The same applies to a copy of the message
7185 that is returned to a non-SMTP sender when a message is rejected.
7187 (b) When a filter file is tested using -bf, no additional Received: header
7188 is added to the test message. After some thought, I decided that this
7191 This change does not affect the value of $received_for. It is still set
7192 after address rewriting, but before local_scan() is called.
7194 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7196 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7197 gave an unhelpful panic error message, and a defer error. I have managed to
7198 change this behaviour so that it now rejects any supplied certificate,
7199 which seems right, as the list of acceptable certificates is empty.
7201 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7202 gave an unhelpful defer error. I have not managed to make this reject any
7203 supplied certificates, but the error message it gives is "no certificate
7204 supplied", which is not helpful.
7206 70. exigrep's output now also includes lines that are not associated with any
7207 message, but which match the given pattern. Implemented by a patch from
7208 Martin Sluka, which also tidied up the Perl a bit.
7210 71. Recipient callout verification, like sender verification, was using <> in
7211 the MAIL FROM command. This isn't really the right thing, since the actual
7212 sender may affect whether the remote host accepts the recipient or not. I
7213 have changed it to use the actual sender in the callout; this means that
7214 the cache record is now keyed on a recipient/sender pair, not just the
7215 recipient address. There doesn't seem to be a real danger of callout loops,
7216 since a callout by the remote host to check the sender would use <>.
7217 [SEE ABOVE: changed after hitting problems.]
7219 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7220 temporary errors. However, in the case of such a code being given after
7221 the end of a data transmission (i.e. after ".") Exim was failing to write
7222 a retry record for the message. (Yes, there was some broken host that was
7223 actually sending 8xx at this point.)
7225 73. An unknown lookup type in a host list could cause Exim to panic-die when
7226 the list was checked. (An example that provoked this was putting <; in the
7227 middle of a list instead of at the start.) If this happened during a DATA
7228 ACL check, a -D file could be left lying around. This kind of configuration
7229 error no longer causes Exim to die; instead it causes a defer error. The
7230 incident is still logged to the main and panic logs.
7232 74. Buglet left over from Exim 3 conversion. The message "too many messages
7233 in one connection" was written to the rejectlog but not the mainlog, except
7234 when address rewriting (yes!) was being logged.
7236 75. Added write_rejectlog option.
7238 76. When a system filter was run not as root (that is, when system_filter_user
7239 was set), the values of the $n variables were not being returned to the
7240 main process; thus, they were not subsequently available in the $sn
7243 77. Added +return_path_on_delivery log selector.
7245 78. A connection timeout was being treated differently from recipients deferred
7246 when testing hosts_max_try with a message that was older than the host's
7247 retry timeout. (The host should not be counted, thus allowing all hosts to
7248 be tried at least once before bouncing.) This may have been the cause of an
7249 occasionally reported bug whereby a message would remain on the queue
7250 longer than the retry timeout, but would be bounced if a delivery was
7251 forced. I say "may" because I never totally pinned down the problem;
7252 setting up timeout/retry tests is difficult. See also the next item.
7254 79. The ultimate address timeout was not being applied to errors that involved
7255 a combination of host plus message (for example, a timeout on a MAIL
7256 command). When an address resolved to a number of possible hosts, and they
7257 were not all tried for each delivery (e.g. because of hosts_max_try), a
7258 message could remain on the queue longer than the retry timeout.
7260 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7263 81. Fixed an obscure SMTP outgoing bug which required at least the following
7264 conditions: (a) there was another message waiting for the same server;
7265 (b) the server returned 5xx to all RCPT commands in the first message so
7266 that the message was not completed; (c) the server dropped the connection
7267 or gave a negative response to the RSET that Exim sends to abort the
7268 transaction. The observed case was a dropped connection after DATA that had
7269 been sent in pipelining mode. That is, the server had advertised PIPELINING
7270 but was not implementing it correctly. The effect of the bug was incorrect
7271 behaviour, such as trying another host, and this could lead to a crash.
7277 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7278 and daemon.c were passed as pointers to ints; they should have been
7279 pointers to socklen_t variables (which are typically unsigned ints).
7281 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7284 3. Fixed a really odd bug that affected only the testing scheme; patching a
7285 certain fixed string in the binary changed the value of another string that
7286 happened to be identical to the end of the original first string.
7288 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7289 name", it returns that address as the IP address. On some operating
7290 systems (e.g. Solaris), it also passes back the IP address string as the
7291 "host name". However, on others (e.g. Linux), it passes back an empty
7292 string. Exim wasn't checking for this, and was changing the host name to an
7293 empty string, assuming it had been canonicalized.
7295 5. Although rare, it is permitted to have more than one PTR record for a given
7296 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7297 all the names associated with an address, because they do in Solaris.
7298 However, it seems that they do not in Linux for data that comes from the
7299 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7300 I found this out when I moved to a new Linux workstation and tried to run
7301 the Exim test suite.
7303 To get round this problem I have changed the code so that it now does its
7304 own call to the DNS to look up PTR records when searching for a host name.
7305 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7306 addresses that are only in /etc/hosts are still found.
7308 This behaviour is, however, controlled by an option called host_lookup_
7309 order, which defaults to "bydns:byaddr". If people want to use the other
7310 order, or indeed, just use one or the other means of lookup, they can
7311 specify it in this variable.
7313 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7314 some operating systems, this comes back from gethostbyaddr() as an empty
7315 string, and this is what Exim used to test for. However, it seems that in
7316 other systems, "." is yielded. Exim now tests for this case too.
7318 7. The values of check_spool_space and check_log_space are now held internally
7319 as a number of kilobytes instead of an absolute number of bytes. If a
7320 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7321 kilobyte. This means that much larger values can be stored.
7323 8. Exim monitor: an attempt to get the action menu when not actually pointing
7324 at a message produces an empty menu entitled "No message selected". This
7325 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7326 no entries in it ("Shell widget menu has zero width and/or height"). So I
7327 have added a single, blank menu entry in this case.
7329 9. Added ${quote_local_part.
7331 10. MIME decoding is now applied to the contents of Subject: header lines when
7334 11. Now that a reference to $sender_host_address automatically causes a reverse
7335 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7336 host lookup before query-style lookups in lists that might use this
7337 variable. This has therefore been abolished, and the "net-" prefix is no
7338 longer necessary for query-style lookups.
7340 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7341 have been a typo for LFLAGS, so it has been changed.
7343 13. The install script calls Exim with "-C /dev/null" in order to find the
7344 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7345 to be output. However, since Exim outputs its version number before the
7346 error, it didn't break the script. It just looked ugly. I fixed this by
7347 always allowing "-C /dev/null" if the caller is root.
7349 14. Ignore overlarge ACL variable number when reading spool file - insurance
7350 against a later release with more variables having written the file.
7352 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7353 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7354 this, as well as the form without the "IPv6" on the front.
7356 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7357 OS/Makefile-Darwin file.
7359 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7360 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7362 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7363 variable. However, some OS use ints for inodes. I've added cast to long int
7364 to get rid of the compiler warning.
7366 19. I had forgotten to lock out "/../" in configuration file names when
7367 ALT_CONFIG_PREFIX was set.
7369 20. Routers used for verification do not need to specify transports. However,
7370 if such a router generated a host list, and callout was configured, Exim
7371 crashed, because it could not find a port number from the (non-existent)
7372 transport. It now assumes port 25 in this circumstance.
7374 21. Added the -t option to exigrep.
7376 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7377 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7378 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7379 them is set, LOOKUP_LSEARCH is forced.
7381 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7383 24. Added sender and host information to the "rejected by local_scan()" log
7384 line; previously there was no indication of these.
7386 25. Added .include_if_exists.
7388 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7389 Linux. It turns out that not all file systems support this. Apparently some
7390 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7391 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7392 not supported on the file descriptor, is now ignored when Exim is trying to
7393 sync a directory. This applies only to Linux.
7395 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7397 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7398 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7401 29. Added support for maildirsize files from supplied patch (modified a bit).
7403 30. The use of :fail: followed by an empty string could lead Exim to respond to
7404 sender verification failures with (e.g.):
7406 550 Verification failed for <xxx>
7407 550 Sender verify failed
7409 where the first response line was missing the '-' that indicates it is not
7410 the final line of the response.
7412 31. The loop for finding the name of the user that called Exim had a hardwired
7413 limit of 10; it now uses the value of finduser_retries, which is used for
7414 all other user lookups.
7416 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7419 33. Exim was neglecting to zero errno before one call of strtol() when
7420 expanding a string and expecting an integer value. On some systems this
7421 resulted in spurious "integer overflow" errors. Also, it was casting the
7422 result into an int without checking.
7424 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7425 did not work. The code looks as if it has *never* worked, though it appears
7426 to have been documented since at least release 1.62. I have made it work.
7428 35. The "timeout_DNS" error in retry rules, also documented since at least
7429 1.62, also never worked. As it isn't clear exactly what this means, and
7430 clearly it isn't a major issue, I have abolished the feature by treating it
7431 as "timeout", and writing a warning to the main and panic logs.
7433 36. The display of retry rules for -brt wasn't always showing the error code
7436 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7437 timeout_connect_A, timeout_connect_MX.
7439 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7440 to the empty sender.
7442 39. The daemon was not analysing the content of -oX till after it had closed
7443 stderr and disconnected from the controlling terminal. This meant that any
7444 syntax errors were only noted on the panic log, and the return code from
7445 the command was 0. By re-arranging the code a little, I've made the
7446 decoding happen first, so such errors now appear on stderr, and the return
7447 code is 1. However, the actual setting up of the sockets still happens in
7448 the disconnected process, so errors there are still only recorded on the
7451 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7452 connections (as happens on some IP stacks) was logged at start up time as
7453 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7454 it from "IPv6 and IPv4", which means that two separate sockets are being
7457 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7458 says whether AF_INET or AF_INET6 was passed as an argument.
7460 42. Exiwhat output was messed up when time zones were included in log
7463 43. Exiwhat now gives more information about the daemon's listening ports,
7464 and whether -tls-on-connect was used.
7466 44. The "port" option of the smtp transport is now expanded.
7468 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7469 silently ignored. Now an error message is written to the main and panic
7472 46. There's a new ACL modifier called "logwrite" which writes to a log file
7473 as soon as it is encountered.
7475 47. Added $local_user_uid and $local_user_gid at routing time.
7477 48. Exim crashed when trying to verify a sender address that was being
7480 49. Exim was recognizing only a space character after ".include". It now also
7481 recognizes a tab character.
7483 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7484 extracting the relevant information from the specification. The man page no
7485 longer contains scrambled data for the -d option, and I've added a section
7486 at the front about calling Exim under different names.
7488 51. Added "extra_headers" argument to the "mail" command in filter files.
7490 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7493 53. Installed eximstats 1.29.
7495 54. Added transport_filter_timeout as a generic transport option.
7497 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7498 Cc: header lines. This was required by RFC 822, but it not required by RFC
7501 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7502 incoming messages that did not have them. Now it does so only if the
7503 message originates locally, that is, if there is no associated remote host
7504 address. When Resent- header lines are present, this applies to the Resent-
7505 lines rather than the non-Resent- lines.
7507 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7508 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7510 58. Messages for configuration errors now include the name of the main
7511 configuration file - useful now that there may be more than one file in a
7512 list (.included file names were always shown).
7514 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7515 for those rare installations that do not start the daemon as root or run it
7516 setuid root. I've cut out the call to initgroups() if the daemon is not
7519 60. The Exim user and group can now be bound into the binary as text strings
7520 that are looked up at the start of Exim's processing.
7522 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7524 62. Added $mailstore_basename variable.
7526 63. Installed patch to sieve.c from Michael Haardt.
7528 64. When Exim failed to open the panic log after failing to open the main log,
7529 the original message it was trying to log was written to stderr and debug
7530 output, but if they were not available (the usual case in production), it
7531 was lost. Now it is written to syslog before the two lines that record the
7532 failures to open the logs.
7534 65. Users' Exim filters run in subprocesses under the user's uid. It is
7535 possible for a "deliver" command or an alias in a "personal" command to
7536 provoke an address rewrite. If logging of address rewriting is configured,
7537 this fails because the process is not running as root or exim. There may be
7538 a better way of dealing with this, but for the moment (because 4.30 needs
7539 to be released), I have disabled address rewrite logging when running a
7540 filter in a non-root, non-exim process.
7546 1. The buildconfig auxiliary program wasn't quoting the value set for
7547 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7548 not defined. This bug was masked in 4.22 by the effect that was fixed in
7551 2. Some messages that were rejected after a message id was allocated were
7552 shown as "incomplete" by exigrep. It no longer does this for messages that
7553 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7555 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7556 have allow_domain_literals set, the ID did not get logged in the <= line.
7557 Domain literals are now always recognized in Message-ID: header lines.
7559 4. The first argument for a ${extract expansion item is the key name or field
7560 number. Leading and trailing spaces in this item were not being ignored,
7561 causing some misleading effects.
7563 5. When deliver_drop_privilege was set, single queue runner processes started
7564 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7565 same command in the process it spins off) were not dropping privilege.
7567 6. When the daemon running as "exim" started a queue runner, it always
7568 re-executed Exim in the spun-off process. This is a waste of effort when
7569 deliver_drop_privilege is set. The new process now just calls the
7570 queue-runner function directly.
7576 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7579 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7580 ignored. Though the use of -oP was forcing the writing of a pid file, it
7581 was always written to the default place.
7583 3. If the message "no IP address found for host xxxx" is generated during
7584 incoming verification, it is now followed by identification of the incoming
7585 connection (so you can more easily find what provoked it).
7587 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7589 5. Added some features to "harden" Exim a bit more against certain attacks:
7591 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7592 be put in Local/Makefile. This is like the never_users runtime option,
7593 but it cannot be overridden. The default setting is "root".
7595 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7596 prefix string with which any file named in a -C command line option
7599 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7600 is retained for -C and -D only if the caller of Exim is root. Without
7601 it, the exim user may also use -C and -D and retain privilege.
7603 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7604 command line option is disabled.
7606 6. Macro names set by the -D option must start with an upper case letter, just
7607 like macro names defined in the configuration file.
7609 7. Added "dereference=" facility to LDAP.
7611 8. Two instances of the typo "uknown" in the source files are fixed.
7613 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7614 the Configure-Makefile script screwed up while processing it.
7616 10. Incorporated PCRE 4.4.
7618 11. The SMTP synchronization check was not operating right at the start of an
7619 SMTP session. For example, it could not catch a HELO sent before the client
7620 waited for the greeting. There is now a check for outstanding input at the
7621 point when the greeting is written. Because of the duplex, asynchronous
7622 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7623 way, but not yet received, when the check is performed.
7625 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7626 on TCP/IP sockets, because this apparently causes some broken clients to
7629 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7630 unchanged) from the Cygwin maintainer.
7632 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7633 was supported instead of testing for mbx. Effectively a typo.
7635 15. The spa authenticator server code was not checking that the input it
7636 received was valid base64.
7638 16. The debug output line for the "set" modifier in ACLs was not showing the
7639 name of the variable that was being set.
7641 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7643 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7644 lookup on its own. Something else was needed to trigger the lookup. For
7645 example, a match in host_lookup or the need for a host name in a host list.
7646 Now, if $sender_host_name is referenced and the host name has not yet been
7647 looked up, a lookup is performed. If the lookup fails, the variable remains
7648 empty, and $host_lookup_failed is set to "1".
7650 19. Added "eqi" as a case-independent comparison operator.
7652 20. The saslauthd authentication condition could segfault if neither service
7653 nor realm was specified.
7655 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7656 error message that was logged was misleading, and incoming SMTP
7657 connections were dropped. The message is now more accurate, and temporary
7658 errors are given to SMTP connections.
7660 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7661 (except RSET) with a 421 error, until QUIT is received. However, it was
7662 failing to send a response to QUIT.
7664 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7665 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7666 ACL is now run after the helo_try_verify_hosts code.
7668 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7669 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7670 case-independent, so other case variants are also recognized). Apparently
7671 some systems use these upper case variants.
7673 25. If more than two messages were waiting for the same host, and a transport
7674 filter was specified for the transport, Exim sent two messages over the
7675 same TCP/IP connection, and then failed with "socket operation on non-
7676 socket" when it tried to send the third.
7678 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7680 27. The extern definition of crypt16() in expand.c was not being excluded when
7681 the OS had its own crypt16() function.
7683 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7684 as a preferred synonym for return_size_limit, both as an option and as an
7687 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7689 30. Changed the default configuration ACL to relax the local part checking rule
7690 for addresses that are not in any local domains. For these addresses,
7691 slashes and pipe symbols are allowed within local parts, but the sequence
7692 /../ is explicitly forbidden.
7694 31. SPA server authentication was not clearing the challenge buffer before
7697 32. log_message in a "warn" ACL statement was writing to the reject log as
7698 well as to the main log, which contradicts the documentation and doesn't
7699 seem right (because no rejection is happening). So I have stopped it.
7701 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7702 However, I am unable to do any testing of this.
7704 34. Fixed an infelicity in the appendfile transport. When checking directories
7705 for a mailbox, to see if any needed to be created, it was accidentally
7706 using path names with one or more superfluous leading slashes; tracing
7707 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7709 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7710 contents are added to the log line that is written for every discarded
7711 recipient. (Previously a log_message setting was ignored.)
7713 36. The ${quote: operator now quotes the string if it is empty.
7715 37. The install script runs exim in order to find its version number. If for
7716 some reason other than non-existence or emptiness, which it checks, it
7717 could not run './exim', it was installing it with an empty version number,
7718 i.e. as "exim-". This error state is now caught, and the installation is
7721 38. An argument was missing from the function that creates an error message
7722 when Exim fails to connect to the socket for saslauthd authentication.
7723 This could cause Exim to crash, or give a corrupted message.
7725 39. Added isip, isip4, and isip6 to ${if conditions.
7727 40. The ACL variables $acl_xx are now saved with the message, and can be
7728 accessed later in routers, transports, and filters.
7730 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7731 strings in the file are not string-expanded.
7733 42. If a MAIL command specified a SIZE value that was too large to fit into an
7734 int variable, the check against message_size_limit failed. Such values are
7735 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7736 day this will have to be increased, but I don't think I want to be around
7737 when emails are that large.
7744 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7745 iconv() is not standard in FreeBSD.
7747 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7748 IPv6 enabled. The observed symptom was a segmentation fault on return from
7749 the function os_common_find_running_interfaces() in src/os.c.
7751 3. In the check_special_case() function in daemon.c I had used "errno" as an
7752 argument name, which causes warnings on some systems. This was basically a
7753 typo, since it was named "eno" in the comments!
7755 4. The code that waits for the clock to tick (at a resolution of some fraction
7756 of a second) so as to ensure message-id uniqueness was always waiting for
7757 at least one whole tick, when it could have waited for less. [This is
7758 almost certainly not relevant at current processor speeds, where it is
7759 unlikely to ever wait at all. But we try to future-proof.]
7761 5. The function that sleeps for a time interval that includes fractions of a
7762 second contained a race. It did not block SIGALRM between setting the
7763 timer, and suspending (a couple of lines later). If the interval was short
7764 and the sigsuspend() was delayed until after it had expired, the suspension
7765 never ended. On busy systems this could lead to processes getting stuck for
7768 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7769 process, before it forks any delivery processes. The open lookup caching
7770 mechanism meant that the open file or database connection was passed into
7771 the delivery process. The problem was that delivery processes always tidy
7772 up cached lookup data. This could cause a problem for the next delivery
7773 process started by the queue runner, because the external queue runner
7774 process does not know about the closure. So the next delivery process
7775 still has data in the lookup cache. In the case of a file lookup, there was
7776 no problem because closing a file descriptor in a subprocess doesn't affect
7777 the parent. However, if the lookup was caching a connection to a database,
7778 the connection was closed, and the second delivery process was likely to
7779 see errors such as "PGSQL: query failed: server closed the connection
7780 unexpectedly". The problem has been fixed by closing all cached lookups
7781 in a queue runner before running a delivery process.
7783 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7784 seem to have the "const" qualifier which it has on other OS. I've
7787 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7788 *of the same type* (client or server) with the same public name that an
7789 error should be diagnosed.
7791 9. When Exim looked up a host name for an IP address, but failed to find the
7792 original IP address when looking up the host name (a safety check), it
7793 output the message "<ip address> does not match any IP for NULL", which was
7794 confusing, to say the least. The bug was that the host name should have
7795 appeared instead of "NULL".
7797 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7798 user that is built into the binary, and the -C or -D options is used, root
7799 privilege is dropped before the configuration file is read. In addition,
7800 logging is switched to stderr instead of the normal log files. If the
7801 configuration then re-defines the Exim user, the unprivileged environment
7802 is probably not what is expected, so Exim logs a panic warning message (but
7805 However, if deliver_drop_privilege is set, the unprivileged state may well
7806 be exactly what is intended, so the warning has been cut out in that case,
7807 and Exim is allowed to try to write to its normal log files.
7813 1. smtp_return_error_details was not giving details for temporary sender
7814 or receiver verification errors.
7816 2. Diagnose a configuration error if two authenticators have the same public
7819 3. Exim used not to create the message log file for a message until the first
7820 delivery attempt. This could be confusing when incoming messages were held
7821 for policy or load reasons. The message log file is now created at the time
7822 the message is received, and an initial "Received" line is written to it.
7824 4. The automatically generated man page for command line options had a minor
7825 bug that caused no ill effects; however, a more serious problem was that
7826 the procedure for building the man page automatically didn't always
7827 operate. Consequently, release 4.20 contains an out-of-date version. This
7828 shouldn't happen again.
7830 5. When building Exim with embedded Perl support, the script that builds the
7831 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7832 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7834 6. The freeze_tell option was not being used for messages that were frozen on
7835 arrival, either by an ACL or by local_scan().
7837 7. Added the smtp_incomplete_transaction log selector.
7839 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7840 was accepting AUTH without a new EHLO.
7842 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7843 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7846 10. Logging of TCP/IP connections (when configured) now happens in the main
7847 daemon process instead of the child process, so that the TCP/IP connection
7848 count is more accurate (but it can never be perfect).
7850 11. The use of "drop" in a nested ACL was not being handled correctly in the
7851 outer ACL. Now, if condition failure induced by the nested "drop" causes
7852 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7853 or "require"), the connection is dropped.
7855 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7856 that yield "discard" can now be used with an "accept" or a "discard" verb,
7857 but an error is generated for any others (because I can't see a useful way
7858 to define what should happen).
7860 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7861 are now processed in the same way as lines in the Exim configuration file.
7862 In particular, continuation lines are supported.
7864 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7866 15. Added -ti meaning -t -i.
7868 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7869 domains, and warn by logging if others are found.
7871 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7872 returned from GIFCONF when Exim is trying to find the list of interfaces on
7873 a host. The code in os.c has been modified to copy each ifreq to an aligned
7874 structure in all cases.
7876 Also, in some cases, the returned ifreq's were being copied to a 'struct
7877 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7878 means the last couple of bytes of an IPv6 address could be chopped if the
7879 ifreq contained only a normal sockaddr (14 bytes storage).
7881 18. Named domain lists were not supported in the hosts_treat_as_local option.
7882 An entry such as +xxxx was not recognized, and was treated as a literal
7885 19. Ensure that header lines added by a DATA ACL are included in the reject log
7886 if the ACL subsequently rejects the message.
7888 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7889 MD5 (which is deprecated).
7891 21. When testing a filter file using -bf, Exim was writing a message when it
7892 took the sender from a "From " line in the message, but it was not doing so
7893 when it took $return_path from a Return-Path: header line. It now does.
7895 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7896 with a valid header line field name (a series of printing characters
7897 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7899 23. Changed "disc" in the source to "disk" to conform to the documentation and
7900 the book and for uniformity.
7902 24. Ignore Sendmail's -Ooption=value command line item.
7904 25. When execve() failed while trying to run a command in a pipe transport,
7905 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7906 could be confused with a return value of 69 from the command itself. This
7907 has been changed to 127, the value the shell returns if it is asked to run
7908 a non-existent command. The wording for the related log line suggests a
7909 non-existent command as the problem.
7911 26. If received_header_text expands to an empty string, do not add a Received:
7912 header line to the message. (Well, it adds a token one on the spool, but
7913 marks it "old" so that it doesn't get used or transmitted.)
7915 27. Installed eximstats 1.28 (addition of -nt option).
7917 28. There was no check for failure on the call to getsockname() in the daemon
7918 code. This can fail if there is a shortage of resources on the system, with
7919 ENOMEM, for example. A temporary error is now given on failure.
7921 29. Contrary to the C standard, it seems that in some environments, the
7922 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7923 program. Exim now does this explicitly; it affects the formatting of
7924 timestamps using strftime().
7926 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7927 complaints. I've now initialized all the variables, to avoid this.
7929 32. Header lines added by a system filter were not being "seen" during
7930 transport-time rewrites.
7932 33. The info_callback() function passed to OpenSSL is set up with type void
7933 (*)(SSL *, int, int), as described somewhere. However, when calling the
7934 function (actually a macro) that sets it up, the type void(*)() is
7935 expected. I've put in a cast to prevent warnings from picky compilers.
7937 34. If a DNS black list lookup found a CNAME record, but there were no A
7938 records associated with the domain it pointed at, Exim crashed.
7940 35. If a DNS black list lookup returned more than one A record, Exim ignored
7941 all but the first. It now scans all returned addresses if a particular IP
7942 value is being sought. In this situation, the contents of the
7943 $dnslist_value variable are a list of all the addresses, separated by a
7946 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7947 to accept a host name and all its aliases if the forward lookup for any of
7948 them yielded the IP address of the incoming connection. Now it accepts only
7949 those names whose forward lookup yields the correct IP address. Any other
7950 names are discarded. This closes a loophole whereby a rogue DNS
7951 administrator could create reverse DNS records to break through a
7952 wildcarded host restriction in an ACL.
7954 37. If a user filter or a system filter that ran in a subprocess used any of
7955 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7956 the wrong values were passed to the pipe command ($thisaddress had the
7957 value of $0, $0 had the value of $1, etc). This bug was introduced by
7958 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7960 38. Improved the line breaking for long SMTP error messages from ACLs.
7961 Previously, if there was no break point between 40 and 75 characters, Exim
7962 left the rest of the message alone. Two changes have been made: (a) I've
7963 reduced the minimum length to 35 characters; (b) if it can't find a break
7964 point between 35 and 75 characters, it looks ahead and uses the first one
7965 that it finds. This may give the occasional overlong line, but at least the
7966 remaining text gets split now.
7968 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7969 file descriptors might be low, and that setting 1000 would always raise it.
7970 It turns out that in some environments, the limit is already over 1000 and
7971 that lowering it causes trouble. So now Exim takes care not to decrease it.
7973 40. When delivering a message, the value of $return_path is set to $sender_
7974 address at the start of routing (routers may change the value). By an
7975 oversight, this default was not being set up when an address was tested by
7976 -bt or -bv, which affected the outcome if any router or filter referred to
7979 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7980 log only when "message" or "log_message" are set. However, if one of the
7981 conditions was an address verification, or a call to a nested ACL, the
7982 messages generated by the underlying test were being passed through. This
7983 no longer happens. The underlying message is available in $acl_verify_
7984 message for both "message" and "log_message" expansions, so it can be
7985 passed through if needed.
7987 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7988 new expansion $bh_ to give the encoded byte string without charset
7989 translation. Translation happens only if iconv() is available; HAVE_ICONV
7990 indicates this at build time. HEADERS_CHARSET gives the charset to
7991 translate to; headers_charset can change it in the configuration, and
7992 "headers charset" can change it in an individual filter file.
7994 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7995 that creates RFC 2047 encoded "words" labels them as that charset instead
7996 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7997 expansion operator; (ii) when Exim creates a From: line for a local
7998 message; (iii) when a header line is rewritten to include a "phrase" part.
8000 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8001 buggy, causing it to skip the first lines of messages whose message ID
8002 ended in 'D'. This would not have bitten before Exim release 4.14, because
8003 message IDs were unlikely to end in 'D' before then. The effect was to have
8004 incorrect size information for certain domains.
8006 45. #include "config.h" was missing at the start of the crypt16.c module. This
8007 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8010 46. If there was a timeout during a "random" callout check, Exim treated it as
8011 a failure of the random address, and carried on sending RSET and the real
8012 address. If the delay was just some slowness somewhere, the response to the
8013 original RCPT would be taken as a response to RSET and so on, causing
8014 mayhem of various kinds.
8016 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8017 when I implemented it. It didn't allow for the fact that some option values
8018 may legitimately be negative (e.g. size_addition), and it didn't even do
8019 the right test for positive values.
8021 48. Domain names in DNS records are case-independent. Exim always looks them up
8022 in lower case. Some resolvers return domain names in exactly the case they
8023 appear in the zone file, that is, they may contain uppercase letters. Not
8024 all resolvers do this - some return always lower case. Exim was treating a
8025 change of case by a resolver as a change of domain, similar to a widening
8026 of a domain abbreviation. This triggered its re-routing code and so it was
8027 trying to route what was effectively the same domain again. This normally
8028 caused routing to fail (because the router wouldn't handle the domain
8029 twice). Now Exim checks for this case specially, and just changes the
8030 casing of the domain that it ultimately uses when it transmits the message
8033 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8036 50. If a filter generated a file delivery with a non-absolute name (possible if
8037 no home directory exists for the router), the forbid_file option was not
8040 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8041 the existing equality matching.
8043 52. Exim was using ints instead of ino_t variables in some places where it was
8044 dealing with inode numbers.
8046 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8047 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8048 called TMPDIR, and if it finds it is different, it changes its value.
8050 54. The smtp_printf() function is now made available to local_scan() so
8051 additional output lines can be written before returning. There is also an
8052 smtp_fflush() function to enable the detection of a dropped connection.
8053 The variables smtp_input and smtp_batched_input are exported to
8056 55. Changed the default runtime configuration: the message "Unknown user"
8057 has been removed from the ACL, and instead placed on the localuser router,
8058 using the cannot_route_message feature. This means that any verification
8059 failures that generate their own messages won't get overridden. Similarly,
8060 the "Unrouteable address" message that was in the ACL for unverifiable
8061 relay addresses has also been removed.
8063 56. Added hosts_avoid_esmtp to the smtp transport.
8065 57. The exicyclog script was not checking for the esoteric option
8066 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8067 will work only if exicyclog is run under the appropriate euid.
8069 58. Following a discussion on the list, the rules by which Exim recognises line
8070 endings on incoming messages have been changed. The -dropcr and drop_cr
8071 options are now no-ops, retained only for backwards compatibility. The
8072 following line terminators are recognized: LF CRLF CR. However, special
8073 processing applies to CR:
8075 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8076 nor a local message in the state where . is a terminator.
8078 (ii) If a bare CR is encountered in a header line, an extra space is added
8079 after the line terminator so as not to end the header. The reasoning
8080 behind this is that bare CRs in header lines are most likely either
8081 to be mistakes, or people trying to play silly games.
8083 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8084 was being incorrectly given as 18 bytes larger than it should have been.
8085 This is a VOB (very old bug).
8087 60. This may never have affected anything current, but just in case it has:
8088 When the local host is found other than at the start of a list of hosts,
8089 the local host, those with the same MX, and any that follow, are discarded.
8090 When the list in question was part of a longer list of hosts, the following
8091 hosts (not currently being processed) were also being discarded. This no
8092 longer happens. I'm not sure if this situation could ever has previously
8095 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8098 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8099 header. This is something that is recommended in a new Internet Draft, and
8100 is something that is documented as being done by Sendmail. There are two
8101 possible values. For messages generated by the autoreply transport, Exim
8104 Auto-Submitted: auto-replied
8106 whereas for all other generated messages (e.g. bounces) it adds
8108 Auto-Submitted: auto-generated
8110 63. The "personal" condition in filters now includes a test for the
8111 Auto-Submitted: header. If it contains the string "auto-" the message it
8112 not considered personal.
8114 64. Added rcpt_include_affixes as a generic transport option.
8116 65. Added queue_only_override (default true).
8118 66. Added the syslog_duplication option.
8120 67. If what should have been the first header line of a message consisted of
8121 a space followed by a colon, Exim was mis-interpreting it as a header line.
8122 It isn't of course - it is syntactically invalid and should therefore be
8123 treated as the start of the message body. The misbehaviour could have
8124 caused a number of strange effects, including loss of data in subsequent
8125 header lines, and spool format errors.
8127 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8128 client host had authenticated. This control can now be exercised by an ACL
8129 for more flexibility.
8131 69. By default, callouts do not happen when testing with -bh. There is now a
8132 variant, -bhc, which does actually run the callout code, including
8133 consulting and updating the callout cache.
8135 70. Added support for saslauthd authentication, courtesy of Alexander
8138 71. If statvfs() failed on the spool or log directories while checking their
8139 size for availability, Exim confusingly gave the error "space shortage".
8140 Furthermore, in debugging mode it crashed with a floating point exception.
8141 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8142 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8143 serious problem, Exim now writes to the main and panic logs when this
8144 happens, with details of the failure. It then refuses to accept the
8145 incoming message, giving the message "spool directory problem" or "log
8146 directory problem" with a 421 code for SMTP messages.
8148 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8149 0, 1, and 2 exist, because some OS complain for execs without them (see
8150 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8151 descriptors. However, the code omitted to check that the open succeeded,
8152 causing mysterious errors if for some reason the permissions on /dev/null
8153 got screwed. Now Exim writes a message to the main and panic logs, and
8154 bombs out if it can't open /dev/null.
8156 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8157 interact so that it is all more flexible. It is supposed to remain
8158 backwards compatible. Also added extra_local_interfaces.
8160 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8161 to bomb out with an assertion failure - to the client this appears as a
8162 connection drop. This problem occurs in the part of the code that was taken
8163 from the Samba project. Fortunately, the assertion is in a very simple
8164 function, so I have fixed this by reproducing the function inline in the
8165 one place where it is called, and arranging for authentication to fail
8166 instead of killing the process with assert().
8168 75. The SPA client code was not working when the server requested OEM rather
8169 than Unicode encoding.
8171 76. Added code to make require_files with a specific uid setting more usable in
8172 the case where statting the file as root fails - usually a non-root-mounted
8173 NFS file system. When this happens and the failure is EACCES, Exim now
8174 forks a subprocess and does the per-uid checking as the relevant uid.
8176 77. Added process_log_path.
8178 78. If log_file_path was not explicitly set, a setting of check_log_space or
8179 check_log_inodes was ignored.
8181 79. If a space check for the spool or log partitions fails, the incident is now
8182 logged. Of course, in the latter case the data may get lost...
8184 80. Added the %p formatting code to string_format() so that it can be used to
8185 print addresses in debug_print(). Adjusted all the address printing in the
8186 debugging in store.c to use %p rather than %d.
8188 81. There was a concern that a line of code in smtp_in.c could overflow a
8189 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8190 initially expressed, the concern was not well-founded, because trailing
8191 spaces are removed early. However, if the trailing spaces were followed by
8192 a NULL, they did not get removed, so the overflow was possible. Two fixes
8195 (a) I re-wrote the offending code in a cleaner fashion.
8196 (b) If an incoming SMTP command contains a NULL character, it is rejected
8199 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8200 runs initgroups(), so that if the Exim user is in any additional groups,
8201 they will be used during message reception.
8207 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob